ويكاموس arwiktionary https://ar.wiktionary.org/wiki/%D9%88%D9%8A%D9%83%D8%A7%D9%85%D9%88%D8%B3:%D8%A7%D9%84%D8%B5%D9%81%D8%AD%D8%A9_%D8%A7%D9%84%D8%B1%D8%A6%D9%8A%D8%B3%D9%8A%D8%A9 MediaWiki 1.47.0-wmf.19 case-sensitive ميديا خاص نقاش مستخدم نقاش المستخدم ويكاموس نقاش ويكاموس ملف نقاش الملف ميدياويكي نقاش ميدياويكي قالب نقاش القالب مساعدة نقاش المساعدة تصنيف نقاش التصنيف ملحق نقاش الملحق مكنز نقاش المكنز نص زمني نقاش النص الزمني وحدة نقاش الوحدة فعالية نقاش فعالية 8 145139 1158950 1089986 2026-09-10T10:31:56Z حبيشان 31560 تنسيق + حذف ResourceLoader وفق التحديث الأخير 1158950 wikitext text/x-wiki == مخفي == * Site [default |hidden |type=styles] |Site.css * VE-wditem-input [hidden |dependencies=ext.visualEditor.core] |VE-wditem-input.js * auto-cite-wikidata [hidden |dependencies=ext.visualEditor.core] |auto-cite-wikidata.js * countdown-timer [default |hidden |dependencies=mediawiki.language |categories=Gadget-Countdown] |countdown-timer.js * SettingsManager [dependencies=mediawiki.user, user.options, mediawiki.api, mediawiki.util |hidden] |SettingsManager.js * libJQuery [hidden] |libJQuery.js |libJQuery.css * LoadJS [default |hidden |dependencies=mediawiki.util] |LoadJS.js == تحرير == * AjaxEdit [dependencies=mediawiki.util,mediawiki.Uri,mediawiki.user,mediawiki.api] |AjaxEdit.js * autocomplete [dependencies=mediawiki.util] |autocomplete.js * BiDiEditing |BiDiEditing.js * Cat-a-lot [rights=patrol |dependencies=mediawiki.util] |Cat-a-lot.js |Cat-a-lot.css * HotCat |HotCat.js * newsectionbottom [skins=vector] |newsectionbottom.js * searchlang [rights=minoredit |skins=vector |dependencies=mediawiki.util] |searchlang.js * speed-deletion [dependencies=mediawiki.util |rights=rollback] |speed-deletion.js * wikEd |wikEd.js * VE-wditem-input-loader [dependencies=ext.visualEditor.targetLoader] |VE-wditem-input-loader.js * auto-cite-wikidata-loader [default |dependencies=ext.visualEditor.targetLoader] |auto-cite-wikidata-loader.js == تصفح == * GreenRedirect [default |rights=minoredit] |GreenRedirect.css * HideExLinkIcon [default |rights=minoredit] |HideExLinkIcon.css * JSL |JSL.js * Linkscount |Linkscount.js * MobileCategories [dependencies=mediawiki.api,mediawiki.util |skins=minerva] |MobileCategories.js * ShortLink |ShortLink.js * userinfo-without-deletedcontribs [dependencies=mediawiki.util] |userinfo-without-deletedcontribs.js * UTCLiveClock [dependencies=mediawiki.util] |UTCLiveClock.js * mobile-sidebar [skins=vector, vector-2022] |mobile-sidebar.js |mobile-sidebar.css * PagePreviews [default |dependencies=mediawiki.api,mediawiki.util,mediawiki.Title |skins=vector,vector-2022,monobook,timeless,modern,cologneblue] |PagePreviews.js * XTools-ArticleInfo [skins=vector, vector-2022, monobook, timeless] |XTools-ArticleInfo.js * Numeral_converter [dependencies=mediawiki.cookie, mediawiki.util, mediawiki.user] |Numeral_converter.js * Numeral-converter-toggle |Numeral-converter-toggle.js |Numeral-converter-toggle.css == تحت-الاختبار== rmt7qc4mfh3refzcgsjp2wsoaxx5p3b 828 188226 1158927 1158923 2026-09-09T12:58:33Z ForzaGreen 28665 #12: block comment -- the third way an orthographic exemption can be written 1158927 Scribunto text/plain local export = {} --[=[ This module implements {{ar-conj}} and provides the underlying conjugation functions for {{ar-verb}} (whose actual formatting is done in [[Module:ar-headword]]). Author: User:Benwing, from an early version (2013-2014) by User:Atitarev, User:ZxxZxxZ. ]=] --[=[ TERMINOLOGY: -- "slot" = A particular combination of tense/mood/person/number/etc. Example slot names for verbs are "past_1s" (past tense first-person singular), "juss_pass_3fp" (non-past jussive passive third-person feminine plural) "ap" (active participle). Each slot is filled with zero or more forms. -- "form" = The conjugated Arabic form representing the value of a given slot. -- "lemma" = The dictionary form of a given Arabic term. For Arabic, normally the third person masculine singular past, although other forms may be used if this form is missing (e.g. in passive-only verbs or verbs lacking the past). ]=] --[=[ FIXME: 1. Finish unimplemented conjugation types. Only IX-final-weak left (extremely rare, possibly only one verb اِعْمَايَ (according to Haywood and Nahmad p. 244, who are very specific about the irregular occurrence of alif + yā instead of expected اِعْمَيَّ with doubled yā). Not in Hans Wehr. NOTE: Not true about this, cf. form IX اِرْعَوَى "to desist, to repent, to see the light". Also note form XII اِخْضَوْضَرَ = form IX اِخْضَرَّ "to be or become green". [DONE except for اِعْمَايَ] 2. Implement irregular verbs as special cases and recognize them, e.g. -- laysa "to not be"; only exists in the past tense, no non-past, no imperative, no participles, no passive, no verbal noun. Irregular alternation las-/lays-. [IMPLEMENTABLE USING OVERRIDES] -- istaḥā yastaḥī "be ashamed of" -- this is complex according to Hans Wehr because there are two verbs, regular istaḥyā yastaḥyī "to spare (someone)'s life" and irregular istaḥyā yastaḥyī "to be ashamed to face (someone)", which is irregular because it has the alternate irregular form istaḥā yastaḥī which only applies to this meaning. Currently we follow Haywood and Nahmad in saying that both varieties can be spelled istaḥyā/istaḥā/istaḥḥā, but we should instead use a variant= param similar to حَيَّ to distinguish the two possibilities, and maybe not include istaḥḥā. -- ʿayya/ʿayiya yaʿayyu/yaʿyā "to not find the right way, be incapable of, stammer, falter, fall ill". This appears to be a mixture of a geminate and final-weak verb. Unclear what the whole paradigm looks like. Do the consonant-ending parts in the past follow the final-weak paradigm? Is it the same in the non-past? Or can you conjugate the non-past fully as either geminate or final-weak? -- اِنْمَحَى inmaḥā or يمَّحَى immaḥā "to be effaced, obliterated; to disappear, vanish" has irregular assimilation of inm- to imm- as an alternative. inmalasa "to become smooth; to glide; to slip away; to escape" also has immalasa as an alternative. The only other form VII verbs in Hans Wehr beginning with -m- are inmalaḵa "to be pulled out, torn out, wrenched" and inmāʿa "to be melted, to melt, to dissolve", which are not listed with imm- alternatives, but might have them; if so, we should handle this generally. [DONE] -- يَرَعَ yaraʕa yariʕu "to be a coward, to be chickenhearted" as an alternative form of يَرِعَ yariʕa yayraʕu (as given in Wehr). [IMPLEMENTABLE USING OVERRIDES] 3. Implement individual override parameters for each paradigm part. See Module:fro-verb for an example of how to do this generally. Note that {{temp|ar-conj-I}} and other of the older templates already had such individual override params. [DONE] Irregular verbs already implemented: -- [ḥayya/ḥayiya yaḥyā "live" -- behaves like a normal final-weak verb (e.g. past first singular ḥayītu) except in the past-tense parts with vowel-initial endings (all the third person except for the third feminine plural). The normal singular and dual endings have -yiya- in them, which compresses to -yya-, with the normal endings the less preferred ones. In masculine third plural, expected ḥayū is replaced by ḥayyū by analogy to the -yy- parts, and the regular form is not given as an alternant in John Mace. Barron's 201 verbs appears to have the regular ḥayū as the part, however. Note also that final -yā appears with tall alif. This appears to be a spelling convention of Arabic, also applying in ḥayyā (form II, "to keep (someone) alive") and 'aḥyā (form IV, "to animate, revive, give birth to, give new life to").] -- implemented -- [ittaxadha yattaxidhu "take"] -- implemented -- [sa'ala yas'alu "ask" with alternative jussive/imperative yasal/sal] -- implemented -- [ra'ā yarā "see"] -- implemented -- ['arā yurī "show"] -- implemented -- ['akala ya'kulu "eat" with imperative kul] -- implemented -- ['axadha ya'xudhu "take" with imperative xudh] -- implemented -- ['amara ya'muru "order" with imperative mur] -- implemented --]=] local force_cat = false -- set to true for debugging -- if true, always maintain manual translit during processing, and compare against full translit at the end local debug_translit = false local lang = require("Module:languages").getByCode("ar") local m_links = require("Module:links") local m_string_utilities = require("Module:string utilities") local m_table = require("Module:table") local ar_utilities = require("Module:ar-utilities") local iut = require("Module:inflection utilities") local put = require("Module:parse utilities") local list_to_text = mw.text.listToText local rfind = m_string_utilities.find local rsubn = m_string_utilities.gsub local rmatch = m_string_utilities.match local rsplit = m_string_utilities.split local usub = m_string_utilities.sub local ulen = m_string_utilities.len local u = m_string_utilities.char local unpack = unpack or table.unpack -- Lua 5.2 compatibility local dump = mw.dumpObject -- Within this module, conjugations are the functions that do the actual -- conjugating by creating the parts of a basic verb. -- They are defined further down. local conjugations = {} -- hamza variants local HAMZA = u(0x0621) -- hamza on the line (stand-alone hamza) = ء local HAMZA_ON_ALIF = u(0x0623) local HAMZA_ON_W = u(0x0624) local HAMZA_UNDER_ALIF = u(0x0625) local HAMZA_ON_Y = u(0x0626) local HAMZA_ANY = "[" .. HAMZA .. HAMZA_ON_ALIF .. HAMZA_UNDER_ALIF .. HAMZA_ON_W .. HAMZA_ON_Y .. "]" local HAMZA_PH = u(0xFFF0) -- hamza placeholder local BAD = u(0xFFF1) local BORDER = u(0xFFF2) local KEEP_W_PH = u(0xFFF3) -- retained-wāw placeholder; see postprocess_term() local KEEP_HAMZA_PH = u(0xFFF4) -- quiescent-hamza placeholder; see postprocess_term() local WAW_SEAT_PH = u(0xFFF5) -- pre-seated hamza-on-wāw placeholder; see postprocess_term() local NO_MADDA_PH = u(0xFFF6) -- hamza held back from the madda rule; see postprocess_term() -- diacritics local A = u(0x064E) -- fatḥa local AN = u(0x064B) -- fatḥatān (fatḥa tanwīn) local U = u(0x064F) -- ḍamma local UN = u(0x064C) -- ḍammatān (ḍamma tanwīn) local I = u(0x0650) -- kasra local IN = u(0x064D) -- kasratān (kasra tanwīn) local SK = u(0x0652) -- sukūn = no vowel local SH = u(0x0651) -- šadda = gemination of consonants local DAGGER_ALIF = u(0x0670) local DIACRITIC_ANY_BUT_SH = "[" .. A .. I .. U .. AN .. IN .. UN .. SK .. DAGGER_ALIF .. "]" -- Pattern matching short vowels local AIU = "[" .. A .. I .. U .. "]" -- Pattern matching short vowels or sukūn local AIUSK = "[" .. A .. I .. U .. SK .. "]" -- Pattern matching any diacritics that may be on a consonant local DIACRITIC = SH .. "?" .. DIACRITIC_ANY_BUT_SH -- translit_patterns local vowels = "aeiouāēīōū" local NV = "[^" .. vowels .. "]" local dia = {a = A, i = I, u = U} local undia = {[A] = "a", [I] = "i", [U] = "u", ["-"] = "-"} -- various letters and signs local ALIF = u(0x0627) -- ʾalif = ا local AMAQ = u(0x0649) -- ʾalif maqṣūra = ى local AMAD = u(0x0622) -- ʾalif madda = آ local TAM = u(0x0629) -- tāʾ marbūṭa = ة local T = u(0x062A) -- tāʾ = ت local HYPHEN = u(0x0640) local N = u(0x0646) -- nūn = ن local W = u(0x0648) -- wāw = و local Y = u(0x064A) -- yāʾ = ي local S = "س" local M = "م" local LRM = u(0x200e) -- left-to-right mark -- common combinations local AH = A .. TAM local AT = A .. T local AA = A .. ALIF local AAMAQ = A .. AMAQ local AAH = AA .. TAM local AAT = AA .. T local II = I .. Y local UU = U .. W local AY = A .. Y local AW = A .. W local AYSK = AY .. SK local AWSK = AW .. SK local NA = N .. A local NI = N .. I local AAN = AA .. N local AANI = AA .. NI local AYNI = AYSK .. NI local AWNA = AWSK .. NA local AYNA = AYSK .. NA local AYAAT = AY .. AAT local UNU = "[" .. UN .. U .. "]" local MA = M .. A local MU = M .. U local TA = T .. A local TU = T .. U local _I = ALIF .. I local _U = ALIF .. U --W: Morphological pattern (الوزن الصرفي) to verb form mapping local MORPHOLOGICAL_PATTERN_TO_FORM = { ["فعَل يفعُل"] = "I/a~u", ["فعَل يفعِل"] = "I/a~i", ["فعَل يفعَل"] = "I/a~a", ["فعُل يفعُل"] = "I/u~u", ["فعِل يفعَل"] = "I/i~a", ["فعِل يفعِل"] = "I/i~i", ["فعّل"] = "II", ["فاعل"] = "III", ["أفعل"] = "IV", ["تفعّل"] = "V", ["تفاعل"] = "VI", ["انفعل"] = "VII", ["افتعل"] = "VIII", ["افعلّ"] = "IX", ["استفعل"] = "X", ["افعالّ"] = "XI", ["افعوعل"] = "XII", ["افعوّل"] = "XIII", ["فعلل"] = "Iq", ["تفعلل"] = "IIq", ["افعنلل"] = "IIIq", -- "XIV" should not be used ["افعللّ"] = "IVq", } --W: Create reverse mapping from form to morphological pattern local FORM_TO_MORPHOLOGICAL_PATTERN = {} for pattern, form in pairs(MORPHOLOGICAL_PATTERN_TO_FORM) do FORM_TO_MORPHOLOGICAL_PATTERN[form] = pattern end --W: local MORPHOLOGICAL_PATTERN_TO_BASIC_DERIV = { ["فعَل يفعُل"] = "مُجرَّد", ["فعَل يفعِل"] = "مُجرَّد", ["فعَل يفعَل"] = "مُجرَّد", ["فعُل يفعُل"] = "مُجرَّد", ["فعِل يفعَل"] = "مُجرَّد", ["فعِل يفعِل"] = "مُجرَّد", ["فعّل"] = "مزيد بحرف", ["فاعل"] = "مزيد بحرف", ["أفعل"] = "مزيد بحرف", ["تفعّل"] = "مزيد بحرفين", ["تفاعل"] = "مزيد بحرفين", ["انفعل"] = "مزيد بحرفين", ["افتعل"] = "مزيد بحرفين", ["افعلّ"] = "مزيد بحرفين", ["استفعل"] = "مزيد بثلاثة أحرف", ["افعالّ"] = "مزيد بثلاثة أحرف", ["افعوعل"] = "مزيد بثلاثة أحرف", ["افعوّل"] = "مزيد بثلاثة أحرف", ["فعلل"] = "مُجرَّد", ["تفعلل"] = "مزيد بحرف", ["افعنلل"] = "مزيد بحرفين", ["افعللّ"] = "مزيد بحرفين", } local translit_cache = { -- hamza variants [HAMZA] = "ʔ", [HAMZA_ON_ALIF] = "ʔ", [HAMZA_ON_W] = "ʔ", [HAMZA_UNDER_ALIF] = "ʔ", [HAMZA_ON_Y] = "ʔ", [HAMZA_PH] = "ʔ", -- diacritics [A] = "a", [AN] = "an", [U] = "u", [UN] = "un", [I] = "i", [IN] = "in", [SK] = "", [SH] = "*", -- handled specially [DAGGER_ALIF] = "ā", -- various letters and signs [""] = "", [ALIF] = BAD, -- we should never be transliterating ALIF by itself, as its translit in isolation is ambiguous [AMAQ] = BAD, [AMAD] = "ʔā", [TAM] = "", [T] = "t", [N] = "n", [W] = "w", [Y] = "y", [S] = "s", [M] = "m", [LRM] = "", -- common combinations [AH] = "a", [AT] = "at", [AA] = "ā", [AAMAQ] = "ā", [AAH] = "āh", [AAT] = "āt", [II] = "ī", [UU] = "ū", [AY] = "ay", [AW] = "aw", [AYSK] = "ay", [AWSK] = "aw", [NA] = "na", [NI] = "ni", [AAN] = "ān", [AANI] = "āni", [AYNI] = "ayni", [AWNA] = "awna", [AYNA] = "ayna", [AYAAT] = "ayāt", [MA] = "ma", [MU] = "mu", [TA] = "ta", [TU] = "tu", [_I] = "i", [_U] = "u", } local function transliterate(text) local cached = translit_cache[text] if cached then if cached == BAD then error(("Internal error: Unable to transliterate %s because explicitly marked as BAD"):format(text)) end return cached end local tr = (lang:transliterate(text)) if not tr then error(("Internal error: Unable to transliterate: %s"):format(text)) end translit_cache[text] = tr return tr end local all_person_number_list = { "1s", "2ms", "2fs", "3ms", "3fs", "2d", "3md", "3fd", "1p", "2mp", "2fp", "3mp", "3fp" } local function make_person_number_slot_accel_list(list) local slot_accel_list = {} return slot_accel_list end local imp_person_number_list = {} for _, pn in ipairs(all_person_number_list) do if pn:find("^2") then table.insert(imp_person_number_list, pn) end end local passive_types = m_table.listToSet { "pass", -- verb has both active and passive "ipass", -- verb is active with impersonal passive "nopass", -- verb is active-only "onlypass", -- verb is passive-only "onlypass-impers", -- verb itself is impersonal, meaning passive-only with impersonal passive } local indicator_flags = m_table.listToSet { "nopast", "no_nonpast", "noimp", "nocat", -- don't categorize or include annotations about this; useful in suppletive parts of verbs "reduced", -- verb has assimilation/reduction of initial coronals "altgem", -- form X with alternative past geminate forms with final-weak endings } export.potential_lemma_slots = {"past_3ms", "past_pass_3ms", "ind_3ms", "ind_pass_3ms", "imp_2ms"} export.unsettable_slots = {} for _, potential_lemma_slot in ipairs(export.potential_lemma_slots) do table.insert(export.unsettable_slots, potential_lemma_slot .. "_linked") end -- We don't set the active participle directly for form I because we don't want stative verbs (with past vowel i or u) -- to default to فَاعِل. Instead we set the special slot 'ap1' and later copy it to 'ap' for non-stative verbs. The user -- meanwhile can explicitly request the فَاعِل form for active participles for stative verbs using `ap:+`. table.insert(export.unsettable_slots, "ap1") -- primary default فَاعِل for form I active participles table.insert(export.unsettable_slots, "ap2") -- secondary default فَعِيل for form I active participles (stative I) table.insert(export.unsettable_slots, "ap3") -- secondary default فَعِل for form I active participles (stative II) table.insert(export.unsettable_slots, "apcd") -- secondary default أَفْعَل for form I active participles (color/defect) table.insert(export.unsettable_slots, "apan") -- secondary default فَعْلَان for form I active participles (in -ān) table.insert(export.unsettable_slots, "pp2") -- secondary default فَعِيل for form I passive participles (same as ap2) table.insert(export.unsettable_slots, "vn2") -- secondary default فِعَال for form III verbal nouns export.unsettable_slots_set = m_table.listToSet(export.unsettable_slots) local default_indicator_to_active_participle_slot = { ["+"] = "ap1", ["++"] = "ap2", ["+++"] = "ap3", ["+cd"] = "apcd", ["+an"] = "apan", } local slots_that_may_be_uncertain = { vn = "verbal noun", ap = "active participle", } -- Initialize all the slots for which we generate forms. local function add_slots(alternant_multiword_spec) alternant_multiword_spec.verb_slots = { {"ap", "act|part"}, {"pp", "pass|part"}, {"vn", "vnoun"}, } for _, unsettable_slot in ipairs(export.unsettable_slots) do table.insert(alternant_multiword_spec.verb_slots, {unsettable_slot, "-"}) end -- Add entries for a slot with person/number variants. -- `slot_prefix` is the prefix of the slot, typically specifying the tense/aspect. -- `tag_suffix` is a string listing the set of inflection tags to add after the person/number tags. -- `person_number_list` is a list of the person/number slot suffixes to add to `slot_prefix`. local function add_personal_slot(slot_prefix, tag_suffix, person_number_list) for _, persnum in ipairs(person_number_list) do local slot = slot_prefix .. "_" .. persnum local accel = persnum:gsub("(.)", "%1|") .. tag_suffix table.insert(alternant_multiword_spec.verb_slots, {slot, accel}) end end local tenses = { {"past", "past|%s"}, {"ind", "non-past|%s|ind"}, {"sub", "non-past|%s|sub"}, {"juss", "non-past|%s|juss"}, } for _, slot_accel in ipairs(tenses) do local slot, accel = unpack(slot_accel) for _, voice in ipairs {"act", "pass"} do add_personal_slot(voice == "act" and slot or slot .. "_pass", accel:format(voice), all_person_number_list) end end add_personal_slot("imp", "imp", imp_person_number_list) alternant_multiword_spec.verb_slots_map = {} for _, slot_accel in ipairs(alternant_multiword_spec.verb_slots) do local slot, accel = unpack(slot_accel) alternant_multiword_spec.verb_slots_map[slot] = accel end end local overridable_stems = {} local slot_override_param_mods = { footnote = { item_dest = "footnotes", store = "insert", }, alt = {}, t = { -- [[Module:links]] expects the gloss in "gloss". item_dest = "gloss", }, gloss = {}, g = { -- [[Module:links]] expects the genders in "g". `sublist = true` automatically splits on comma (optionally -- with surrounding whitespace). item_dest = "genders", sublist = true, }, pos = {}, lit = {}, id = {}, --W: The q/qq/l/ll (qualifier and label) modifiers were removed: rendering them -- needs [[Module:pron qualifier]], which does not exist on ar.wiktionary, so any -- use of them raised a script error. Accepting them here again requires importing -- that module first and restoring the block in generate_link() below. } local function generate_obj(formval, parse_err, prefix, is_slot_override) local val, uncertain = formval:match("^(.*)(%?)$") val = val or formval uncertain = not not uncertain local ar, translit = val:match("^(.*)//(.*)$") if not ar then ar = val end if ar == "" then if uncertain then ar = "?" else error(("Can't specify blank value for override for %s override '%s'"):format( is_slot_override and "slot" or "stem", prefix)) end end return {form = ar, translit = translit, uncertain = uncertain} end local function parse_inline_modifiers(comma_separated_group, parse_err, prefix, is_slot_override) local function this_generate_obj(formval, parse_err) return generate_obj(formval, parse_err, prefix, is_slot_override) end return put.parse_inline_modifiers_from_segments { group = comma_separated_group, props = { param_mods = slot_override_param_mods, parse_err = parse_err, generate_obj = this_generate_obj, pre_normalize_modifiers = function(data) local modtext = data.modtext modtext = modtext:match("^(%[.*%])$") if modtext then return ("<footnote:%s>"):format(modtext) end return data.modtext end, }, } end local function allow_multiple_values_for_override(comma_separated_groups, data, is_slot_override) local retvals = {} for _, comma_separated_group in ipairs(comma_separated_groups) do local retval if is_slot_override then --W: pass `prefix` and `is_slot_override` through; without them the error -- raised for a blank override value formatted a nil prefix, which is a -- crash on Lua 5.1 rather than the intended message. Reachable from e.g. -- <I/a~u.vn:[[كتابة]]>, where the [[...]] is consumed as a footnote and -- leaves the value empty. retval = parse_inline_modifiers(comma_separated_group, data.parse_err, data.prefix, is_slot_override) else retval = generate_obj(comma_separated_group[1], data.parse_err, data.prefix, is_slot_override) retval.footnotes = data.fetch_footnotes(comma_separated_group) end table.insert(retvals, retval) end for _, form in ipairs(retvals) do if form.form == "+" or default_indicator_to_active_participle_slot[form.form] then if form.form ~= "+" and default_indicator_to_active_participle_slot[form.form] and not is_slot_override then error(("Stem override '%s' cannot use %s to request a secondary default"):format( data.prefix, form.form)) end data.base.slot_override_uses_default[data.prefix] = true end end for _, form in ipairs(retvals) do if form.form == "-" then data.base.slot_explicitly_missing[data.prefix] = true break end end if data.base.slot_explicitly_missing[data.prefix] then for _, form in ipairs(retvals) do if form.form ~= "-" then data.parse_err(("For slot or stem '%s', saw both - and a value other than -, which isn't allowed"): format(data.prefix)) end end return nil end return retvals end local function simple_choice(choices) return function(separated_groups, data) if #separated_groups > 1 then data.parse_err("For spec '" .. data.prefix .. ":', only one value currently allowed") end if #separated_groups[1] > 1 then data.parse_err("For spec '" .. data.prefix .. ":', no footnotes currently allowed") end local choice = separated_groups[1][1] if not m_table.contains(choices, choice) then data.parse_err("For spec '" .. data.prefix .. ":', saw value '" .. choice .. "' but expected one of '" .. table.concat(choices, "،") .. "'") end return choice end end for _, overridable_stem in ipairs { "past", "past_v", "past_c", "past_pass", "past_pass_v", "past_pass_c", "nonpast", "nonpast_v", "nonpast_c", "nonpast_pass", "nonpast_pass_v", "nonpast_pass_c", "imp", "imp_v", "imp_c", } do overridable_stems[overridable_stem] = allow_multiple_values_for_override end overridable_stems.past_final_weak_vowel = simple_choice { "ay", "aw", "ī", "ū" } overridable_stems.past_pass_final_weak_vowel = simple_choice { "ay", "aw", "ī", "ū" } overridable_stems.nonpast_final_weak_vowel = simple_choice { "ā", "ī", "ū" } overridable_stems.nonpast_pass_final_weak_vowel = simple_choice { "ā", "ī", "ū" } ------------------------------------------------------------------------------- -- Utility functions -- ------------------------------------------------------------------------------- -- version of rsubn() that discards all but the first return value local function rsub(term, foo, bar) return (rsubn(term, foo, bar)) end -- version of rsubn() that returns a 2nd argument boolean indicating whether a substitution was made. local function rsubb(term, foo, bar) local retval, nsubs = rsubn(term, foo, bar) return retval, nsubs > 0 end -- Concatenate one or more strings or form objects. local function q(...) local not_all_strings = debug_translit local has_manual_translit = debug_translit for i = 1, select("#", ...) do local argt = select(i, ...) if not argt then error(("Internal error: Saw nil at index %s: %s"):format(i, dump({...}))) end if type(argt) ~= "string" then not_all_strings = true if argt.translit then has_manual_translit = true break end end end if not not_all_strings then -- just strings, concatenate directly return table.concat({...}) end local formvals = {} local translit = has_manual_translit and {} or nil local footnotes for i = 1, select("#", ...) do local argt = select(i, ...) if type(argt) == "string" then formvals[i] = argt if has_manual_translit then translit[i] = transliterate(argt) end else formvals[i] = argt.form if has_manual_translit then translit[i] = argt.translit or transliterate(argt.form) end footnotes = iut.combine_footnotes(footnotes, argt.footnotes) end end -- FIXME: Do we want to support other properties? return { form = table.concat(formvals), translit = has_manual_translit and table.concat(translit) or nil, footnotes = footnotes, } end -- Return the formval associated with `rad` (a radical or past/non-past vowel, either a string or form object). local function rget(rad) if type(rad) == "string" then return rad elseif type(rad) == "table" then return rad.form else error(("Internal error: Unexpected type for radical or past/non-past vowel: %s"):format(dump(rad))) end end export.rget = rget -- for use in [[Module:ar-headword]] -- Return the footnotes associated with `rad` (a radical or past/non-past vowel, either a string or form object). local function rget_footnotes(rad) if type(rad) == "string" then return nil elseif type(rad) == "table" then return rad.footnotes else error(("Internal error: Unexpected type for radical or past/non-past vowel: %s"):format(dump(rad))) end end -- Return true if the formval associated with `rad` (a radical or past/non-past vowel, either a string or form object) -- is `val`. local function req(rad, val) return rget(rad) == val end -- True if a form-I verb whose first radical is wāw elides that wāw in the non-past (يَعِدُ for -- *يَوْعِدُ) rather than keeping it (يَوْجَلُ). The elision belongs to the kasra non-past -- a~i -- وَعَدَ يَعِدُ, i~i وَرِثَ يَرِثُ -- and to the fatḥa non-past that is a throat-letter lowering of -- that same underlying kasra: a~a وَضَعَ يَضَعُ, وَهَبَ يَهَبُ. Everywhere else the wāw stays: -- i~a وَجِلَ يَوْجَلُ، وَنِيَ يَوْنَى (a fatḥa non-past that is the regular partner of a kasra past, -- not a lowered kasra) -- *~u وَجَلَ يَوْجُلُ، وَجُهَ يَوْجُهُ (a ḍamma non-past is never a lowered kasra) -- The two elided-anyway lexemes with a kasra past (وَسِعَ يَسَعُ, وَطِئَ يَطَأُ) are lexical and are -- what the explicit `I-assimilated` weakness exists to express -- see the comment on it in -- detect_indicator_spec(). -- -- The u-non-past half of this is attested directly: أنطوان الدحداح، معجم تصريف الأفعال العربية -- prints يَوْجُلُ/اُوْجُلْ for وَجَلَ (ص136) and says so in the page's own note -- "لا تُحذَفُ واوُ -- الفعلِ في المُضارعِ ولا تُقلَبُ في الأمرِ" -- and يَوْنَى/اِوْنَ for وَنِيَ (ص230). -- -- Returns true when the vowels aren't known, which keeps the historical default for callers -- that have radicals but no vowel pattern. local function form_i_w_assimilated(past_vowel, nonpast_vowel) if not past_vowel or not nonpast_vowel then return true end return req(nonpast_vowel, I) or req(nonpast_vowel, A) and req(past_vowel, A) end --W: Geminate form-I roots whose degemination vowel follows the مضارع by إتباع. -- When gemination breaks under a consonant-initial suffix, the vowel that surfaces on the -- second radical (مَدَدْتَ's second د) is normally the verb's own past vowel, which the وزن -- fixes at one value per measure: فتح for فعَل يفعُل (مَدَدْتَ، فَرَرْتَ، عَضَضْتَ), كسر for -- فعِل يفعَل (ظَلِلْتَ). Classical grammar also attests a minority إتباع variant on the -- فعَل يفعُل measure, where that vowel instead follows the مضارع's own ضمة (هَمُمْتَ). Which -- of the two a given verb takes is a fact about the lexeme, not about its وزن -- هَمَّ and -- مَدَّ share the measure and differ here -- so it cannot be derived from جذر + وزن and is -- listed, the same way SOUND_FORM_I_UU_ROOTS lists هَيُؤَ. Keyed rad1..rad2..rad3 and gated -- on the a~u pattern, so a listed root still degeminates normally under any other pattern. local GEMINATE_ITBAA_ROOTS = { ["همم"] = true, -- هَمَّ ـُ: أنطوان الدحداح، معجم تصريف الأفعال العربية، ص192 } -- The vowel that surfaces on the second radical of a geminate form-I verb once gemination -- breaks. `past_vowel` unless the root is an attested إتباع lexeme; see GEMINATE_ITBAA_ROOTS. local function geminate_degemination_vowel(rad1, rad2, rad3, past_vowel, nonpast_vowel) if past_vowel and nonpast_vowel and req(past_vowel, A) and req(nonpast_vowel, U) and GEMINATE_ITBAA_ROOTS[rad1 .. rad2 .. rad3] then return nonpast_vowel end return past_vowel end --W: Form-I final-weak roots that keep rad3 as a consonant in the 2fs non-past and -- imperative (تَسْهُوِينَ، اُسْهُوِي) instead of contracting it away (تَسْهِينَ، اِسْهِي). -- -- This was briefly gated on the فعُل يفعُل measure alone, then narrowed to a root table, -- then widened back to the measure -- and the middle step is why. The argument for the -- measure was always that the measure is what keeps rad3 consonantal in the first place -- (its past 3ms is سَهُوَ, not the سَها an a~u root like دَعَا gives), but at the time the -- corpus held exactly one فعُل يفعُل final-weak page, so there was no second paradigm to -- check it against and the rule it displaces -- contract before the 2fs -ī, as دَعَا does -- -- was the one the module had applied for years. Generalising an unverified reading over a -- whole class of live entries is the wrong direction to be wrong in, so it went back to -- listing سهو alone, keyed and gated exactly like SOUND_FORM_I_UU_ROOTS. -- -- Checkpoint 4 supplied the second page, and it answers the question. p222 (model 147) -- prints TWO فعُل يفعُل final-weak paradigms, حَلُوَ and سَرُوَ, and both retain rad3 in both -- eligible slots: تَحْلُوِينَ / اُحْلُوِي and تَسْرُوِينَ / اُسْرُوِي, the wāw carrying its kasra -- under it, all four read at 6x off the strip against the same columns' يَحْلُونَ / يَسْرُونَ. -- Two independent roots, one page, agreeing with سهو on p204 -- and سَرُوَ is precisely the -- root the narrowing had put back to تَسْرِينَ, so the cautious reading is the one the ink -- refutes. So the gate is the vowel pattern after all and the table is gone. local function final_weak_uu_retains_rad3(rad1, rad2, rad3, past_ending_vowel, nonpast_ending_vowel) return past_ending_vowel == "ū" and nonpast_ending_vowel == "ū" end -- Map `vow` (a past/non-past vowel, either a string or form object without translit) by passing the formval through -- `fn`. Don't call this on radicals because they may have manual translit and it isn't clear how to handle that. local function map_vowel(vow, fn) if type(vow) == "string" then return fn(vow) elseif type(vow) == "table" then return {form = fn(vow.form), footnotes = vow.footnotes} else error(("Internal error: Unexpected type for past/non-past vowel: %s"):format(dump(vow))) end end local function get_radicals_3(vowel_spec) return vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3, vowel_spec.past, vowel_spec.nonpast end local function get_radicals_4(vowel_spec) return vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3, vowel_spec.rad4 end local function is_final_weak(base, vowel_spec) return vowel_spec.weakness == "final-weak" or base.form == "XV" end --W: Create a simple link local function link_term_simple(text) return "[[" .. text .. "]]" end local function link_term(text, face, id) return m_links.full_link({lang = lang, term = text, tr = "-", id = id}, face) end local function tag_text(text, tag, class) return m_links.full_link({lang = lang, alt = text, tr = "-"}) end local function track(page) require("Module:debug/track")("ar-verb/" .. page) return true end local function track_if_ar_conj(base, page) if base.alternant_multiword_spec.source_template == "ar-conj" then require("Module:debug/track")("ar-verb/" .. page) end return true end local function reorder_shadda(word) -- shadda+short-vowel (including tanwīn vowels, i.e. -an -in -un) gets -- replaced with short-vowel+shadda during NFC normalisation, which -- MediaWiki does for all Unicode strings; however, it makes various -- processes inconvenient, so undo it. word = rsub(word, "(" .. DIACRITIC_ANY_BUT_SH .. ")" .. SH, SH .. "%1") return word end local function apply_nfc_shadda(word) -- Apply NFC-style normalization: reorder shadda+vowel to vowel+shadda -- to match MediaWiki's automatic NFC normalization. This reverses the -- internal shadda+vowel format used during processing. word = rsub(word, SH .. "(" .. DIACRITIC_ANY_BUT_SH .. ")", "%1" .. SH) return word end ------------------------------------------------------------------------------- -- Basic functions to inflect tenses -- ------------------------------------------------------------------------------- local function skip_slot(base, slot, allow_overrides) if base.slot_explicitly_missing[slot] then return true end if not allow_overrides and base.slot_overrides[slot] and not base.slot_override_uses_default[slot] then -- Skip any slots for which there are overrides, except those that request the default value using +, ++, etc. return true end if base.passive == "nopass" and (slot == "pp" or slot:find("_pass")) then return true elseif base.passive == "onlypass" and slot ~= "pp" and slot ~= "vn" and not slot:find("_pass") then return true elseif base.passive == "ipass" and slot:find("_pass") and not slot:find("3ms") then return true elseif base.passive == "onlypass-impers" and slot ~= "pp" and slot ~= "vn" and (not slot:find("_pass") or slot:find("_pass") and not slot:find("3ms")) then return true end if base.nopast and slot:find("^past_") then return true end if base.noimp and slot:find("^imp_") then return true end if base.no_nonpast and (slot:find("^ind_") or slot:find("^sub_") or slot:find("^juss")) then return true end return false end local function basic_combine_stem_ending(stem, ending) return stem .. ending end local function basic_combine_stem_ending_tr(stem, ending) return stem .. ending end -- Concatenate `prefixes`, `stems` and `endings` (any of which may be an abbreviate form list, i.e. strings, form -- objects or lists of strings or form objects) and store into `slot`. If a user-supplied override exists for the slot, -- nothing will happen unless `allow_overrides` is provided. local function add3(base, slot, prefixes, stems, endings, allow_overrides) if skip_slot(base, slot, allow_overrides) then return end -- Optimization since the prefixes are almost always single strings. if type(prefixes) == "string" then local function do_combine_stem_ending(stem, ending) return prefixes .. stem .. ending end local function do_combine_stem_ending_tr(stem, ending) return transliterate(prefixes) .. stem .. ending end iut.add_forms(base.forms, slot, stems, endings, do_combine_stem_ending, transliterate, do_combine_stem_ending_tr, base.form_footnotes) else iut.add_multiple_forms(base.forms, slot, {prefixes, stems, endings}, basic_combine_stem_ending, transliterate, basic_combine_stem_ending_tr, base.form_footnotes) end end -- Insert one or more forms in `form_or_forms` into `slot`. `form_or_forms` is an abbreviated form list (see comment at -- top of [[Module:inflection utilities]]). If a user-supplied override exists for the slot, nothing will happen unless -- `allow_overrides` is provided. BEWARE: One form object should never occur in two different slots, or twice in a given -- slot; if taking a form object from an existing slot, make sure to shallowCopy() it. local function insert_form_or_forms(base, slot, form_or_forms, allow_overrides, uncertain) if not skip_slot(base, slot, allow_overrides) then -- Some optimizations of the most common case of inserting a single string. if type(form_or_forms) == "string" and not base.form_footnotes then form_or_forms = {form = form_or_forms, uncertain = uncertain} iut.insert_form(base.forms, slot, form_or_forms) else local list = iut.convert_to_general_list_form(form_or_forms, base.form_footnotes) if uncertain then for _, formobj in ipairs(list) do formobj.uncertain = true end end iut.insert_forms(base.forms, slot, list) end end end -- Insert `string_or_form` into both the ap2 and pp2 slots, shallowCopying a form object to make sure no form objects -- occur in two slots. local function insert_ap2_pp2(base, string_or_form) insert_form_or_forms(base, "ap2", string_or_form) if type(string_or_form) == "table" then string_or_form = m_table.shallowCopy(string_or_form) end insert_form_or_forms(base, "pp2", string_or_form) end -- Convert `stemforms` (a string, a form object, or a list of strings and/or form objects) into "general form" (a list -- of form objects) and map `fn` over the list of objects. `fn` is passed two arguments (form value and translit) and -- should likewise return the new form value and translit. Footnotes will be preserved. FIXME: Preserve other metadata. local function map_general(stemforms, fn) return iut.map_forms(iut.convert_to_general_list_form(stemforms), fn) end -- Similar to map_general() except that `fn` should return a single value (one or more strings or form objects), instead -- of two values (form value and translit), and the resulting value(s) from all calls to `fn` will be flattened to -- construct the overall return value. Footnotes will be preserved. FIXME: Preserve other metadata. local function flatmap_general(stemforms, fn) return iut.flatmap_forms(iut.convert_to_general_list_form(stemforms), fn) end -- Given user-supplied stem overrides in `base`, construct any derived stem overrides (e.g. vowel-specific or -- consonant-specific variants), and truncate initial y-/ي- in any non-past overrides. local function construct_stems(base) local stems = base.stem_overrides stems.past_v = stems.past_v or stems.past stems.past_c = stems.past_c or stems.past stems.past_pass_v = stems.past_pass_v or stems.past_pass stems.past_pass_c = stems.past_pass_c or stems.past_pass stems.nonpast_v = stems.nonpast_v or stems.nonpast stems.nonpast_c = stems.nonpast_c or stems.nonpast stems.nonpast_pass_v = stems.nonpast_pass_v or stems.nonpast_pass stems.nonpast_pass_c = stems.nonpast_pass_c or stems.nonpast_pass stems.imp_v = stems.imp_v or stems.imp stems.imp_c = stems.imp_c or stems.imp local function truncate_nonpast_initial_cons(stem_type, form, translit) if form == "+" then return form, translit end if not form:find("^" .. Y) then error(("Form value %s for stem type '%s' should begin with ي"):format(form, stem_type)) end form = form:gsub("^" .. Y, "") if translit then if not translit:find("^y") then error(("Translit value %s for stem type '%s' should begin with y"):format(translit, stem_type)) end translit = translit:gsub("^y", "") end return form, translit end for _, nonpast_stem_type in ipairs { "nonpast_v", "nonpast_c", "nonpast_pass_v", "nonpast_pass_c" } do if stems[nonpast_stem_type] then stems[nonpast_stem_type] = map_general(stems[nonpast_stem_type], function(form, translit) return truncate_nonpast_initial_cons(nonpast_stem_type, form, translit) end) end end end -- Given user-specified overrides for stem `stemname`, return overrides with occurrences of + replaced by -- `default_stem`. If no overrides, return `default_stem`, or {} if no default. local function override_stem_if_needed(base, stemname, default_stem) local overrides = base.stem_overrides[stemname] if not overrides then return default_stem or {} end return map_general(overrides, function(form, translit) if form ~= "+" and default_indicator_to_active_participle_slot[form] then error(("Stem overrides cannot use secondary default indicators but saw %s in stem override '%s'"):format( form, stemname)) end if form == "+" then if translit then error(("Cannot supply manual translit along with + for stem override '%s'"):format(stemname)) end if not default_stem then error(("Cannot use + for stem override '%s' because no default is available"):format(stemname)) end if type(default_stem) ~= "string" then error(("Internal error: Default stem for '%s' is not a string: %s"):format(stemname, dump(default_stem))) end return default_stem end return form, translit end) end ------------------------------------------------------------------------------- -- Properties of different verbal forms -- ------------------------------------------------------------------------------- local allowed_vforms = {"I", "II", "III", "IV", "V", "VI", "VII", "VIII", "IX", "X", "XI", "XII", "XIII", "XIV", "XV", "Iq", "IIq", "IIIq", "IVq"} local allowed_vforms_set = m_table.listToSet(allowed_vforms) local allowed_vforms_with_weakness = m_table.shallowCopy(allowed_vforms) -- The user needs to be able to explicitly specify that a form-I verb (specifically one whose initial radical is و) is -- sound. Cf. wajiʕa yawjaʕu (not #yajaʕu) "to ache, to hurt". In general, i~a and u~u verbs whose initial radical is و -- seem to not assimilate the first radical; cf. وقح "to be shameless", variously waqaḥa~yaqiḥu, waquḥa~yawquḥu and -- waqiḥa~yawqaḥu, whereas a~i verbs (wafaḍa~yafiḍu "to rush"), i~i verbs (wafiqa~yafiqu "to be proper, to be suitable") -- and a~a verbs (waḍaʕa~yaḍaʕu "to set down, to place") do assimilate. But there are naturally exceptions, e.g. -- waṭiʔa~yaṭaʔu "to tread, to trample"; wasiʕa~yasaʕu "to be spacious; to be well-off"; waṯiʔa~yaṯaʔu "to get bruised, -- to be sprained". Also beware of waniya~yawnā "to be faint; to languish", which is sound in the first radical and -- final-weak in the last radical. Nonetheless, the regularity of the patterns mentioned above suggest we should provide -- them as defaults. -- Note that there are other cases of unexpectedly sound verbs, e.g. izdawaja~yazdawiju "to be in pairs", layisa~yalyasu -- "to be valiant, to be brave", ʔaḥwaja~yuḥwiju "to need", istahwana~yastahwinu "to consider easy", sawisa~yaswasu "to -- be or become moth-eaten or worm-eaten" (vs. sāsa~yasūsu "to govern, to rule" from the same radicals), ʕawira~yaʕwaru -- "to be one-eyed", istajwaba~yastajwibu "to interrogate", etc. But in these cases there is no need for explicit user -- specification as the lemma itself specifies the unexpected soundness. for _, form_with_weakness in ipairs { "I-sound", "I-assimilated", "none-sound", "none-hollow", "none-geminate", "none-final-weak" } do table.insert(allowed_vforms_with_weakness, form_with_weakness) end local allowed_vforms_with_weakness_set = m_table.listToSet(allowed_vforms_with_weakness) local function vform_supports_final_weak(vform) return vform ~= "XI" and vform ~= "XV" and vform ~= "IVq" end local function vform_supports_geminate(vform) return vform == "I" or vform == "III" or vform == "IV" or vform == "VI" or vform == "VII" or vform == "VIII" or vform == "X" end local function vform_supports_hollow(vform) return vform == "I" or vform == "IV" or vform == "VII" or vform == "VIII" or vform == "X" end local function vform_probably_impersonal_passive(vform, weakness, past_vowel, nonpast_vowel) return vform == "I" and req(past_vowel, I) or vform == "V" or vform == "VI" or vform == "X" or vform == "IIq" end local function vform_probably_full_passive(vform) return vform == "II" or vform == "III" or vform == "IV" or vform == "Iq" end local function vform_probably_no_passive(vform, weakness, past_vowel, nonpast_vowel) return vform == "I" and req(past_vowel, U) or vform == "VII" or vform == "IX" or vform == "XI" or vform == "XII" or vform == "XIII" or vform == "XIV" or vform == "XV" or vform == "IIIq" or vform == "IVq" end -- Active vforms II, III, IV, Iq use non-past prefixes in -u- instead of -a-. local function prefix_vowel_from_vform(vform) if vform == "II" or vform == "III" or vform == "IV" or vform == "Iq" then return "u" else return "a" end end -- True if the active non-past takes a-vocalization rather than i-vocalization in its last syllable. local function vform_nonpast_a_vowel(vform) return vform == "V" or vform == "VI" or vform == "XV" or vform == "IIq" end -- True if the `passive` spec indicates a passive-only verb. local function is_passive_only(passive) return passive == "onlypass" or passive == "onlypass-impers" end export.is_passive_only = is_passive_only -- for use in [[Module:ar-headword]] ------------------------------------------------------------------------------- -- Properties of specific sounds -- ------------------------------------------------------------------------------- -- Is radical wāw (و) or yāʾ (ي)? local function is_waw_ya(rad) return req(rad, W) or req(rad, Y) end -- Check that radical is wāw (و) or yāʾ (ي), error if not local function check_waw_ya(rad) if not is_waw_ya(rad) then error("Expecting weak radical: '" .. rget(rad) .. "' should be " .. W .. " or " .. Y) end end -- Form-I verb حيّ or حيي and form-X verb استحيا or استحى local function hayy_radicals(rad1, rad2, rad3) return req(rad1, "ح") and req(rad2, Y) and is_waw_ya(rad3) end -- FUCK ME HARD. "Lua error at line 1514: main function has more than 200 local variables". local function create_conjugations() ------------------------------------------------------------------------------- -- Radicals associated with various irregular verbs -- ------------------------------------------------------------------------------- -- Form-I verb أخذ or form-VIII verb اتخذ local function axadh_radicals(rad1, rad2, rad3) return req(rad1, HAMZA) and req(rad2, "خ") and req(rad3, "ذ") end -- Form-I verb whose imperative has a reduced form: أكل and أخذ and أمر. Return "shortonly" if only -- short-form imperatives exist (أكل and أخذ) or "shortlong" if long-form imperatives also exist (أمر); -- they are used after a clitic like فَ and وَ. local function reduced_imperative_verb(rad1, rad2, rad3) return axadh_radicals(rad1, rad2, rad3) and "shortonly" or req(rad1, HAMZA) and req(rad2, "ك") and req(rad3, "ل") and "shortonly" or req(rad1, HAMZA) and req(rad2, "م") and req(rad3, "ر") and "shortlong" end -- Form-I verb رأى and form-IV verb أرى local function raa_radicals(rad1, rad2, rad3) return req(rad1, "ر") and req(rad2, HAMZA) and is_waw_ya(rad3) end -- Form-I verb سأل local function saal_radicals(rad1, rad2, rad3) return req(rad1, "س") and req(rad2, HAMZA) and req(rad3, "ل") end -- Form-I verb كان local function kaan_radicals(rad1, rad2, rad3) return req(rad1, "ك") and req(rad2, W) and req(rad3, N) end ------------------------------------------------------------------------------- -- Sets of past endings -- ------------------------------------------------------------------------------- -- The 13 endings of the sound/hollow/geminate past tense. local past_endings = { -- singular SK .. TU, SK .. TA, SK .. "تِ", A, A .. "تْ", --dual SK .. "تُمَا", AA, A .. "تَا", -- plural SK .. "نَا", SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this SK .. "تُن" .. SH .. A, UU .. ALIF, SK .. "نَ" } -- Make endings for final-weak past in -aytu or -awtu. AYAW is AY or AW as appropriate. Note that AA and AW are -- global variables. local function make_past_endings_ay_aw(ayaw, third_sg_masc) return { -- singular ayaw .. SK .. TU, ayaw .. SK .. TA, ayaw .. SK .. "تِ", third_sg_masc, A .. "تْ", --dual ayaw .. SK .. "تُمَا", ayaw .. AA, A .. "تَا", -- plural ayaw .. SK .. "نَا", ayaw .. SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this ayaw .. SK .. "تُن" .. SH .. A, AW .. SK .. ALIF, ayaw .. SK .. "نَ" } end -- past final-weak -aytu endings local past_endings_ay = make_past_endings_ay_aw(AY, AAMAQ) -- past final-weak -awtu endings local past_endings_aw = make_past_endings_ay_aw(AW, AA) -- used for alternative endings for form-X geminate verbs like اِسْتَمَرَّ local past_endings_ay_12_person_only = { -- singular AY .. SK .. TU, AY .. SK .. TA, AY .. SK .. "تِ", {}, {}, --dual AY .. SK .. "تُمَا", {}, {}, -- plural AY .. SK .. "نَا", AY .. SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this AY .. SK .. "تُن" .. SH .. A, {}, {}, } -- Make endings for final-weak past in -ītu or -ūtu. IIUU is ī or ū as appropriate. Note that AA and UU are global -- variables. local function make_past_endings_ii_uu(iiuu) return { -- singular iiuu .. TU, iiuu .. TA, iiuu .. "تِ", iiuu .. A, iiuu .. A .. "تْ", --dual iiuu .. "تُمَا", iiuu .. AA, iiuu .. A .. "تَا", -- plural iiuu .. "نَا", iiuu .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this iiuu .. "تُن" .. SH .. A, UU .. ALIF, iiuu .. "نَ" } end -- past final-weak -ītu endings local past_endings_ii = make_past_endings_ii_uu(II) -- past final-weak -ūtu endings local past_endings_uu = make_past_endings_ii_uu(UU) ------------------------------------------------------------------------------- -- Sets of non-past prefixes and endings -- ------------------------------------------------------------------------------- local nonpast_prefix_consonants = { -- singular HAMZA, T, T, Y, T, -- dual T, Y, T, -- plural N, T, T, Y, Y } -- There are only five distinct endings in all non-past verbs. Make any set of non-past endings given these five -- distinct endings. local function make_nonpast_endings(null, fem, dual, pl, fempl) return { -- singular null, null, fem, null, null, -- dual dual, dual, dual, -- plural null, pl, fempl, pl, fempl } end -- endings for non-past indicative local ind_endings = make_nonpast_endings( U, II .. NA, AANI, UU .. NA, SK .. NA ) -- Make the endings for non-past subjunctive/jussive, given the vowel diacritic used in "null" endings -- (1s/2ms/3ms/3fs/1p). local function make_sub_juss_endings(dia_null) return make_nonpast_endings( dia_null, II, AA, UU .. ALIF, SK .. NA ) end -- endings for non-past subjunctive local sub_endings = make_sub_juss_endings(A) -- endings for non-past jussive local juss_endings = make_sub_juss_endings(SK) -- endings for alternative geminate non-past jussive in -a; same as subjunctive local juss_endings_alt_a = sub_endings -- endings for alternative geminate non-past jussive in -i local juss_endings_alt_i = make_sub_juss_endings(I) -- Endings for final-weak non-past indicative in -ā. Note that AY, AW and AAMAQ are global variables. local ind_endings_aa = make_nonpast_endings( AAMAQ, AYSK .. NA, AY .. AANI, AWSK .. NA, AYSK .. NA ) -- Make endings for final-weak non-past indicative in -ī or -ū; IIUU is ī or ū as appropriate. Note that II and UU -- are global variables. local function make_ind_endings_ii_uu(iiuu) return make_nonpast_endings( iiuu, II .. NA, iiuu .. AANI, UU .. NA, iiuu .. NA ) end -- endings for final-weak non-past indicative in -ī local ind_endings_ii = make_ind_endings_ii_uu(II) -- endings for final-weak non-past indicative in -ū local ind_endings_uu = make_ind_endings_ii_uu(UU) -- Endings for final-weak non-past subjunctive in -ā. Note that AY, AW, ALIF, AAMAQ are global variables. local sub_endings_aa = make_nonpast_endings( AAMAQ, AYSK, AY .. AA, AWSK .. ALIF, AYSK .. NA ) -- Make endings for final-weak non-past subjunctive in -ī or -ū. IIUU is ī or ū as appropriate. Note that AA, II, -- UU, ALIF are global variables. local function make_sub_endings_ii_uu(iiuu) return make_nonpast_endings( iiuu .. A, II, iiuu .. AA, UU .. ALIF, iiuu .. NA ) end -- endings for final-weak non-past subjunctive in -ī local sub_endings_ii = make_sub_endings_ii_uu(II) -- endings for final-weak non-past subjunctive in -ū local sub_endings_uu = make_sub_endings_ii_uu(UU) -- endings for final-weak non-past jussive in -ā local juss_endings_aa = make_nonpast_endings( A, AYSK, AY .. AA, AWSK .. ALIF, AYSK .. NA ) -- Make endings for final-weak non-past jussive in -ī or -ū. IU is short i or u, IIUU is long ī or ū as appropriate. -- Note that AA, II, UU, ALIF are global variables. local function make_juss_endings_ii_uu(iu, iiuu) return make_nonpast_endings( iu, II, iiuu .. AA, UU .. ALIF, iiuu .. NA ) end -- endings for final-weak non-past jussive in -ī local juss_endings_ii = make_juss_endings_ii_uu(I, II) -- endings for final-weak non-past jussive in -ū local juss_endings_uu = make_juss_endings_ii_uu(U, UU) ------------------------------------------------------------------------------- -- Sets of imperative endings -- ------------------------------------------------------------------------------- -- Extract the second person jussive endings to get corresponding imperative endings. local function imperative_endings_from_jussive(endings) return {endings[2], endings[3], endings[6], endings[10], endings[11]} end -- normal imperative endings local imp_endings = imperative_endings_from_jussive(juss_endings) -- alternative geminate imperative endings in -a local imp_endings_alt_a = imperative_endings_from_jussive(juss_endings_alt_a) -- alternative geminate imperative endings in -i local imp_endings_alt_i = imperative_endings_from_jussive(juss_endings_alt_i) -- final-weak imperative endings in -ā local imp_endings_aa = imperative_endings_from_jussive(juss_endings_aa) -- final-weak imperative endings in -ī local imp_endings_ii = imperative_endings_from_jussive(juss_endings_ii) -- final-weak imperative endings in -ū local imp_endings_uu = imperative_endings_from_jussive(juss_endings_uu) ------------------------------------------------------------------------------- -- Basic functions to inflect tenses -- ------------------------------------------------------------------------------- -- Add to `base` the inflections for the tense indicated by `tense` (the prefix in the slot names, e.g. 'past' -- or 'juss_pass'), formed by combining the `prefixes`, `stems` and `endings`. Each of `prefixes`, `stems` and -- `endings` is either a sequence of 5 (for the imperative) or 13 (for other tenses) abbreviated form lists (each of -- which is either a string, a form object, or a list of strings and/or form objects; see -- [[Module:inflection utilities]] for more info). Alternatively, any of `prefixes`, `stems` or `endings` can be a -- single-element list containing an abbreviated form list, with an additional key `all_same` set to true, or (as a -- special case) a single string; in the latter cases, the same value is used for all 5 or 13 slots. If existing -- inflections already exist, they will be added to, not overridden. `pnums` is the list of person/number slot name -- suffixes, which must match up with the elements in `prefixes`, `stems` and `endings` (i.e. 5 for imperative, 13 -- otherwise). local function inflect_tense_1(base, tense, prefixes, stems, endings, pnums) if not prefixes or not stems or not endings then return end local function verify_affixes(affixname, affixes) local function interr(msg) error(("Internal error: For tense '%s', '%s' %s: %s"):format(tense, affixname, msg, dump(affixes))) end if type(affixes) == "string" then -- do nothing elseif type(affixes) ~= "table" then interr("is not a table or string") elseif affixes.all_same then if #affixes ~= 1 then interr(("with all_same = true should have length 1 but has length %s"):format(#affixes)) end else if #affixes ~= #pnums then interr(("should have length %s but has length %s"):format(#pnums, #affixes)) end end end verify_affixes("prefixes", prefixes) verify_affixes("stems", stems) verify_affixes("endings", endings) local function get_affix(affixes, i) if type(affixes) == "string" then return affixes elseif affixes.all_same then return affixes[1] else return affixes[i] end end for i, pnum in ipairs(pnums) do local prefix = get_affix(prefixes, i) local stem = get_affix(stems, i) local ending = get_affix(endings, i) local slot = tense .. "_" .. pnum add3(base, slot, prefix, stem, ending) end end -- Add to `base` the inflections for the tense indicated by `tense` (the prefix in the slot names, e.g. 'past' -- or 'juss_pass'), formed by combining the `prefixes`, `stems` and `endings`. This is a simple wrapper around -- inflect_tense_1() that applies to all tenses other than the imperative; see inflect_tense_1() for more -- information about the parameters. local function inflect_tense(base, tense, prefixes, stems, endings) inflect_tense_1(base, tense, prefixes, stems, endings, all_person_number_list) end -- Like inflect_tense() but for the imperative, which has only five parts instead of 13 and no prefixes. local function inflect_tense_imp(base, stems, endings) inflect_tense_1(base, "imp", "", stems, endings, imp_person_number_list) end ------------------------------------------------------------------------------- -- Functions to inflect the past tense -- ------------------------------------------------------------------------------- -- Generate past verbs using specified vowel and consonant stems; works for sound, assimilated, hollow, and geminate -- verbs, active and passive. local function past_2stem_conj(base, tense, v_stem, c_stem, footnote_12) local passive = tense:find("_pass") and "_pass" or "" -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "past" .. passive .. "_v", v_stem) local c_stem_12 = c_stem if footnote_12 then c_stem_12 = iut.combine_form_and_footnotes(c_stem_12, footnote_12) end c_stem_12 = override_stem_if_needed(base, "past" .. passive .. "_c", c_stem_12) local c_stem_3 = override_stem_if_needed(base, "past" .. passive .. "_c", c_stem) inflect_tense(base, tense, "", { -- singular c_stem_12, c_stem_12, c_stem_12, v_stem, v_stem, --dual c_stem_12, v_stem, v_stem, -- plural c_stem_12, c_stem_12, c_stem_12, v_stem, c_stem_3 }, past_endings) end -- Generate past verbs using single specified stem; works for sound and assimilated verbs, active and passive. local function past_1stem_conj(base, tense, stem) past_2stem_conj(base, tense, stem, stem) end ------------------------------------------------------------------------------- -- Functions to inflect non-past tenses -- ------------------------------------------------------------------------------- -- Generate non-past conjugation, with two stems, for vowel-initial and consonant-initial endings, respectively. -- Useful for active and passive; for all forms; for all weaknesses (sound, assimilated, hollow, final-weak and -- geminate) and for all types of non-past (indicative, subjunctive, jussive) except for the imperative. (There is a -- separate wrapper function below for geminate jussives because they have three alternants.) Both stems may be the -- same, e.g. for sound verbs. -- `prefix_vowel` will be either "a" or "u". `endings` should be an array of 13 items. If `endings` is nil or -- omitted, infer the endings from the tense. If `jussive` is true, or `endings` is nil and `tense` indicatives -- jussive, use the jussive pattern of vowel/consonant stems (different from the normal ones). local function nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, endings, jussive) local passive = tense:find("_pass") and "_pass" or "" -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "nonpast" .. passive .. "_v", v_stem and q(dia[prefix_vowel], v_stem) or nil) c_stem = override_stem_if_needed(base, "nonpast" .. passive .. "_c", c_stem and q(dia[prefix_vowel], c_stem) or nil) if not endings then if tense:find("^ind") then endings = ind_endings elseif tense:find("^sub") then endings = sub_endings elseif tense:find("^juss") then jussive = true endings = juss_endings else error("Internal error: Unrecognized tense '" .. tense .."'") end end if not jussive then inflect_tense(base, tense, nonpast_prefix_consonants, { -- singular v_stem, v_stem, v_stem, v_stem, v_stem, --dual v_stem, v_stem, v_stem, -- plural v_stem, v_stem, c_stem, v_stem, c_stem }, endings) else inflect_tense(base, tense, nonpast_prefix_consonants, { -- singular -- 'adlul, tadlul, tadullī, yadlul, tadlul c_stem, c_stem, v_stem, c_stem, c_stem, --dual -- tadullā, yadullā, tadullā v_stem, v_stem, v_stem, -- plural -- nadlul, tadullū, tadlulna, yadullū, yadlulna c_stem, v_stem, c_stem, v_stem, c_stem }, endings) end end -- Generate non-past conjugation with one stem (no distinct stems for vowel-initial and consonant-initial endings). -- See nonpast_2stem_conj(). local function nonpast_1stem_conj(base, tense, prefix_vowel, stem, endings, jussive) nonpast_2stem_conj(base, tense, prefix_vowel, stem, stem, endings, jussive) end -- Generate active/passive jussive geminative. There are three alternants, two with terminations -a and -i and one -- in a null termination with a distinct pattern of vowel/consonant stem usage. See nonpast_2stem_conj() for a -- description of the arguments. local function jussive_gem_conj(base, tense, prefix_vowel, v_stem, c_stem) -- alternative in -a nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings_alt_a) -- alternative in -i nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings_alt_i) -- alternative in -null; requires different combination of v_stem and -- c_stem since the null endings require the c_stem (e.g. "tadlul" here) -- whereas the corresponding endings above in -a or -i require the v_stem -- (e.g. "tadulla, tadulli" above) nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings, "jussive") end ------------------------------------------------------------------------------- -- Functions to inflect the imperative -- ------------------------------------------------------------------------------- -- Generate imperative conjugation, with two stems, for vowel-initial and consonant-initial endings, respectively. -- Useful for all forms, and for all weaknesses other than final-weak. Note that the two stems may be the same -- (specifically for sound and assimilated verbs). If `endings` is nil or omitted, use `imp_endings`. If `alt_gem` -- is specified, use the pattern of vowel and consonant stems appropriate for the alternative geminate imperatives -- that use a null ending of -a or -i instead of an empty ending. local function make_2stem_imperative(base, v_stem, c_stem, endings, alt_gem) endings = endings or imp_endings -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "imp_v", v_stem) c_stem = override_stem_if_needed(base, "imp_c", c_stem) if alt_gem then inflect_tense_imp(base, {v_stem, v_stem, v_stem, v_stem, c_stem}, endings) else inflect_tense_imp(base, {c_stem, v_stem, v_stem, v_stem, c_stem}, endings) end end -- Generate imperative parts for sound or assimilated verbs. local function make_1stem_imperative(base, stem) make_2stem_imperative(base, stem, stem) end -- Generate imperative parts for geminate verbs form I (also IV, VII, VIII, X). local function make_gem_imperative(base, v_stem, c_stem) make_2stem_imperative(base, v_stem, c_stem, imp_endings_alt_a, "alt gem") make_2stem_imperative(base, v_stem, c_stem, imp_endings_alt_i, "alt gem") make_2stem_imperative(base, v_stem, c_stem) end ------------------------------------------------------------------------------- -- Functions to inflect entire verbs -- ------------------------------------------------------------------------------- -- Generate finite parts of a sound verb (also works for assimilated verbs) from five stems (past and non-past, -- active and passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u"). local function make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel) past_1stem_conj(base, "past", past_stem) past_1stem_conj(base, "past_pass", past_pass_stem) nonpast_1stem_conj(base, "ind", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "sub", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "juss", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem) make_1stem_imperative(base, imp_stem) end local function past_final_weak_endings_from_vowel(vowel) if vowel == "ay" then return past_endings_ay elseif vowel == "aw" then return past_endings_aw elseif vowel == "ī" then return past_endings_ii elseif vowel == "ū" then return past_endings_uu elseif not vowel then return nil else error(("Internal error: Unrecognized past final-weak vowel spec '%s'"):format(vowel)) end end local function nonpast_final_weak_endings_from_vowel(vowel) if vowel == "ā" then return ind_endings_aa, sub_endings_aa, juss_endings_aa, imp_endings_aa elseif vowel == "ī" then return ind_endings_ii, sub_endings_ii, juss_endings_ii, imp_endings_ii elseif vowel == "ū" then return ind_endings_uu, sub_endings_uu, juss_endings_uu, imp_endings_uu elseif not vowel then return nil else error(("Internal error: Unrecognized non-past final-weak vowel spec '%s'"):format(vowel)) end end -- Generate finite parts of a final-weak verb from five stems (past and non-past, active and passive, plus -- imperative), the past active ending vowel (ay, aw, ī or ū), the non-past active ending vowel (ā, ī or ū) and the -- prefix vowel in the active non-past (a or u). All five imperative persons share `imp_stem`, so the -- hamzat wasl carries the verb's own vowel class throughout -- اُدْعُ and اُدْعِي alike. The 2fs used to -- take a kasra one of its own (finding #5, now #11): its ending is -ī regardless of vowel class, so the -- stem vowel surfaces as a kasra there, and #5 read that as pulling the hamza with it. The book prints -- the ḍamma -- اُدْعِي (p.141), اُعْلِي and اُغْزِي (p159) -- so the surfacing kasra is عارض and leaves the -- hamza alone, exactly as the ḍamma of اِمْشُوا leaves it alone in the mirror case. -- -- `retain_rad3_2fs`, if true, keeps rad3 as a consonant before that -ī instead of letting it -- contract away: تَسْهُوِينَ / اُسْهُوِي rather than تَسْهِينَ / اِسْهِي. See README "Findings" #8 -- and final_weak_uu_retains_rad3(). local function make_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, past_ending_vowel, nonpast_ending_vowel, prefix_vowel, retain_rad3_2fs) past_stem = override_stem_if_needed(base, "past", past_stem) past_pass_stem = override_stem_if_needed(base, "past_pass", past_pass_stem) -- Don't call override_stem_if_needed() here for non-past stems; it's called in nonpast_2stem_conj(). imp_stem = override_stem_if_needed(base, "imp", imp_stem) -- + not supported for ending vowel overrides past_ending_vowel = base.stem_overrides.past_final_weak_vowel or past_ending_vowel local past_pass_ending_vowel = base.stem_overrides.past_pass_final_weak_vowel or "ī" nonpast_ending_vowel = base.stem_overrides.nonpast_final_weak_vowel or nonpast_ending_vowel local nonpast_pass_ending_vowel = base.stem_overrides.nonpast_pass_final_weak_vowel or "ā" local past_endings = past_final_weak_endings_from_vowel(past_ending_vowel) local past_pass_endings = past_final_weak_endings_from_vowel(past_pass_ending_vowel) local ind_endings, sub_endings, juss_endings, imp_endings = nonpast_final_weak_endings_from_vowel(nonpast_ending_vowel) local ind_pass_endings, sub_pass_endings, juss_pass_endings = nonpast_final_weak_endings_from_vowel(nonpast_pass_ending_vowel) if retain_rad3_2fs then -- rad3 surfaces as a consonant carrying the class's own ḍamma, then the regular -ī: -- سْهـ + ُوِينَ, not سْهـ + ِينَ. The 2fs sits at index 3 of a make_nonpast_endings() -- table (`fem`) and at index 2 of the five imperative endings that -- imperative_endings_from_jussive() extracts from it. Copied rather than mutated -- -- these ending tables are module-level constants shared by every final-weak verb. -- Written out inline, not via a helper: create_conjugations() is at Lua's 60-upvalue -- ceiling, and one more captured local stops the module loading at all. local function set_2fs(endings, index, ending) local out = {} for i, v in ipairs(endings) do out[i] = v end out[index] = ending return out end ind_endings = set_2fs(ind_endings, 3, UU .. II .. NA) sub_endings = set_2fs(sub_endings, 3, UU .. II) juss_endings = set_2fs(juss_endings, 3, UU .. II) imp_endings = set_2fs(imp_endings, 2, UU .. II) end inflect_tense(base, "past", "", {past_stem, all_same = 1}, past_endings) inflect_tense(base, "past_pass", "", {past_pass_stem, all_same = 1}, past_pass_endings) nonpast_1stem_conj(base, "ind", prefix_vowel, nonpast_stem, ind_endings) nonpast_1stem_conj(base, "sub", prefix_vowel, nonpast_stem, sub_endings) nonpast_1stem_conj(base, "juss", prefix_vowel, nonpast_stem, juss_endings) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem, ind_pass_endings) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem, sub_pass_endings) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem, juss_pass_endings) inflect_tense_imp(base, {imp_stem, all_same = 1}, imp_endings) end -- Generate finite parts of an augmented (form II+) final-weak verb from five stems (past and non-past, active and -- passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u") and a flag indicating if it -- behaves like a form V/VI verb in taking non-past endings in -ā instead of -ī. local function make_augmented_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel, form56) make_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, "ay", form56 and "ā" or "ī", prefix_vowel) end -- Generate finite parts of an augmented (form II+) sound or final-weak verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (active past stem minus last syllable (= -al or -ā)); -- * `nonpast_stem_base` (non-past stem minus last syllable (= -al/-il or -ā/-ī); -- * `past_pass_stem_base` (passive past stem minus last syllable (= -il or -ī)); -- * `vn` (verbal noun). local function make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) insert_form_or_forms(base, "vn", vn) local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local final_weak = is_final_weak(base, vowel_spec) local prefix_vowel = prefix_vowel_from_vform(base.verb_form) local form56 = vform_nonpast_a_vowel(base.verb_form) local a_base_suffix = final_weak and "" or q(A, lastrad) local i_base_suffix = final_weak and "" or q(I, lastrad) -- past and non-past stems, active and passive local past_stem = q(past_stem_base, a_base_suffix) -- In forms 5 and 6, non-past has /a/ as last stem vowel in the non-past -- in both active and passive, but /i/ in the active participle and /a/ -- in the passive participle. Elsewhere, consistent /i/ in active non-past -- and participle, consistent /a/ in passive non-past and participle. -- Hence, forms 5 and 6 differ only in the non-past active (but not -- active participle), so we have to split the finite non-past stem and -- active participle stem. local nonpast_stem = q(nonpast_stem_base, form56 and a_base_suffix or i_base_suffix) local ap_stem = q(nonpast_stem_base, i_base_suffix) local past_pass_stem = q(past_pass_stem_base, i_base_suffix) local nonpast_pass_stem = q(nonpast_stem_base, a_base_suffix) -- imperative stem local imp_stem = q(past_stem_base, form56 and a_base_suffix or i_base_suffix) -- make parts if final_weak then make_augmented_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel, form56) else make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel) end -- active and passive participle if final_weak then insert_form_or_forms(base, "ap", q(MU, ap_stem, IN)) insert_form_or_forms(base, "pp", q(MU, nonpast_pass_stem, AN, AMAQ)) else insert_form_or_forms(base, "ap", q(MU, ap_stem)) insert_form_or_forms(base, "pp", q(MU, nonpast_pass_stem)) end end -- Generate finite parts of a hollow or geminate verb from ten stems (vowel and consonant stems for each of past and -- non-past, active and passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u"), plus a -- flag indicating if we are a geminate verb. local function make_hollow_geminate_verb(base, geminate, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel, altgem_note) past_2stem_conj(base, "past", past_v_stem, past_c_stem, altgem_note) past_2stem_conj(base, "past_pass", past_pass_v_stem, past_pass_c_stem) nonpast_2stem_conj(base, "ind", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "sub", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "ind_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) nonpast_2stem_conj(base, "sub_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) if geminate then jussive_gem_conj(base, "juss", prefix_vowel, nonpast_v_stem, nonpast_c_stem) jussive_gem_conj(base, "juss_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) make_gem_imperative(base, imp_v_stem, imp_c_stem) else nonpast_2stem_conj(base, "juss", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "juss_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) make_2stem_imperative(base, imp_v_stem, imp_c_stem) end end -- Generate finite parts of an augmented (form II+) hollow verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (invariable part of active past stem); -- * `nonpast_stem_base` (invariable part of nonpast stem); -- * `past_pass_stem_base` (invariable part of passive past stem); -- * `vn` (verbal noun). local function make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) insert_form_or_forms(base, "vn", vn) local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local form410 = base.verb_form == "IV" or base.verb_form == "X" local prefix_vowel = prefix_vowel_from_vform(base.verb_form) local a_base_suffix_v, a_base_suffix_c local i_base_suffix_v, i_base_suffix_c a_base_suffix_v = q(AA, lastrad) -- 'af-āl-a, inf-āl-a a_base_suffix_c = q(A, lastrad) -- 'af-al-tu, inf-al-tu i_base_suffix_v = q(II, lastrad) -- 'uf-īl-a, unf-īl-a i_base_suffix_c = q(I, lastrad) -- 'uf-il-tu, unf-il-tu -- past and non-past stems, active and passive, for vowel-initial and -- consonant-initial endings local past_v_stem = q(past_stem_base, a_base_suffix_v) local past_c_stem = q(past_stem_base, a_base_suffix_c) -- yu-f-īl-u, ya-staf-īl-u but yanf-āl-u, yaft-āl-u local nonpast_v_stem = q(nonpast_stem_base, form410 and i_base_suffix_v or a_base_suffix_v) local nonpast_c_stem = q(nonpast_stem_base, form410 and i_base_suffix_c or a_base_suffix_c) local past_pass_v_stem = q(past_pass_stem_base, i_base_suffix_v) local past_pass_c_stem = q(past_pass_stem_base, i_base_suffix_c) local nonpast_pass_v_stem = q(nonpast_stem_base, a_base_suffix_v) local nonpast_pass_c_stem = q(nonpast_stem_base, a_base_suffix_c) -- imperative stem local imp_v_stem = q(past_stem_base, form410 and i_base_suffix_v or a_base_suffix_v) local imp_c_stem = q(past_stem_base, form410 and i_base_suffix_c or a_base_suffix_c) -- make parts make_hollow_geminate_verb(base, false, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel) -- active participle insert_form_or_forms(base, "ap", q(MU, nonpast_v_stem)) -- passive participle insert_form_or_forms(base, "pp", q(MU, nonpast_pass_v_stem)) end -- Generate finite parts of an augmented (form II+) geminate verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (invariable part of active past stem; this and the stem bases below will end with a consonant -- for forms IV, X, IVq, and a short vowel for the others); -- * `nonpast_stem_base` (invariable part of nonpast stem); -- * `past_pass_stem_base` (invariable part of passive past stem); -- * `vn` (verbal noun); -- * `altgem_note` (footnote to add to active past 1/2-person forms, when alternative forms are supplied [form X]). local function make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn, altgem_note) insert_form_or_forms(base, "vn", vn) local vform = base.verb_form local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local prefix_vowel = prefix_vowel_from_vform(vform) local a_base_suffix_v, a_base_suffix_c local i_base_suffix_v, i_base_suffix_c if vform == "IV" or vform == "X" or vform == "IVq" then a_base_suffix_v = q(A, lastrad, SH) -- 'af-all a_base_suffix_c = q(SK, lastrad, A, lastrad) -- 'af-lal i_base_suffix_v = q(I, lastrad, SH) -- yuf-ill i_base_suffix_c = q(SK, lastrad, I, lastrad) -- yuf-lil else a_base_suffix_v = q(lastrad, SH) -- fā-ll, infa-ll a_base_suffix_c = q(lastrad, A, lastrad) -- fā-lal, infa-lal i_base_suffix_v = q(lastrad, SH) -- yufā-ll, yanfa-ll i_base_suffix_c = q(lastrad, I, lastrad) -- yufā-lil, yanfa-lil end -- past and non-past stems, active and passive, for vowel-initial and -- consonant-initial endings local past_v_stem = q(past_stem_base, a_base_suffix_v) local past_c_stem = q(past_stem_base, a_base_suffix_c) local nonpast_v_stem = q(nonpast_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_v or i_base_suffix_v) local nonpast_c_stem = q(nonpast_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_c or i_base_suffix_c) -- NOTE: Formerly had a comment that "vform III and VI passive past do not have contracted parts, only -- uncontracted parts, which are added separately by those functions". This is based on Mace -- "Arabic Verbs and Essential Grammar" (1999) entry 63 (continued), which shows passive ḥūjija but no ḥūjja; -- but that is apparently a mistake, as (1) verb tables in other books do show contracted passive parts for -- these forms; (2) there is no mention of such an exception on p. 99, which explains how geminate ("doubled") -- verbs work (on the contrary, it says "The contracted and uncontracted pairs (see above) are found all -- over Forms III and VI of the doubled verbs"). local past_pass_v_stem = q(past_pass_stem_base, i_base_suffix_v) local past_pass_c_stem = q(past_pass_stem_base, i_base_suffix_c) local nonpast_pass_v_stem = q(nonpast_stem_base, a_base_suffix_v) local nonpast_pass_c_stem = q(nonpast_stem_base, a_base_suffix_c) -- imperative stem local imp_v_stem = q(past_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_v or i_base_suffix_v) local imp_c_stem = q(past_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_c or i_base_suffix_c) -- make parts make_hollow_geminate_verb(base, "geminate", past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel, altgem_note) -- active participle insert_form_or_forms(base, "ap", q(MU, nonpast_v_stem)) -- passive participle insert_form_or_forms(base, "pp", q(MU, nonpast_pass_v_stem)) end ------------------------------------------------------------------------------- -- Conjugation functions for specific conjugation types -- ------------------------------------------------------------------------------- local function form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1) local imp_vowel = map_vowel(nonpast_vowel, function(vow) if vow == A or vow == I then return I elseif vow == U then return U elseif not skip_slot(base, "imp_2ms") then error(("Internal error: Non-past vowel %s isn't a, i, or u, should have been caught earlier"):format( dump(nonpast_vowel))) else -- Passive-only; imperative won't ever be displayed so it doesn't matter. return I end end) -- Mace ("Arabic Verbs and Essentials of Grammar" p. 63: [https://archive.org/details/arabicverbsessen00john/page/62/mode/2up]) -- claims that initial hamza is assimilated/elided into a long vowel in the form-I imperative, but apparently -- this isn't corrrect. local vowel_on_alif = map_vowel(imp_vowel, function(vow) return ALIF .. vow end) return q(vowel_on_alif, rad1, SK) end -- Implement form-I sound or assimilated verb. ASSIMILATED is true for assimilated verbs. local function make_form_i_sound_assimilated_verb(base, vowel_spec, assimilated) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied -- past and non-past stems, active and passive local past_stem = q(rad1, A, rad2, past_vowel, rad3) local nonpast_stem = assimilated and q(rad2, nonpast_vowel, rad3) or q(rad1, SK, rad2, nonpast_vowel, rad3) local past_pass_stem = q(rad1, U, rad2, I, rad3) local nonpast_pass_stem = q(rad1, SK, rad2, A, rad3) -- imperative stem -- check for irregular verb with reduced imperative (أَخَذَ or أَكَلَ or أَمَرَ) local reducedimp = reduced_imperative_verb(rad1, rad2, rad3) if reducedimp then base.irregular = true end local imp_stem_suffix = q(rad2, nonpast_vowel, rad3) local long_imp_stem_base = form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1) local short_imp_stem_base = "" local imp_stem = q((assimilated or reducedimp) and "" or long_imp_stem_base, imp_stem_suffix) -- A form-I mithal that keeps its wāw spells its imperative on a hamzat wasl, -- اِوْجَلْ / اِوْنَ. Tell postprocess_term() not to reduce that اِوْ to اِي -- see #10 -- there for why the substitution list can't decide this for itself. if not assimilated and req(rad1, W) then base.orth.keep_initial_w = true end -- make parts make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, "a") if reducedimp == "shortlong" then make_1stem_imperative(base, iut.combine_form_and_footnotes(q(long_imp_stem_base, imp_stem_suffix), mw.getCurrentFrame():preprocess("[used especially with a clitic such as {{m|ar|فَ}} or {{m|ar|وَ}}]"))) end -- Check for irregular verb سَأَلَ with alternative jussive and imperative. Calling this after make_sound_verb() -- adds additional entries to the paradigm parts. if saal_radicals(rad1, rad2, rad3) then base.irregular = true nonpast_1stem_conj(base, "juss", "a", "سَل") nonpast_1stem_conj(base, "juss_pass", "u", "سَل") make_1stem_imperative(base, "سَل") end -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, I, rad3)) -- Insert alternative active participle (stative type I) فَعِيل. Since not all verbs have this, we require that -- verbs that do have it specify it explicitly; a shortcut ++ is provided to make this easier (e.g. <ap:++> to -- indicate that the alternative form should be used for the active participle, <ap:+,++> to indicate that both -- forms can be used, and <ap:-> to indicate that there is no active participle). The same form is used for -- secondary default passive participle. insert_ap2_pp2(base, q(rad1, A, rad2, II, rad3)) -- Active participle, stative type II فَعِل (+++). insert_form_or_forms(base, "ap3", q(rad1, A, rad2, I, rad3)) -- Active participle, color/defect أَفْعَل (+cd). insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, A, rad3)) -- Active participle, -ān فَعْلَان (+an). insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, UU, rad3)) end conjugations["I-sound"] = function(base, vowel_spec) make_form_i_sound_assimilated_verb(base, vowel_spec, false) end conjugations["none-sound"] = function(base, vowel_spec) -- All default stems are nil. make_sound_verb(base) end conjugations["none-hollow"] = function(base, vowel_spec) -- All default stems are nil. make_hollow_geminate_verb(base, false) end conjugations["none-geminate"] = function(base, vowel_spec) -- All default stems are nil. make_hollow_geminate_verb(base, "geminate") end conjugations["none-final-weak"] = function(base, vowel_spec) -- All default stems are nil. make_final_weak_verb(base) end conjugations["I-assimilated"] = function(base, vowel_spec) make_form_i_sound_assimilated_verb(base, vowel_spec, "assimilated") end local function make_form_i_hayy_verb(base, vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied base.irregular = true -- past and non-past stems, active and passive, and imperative stem local past_c_stem = "حَيِي" local past_v_stem_long = past_c_stem local past_v_stem_short = "حَيّ" -- 3mp of the long variant (حَيِيَ) is حَيُوا, not the geminated حَيُّوا of the short -- variant (حَيَّ) -- rad2 (ي) here is a plain radical, not doubled; see README -- "Open findings" #4. Needs its own stem, distinct from both the long and short ones. local past_v_stem_3mp = "حَي" local past_pass_c_stem = "حُيِي" local past_pass_v_stem_long = past_pass_c_stem local past_pass_v_stem_short = "حُيّ" local past_pass_v_stem_3mp = "حُي" local nonpast_stem = "حْي" local nonpast_pass_stem = nonpast_stem local imp_stem = _I .. nonpast_stem -- make parts past_2stem_conj(base, "past", {}, past_c_stem) past_2stem_conj(base, "past_pass", {}, past_pass_c_stem) local variant = vowel_spec.variant or "both" if variant == "short" or variant == "both" then past_2stem_conj(base, "past", past_v_stem_short, {}) past_2stem_conj(base, "past_pass", past_pass_v_stem_short, {}) end function inflect_long_variant(tense, long_stem, mp_stem) inflect_tense_1(base, tense, "", {long_stem, long_stem, long_stem, long_stem, mp_stem}, {past_endings[4], past_endings[5], past_endings[7], past_endings[8], past_endings[12]}, {"3ms", "3fs", "3md", "3fd", "3mp"}) end if variant == "long" or variant == "both" then inflect_long_variant("past", past_v_stem_long, past_v_stem_3mp) inflect_long_variant("past_pass", past_pass_v_stem_long, past_pass_v_stem_3mp) end nonpast_1stem_conj(base, "ind", "a", nonpast_stem, ind_endings_aa) nonpast_1stem_conj(base, "sub", "a", nonpast_stem, sub_endings_aa) nonpast_1stem_conj(base, "juss", "a", nonpast_stem, juss_endings_aa) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem, ind_endings_aa) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem, sub_endings_aa) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem, juss_endings_aa) inflect_tense_imp(base, {imp_stem, all_same = 1}, imp_endings_aa) -- active and passive participles apparently do not exist for this verb end -- Implement form-I final-weak assimilated+final-weak verb. ASSIMILATED is true for assimilated verbs. local function make_form_i_final_weak_verb(base, vowel_spec, assimilated) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- حَيَّ or حَيِيَ is weird enough that we handle it as a separate function. if hayy_radicals(rad1, rad2, rad3) then make_form_i_hayy_verb(base, vowel_spec) return end -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- Past and non-past stems, active and passive, and imperative stem. local past_stem = q(rad1, A, rad2) local past_pass_stem = q(rad1, U, rad2) local nonpast_stem, nonpast_pass_stem, imp_stem if raa_radicals(rad1, rad2, rad3) then base.irregular = true nonpast_stem = rad1 nonpast_pass_stem = rad1 imp_stem = rad1 else nonpast_pass_stem = q(rad1, SK, rad2) if assimilated then nonpast_stem = rad2 imp_stem = rad2 else nonpast_stem = nonpast_pass_stem imp_stem = q(form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1), rad2) end end -- A form-I mithal that keeps its wāw spells its imperative on a hamzat wasl, -- اِوْجَلْ / اِوْنَ. Tell postprocess_term() not to reduce that اِوْ to اِي -- see #10 -- there for why the substitution list can't decide this for itself. if not assimilated and req(rad1, W) then base.orth.keep_initial_w = true end -- Make parts. local past_ending_vowel = req(rad3, Y) and req(past_vowel, A) and "ay" or req(rad3, W) and req(past_vowel, A) and "aw" or req(past_vowel, I) and "ī" or "ū" -- Try to preserve footnotes attached to the third radical and/or past and/or non-past vowels. local past_footnotes = iut.combine_footnotes(rget_footnotes(rad3), rget_footnotes(past_vowel)) local nonpast_ending_vowel = req(nonpast_vowel, A) and "ā" or req(nonpast_vowel, I) and "ī" or "ū" local nonpast_footnotes = iut.combine_footnotes(rget_footnotes(rad3), rget_footnotes(nonpast_vowel)) make_final_weak_verb(base, iut.combine_form_and_footnotes(past_stem, past_footnotes), iut.combine_form_and_footnotes(past_pass_stem, past_footnotes), iut.combine_form_and_footnotes(nonpast_stem, nonpast_footnotes), iut.combine_form_and_footnotes(nonpast_pass_stem, nonpast_footnotes), iut.combine_form_and_footnotes(imp_stem, nonpast_footnotes), past_ending_vowel, nonpast_ending_vowel, "a", final_weak_uu_retains_rad3(rad1, rad2, rad3, past_ending_vowel, nonpast_ending_vowel)) -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, IN)) -- Active participle, stative type I فَعِيّ (++). FIXME: Is this correct when rad3 is W? insert_ap2_pp2(base, q(rad1, A, rad2, II, SH)) -- Active participle, stative type II فَعٍ (+++). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "ap3", q(rad1, A, rad2, IN)) -- Active participle, color/defect أَفْعَى (+cd). insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, AAMAQ)) -- Active participle, -ān فَعْيَان or فَعْوَان (+an). -- FIXME: Any examples of this for both rad3 = W and y to verify it's correct? insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, req(rad3, Y) and II or UU, SH)) end conjugations["I-final-weak"] = function(base, vowel_spec) make_form_i_final_weak_verb(base, vowel_spec, false) end conjugations["I-assimilated+final-weak"] = function(base, vowel_spec) make_form_i_final_weak_verb(base, vowel_spec, "assimilated") end conjugations["I-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- In some sense, hollow vowels i~i and u~u are more "correct" than a~i and a~u, but the latter follow the -- pattern of other form-I verbs, so we map i~i to a~i and u~u to a~u in infer_radicals(). Now however we have -- to undo this to get the actual past vowel based on the non-past vowel. if req(past_vowel, A) then past_vowel = map_vowel(past_vowel, function(vow) return req(nonpast_vowel, A) and I or rget(nonpast_vowel) end) end local lengthened_nonpast = map_vowel(nonpast_vowel, function(vow) return vow == U and UU or vow == I and II or AA end) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- active past stems - vowel (v) and consonant (c) local past_v_stem = q(rad1, AA, rad3) local past_c_stem = q(rad1, past_vowel, rad3) -- active non-past stems - vowel (v) and consonant (c) local nonpast_v_stem = q(rad1, lengthened_nonpast, rad3) local nonpast_c_stem = q(rad1, nonpast_vowel, rad3) -- passive past stems - vowel (v) and consonant (c) -- 'ufīla always; the contracted stem takes whichever of i/u the active past_c_stem -- *isn't* -- xāfa/xiftu (i) -> passive xufta (u), qāla/qultu (u) -> passive qilta (i). -- Contraction drops rad2, so this vowel is the only thing distinguishing active from -- passive in these forms, hence the two roots' behaviour is opposite, not independent. -- Confirmed on two printed pages (خافَ p.224 "خُفْتَ", قالَ p.138 "قِلْتَ"); previously -- hardcoded to u from خافَ alone, which is only half the pattern -- see README. local past_pass_v_stem = q(rad1, II, rad3) local past_pass_c_stem = q(rad1, map_vowel(past_vowel, function(vow) return vow == I and U or I end), rad3) -- passive non-past stems - vowel (v) and consonant (c) -- yufāla/yufalna -- stem is built differently but conjugation is identical to sound verbs local nonpast_pass_v_stem = q(rad1, AA, rad3) local nonpast_pass_c_stem = q(rad1, A, rad3) -- imperative stem local imp_v_stem = nonpast_v_stem local imp_c_stem = nonpast_c_stem -- make parts make_hollow_geminate_verb(base, false, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, "a") if kaan_radicals(rad1, rad2, rad3) then local endings = make_nonpast_endings(U, {}, {}, {}, {}) inflect_tense(base, "juss", nonpast_prefix_consonants, q(A, rad1), endings) base.irregular = true end -- Active participle. insert_form_or_forms(base, "ap1", req(rad3, HAMZA) and q(rad1, AA, HAMZA, IN) or q(rad1, AA, HAMZA, I, rad3)) -- Active participle, stative type I فَيِّد (++). FIXME: Any examples of this to verify it's correct? insert_ap2_pp2(base, q(rad1, A, Y, SH, I, rad3)) -- Active participle, stative type II فَيِد (+++). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "ap3", q(rad1, A, Y, I, rad3)) -- Active participle, color/defect أَفّيَد or أَفّوَد (+cd). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, A, rad3)) -- Active participle, -ān فَيْدَان or فَوْدَان (+an). Example: جَاعَ "to be hungry", act part جَوْعَان insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, req(rad2, Y) and II or UU, rad3)) end conjugations["I-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- active past stems - vowel (v) and consonant (c) -- The contracted stem's vowel is the past vowel for most roots (مَدَدْتَ) but follows the -- مضارع for the إتباع lexemes (هَمُمْتَ); see geminate_degemination_vowel(). local past_v_stem = q(rad1, A, rad2, SH) local past_c_stem = q(rad1, A, rad2, geminate_degemination_vowel(rad1, rad2, rad3, past_vowel, nonpast_vowel), rad2) -- active non-past stems - vowel (v) and consonant (c) local nonpast_v_stem = q(rad1, nonpast_vowel, rad2, SH) local nonpast_c_stem = q(rad1, SK, rad2, nonpast_vowel, rad2) -- passive past stems - vowel (v) and consonant (c) -- dulla/dulilta local past_pass_v_stem = q(rad1, U, rad2, SH) local past_pass_c_stem = q(rad1, U, rad2, I, rad2) -- passive non-past stems - vowel (v) and consonant (c) --yudallu/yudlalna -- stem is built differently but conjugation is identical to sound verbs local nonpast_pass_v_stem = q(rad1, A, rad2, SH) local nonpast_pass_c_stem = q(rad1, SK, rad2, A, rad2) -- imperative stem local imp_v_stem = q(rad1, nonpast_vowel, rad2, SH) local imp_c_stem = q(form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1), rad2, nonpast_vowel, rad2) -- make parts make_hollow_geminate_verb(base, "geminate", past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, "a") -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, SH)) -- Active participle, stative type I فَعِيع (++). FIXME: Any examples of this to verify it's correct? insert_ap2_pp2(base, q(rad1, A, rad2, II, rad2)) -- Active participle, stative type II فَعّ (+++). Example: بَرَّ "to be pious", active participle بَرّ insert_form_or_forms(base, "ap3", q(rad1, A, rad2, SH)) -- Active participle, color/defect أَفَعّ (+cd). -- Example: لَصَّ "to be thievish, to steal repeatedly", active participle أَلَصّ. insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, A, rad2, SH)) -- Active participle, -ān فَعَّان (+an). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "apan", q(rad1, A, rad2, SH, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, UU, rad2)) end -- Return the ta- (active, past and non-past) and tu- (passive past) prefixes for a form II/III/V/VI verb. -- Form V and VI verbs normally use ta- and tu-, but reduced (base.reduced) verbs use different prefixes. Form II -- and III verbs have no prefix. local function form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vform = base.verb_form if vform == "V" or vform == "VI" then if base.reduced then -- To simplify the code, we generate two rad1's with a sukūn between them, which is cleaned up in -- postprocessing. return q(_I, rad1, SK), q(rad1, SK), q(_U, rad1, SK) else return TA, TA, TU end else return "", "", "" end end -- Make form II or V sound or final-weak verb. local function make_form_ii_v_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vn = vform == "V" and q(ta_past_prefix, rad1, A, rad2, SH, final_weak and IN or q(U, rad3)) or q(TA, rad1, SK, rad2, II, final_weak and AH or rad3) -- various stem bases local past_stem_base = q(ta_past_prefix, rad1, A, rad2, SH) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, A, rad2, SH) local past_pass_stem_base = q(tu_past_prefix, rad1, U, rad2, SH) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["II-sound"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["II-final-weak"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end local function make_form_iii_alt_vn(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) -- Insert alternative verbal noun فِعَال. Since not all verbs have this, we require that verbs that do have it -- specify it explicitly; a shortcut ++ is provided to make this easier (e.g. <vn:+,++> to indicate that -- both the normal verbal noun مُفَاعَلَة and secondary verbal noun فِعَال are available). insert_form_or_forms(base, "vn2", q(rad1, I, rad2, AA, final_weak and HAMZA or rad3)) end -- Make form III or VI sound or final-weak verb. local function make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vn = vform == "VI" and q(ta_past_prefix, rad1, AA, rad2, final_weak and IN or q(U, rad3)) or q(MU, rad1, AA, rad2, final_weak and AAH or q(A, rad3, AH)) -- various stem bases local past_stem_base = q(ta_past_prefix, rad1, AA, rad2) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, AA, rad2) local past_pass_stem_base = q(tu_past_prefix, rad1, UU, rad2) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) if vform == "III" then make_form_iii_alt_vn(base, vowel_spec) end end conjugations["III-sound"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["III-final-weak"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end -- Make form III or VI geminate verb. local function make_form_iii_vi_geminate_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) -- Alternative verbal noun فِعَال will be inserted when we add sound parts below. local vn = vform == "VI" and q(ta_past_prefix, rad1, AA, rad2, SH) or q(MU, rad1, AA, rad2, SH, AH) -- Various stem bases. local past_stem_base = q(ta_past_prefix, rad1, AA) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, AA) local past_pass_stem_base = q(tu_past_prefix, rad1, UU) -- Make parts. local variant = vowel_spec.variant or "short" if variant == "short" or variant == "both" then make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Also add alternative sound (non-compressed) parts. This will lead to some duplicate entries, but they are -- removed during addition. if variant == "long" or variant == "both" then make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) elseif vform == "III" then -- Still need to add the alternative form-III verbal noun. make_form_iii_alt_vn(base, vowel_spec) end end conjugations["III-geminate"] = function(base, vowel_spec) make_form_iii_vi_geminate_verb(base, vowel_spec) end -- Make form IV sound or final-weak verb. local function make_form_iv_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) -- core of stem base, minus stem prefixes local stem_core -- check for irregular verb أَرَى local is_raa = raa_radicals(rad1, rad2, rad3) if is_raa then base.irregular = true stem_core = rad1 else stem_core = q(rad1, SK, rad2) end -- verbal noun local vn = is_raa and q(HAMZA, I, stem_core, AA, HAMZA, AH) or q(HAMZA, I, stem_core, AA, final_weak and HAMZA or rad3) -- various stem bases local past_stem_base = q(HAMZA, A, stem_core) local nonpast_stem_base = stem_core local past_pass_stem_base = q(HAMZA, U, stem_core) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IV-sound"] = function(base, vowel_spec) make_form_iv_sound_final_weak_verb(base, vowel_spec) end conjugations["IV-final-weak"] = function(base, vowel_spec) make_form_iv_sound_final_weak_verb(base, vowel_spec) end conjugations["IV-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- verbal noun local vn = q(HAMZA, I, rad1, AA, rad3, AH) -- various stem bases local past_stem_base = q(HAMZA, A, rad1) local nonpast_stem_base = rad1 local past_pass_stem_base = q(HAMZA, U, rad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IV-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(HAMZA, I, rad1, SK, rad2, AA, rad2) -- various stem bases local past_stem_base = q(HAMZA, A, rad1) local nonpast_stem_base = rad1 local past_pass_stem_base = q(HAMZA, U, rad1) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["V-sound"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["V-final-weak"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-sound"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-final-weak"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-geminate"] = function(base, vowel_spec) make_form_iii_vi_geminate_verb(base, vowel_spec) end -- Make a verbal noun of the general form that applies to forms VII and above. RAD12 is the first consonant cluster -- (after initial اِ) and RAD34 is the second consonant cluster. RAD5 is the final consonant. local function high_form_verbal_noun(rad12, rad34, rad5) return q(_I, rad12, I, rad34, AA, rad5) end -- Populate a sound or final-weak verb for any of the various high-numbered augmented forms (form VII and up) that -- have up to 5 consonants in two clusters in the stem and the same pattern of vowels between. Some of these -- consonants in certain verb parts are w's, which leads to apparent anomalies in certain stems of these parts, but -- these anomalies are handled automatically in postprocessing, where we resolve sequences of iwC -> īC, uwC -> ūC, -- w + sukūn + w -> w + shadda. -- RAD12 is the first consonant cluster (after initial اِ) and RAD34 is the second consonant cluster. RAD5 is the -- final consonant. local function make_high_form_sound_final_weak_verb(base, vowel_spec, rad12, rad34, rad5) local final_weak = is_final_weak(base, vowel_spec) local vn = high_form_verbal_noun(rad12, rad34, final_weak and HAMZA or rad5) -- various stem bases local nonpast_stem_base = q(rad12, A, rad34) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad12, U, rad34) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end local function form_vii_nrad1(base, rad1) if base.reduced then if not req(rad1, M) then error(("Internal error: Form VII first radical %s is not م but .reduced specified; should have been caught earlier"): format(rget(rad1))) end return M .. SH else return q("نْ", rad1) end end -- Make form VII sound or final-weak verb. local function make_form_vii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high_form_sound_final_weak_verb(base, vowel_spec, form_vii_nrad1(base, rad1), rad2, rad3) end conjugations["VII-sound"] = function(base, vowel_spec) make_form_vii_sound_final_weak_verb(base, vowel_spec) end conjugations["VII-final-weak"] = function(base, vowel_spec) make_form_vii_sound_final_weak_verb(base, vowel_spec) end conjugations["VII-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local nrad1 = form_vii_nrad1(base, rad1) local vn = high_form_verbal_noun(nrad1, Y, rad3) -- various stem bases local nonpast_stem_base = nrad1 local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nrad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["VII-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local nrad1 = form_vii_nrad1(base, rad1) local vn = high_form_verbal_noun(nrad1, rad2, rad2) -- various stem bases local nonpast_stem_base = q(nrad1, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nrad1, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Return Form VIII verbal noun. local function form_viii_verbal_noun(base, vowel_spec, rad1, rad2, rad3) local final_weak = is_final_weak(base, vowel_spec) rad3 = final_weak and HAMZA or rad3 return {high_form_verbal_noun(vowel_spec.form_viii_assim, rad2, rad3)} end -- Make form VIII sound or final-weak verb. local function make_form_viii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- check for irregular verb اِتَّخَذَ if axadh_radicals(rad1, rad2, rad3) then base.irregular = true rad1 = T end make_high_form_sound_final_weak_verb(base, vowel_spec, vowel_spec.form_viii_assim, rad2, rad3) end conjugations["VIII-sound"] = function(base, vowel_spec) make_form_viii_sound_final_weak_verb(base, vowel_spec) end conjugations["VIII-final-weak"] = function(base, vowel_spec) make_form_viii_sound_final_weak_verb(base, vowel_spec) end conjugations["VIII-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = form_viii_verbal_noun(base, vowel_spec, rad1, Y, rad3) -- various stem bases local nonpast_stem_base = vowel_spec.form_viii_assim local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nonpast_stem_base) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["VIII-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = form_viii_verbal_noun(base, vowel_spec, rad1, rad2, rad2) -- various stem bases local nonpast_stem_base = q(vowel_spec.form_viii_assim, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, vowel_spec.form_viii_assim, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IX-sound"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(_I, rad1, SK, rad2, I, rad3, AA, rad3) -- various stem bases local nonpast_stem_base = q(rad1, SK, rad2, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad1, SK, rad2, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IX-final-weak"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high_form_sound_final_weak_verb(base, vowel_spec, q(rad1, SK, rad2), rad3, rad3) end -- Populate a sound or final-weak verb for any of the various high-numbered -- augmented forms that have 5 consonants in the stem and the same pattern of -- vowels. Some of these consonants in certain verb parts are w's, which leads to -- apparent anomalies in certain stems of these parts, but these anomalies -- are handled automatically in postprocessing, where we resolve sequences of -- iwC -> īC, uwC -> ūC, w + sukūn + w -> w + shadda. local function make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, rad3, rad4, rad5) make_high_form_sound_final_weak_verb(base, vowel_spec, q(rad1, SK, rad2), q(rad3, SK, rad4), rad5) end -- Make form X sound or final-weak verb. local function make_form_x_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- check for irregular verb اِسْتَحْيَا (also اِسْتَحَى) local is_hayy = hayy_radicals(rad1, rad2, rad3) local variant = vowel_spec.variant or "both" if not is_hayy or variant == "long" or variant == "both" then make_high5_form_sound_final_weak_verb(base, vowel_spec, S, T, rad1, rad2, rad3) end if is_hayy and (variant == "short" or variant == "both") then base.irregular = true -- Add alternative entries to the verbal paradigms. Any duplicates are removed during addition. make_high_form_sound_final_weak_verb(base, vowel_spec, S .. SK .. T, rad1, rad3) end end conjugations["X-sound"] = function(base, vowel_spec) make_form_x_sound_final_weak_verb(base, vowel_spec) end conjugations["X-final-weak"] = function(base, vowel_spec) make_form_x_sound_final_weak_verb(base, vowel_spec) end conjugations["X-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(base.reduced and "اِسْ" or "اِسْتِ", rad1, AA, rad3, AH) -- various stem bases local past_stem_base = q(base.reduced and "اِسْ" or "اِسْتَ", rad1) local nonpast_stem_base = q(base.reduced and "سْ" or "سْتَ", rad1) local past_pass_stem_base = q(base.reduced and "اُسْ" or "اُسْتُ", rad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["X-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q("اِسْتِ", rad1, SK, rad2, AA, rad2) -- various stem bases local past_stem_base = q("اِسْتَ", rad1) local nonpast_stem_base = q("سْتَ", rad1) local past_pass_stem_base = q("اُسْتُ", rad1) -- make parts if base.altgem then inflect_tense(base, "past", "", {q(past_stem_base, A, rad2, SH), all_same = 1}, past_endings_ay_12_person_only) end make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn, base.altgem and "[uncommon]" or nil) end conjugations["XI-sound"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(_I, rad1, SK, rad2, II, rad3, AA, rad3) -- various stem bases local nonpast_stem_base = q(rad1, SK, rad2, AA) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad1, SK, rad2, UU) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Probably no form XI final-weak, since already geminate in form; would behave as XI-sound. -- Make form XII sound or final-weak verb. local function make_form_xii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, W, rad2, rad3) end conjugations["XII-sound"] = function(base, vowel_spec) make_form_xii_sound_final_weak_verb(base, vowel_spec) end conjugations["XII-final-weak"] = function(base, vowel_spec) make_form_xii_sound_final_weak_verb(base, vowel_spec) end -- Make form XIII sound or final-weak verb. local function make_form_xiii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, W, W, rad3) end conjugations["XIII-sound"] = function(base, vowel_spec) make_form_xiii_sound_final_weak_verb(base, vowel_spec) end conjugations["XIII-final-weak"] = function(base, vowel_spec) make_form_xiii_sound_final_weak_verb(base, vowel_spec) end -- Make a form XIV or XV sound or final-weak verb. Last radical appears twice (if`anlala / yaf`anlilu) so if it were -- w or y you'd get if`anwā / yaf`anwī or if`anyā / yaf`anyī, i.e. unlike for most augmented verbs, the identity of -- the radical matters. local function make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local lastrad = base.verb_form == "XV" and Y or rad3 make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, N, rad3, lastrad) end conjugations["XIV-sound"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end conjugations["XIV-final-weak"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end conjugations["XV-sound"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end -- Probably no form XV final-weak, since already final-weak in form; would behave as XV-sound. -- Make form Iq or IIq sound or final-weak verb. local function make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local vn = vform == "IIq" and q(TA, rad1, A, rad2, SK, rad3, (final_weak and IN or q(U, rad4))) or q(rad1, A, rad2, SK, rad3, (final_weak and AAH or q(A, rad4, AH))) local ta_pref = vform == "IIq" and TA or "" local tu_pref = vform == "IIq" and TU or "" -- various stem bases local past_stem_base = q(ta_pref, rad1, A, rad2, SK, rad3) local nonpast_stem_base = past_stem_base local past_pass_stem_base = q(tu_pref, rad1, U, rad2, SK, rad3) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["Iq-sound"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["Iq-final-weak"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIq-sound"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIq-final-weak"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end -- Make form IIIq sound or final-weak verb. local function make_form_iiiq_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, N, rad3, rad4) end conjugations["IIIq-sound"] = function(base, vowel_spec) make_form_iiiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIIq-final-weak"] = function(base, vowel_spec) make_form_iiiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IVq-sound"] = function(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local vn = q(_I, rad1, SK, rad2, I, rad3, SK, rad4, AA, rad4) -- various stem bases local past_stem_base = q(_I, rad1, SK, rad2, A, rad3) local nonpast_stem_base = q(rad1, SK, rad2, A, rad3) local past_pass_stem_base = q(_U, rad1, SK, rad2, U, rad3) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Probably no form IVq final-weak, since already geminate in form; would behave as IVq-sound. end create_conjugations() ------------------------------------------------------------------------------- -- Guts of main conjugation function -- ------------------------------------------------------------------------------- -- Given form, weakness and radicals, check to make sure the radicals present are allowable for the weakness. Hamzas on -- alif/wāw/yāʾ seats are never allowed (should always appear as hamza-on-the-line), and various weaknesses have various -- strictures on allowable consonants. local function check_radicals(form, weakness, rad1, rad2, rad3, rad4) local function hamza_check(index, rad) if rad == HAMZA_ON_ALIF or rad == HAMZA_UNDER_ALIF or rad == HAMZA_ON_W or rad == HAMZA_ON_Y then error("Radical " .. index .. " is " .. rad .. " but should be ء (hamza on the line)") end end local function check_waw_ya(index, rad) if not is_waw_ya(rad) then error("Radical " .. index .. " is " .. rad .. " but should be و or ي") end end local function check_not_waw_ya(index, rad) if is_waw_ya(rad) then error("In a sound verb, radical " .. index .. " should not be و or ي") end end hamza_check(rad1) hamza_check(rad2) hamza_check(rad3) hamza_check(rad4) if weakness == "assimilated" or weakness == "assimilated+final-weak" then if rad1 ~= W then error("Radical 1 is " .. rad1 .. " but should be و") end -- don't check that non-assimilated form I verbs don't have wāw as their -- first radical because some form-I verbs exist where a first-radical wāw -- behaves as sound, e.g. wajuha yawjuhu "to be distinguished". end if weakness == "final-weak" or weakness == "assimilated+final-weak" then if rad4 then check_waw_ya(4, rad4) else check_waw_ya(3, rad3) end elseif vform_supports_final_weak(form) then -- non-final-weak verbs cannot have weak final radical if there's a corresponding -- final-weak verb category. I think this is safe. We may have problems with -- ḥayya/ḥayiya yaḥyā if we treat it as a geminate verb. if rad4 then check_not_waw_ya(4, rad4) else check_not_waw_ya(3, rad3) end end if weakness == "hollow" then check_waw_ya(2, rad2) -- don't check that non-hollow verbs in forms that support hollow verbs -- don't have wāw or yāʾ as their second radical because some verbs exist -- where a middle-radical wāw/yāʾ behaves as sound, e.g. form-VIII izdawaja -- "to be in pairs". end if weakness == "geminate" then if rad4 then error("Internal error: No geminate quadrilaterals, should not be seen") end if rad2 ~= rad3 then error("Weakness is geminate; radical 3 is " .. rad3 .. " but should be same as radical 2 " .. rad2) end elseif vform_supports_geminate(form) then -- non-geminate verbs cannot have second and third radical same if there's -- a corresponding geminate verb category. I think this is safe. We -- don't fuss over double wāw or double yāʾ because this could legitimately -- be a final-weak verb with middle wāw/yāʾ, treated as sound. if rad4 then error("Internal error: No quadrilaterals should support geminate verbs") end if rad2 == rad3 and not is_waw_ya(rad2) then error("Weakness is '" .. weakness .. "'; radical 2 and 3 are same at " .. rad2 .. " but should not be; consider making weakness 'geminate'") end end end -- array of substitutions; each element is a 2-entry array FROM, TO; do it -- this way so the concatenations only get evaluated once local postprocess_subs = { -- reorder short-vowel + shadda -> shadda + short-vowel for easier processing {"(" .. AIU .. ")" .. SH, SH .. "%1"}, ----------same letter separated by sukūn should instead use shadda--------- ------------happens e.g. in kun-nā "we were".----------------- {"(.)" .. SK .. "%1", "%1" .. SH}, ---------------------------- assimilated verbs ---------------------------- -- iw, iy -> ī (assimilated verbs) {I .. W .. SK, II}, {I .. Y .. SK, II}, -- uw, uy -> ū (assimilated verbs) {U .. W .. SK, UU}, {U .. Y .. SK, UU}, -------------- final -yā uses tall alif not alif maqṣūra ------------------ {"(" .. Y .. SH .. "?" .. A .. ")" .. AMAQ, "%1" .. ALIF}, ----------------------- handle hamza assimilation ------------------------- -- initial hamza + short-vowel + hamza + sukūn -> hamza + long vowel {HAMZA .. A .. HAMZA .. SK, HAMZA .. A .. ALIF}, {HAMZA .. I .. HAMZA .. SK, HAMZA .. I .. Y}, {HAMZA .. U .. HAMZA .. SK, HAMZA .. U .. W} } local postprocess_tr_subs = { {"ī([" .. vowels .. "y*])", "iy%1"}, {"ū([" .. vowels .. "w*])", "uw%1"}, {"(.)%*", "%1%1"}, -- implement shadda ---------------------------- assimilated verbs ---------------------------- -- iw, iy -> ī (assimilated verbs) {"iw([^" .. vowels .. "w])", "ī%1"}, {"iy([^" .. vowels .. "y])", "ī%1"}, -- uw, uy -> ū (assimilated verbs) {"uw([^" .. vowels .. "w])", "ū%1"}, {"uy([^" .. vowels .. "y])", "ū%1"}, ----------------------- handle hamza assimilation ------------------------- -- initial hamza + short-vowel + hamza + sukūn -> hamza + long vowel {"ʔaʔ(" .. NV .. ")", "ʔā%1"}, {"ʔiʔ(" .. NV .. ")", "ʔī%1"}, {"ʔuʔ(" .. NV .. ")", "ʔū%1"}, } -- Post-process verb parts to eliminate phonological anomalies. Many of the changes, particularly the tricky ones, -- involve converting hamza to have the proper seat. The rules for this are complicated and are documented on the -- [[w:Hamza]] Wikipedia page. In some cases there are alternatives allowed, and we handle them below by returning -- multiple possibilities. -- -- ORTHOGRAPHIC EXEMPTIONS (`orth`, set on `base` by the code that knows the fact) -- -- postprocess_subs above and ar_utilities.process_hamza apply general Arabic orthography -- to a finished word. A handful of the reference's verb spellings are decisions neither -- pass can make, because what settles them is not in the string: that a hamza is a -- *radical*, that a short vowel is the stem's rather than iʿrāb, that a verb is form VIII -- rather than form I. Each of the three has both halves in the corpus. Form I أَتَى and -- form VIII اِئْتَمَنَ reach this function's 1s مضارع as the same ءَ + ءْ + ت, and the book -- prints آتِي for the one and أَأْتَمِنُ for the other (#14). شَأَى's imperative اُشْؤُ and -- جَرُؤَ's past both end in hamza + a short vowel, and only the first of the two vowels -- belongs to the stem (#12). -- -- module/ar-utilities.lua is byte-identical to module/vendor/ar-utilities.lua, and its -- other callers are on-wiki nominal templates that cannot be regression-tested from this -- repo, so none of this is fixed by editing it. Each case is instead handled on the -- caller's side of the shared passes, in one of three ways: -- -- * park the sequence on a placeholder so a shared rule cannot see it, and restore it -- afterwards -- #10's original mechanism, now also #14's; -- * add a spelling beside whatever the shared code returns, because the reference -- attests both and this module already carries alternatives elsewhere -- #16, #17; or -- * rewrite the sequence, either before the shared pass so its rule never matches -- (#12's past clause writes the ā as an ألف مقصورة, which the madda rule does not -- look for) or after it, where the shared code has made a decision the caller's fact -- overrides (#12's seating clause, #18's). -- -- Adding one is three things: a key in `base.orth` set by the builder that knows the -- fact, a clause here naming the finding, and the XFAIL entries in tests/run.lua deleted -- in the same commit. -- -- Three of the clauses below are unconditional, and deliberately so: they repair a -- spelling the shared code produces that Arabic does not use (#18), restore one it -- discards (#17), or extend one of its rules to a seat it never reaches (#15). None of -- the three needs a fact from the caller -- only a place to live that is not the shared -- file. local function postprocess_term(term, orth) if term == "?" then return "?" end orth = orth or {} -- Phase 1: the reductions in postprocess_subs, with the parked exemptions held out -- across them. `park_w` is a parameter rather than read from `orth` because #16 runs -- this phase both ways -- see below. local function reduce(t, park_w) -- Add BORDER at text boundaries. t = BORDER .. t .. BORDER -- #10: a word-initial hamzat wasl + wāw + sukūn is how a form-I mithal that keeps -- its wāw spells its imperative (اِوْنَ، اِوْجَلْ). The iw -> ī reduction is right -- everywhere else and cannot simply be dropped -- it is what turns the form VIII -- mithal past اِوْتَعَدَ into اِيتَعَدَ, in the same word-initial position, so position -- alone doesn't separate the two cases. if park_w then t = rsub(t, BORDER .. ALIF .. I .. W .. SK, BORDER .. ALIF .. I .. KEEP_W_PH .. SK) end -- #14: the reference spells the 1s مضارع of a hamza-initial form VIII analytically, -- أَأْتَمِنُ -- the prefix hamza, then the quiescent radical hamza on its own seat -- -- where the أَ + أْ -> آ contraction below and the alif-madda rule in process_hamza -- between them return آتَمِنُ. Form I of a hamza-initial root arrives here with the -- identical sequence and *is* printed contracted (آكُلُ، آخُذُ، آتِي), so the string -- cannot decide this and the flag does. if orth.analytic_hamza then t = rsub(t, BORDER .. HAMZA .. A .. HAMZA .. SK, BORDER .. HAMZA .. A .. KEEP_HAMZA_PH .. SK) end -- Do the main post-processing, based on the pattern substitutions in postprocess_subs. for _, sub in ipairs(postprocess_subs) do t = rsub(t, sub[1], sub[2]) end -- #12, the past half: a form-I final-weak verb on a wāw rad3 builds its past 3ms ā -- as a real alif (دَعَا), and after a hamza rad2 that puts hamza-on-alif + fatḥa + -- alif in front of process_hamza's madda rule (أَا -> آ), which swallows the seat: -- شَأَى comes out شَآ. The rule itself is right -- it is what gives قَرَأَ its dual قَرَآ -- -- and it keys on ALIF, not on AMAQ, which is why the module already gets رَأَى and -- نَأَى right: a yāʾ rad3 never produces the second alif. So the ā is written as an -- ألف مقصورة here instead, which sidesteps the madda rule rather than suppressing -- it. Only the past 3ms of an -aw verb reaches this sequence; every derived form -- already ends its past 3ms in AMAQ. if orth.hamza_rad2_final_weak then t = rsub(t, HAMZA .. A .. ALIF .. BORDER, HAMZA .. A .. AMAQ .. BORDER) end t = t:gsub(KEEP_W_PH, W) -- Restored as a plain hamza, not a seat: the seating pass below still has to place it, -- and gets it right (a quiescent medial hamza after a fatḥa takes the alif). t = t:gsub(KEEP_HAMZA_PH, HAMZA) return (t:gsub(BORDER, "")) end -- Phase 2: hamza seats. Returns the list of spellings for one reduced term. local function seat(t) if not rfind(t, HAMZA) then return { t } end -- #17: after a long ū, process_hamza treats و + ؤُ + و as having no acceptable -- original spelling and returns only its two alternatives (ئُو، ءُو) -- while the -- same function keeps the original beside the alternative after a long ā, which is -- what gives سَاؤُوا. The reference prints the wāw seat after a long ū throughout -- (يَسُوؤُونَ، تَسُوؤُونَ، سُوؤُوا, against its own تَسُوئِينَ / سُوئِي for the kasra: the -- seat follows the hamza's own vowel). So the discarded spelling is added back -- here, after theirs rather than instead of them -- the same order the long-ā -- branch of that function already uses. local waw_seat if rfind(t, W .. HAMZA .. U .. W) then local parked = t:gsub(W .. HAMZA .. U .. W, W .. WAW_SEAT_PH .. U .. W) waw_seat = ar_utilities.process_hamza(parked:gsub(HAMZA, HAMZA_PH))[1] waw_seat = waw_seat:gsub(WAW_SEAT_PH, HAMZA_ON_W) end local out = ar_utilities.process_hamza(t:gsub(HAMZA, HAMZA_PH)) if waw_seat then out[#out + 1] = waw_seat end -- #12, the imperative/jussive half: apocopation strands the radical hamza at the -- end of the word, where process_hamza takes its seat from the *preceding* letter's -- vowel and, finding a sukūn, drops it on the line -- اُشْءُ, أَفْءِ. That rule is right -- for a hamza whose own vowel is iʿrāb (جُزْء), and cannot be made position-sensitive -- inside the shared file: the fatḥa of جَرُؤَ is the ماضٍ ending and the ḍamma before -- it rightly wins the seat. Here the short vowel left after rad3 goes is the stem's -- own, which is the fact this flag carries. The reference seats it: اُشْؤُ (p160, شأو, -- form I) and أَفْإِ (p300, فأو, form IV) -- two roots, two measures, two pages, and -- the two vowels between them. -- -- The seat is the wāw for a ḍamma and the alif otherwise; #18's substitution below -- then writes the alif's hamza underneath when the vowel is a kasra, which is how -- the alif branch comes out as إ. That is not a third rule invented here: the same -- إ is what p368's اِرْتَإِ and p182's imperative إ already print, and what the module -- already produces for both, having reached the alif seat by its own route (a -- preceding fatḥa, and word-initial position). Only p300 needed the seat itself. -- -- Gated on a preceding sukūn, which is the whole of what the two pages attest. The -- hamza after a *long* vowel is left alone (وأي's passive jussive تُوءَ), where the -- shared code also leaves it on the line and no page here prints the slot; and the -- pattern wants the vowel directly after the hamza, so a šadda'd seat (p271's رَأِّ) -- is not matched, the same gate #18 gets for free. if orth.hamza_rad2_final_weak then for i, v in ipairs(out) do v = rsub(BORDER .. v .. BORDER, SK .. HAMZA .. "(" .. AIU .. ")" .. BORDER, function(vowel) return SK .. (vowel == U and HAMZA_ON_W or HAMZA_ON_ALIF) .. vowel .. BORDER end) out[i] = (v:gsub(BORDER, "")) end end -- #15: a hamza left on the line, carrying a fatḥa, with an alif after it, is -- printed as a madda -- but not when that would end the word directly after an -- alif. The reference sets يَداآنِ / تَداآنِ (p201, دوء), يَهاآنِ / تَهاآنِ (p242, -- هوء) and يَسُوآنِ / تَسُوآنِ (p157, سوء) for the مضارع duals, and سُوآ for p157's -- imperative dual, while the same three verbs print their *word-final* duals -- داءا / هاءا / سَاءَا with the hamza on the line -- so the exception is carried by -- the preceding alif, not by the position alone: سُوآ is word-final too and does -- contract, after a long ū. Ten cells, three roots, three pages, and the blocked -- case attested on all three. process_hamza applies this rule only to a hamza it -- has seated on an alif (which is what gives تَدَرْبَآ); after a long vowel it -- leaves the seat on the line and the rule never fires, so the extension is here. -- -- Confined to a hamza after a *long vowel*, which is the whole of what those three -- pages attest. Written without that gate it also contracts the فَعْلان participle -- سَوْءَان to سَوْآن -- a hamza after the sukūn of a diphthong, not after a long vowel, -- in a slot no fixture covers and no page here prints. The mater carries no sukūn -- in a long vowel as this module builds one, so requiring the two to be adjacent -- is the gate. for i, v in ipairs(out) do v = BORDER .. v .. BORDER v = rsub(v, ALIF .. HAMZA .. A .. ALIF .. BORDER, ALIF .. NO_MADDA_PH .. A .. ALIF .. BORDER) v = rsub(v, "([" .. AMAD .. ALIF .. W .. Y .. "])" .. HAMZA .. A .. ALIF, "%1" .. AMAD) v = v:gsub(NO_MADDA_PH, HAMZA) out[i] = (v:gsub(BORDER, "")) end -- #18: a hamza seated on an alif and carrying a bare kasra is written *below* the -- alif. The seat is chosen from the preceding letter's vowel, which is right when -- the hamza's own vowel is iʿrāb (مَلْجَأٍ) and is why the rule lives in the shared -- code; in a verb there is no iʿrāb, and the kasra that follows an apocopated -- radical hamza is the stem's own -- so اِرْتَأَى's jussive is اِرْتَإِ. -- -- Not when the alif also carries a šadda, and the corpus is what says so: رَأَّى -- (p271, form II of the *same* root رأو as p368's اِرْتَأَى) prints its apocopated -- imperative رَأِّ, hamza above, where the doubled radical puts a šadda on the seat. -- Written unconditionally first, this rule broke that cell and nothing else in 379 -- paradigms, which is both the counter-example and its own control: one root, two -- forms, two pages, and the presence of the šadda the only difference between them. -- No test for the šadda is needed: postprocess_subs has already reordered it in -- front of the vowel, so أ + šadda + kasra simply isn't the sequence below. for i, v in ipairs(out) do out[i] = v:gsub(HAMZA_ON_ALIF .. I, HAMZA_UNDER_ALIF .. I) end return out end local out = seat(reduce(term, orth.keep_initial_w)) -- #16: a retained-wāw مثال reduces اِوْ to اِيْ in the imperative on p240 (اِيْجَعْ, all -- five cells) and does not on p247 (اِوْنَ) -- the book carries both spellings and -- grammar allows both, so the module carries both too, the way it already carries -- سَائُوا | سَاؤُوا. #10's park is still right; what was missing is the second form -- beside the first. Widening loose() instead would retire the suite's only check on -- the mithal-wāwī / mithal-yāʾī imperative distinction, which is why this is here. if orth.keep_initial_w then for _, reduced in ipairs(seat(reduce(term, false))) do local seen = false for _, v in ipairs(out) do if v == reduced then seen = true break end end if not seen then out[#out + 1] = reduced end end end if #out == 1 then return out[1] end return out end local function postprocess_translit(translit) if translit == "?" then return "?" end -- Add BORDER at text boundaries. translit = BORDER .. translit .. BORDER -- Do the main post-processing, based on the pattern substitutions in postprocess_tr_subs. for _, sub in ipairs(postprocess_tr_subs) do translit = rsub(translit, sub[1], sub[2]) end translit = translit:gsub(BORDER, "") return translit end -- #13: نقل الحركة + إدغام in a لفيف مقرون verb outside form I. Where the paradigm brings -- two vowelled yāʾs together on a *quiescent* radical, the reference moves the first -- yāʾ's vowel back onto that radical and geminates the pair: أَحْيَيَا -> أَحَيَّا, -- يُحْيِيَانِ / تُحْيِيَانِ -> يُحِيَّانِ / تُحِيَّانِ, أَحْيِيَا -> أَحِيَّا (p302, form IV on حيي), -- and the same five slots of اِسْتَحْيَا (p389, form X on the same root). -- -- Only the نقل is done here. The إدغام is postprocess_subs' own "same letter separated -- by a sukūn -> šadda" rule, which already puts the šadda in front of the vowel the way -- the rest of the module writes it -- so this has to run *before* postprocess_forms(), -- and does, from conjugate_verb(). -- -- The gate is that sukūn, and the reference's pages calibrate it from both sides: -- * a *long* vowel in front of the first yāʾ can host nothing, and عايا (p287, form -- III on عيي) prints يُعايِيانِ، تُعايِيانِ and عايِيا with both yāʾs. The pattern below -- does not match them because there is no consonant + sukūn there to match; -- * the second yāʾ has to carry a short vowel of its own. The control is in the same -- columns of the pages that do contract: يُحْيِي (3ms), يُحْيِينَ (3fp) and تُحْيِينَ -- (2fs), where it is the mater of a long ī, are printed uncontracted, and the -- pattern does not match those either. -- Between them the two conditions pick out exactly the five slots of p302 and p389 that -- contract, against the twenty-odd cells of the same two paradigms that do not. -- -- Form I is excluded, and by attestation rather than by phonology: حَيِيَ (p229, 017_159) -- and عَيِيَ (p250, 238_159) put the identical consonant + sukūn + yāʾ sequence in their -- مضارع and imperative duals and both print them uncontracted -- يَحْيَيانِ، اِحْيَيا، -- يَعْيَيانِ، اِعْيَيا. Form I has its own version of this alternation, حَيَّ beside حَيِيَ, -- and it is a different mechanism: the vowel is dropped rather than moved, which is why -- make_form_i_hayy_verb() spells it out as a stem pair instead. -- -- Added beside the uncontracted form and never instead of it: both spellings are -- current, and form X already carries the two the module builds from its own stems -- (اِسْتَحْيَيَا | اِسْتَحَيَا), so this is a third entry in a slot that already had a choice. local NAQL_FROM = "(.)" .. SK .. Y .. "(" .. AIU .. ")" .. Y .. "(" .. AIU .. ")" local NAQL_TO = "%1%2" .. Y .. SK .. Y .. "%3" local function add_naql_idgham_forms(base) -- "none" is a user-supplied paradigm with no verb form to reason from; form I is -- excluded on the evidence above. if base.verb_form == "I" or base.verb_form == "none" then return end -- Collected first and inserted afterwards: inserting into base.forms while iterating -- over it would let a contracted form be re-read as a source in the same pass. local additions for slot, forms in pairs(base.forms) do for _, form in ipairs(forms) do local naql, changed = rsubb(form.form, NAQL_FROM, NAQL_TO) if changed then additions = additions or {} additions[#additions + 1] = {slot, {form = naql, footnotes = form.footnotes}} end end end if additions then for _, addition in ipairs(additions) do iut.insert_form(base.forms, addition[1], addition[2]) end end end local function postprocess_forms(base) local converted_values = {} for slot, forms in pairs(base.forms) do local need_dedup = false for i, form in ipairs(forms) do local term = postprocess_term(form.form, base.orth) local translit = form.translit and postprocess_translit(form.translit) or nil if term ~= form.form or translit ~= form.translit then need_dedup = true end converted_values[i] = {term, translit} end if need_dedup then local temp_dedup = {} for i = 1, #forms do local new_term, new_translit = unpack(converted_values[i]) if type(new_term) == "table" then for _, nt in ipairs(new_term) do local new_formobj = { form = nt, translit = new_translit, footnotes = forms[i].footnotes, } iut.insert_form(temp_dedup, "temp", new_formobj) end else local new_formobj = { form = new_term, translit = new_translit, footnotes = forms[i].footnotes, } iut.insert_form(temp_dedup, "temp", new_formobj) end end base.forms[slot] = temp_dedup.temp end end end local function process_slot_overrides(base) for slot, forms in pairs(base.slot_overrides) do local existing_values = base.forms[slot] base.forms[slot] = nil for _, form in ipairs(forms) do -- + in active participle for form I requests slot ap1 if form.form == "+" and (base.verb_form ~= "I" or slot ~= "ap") then if not existing_values then error(("Slot '%s' requested the default value but no such value available"):format(slot)) end -- We maintain an invariant that no two slots share a form object (although they may share the footnote -- lists inside the form objects). However, there is no need to copy the form objects here because there -- is a one-to-one correspondence between slots and slot overrides, i.e. you can't have a default value -- go into two slots. insert_form_or_forms(base, slot, existing_values, "allow overrides", form.uncertain) elseif default_indicator_to_active_participle_slot[form.form] then if form.form == "++" then if slot ~= "vn" and slot ~= "ap" and slot ~= "pp" then error(("Secondary default value request '++' only applicable to verbal nouns and pariciples, but found in slot '%s'"): format(slot)) end else if slot ~= "ap" then error(("Secondary default value request '%s' only applicable to active pariciples, but found in slot '%s'"): format(form.form, slot)) end end local secondary_default_slot = slot == "vn" and "vn2" or slot == "pp" and "pp2" or default_indicator_to_active_participle_slot[form.form] local existing_values = base.forms[secondary_default_slot] if not existing_values then error(("Slot '%s' requested a secondary default value using '%s' but no such value available"): format(slot, form.form)) end -- See comment above about the lack of need to copy the form objects. insert_form_or_forms(base, slot, existing_values, "allow overrides", form.uncertain) -- To make sure there aren't shared form objects. base.forms[secondary_default_slot] = nil else insert_form_or_forms(base, slot, form, "allow overrides", form.uncertain) end end end -- Now, for non-stative form-I verbs, fill the active participle slot from ap1 unless it should be missing (e.g. -- passive-only or user specified 'ap:-'). if base.verb_form == "I" and not base.forms.ap and base.forms.ap1 and not skip_slot(base, "ap") then local saw_non_stative = false for _, vowel_spec in ipairs(base.conj_vowels) do if req(vowel_spec.past, A) then saw_non_stative = true break end end if saw_non_stative then base.forms.ap = base.forms.ap1 -- To make sure there aren't shared form objects. base.forms.ap1 = nil end end end local function handle_lemma_linked(base) -- Compute linked versions of potential lemma slots, for use in {{ar-verb}}. We substitute the original lemma -- (before removing links) for forms that are the same as the lemma, if the original lemma has links. for _, slot in ipairs(export.potential_lemma_slots) do if base.forms[slot] then insert_form_or_forms(base, slot .. "_linked", iut.map_forms(base.forms[slot], function(form) if form == base.lemma and rfind(base.linked_lemma, "%[%[") then return base.linked_lemma else return form end end)) end end end -- Process specs given by the user using 'addnote[SLOTSPEC][FOOTNOTE][FOOTNOTE][...]'. local function process_addnote_specs(base) for _, spec in ipairs(base.addnote_specs) do for _, slot_spec in ipairs(spec.slot_specs) do slot_spec = "^" .. slot_spec .. "$" for slot, forms in pairs(base.forms) do if rfind(slot, slot_spec) then -- To save on memory, side-effect the existing forms. for _, form in ipairs(forms) do form.footnotes = iut.combine_footnotes(form.footnotes, spec.footnotes) end end end end end end local function add_missing_links_to_forms(base) -- Any forms without links should get them now. Redundant ones will be stripped later. for slot, forms in pairs(base.forms) do for _, form in ipairs(forms) do if not form.form:find("%[%[") then form.form = "[[" .. form.form .. "]]" end end end end local function conjugate_verb(base) construct_stems(base) -- Orthographic exemptions the shared hamza/reduction passes can't decide for -- themselves; see the block comment on postprocess_term(). Most are set by the -- stem builder that knows the fact, so this table has to exist before dispatch. base.orth = base.orth or {} for _, vowel_spec in ipairs(base.conj_vowels) do -- Reconstruct conjugation type from verb form and (possibly inferred) weakness. conj_type = base.verb_form .. "-" .. vowel_spec.weakness -- #14: a hamza-initial form VIII spells its 1s مضارع analytically (أَأْتَمِنُ), where -- form I of a hamza-initial root contracts to a madda (آتِي) from the identical -- sequence. Declared here rather than in the four VIII builders because it holds -- for all of them and depends on nothing else they compute. if base.verb_form == "VIII" and vowel_spec.rad1 and rfind(rget(vowel_spec.rad1) or "", HAMZA_ANY) then base.orth.analytic_hamza = true end -- #12: a final-weak verb whose rad2 is a hamza. Two of the shared passes read that -- radical as an ordinary word-final hamza and get it wrong in the two slots where -- rad3 stops being a consonant -- see the two clauses in postprocess_term(). Like -- #14 this is declared here rather than in a builder: it holds for every form the -- root can take (شأو, فأو, رأو, وأي in the corpus, forms I, II, IV and VIII) and -- depends on nothing else the builders compute. if rfind(vowel_spec.weakness or "", "final%-weak") and vowel_spec.rad2 and rfind(rget(vowel_spec.rad2) or "", HAMZA_ANY) then base.orth.hamza_rad2_final_weak = true end -- Check that the conjugation type is recognized. if not conjugations[conj_type] then error("Unknown conjugation type '" .. conj_type .. "'") end -- The way the conjugation functions work is they always add entries to the appropriate parts of the paradigm -- (each of which is an array), rather than setting the values. This makes it possible to call more than one -- conjugation function and essentially get a paradigm of the "either A or B" kind. Doing this may insert -- duplicate entries into a particular paradigm part, but this is not a problem because we check for duplicate -- entries when adding them, and don't insert in that case. conjugations[conj_type](base, vowel_spec) end -- #13: before postprocess_forms(), which supplies the إدغام half of the alternation. add_naql_idgham_forms(base) postprocess_forms(base) process_slot_overrides(base) -- This should happen before add_missing_links_to_forms() so that the comparison `form == base.lemma` in -- handle_lemma_linked() works correctly and compares unlinked forms to unlinked forms. handle_lemma_linked(base) process_addnote_specs(base) if not base.alternant_multiword_spec.args.noautolinkverb then add_missing_links_to_forms(base) end end local function parse_indicator_spec(angle_bracket_spec) -- Store the original angle bracket spec so we can reconstruct the overall conj spec with the lemma(s) in them. local base = { angle_bracket_spec = angle_bracket_spec, conj_vowels = {}, root_consonants = {}, user_stem_overrides = {}, user_slot_overrides = {}, slot_explicitly_missing = {}, slot_uncertain = {}, slot_override_uses_default = {}, addnote_specs = {}, } local function parse_err(msg) error(msg .. ": " .. angle_bracket_spec) end local function fetch_footnotes(separated_group) local footnotes for j = 2, #separated_group - 1, 2 do if separated_group[j + 1] ~= "" then parse_err("Extraneous text after bracketed footnotes: '" .. table.concat(separated_group) .. "'") end if not footnotes then footnotes = {} end table.insert(footnotes, separated_group[j]) end return footnotes end local inside = angle_bracket_spec:match("^<(.*)>$") assert(inside) local segments = put.parse_multi_delimiter_balanced_segment_run(inside, {{"[", "]"}, {"<", ">"}}) local dot_separated_groups = put.split_alternating_runs_and_strip_spaces(segments, "%.") -- The first dot-separated element must specify the verb form, e.g. IV or IIq. If the form is I, it needs to include -- the the past and non-past vowels, e.g. I/a~u for kataba ~ yaktubu. More than one vowel can be given, -- comma-separated, and more than one past~non-past pair can be given, slash-separated, e.g. I/a,u~u/i~a for form I -- كمل, which can be conjugated as kamala/kamula ~ yakmulu or kamila ~ yakmalu. An individual vowel spec must be one -- of a, i or u and in general (a) at least one past~non-past pair most be given, and (b) both past and non-past -- vowels must be given even though sometimes the vowel can be determined from the unvocalized form. An exception is -- passive-only verbs, where the vowels can't in general be determined (except indirectly in some cases by looking -- at an associated non-passive verb); in that case, the vowel~vowel spec can left out. local slash_separated_groups = put.split_alternating_runs_and_strip_spaces(dot_separated_groups[1], "/") local form_spec = slash_separated_groups[1] base.form_footnotes = fetch_footnotes(form_spec) if form_spec[1] == "" then parse_err("Missing verb form") end if not allowed_vforms_with_weakness_set[form_spec[1]] then parse_err(("Unrecognized verb form '%s', should be one of %s"):format( form_spec[1], list_to_text(allowed_vforms, nil, " or "))) end if form_spec[1]:find("%-") then base.verb_form, base.explicit_weakness = form_spec[1]:match("^(.-)%-(.*)$") else base.verb_form = form_spec[1] end if #slash_separated_groups > 1 then if base.verb_form ~= "I" then parse_err(("Past~non-past vowels can only be specified when verb form is I, but saw form '%s'"):format( base.verb_form)) end for i = 2, #slash_separated_groups do local slash_separated_group = slash_separated_groups[i] local tilde_separated_groups = put.split_alternating_runs_and_strip_spaces(slash_separated_group, "~") if #tilde_separated_groups ~= 2 then parse_err(("Expected two tilde-separated vowel specs: %s"):format(table.concat(slash_separated_group))) end local function parse_conj_vowels(tilde_separated_group, vtype) local conj_vowel_objects = {} local comma_separated_groups = put.split_alternating_runs_and_strip_spaces(tilde_separated_group, ",") for _, comma_separated_group in ipairs(comma_separated_groups) do local conj_vowel = comma_separated_group[1] if conj_vowel ~= "a" and conj_vowel ~= "i" and conj_vowel ~= "u" then parse_err(("Expected %s conjugation vowel '%s' to be one of a, i or u in %s"):format( vtype, conj_vowel, table.concat(slash_separated_group))) end conj_vowel = dia[conj_vowel] local conj_vowel_footnotes = fetch_footnotes(comma_separated_group) -- Try to use strings when possible as it makes q() significantly more efficient. if conj_vowel_footnotes then table.insert(conj_vowel_objects, {form = conj_vowel, footnotes = conj_vowel_footnotes}) else table.insert(conj_vowel_objects, conj_vowel) end end return conj_vowel_objects end local conj_vowel_spec = { past = parse_conj_vowels(tilde_separated_groups[1], "past"), nonpast = parse_conj_vowels(tilde_separated_groups[2], "non-past"), } table.insert(base.conj_vowels, conj_vowel_spec) end end for i = 2, #dot_separated_groups do local dot_separated_group = dot_separated_groups[i] local first_element = dot_separated_group[1] if first_element == "addnote" then local spec_and_footnotes = fetch_footnotes(dot_separated_group) if #spec_and_footnotes < 2 then parse_err("Spec with 'addnote' should be of the form 'addnote[SLOTSPEC][FOOTNOTE][FOOTNOTE][...]'") end local slot_spec = table.remove(spec_and_footnotes, 1) local slot_spec_inside = rmatch(slot_spec, "^%[(.*)%]$") if not slot_spec_inside then parse_err("Internal error: slot_spec " .. slot_spec .. " should be surrounded with brackets") end local slot_specs = rsplit(slot_spec_inside, ",") -- FIXME: Here, [[Module:it-verb]] called strip_spaces(). Generally we don't do this. Should we? table.insert(base.addnote_specs, {slot_specs = slot_specs, footnotes = spec_and_footnotes}) elseif first_element:find("^var:") then if #dot_separated_group > 1 then parse_err(("Can't attach footnotes to 'var:' spec '%s'"):format(first_element)) end base.variant = first_element:match("^var:(.*)$") elseif first_element:find("^I+V?:") then local root_cons, root_cons_value = first_element:match("^(I+V?):(.*)$") local root_index if root_cons == "I" then root_index = 1 elseif root_cons == "II" then root_index = 2 elseif root_cons == "III" then root_index = 3 elseif root_cons == "IV" then root_index = 4 if not base.verb_form:find("q$") then parse_err(("Can't specify root consonant IV for non-quadriliteral verb form '%s': %s"):format( base.verb_form, first_element)) end end local cons, translit = root_cons_value:match("^(.*)//(.*)$") if not cons then cons = root_cons_value end local root_footnotes = fetch_footnotes(dot_separated_group) if not translit and not root_footnotes then base.root_consonants[root_index] = cons else base.root_consonants[root_index] = {form = cons, translit = translit, footnotes = root_footnotes} end elseif first_element:find("^[a-z][a-z0-9_]*:") then local slot_or_stem, remainder = first_element:match("^(.-):(.*)$") dot_separated_group[1] = remainder local comma_separated_groups = put.split_alternating_runs_and_strip_spaces(dot_separated_group, "[,،]") if overridable_stems[slot_or_stem] then if base.user_stem_overrides[slot_or_stem] then parse_err("Overridable stem '" .. slot_or_stem .. "' specified twice") end base.user_stem_overrides[slot_or_stem] = overridable_stems[slot_or_stem](comma_separated_groups, {prefix = slot_or_stem, base = base, parse_err = parse_err, fetch_footnotes = fetch_footnotes}) else -- assume a form override; we validate further later when the possible slots are available if base.user_slot_overrides[slot_or_stem] then parse_err("Form override '" .. slot_or_stem .. "' specified twice") end base.user_slot_overrides[slot_or_stem] = allow_multiple_values_for_override(comma_separated_groups, {prefix = slot_or_stem, base = base, parse_err = parse_err, fetch_footnotes = fetch_footnotes}, "is form override") end elseif indicator_flags[first_element] then if #dot_separated_group > 1 then parse_err("No footnotes allowed with '" .. first_element .. "' spec") end if base[first_element] then parse_err("Spec '" .. first_element .. "' specified twice") end base[first_element] = true else local passive, uncertain = first_element:match("^(.*)(%?)$") passive = passive or first_element uncertain = not not uncertain if passive_types[passive] then if #dot_separated_group > 1 then parse_err("No footnotes allowed with '" .. passive .. "' spec") end if base.passive then parse_err("Value for passive type specified twice") end base.passive = passive base.passive_uncertain = uncertain else parse_err("Unrecognized spec '" .. first_element .. "'") end end end return base end -- Normalize all lemmas, substituting the pagename for blank lemmas and adding links to multiword lemmas. local function normalize_all_lemmas(alternant_multiword_spec, head) -- (1) Add links to all before and after text. Remember the original text so we can reconstruct the verb spec later. if not alternant_multiword_spec.args.noautolinktext then iut.add_links_to_before_and_after_text(alternant_multiword_spec, "remember original") end -- (2) Remove any links from the lemma, but remember the original form so we can use it below in the 'lemma_linked' -- form. iut.map_word_specs(alternant_multiword_spec, function(base) if base.lemma == "" then base.lemma = head end base.user_specified_lemma = base.lemma base.lemma = m_links.remove_links(base.lemma) base.user_specified_verb = base.lemma base.verb = base.user_specified_verb local linked_lemma if alternant_multiword_spec.args.noautolinkverb or base.user_specified_lemma:find("%[%[") then linked_lemma = base.user_specified_lemma else -- Add links to the lemma so the user doesn't specifically need to, since we preserve -- links in multiword lemmas and include links in non-lemma forms rather than allowing -- the entire form to be a link. linked_lemma = iut.add_links(base.user_specified_lemma) end base.linked_lemma = linked_lemma end) end --W: Lexical exceptions — yāʾ-middle roots whose Form IV (أفعل) keeps the yāʾ SOUND -- instead of undergoing the regular إعلال to a hollow verb (أَفْيَلَ, not أَفَالَ). -- These are attested verbs (not the صفة مشبهة like أَخْيَف/خَيْفاء); each is sourced below. -- Only Form IV is affected; forms VII/VIII/X of these roots remain hollow. -- Keyed by rad1..rad2..rad3. local SOUND_FORM_IV_ROOTS = { ["خيل"] = true, -- أَخْيَلَ: أخيلت السماءُ تهيأت للمطر — المعجم الوسيط ج1 ص266 (وبه أَخالت أيضًا) ["غيل"] = true, -- أَغْيَلَ: أغيلته فهي مُغْيِل، سقته الغَيْل — لسان العرب ج11 ص111 (وبه أغالت أيضًا) ["حيج"] = true, -- أَحْيَجَ: أحيجت الأرضُ، على خلاف القياس — تاج العروس ج5 ص500 ["حين"] = true, -- أَحْيَنَ: أحين القومُ آن لهم أن يبلغوا ما أملوه — المعجم الوسيط ج1 ص212 ["خيف"] = true, -- أَخْيَفَ: أخيف الحاجُّ نزلوا خيف منى — كتاب الأفعال لابن القوطية ص35 ["ريف"] = true, -- أَرْيَفَ: أريفت الأرضُ إريافًا (خصبت) — المحيط في اللغة ج10 ص256 ["زين"] = true, -- أَزْيَنَ: أزينت الأرضُ، حسّنها عشبها — مقاييس اللغة ج3 ص41 } --W: Lexical exceptions — hollow roots in the Form I فَعُلَ pattern (past~non-past u~u) -- that keep the middle radical SOUND instead of contracting to a long vowel (هَيُؤَ / -- يَهْيُؤُ, not هَاءَ / يَهُوءُ). This is not the general rule for u~u hollow verbs -- -- طَالَ / يَطُولُ (ط-و-ل) contracts as normal -- so each entry here is a specific -- attested exception, not a phonological class. -- Keyed by rad1..rad2..rad3. -- Keyed with bare hamza (ء) for rad3, matching normalize_root()'s canonicalization of root radicals -- (HAMZA_ON_ALIF etc. all fold to HAMZA) -- the raw root "هيأ" normalizes to radicals ه, ي, ء. local SOUND_FORM_I_UU_ROOTS = { ["هي" .. HAMZA] = true, -- هَيُؤَ ـُ: أنطوان الدحداح، معجم تصريف الأفعال العربية، مكتبة لبنان، ط2، 1995، ص203 } -- Determine weakness from radicals. Used when root given in place of lemma (e.g. for {{ar-verb forms}}). -- `past_vowel`/`nonpast_vowel`, if given, are the past/non-past vowel of a form-I verb (a string or -- footnote-bearing form object, per `req()`); they let SOUND_FORM_I_UU_ROOTS single out the specific -- vowel pattern under which a hollow root is attested as sound, without affecting its other patterns -- (e.g. a root could still contract normally under a different past~non-past pairing). local function weakness_from_radicals(form, rad1, rad2, rad3, rad4, past_vowel, nonpast_vowel) local weakness = nil local quadlit = form:find("q$") -- If weakness unspecified, derive from radicals. if not quadlit then if is_waw_ya(rad3) and rad1 == W and form == "I" then weakness = form_i_w_assimilated(past_vowel, nonpast_vowel) and "assimilated+final-weak" or "final-weak" elseif is_waw_ya(rad3) and vform_supports_final_weak(form) then weakness = "final-weak" elseif rad2 == rad3 and vform_supports_geminate(form) then weakness = "geminate" elseif is_waw_ya(rad2) and vform_supports_hollow(form) then --W: Both wāw (و) and yāʾ (ي) in the middle position create hollow verbs, in form I -- (قَالَ from ق-و-ل, بَاعَ from ب-ي-ع) and in the augmented forms IV, VII, VIII, X -- (أَتَاسَ from ت-ي-س, أَبَانَ from ب-ي-ن, أَطَاحَ from ط-ي-ح — معجم متن اللغة). if form == "IV" and rad2 == Y and SOUND_FORM_IV_ROOTS[rad1 .. rad2 .. rad3] then -- Lexical exceptions whose Form IV keeps the yāʾ sound (أَفْيَلَ); see SOUND_FORM_IV_ROOTS. weakness = "sound" elseif form == "I" and past_vowel and nonpast_vowel and req(past_vowel, U) and req(nonpast_vowel, U) and SOUND_FORM_I_UU_ROOTS[rad1 .. rad2 .. rad3] then -- Lexical exceptions whose Form I فَعُلَ keeps the middle radical sound (هَيُؤَ); see -- SOUND_FORM_I_UU_ROOTS. Gated to u~u specifically since that's the attested pattern. weakness = "sound" else weakness = "hollow" end elseif rad1 == W and form == "I" then -- Vowel-conditioned, like the lemma path below: a wāw-initial form-I verb only -- elides its wāw under a kasra (or lowered-kasra) non-past. Before this, the two -- inference paths in this module disagreed -- infer_radicals() kept the wāw for -- i~a and u~u while this one dropped it for every vowel pattern. weakness = form_i_w_assimilated(past_vowel, nonpast_vowel) and "assimilated" or "sound" else weakness = "sound" end else if is_waw_ya(rad4) then weakness = "final-weak" else weakness = "sound" end end return weakness end -- Join the infixed tāʔ (ت) to the first radical in form VIII verbs. This may cause assimilation of the tāʔ to the -- radical or in some cases the radical to the tāʔ. Used when a root is supplied instead of a lemma (which already has -- the appropriate assimilation in it). local function form_viii_join_ta(rad) if rad == W or rad == Y or rad == "ت" then return "تّ" elseif rad == "د" then return "دّ" elseif rad == "ث" then return "ثّ" elseif rad == "ذ" then return "ذّ" elseif rad == "ز" then return "زْد" elseif rad == "ص" then return "صْط" elseif rad == "ض" then return "ضْط" elseif rad == "ط" then return "طّ" elseif rad == "ظ" then return "ظّ" else return rad .. SK .. "ت" end end local function detect_indicator_spec(base) base.forms = {} base.stem_overrides = {} base.slot_overrides = {} if not base.conj_vowels[1] then -- These may be converted to inferred vowels. If not, we throw an error if form I and not passive-only. base.conj_vowels = {{ past = "-", nonpast = "-", }} else -- If multiple vowels specified for a given vowel type (e.g. a,u~u), expand so that each spec in local expansion = {} for _, spec in ipairs(base.conj_vowels) do for _, past in ipairs(spec.past) do for _, nonpast in ipairs(spec.nonpast) do table.insert(expansion, {past = past, nonpast = nonpast}) end end end base.conj_vowels = expansion end local vform = base.verb_form -- check for quadriliteral form (Iq, IIq, IIIq, IVq) base.quadlit = not not vform:find("q$") -- Infer radicals as necessary. We infer a separate set of radicals for each past~non-past vowel combination because -- they may be different (particularly with form-I hollow verbs). for _, vowel_spec in ipairs(base.conj_vowels) do -- NOTE: rad1, rad2, etc. refer to user-specified radicals, which are formobj tables that optionally specify an -- explicit manual translit, whereas ir1, ir2, etc. refer to inferred radicals, which are either strings or -- lists of possible radicals. local rads = base.root_consonants local rad1, rad2, rad3, rad4 = rads[1], rads[2], rads[3], rads[4] -- Default any unspecified radicals to radicals determined from the headword. The returned radicals may be -- lists of possible radicals, where the first radical should be chosen if the user didn't explicitly specify a -- radical but all are allowed. If `ambig = true` is set in the table, the radical is considered ambiguous and -- categories won't be created for weak radicals. local weakness, ir1, ir2, ir3, ir4 if vform ~= "none" then ir1, ir2, ir3 = rmatch(base.lemma, "^([^_])_([^_])_([^_])$") if not ir1 then ir1, ir2, ir3, ir4 = rmatch(base.lemma, "^([^_])_([^_])_([^_])_([^_])$") end if ir1 then -- root given instead of lemma weakness = weakness_from_radicals(vform, ir1, ir2, ir3, ir4, vowel_spec.past, vowel_spec.nonpast) if vform == "VIII" then vowel_spec.form_viii_assim = form_viii_join_ta(ir1) end else local ret = export.infer_radicals { headword = base.lemma, vform = vform, passive = base.passive, past_vowel = vowel_spec.past, nonpast_vowel = vowel_spec.nonpast, is_reduced = base.reduced, } weakness, ir1, ir2, ir3, ir4 = ret.weakness, ret.rad1, ret.rad2, ret.rad3, ret.rad4 vowel_spec.form_viii_assim = ret.form_viii_assim vowel_spec.past = ret.past_vowel vowel_spec.nonpast = ret.nonpast_vowel vowel_spec.variant = base.variant or ret.variant end end -- For most ambiguous radicals, the choice of radical doesn't matter because it doesn't affect the conjugation -- one way or another. For form I hollow verbs, however, it definitely does. In fact, the choice of radical is -- critical even beyond the past and non-past vowels because it affects the form of the passive participle. So, -- check for this and signal an error if the radical could not be inferred and is not given explicitly. if vform == "I" and type(ir2) == "table" and ir2.need_radical and not rad2 then error("Unable to guess middle radical of hollow form I verb; need to specify radical explicitly") end if vform == "I" and not is_passive_only(base.passive) and ( rget(vowel_spec.past) == "-" or rget(vowel_spec.nonpast) == "-") then error("Form I verb that isn't passive-only or final-weak must have past~non-past vowels specified") end -- Convert ambiguous radicals. local function regularize_inferred_radical(rad) if type(rad) == "table" then if rad.ambig then return {form = rad[1], ambig = true} else return rad[1] end else return rad end end -- Return the appropriate radical at index `index` (1 through 4), based either on the user-specified radical -- `user_radical` or (if unspecified) `inferred_radical`, inferred from the unvocalized lemma. Two values are -- returned, the "regularized" version of the radical (where ambiguous inferred radicals are converted to their -- most likely actual radical) and the non-regularized version. The returned values are form objects rather than -- strings. local function fetch_radical(user_radical, inferred_radical, index) if not user_radical then return regularize_inferred_radical(inferred_radical), inferred_radical else local rad_formval = rget(user_radical) if type(inferred_radical) == "table" then local allowed_radical_set = m_table.listToSet(inferred_radical) if not allowed_radical_set[rad_formval] then error(("For lemma %s, radical %s ambiguously inferred as %s but user radical incompatibly given as %s"): format(base.lemma, index, list_to_text(inferred_radical, nil, " or "), rad_formval)) end elseif rad_formval ~= inferred_radical then error(("For lemma %s, radical %s inferred as %s but user radical incompatibly given as %s"): format(base.lemma, index, inferred_radical, rad_formval)) end return user_radical, user_radical end end if vform ~= "none" then vowel_spec.rad1, vowel_spec.unreg_rad1 = fetch_radical(rad1, ir1, 1) vowel_spec.rad2, vowel_spec.unreg_rad2 = fetch_radical(rad2, ir2, 2) vowel_spec.rad3, vowel_spec.unreg_rad3 = fetch_radical(rad3, ir3, 3) if base.quadlit then vowel_spec.rad4, vowel_spec.unreg_rad4 = fetch_radical(rad4, ir4, 4) end end if vform == "I" then -- If explicit weakness given using 'I-sound' or 'I-assimilated', we may need to adjust the inferred weakness. if base.explicit_weakness == "sound" then if weakness == "assimilated" then weakness = "sound" elseif weakness == "assimilated+final-weak" then -- Verbs like waniya~yawnā "to be faint; to languish" (although the defaults should handle this -- correctly) weakness = "final-weak" else error(("Can't specify form 'I-sound' when inferred weakness is '%s' for lemma %s"):format( weakness, base.lemma)) end elseif base.explicit_weakness == "assimilated" then if weakness == "sound" then -- i~a verbs like waṭiʔa~yaṭaʔu "to tread, to trample"; wasiʕa~yasaʕu "to be spacious; to be well-off"; -- waṯiʔa~yaṯaʔu "to get bruised, to be sprained", which would default to sound. weakness = "assimilated" elseif weakness == "final-weak" then -- For completeness; not clear if any verbs occur where this is needed. (There are plenty of -- assimilated+final-weak verbs but the defaults should take care of them.) weakness = "assimilated+final-weak" else error(("Can't specify form 'I-assimilated' when inferred weakness is '%s' for lemma %s"):format( weakness, base.lemma)) end elseif base.explicit_weakness then error(("Internal error: Unrecognized value '%s' for base.explicit_weakness"):format(base.explicit_weakness)) end elseif vform == "none" then weakness = base.explicit_weakness elseif base.explicit_weakness then error(("Internal error: Explicit weakness should not be specifiable except with forms I and none, but saw explicit weakness '%s' with verb form '%s'"): format(base.explicit_weakness, vform)) end vowel_spec.weakness = weakness if vform ~= "none" then -- Error if radicals are wrong given the weakness. More likely to happen if the weakness is explicitly given -- rather than inferred. Will also happen if certain incorrect letters are included as radicals e.g. hamza on -- top of various letters, alif maqṣūra, tā' marbūṭa. check_radicals(vform, weakness, rget(vowel_spec.rad1), rget(vowel_spec.rad2), rget(vowel_spec.rad3), base.quadlit and rget(vowel_spec.rad4) or nil) end -- Check the variant value. local form_iii_vi_geminate = (vform == "III" or vform == "VI") and rget(vowel_spec.rad2) == rget(vowel_spec.rad3) and not req(vowel_spec.rad2, Y) local hayy_i_x = hayy_radicals(vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3) and (vform == "I" or vform == "X") if form_iii_vi_geminate or hayy_i_x then if vowel_spec.variant and vowel_spec.variant ~= "long" and vowel_spec.variant ~= "short" and vowel_spec.variant ~= "both" then error(("For form-III/VI geminate verb or form-I/X verb with ح-ي-ي radicals, saw unrecognized 'var:%s' value; should be 'var:long', 'var:short' or 'var:both'"):format( vowel_spec.variant)) end elseif vowel_spec.variant then error(("Variant value 'var:%s' not allowed in this context"):format(vowel_spec.variant)) end end -- If form I, regroup expanded vowels for display purposes. if vform == "I" then local group_by_past = {} for _, vowel_spec in ipairs(base.conj_vowels) do m_table.insertIfNot(group_by_past, { past = undia[rget(vowel_spec.past)], nonpasts = {undia[rget(vowel_spec.nonpast)]}, }, { key = function(obj) return obj.past end, combine = function(obj1, obj2) for _, nonpast in ipairs(obj2.nonpasts) do m_table.insertIfNot(obj1.nonpasts, nonpast) end end, }) end local group_by_nonpast = {} for _, vowel_spec in ipairs(group_by_past) do m_table.insertIfNot(group_by_nonpast, { pasts = {vowel_spec.past}, nonpasts = vowel_spec.nonpasts, }, { key = function(obj) return obj.nonpasts end, combine = function(obj1, obj2) for _, past in ipairs(obj2.pasts) do m_table.insertIfNot(obj1.pasts, past) end end, }) end base.grouped_conj_vowels = group_by_nonpast end -- Set value of passive. If not specified, default is yes for forms II, III, IV and Iq; no but uncertainly for -- forms VII, IX, XI - XV and IIIq - IVq, as well as form I with past vowel u; impersonal but uncertainly for form -- V, VI, X and IIq, as well as form I with past vowel i; and yes but uncertainly for the remainder (form I with -- past vowel only a and form VIII). if not base.passive then base.passive_defaulted = true -- Temporary tracking for defaulted passives by verb form, weakness and (for form I) past/non-past vowels. track_if_ar_conj(base, "passive-defaulted/" .. vform) for _, vowel_spec in ipairs(base.conj_vowels) do track_if_ar_conj(base, "passive-defaulted/" .. vform.. "/" .. vowel_spec.weakness) if vform == "I" then --W: use rget(); the vowel is a form object when the user attached a footnote -- (e.g. <I/a[نادر]~u>), and undia[<table>] is nil, which crashes format() on Lua 5.1. local past_nonpast = ("%s~%s"):format(undia[rget(vowel_spec.past)], undia[rget(vowel_spec.nonpast)]) track_if_ar_conj(base, "passive-defaulted/I/" .. past_nonpast) track_if_ar_conj(base, "passive-defaulted/I/" .. vowel_spec.weakness .. "/" .. past_nonpast) end end if vform_probably_full_passive(vform) then base.passive = "pass" else base.passive_uncertain = true for _, vowel_spec in ipairs(base.conj_vowels) do if vform_probably_no_passive(vform, vowel_spec.weakness, vowel_spec.past, vowel_spec.nonpast) then base.passive = "nopass" break elseif vform_probably_impersonal_passive(vform, vowel_spec.weakness, vowel_spec.past, vowel_spec.nonpast) then base.passive = "ipass" break end end base.passive = base.passive or "pass" end end -- NOTE: Currently there are no built-in stems or form overrides for Arabic; this code is inherited from -- [[Module:ca-verb]], where such things do exist, and is kept for generality in case we decide in the future to -- implement such things. -- Override built-in verb stems and overrides with user-specified ones. for stem, values in pairs(base.user_stem_overrides) do base.stem_overrides[stem] = values end for slot, values in pairs(base.user_slot_overrides) do if not base.alternant_multiword_spec.verb_slots_map[slot] then error("Unrecognized override slot '" .. slot .. "': " .. base.angle_bracket_spec) end if export.unsettable_slots_set[slot] then error("Slot '" .. slot .. "' cannot be set using an override: " .. base.angle_bracket_spec) end if skip_slot(base, slot, "allow overrides") then error("Override slot '" .. slot .. "' would be skipped based on the passive, 'noimp' and/or 'no_nonpast' settings: " .. base.angle_bracket_spec) end base.slot_overrides[slot] = values end if base.verb_form == "none-final-weak" then for _, stem_type in ipairs { "past", "past_pass", "nonpast", "nonpast_pass" } do if base.stem_overrides[stem_type .. "_c"] or base.stem_overrides[stem_type .. "_v"] then error(("Specify past stem for verb type 'none-final-weak' using '%s:...' not '%s_c:...' or '%s_v:...'"): format(stem_type, stem_type, stem_type)) end end for _, stem_type in ipairs { "past", "nonpast" } do if base.stem_overrides[stem_type] or not base.stem_overrides[stem_type .. "_final_weak_vowel"] then error(("For verb type 'none-final-weak', if '%s:...' specified, so must '%s_final_weak_vowel:...'"): format(stem_type, stem_type)) end end end end local function detect_all_indicator_specs(alternant_multiword_spec) add_slots(alternant_multiword_spec) alternant_multiword_spec.verb_forms = {} -- This means at least one individual base had the slot marked as explicitly missing. Another base (e.g. when -- there are multiple alternants) might have a value for the slot. In practice, we only respect this when there are -- no overall values in the slot and `slot_uncertain` isn't set; in this case, we display "no ..." for the slot -- instead of simply not displaying anything for the slot. alternant_multiword_spec.slot_explicitly_missing = {} -- This means at least one individual base had no values for the slot and the slot marked as explicitly uncertain. -- Note that this is different from a value being present but marked as uncertain (e.g. if an override was given -- with a ? after it); this causes the form object for the value to have `uncertain = true` set. If there are no -- overall values in the slot and `slot_uncertain` is set, we display this in the headword. alternant_multiword_spec.slot_uncertain = {} iut.map_word_specs(alternant_multiword_spec, function(base) -- So arguments, etc. can be accessed. WARNING: Creates circular reference. base.alternant_multiword_spec = alternant_multiword_spec detect_indicator_spec(base) if not base.nocat then m_table.insertIfNot(alternant_multiword_spec.verb_forms, base.verb_form) end if base.passive then alternant_multiword_spec.passive = base.passive end if base.passive_uncertain then alternant_multiword_spec.passive_uncertain = true end for slot, _ in pairs(base.slot_explicitly_missing) do alternant_multiword_spec.slot_explicitly_missing[slot] = true end end) end local function determine_slot_uncertainty_from_forms(alternant_multiword_spec) iut.map_word_specs(alternant_multiword_spec, function(base) -- If no verbal noun and verb form is not 'none' (manually-specified stems) — which currently only happens for -- form I — and the verbal noun wasn't explicitly indicated as missing using <vn:->, we assume it's just -- unknown/unspecified rather than missing. Same with active participles. for uncertain_slot, _ in pairs(slots_that_may_be_uncertain) do if not base.forms[uncertain_slot] and vform ~= "none" and not skip_slot(base, uncertain_slot) then base.slot_uncertain[uncertain_slot] = true end end -- Propagate slot uncertainty up. Currently only the verbal noun can have this set but we write the code -- generally. for slot, _ in pairs(base.slot_uncertain) do alternant_multiword_spec.slot_uncertain[slot] = true end end) -- If slot is uncertain and has no value, explicitly set its value to "?". for uncertain_slot, _ in pairs(slots_that_may_be_uncertain) do if not alternant_multiword_spec.forms[uncertain_slot] and alternant_multiword_spec.slot_uncertain[uncertain_slot] then alternant_multiword_spec.forms[uncertain_slot] = {{form = "?"}} end end end -- Determine certain properties of the verb from the overall forms, such as whether the verb is active-only or -- passive-only, is impersonal, lacks an imperative, etc. local function determine_verb_properties_from_forms(alternant_multiword_spec) alternant_multiword_spec.has_active = false alternant_multiword_spec.has_passive = false alternant_multiword_spec.has_non_impers_active = false alternant_multiword_spec.has_non_impers_passive = false alternant_multiword_spec.has_imp = false alternant_multiword_spec.has_past = false alternant_multiword_spec.has_nonpast = false for slot, _ in pairs(alternant_multiword_spec.forms) do if slot == "ap" or slot:find("[123]") and not slot:find("_pass") then alternant_multiword_spec.has_active = true end if slot == "pp" or slot:find("[123]") and slot:find("_pass") then alternant_multiword_spec.has_passive = true end if slot:find("[123]") and not slot:find("pass_[123]") and not slot:find("3ms") then alternant_multiword_spec.has_non_impers_active = true end if slot:find("pass_[123]") and not slot:find("3ms") then alternant_multiword_spec.has_non_impers_passive = true end if slot:find("^imp_") then alternant_multiword_spec.has_imp = true end if slot:find("^past_") then alternant_multiword_spec.has_past = true end if slot:find("^ind_") or slot:find("^sub_") or slot:find("^juss_") then alternant_multiword_spec.has_nonpast = true end end end local function add_categories_and_annotation(alternant_multiword_spec, base, multiword_lemma, insert_ann, insert_cat) -- Useful e.g. in constructing suppletive verbs out of parts. For a verb like جاء or أتى whose imperative comes -- from the unrelated verb تعالى, we don't want the latter verb showing up in categories or annotations. if base.nocat then return end local vform = base.verb_form if vform ~= "none" then insert_ann("form", vform) insert_cat("form-" .. vform .. " verbs") --W: Add morphological pattern corresponding to the form local morph_pattern = FORM_TO_MORPHOLOGICAL_PATTERN[vform] if morph_pattern then insert_ann("morph_pattern", morph_pattern) end end if base.reduced then insert_ann("reduced", "reduced") if vform ~= "none" then insert_cat("form-" .. vform .. " reduced verbs") end end if base.quadlit then insert_cat("verbs with quadriliteral roots") end if base.passive_defaulted then insert_cat("verbs with defaulted passive") end for _, vowel_spec in ipairs(base.conj_vowels) do local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local weakness = vowel_spec.weakness -- We have to distinguish weakness by form and weakness by conjugation. Weakness by form merely indicates the -- presence of weak letters in certain positions in the radicals. Weakness by conjugation is related to how the -- verbs are conjugated. For example, form-II verbs that are "hollow by form" (middle radical is wāw or yāʾ) are -- conjugated as sound verbs. Another example: form-I verbs with initial wāw are "assimilated by form" and most -- are assimilated by conjugation as well, but a few are sound by conjugation, e.g. wajuha yawjuhu "to be -- distinguished" (rather than wajuha yajuhu); similarly for some hollow-by-form verbs in various forms, e.g. -- form VIII izdawaja yazdawiju "to be in pairs" (rather than izdāja yazdāju). Categories referring to weakness -- always refer to weakness by conjugation; weakness by form is distinguished only by categories such as -- [[:Category:Arabic form-III verbs with و as second radical]]. insert_ann("weakness", weakness) if vform ~= "none" then insert_cat(("%s form-%s verbs"):format(weakness, vform)) end local function radical_is_ambiguous(rad) return type(rad) == "table" and rad.ambig end local function radical_is_unambiguous_weak(rad) return not radical_is_ambiguous(rad) and (is_waw_ya(rad) or req(rad, HAMZA)) end if vform ~= "none" then local ur1, ur2, ur3, ur4 = vowel_spec.unreg_rad1, vowel_spec.unreg_rad2, vowel_spec.unreg_rad3, vowel_spec.unreg_rad4 -- Create headword categories based on the radicals. Do the following before -- converting the Latin radicals into Arabic ones so we distinguish -- between ambiguous and non-ambiguous radicals. if radical_is_ambiguous(ur1) or radical_is_ambiguous(ur2) or radical_is_ambiguous(ur3) or ur4 and radical_is_ambiguous(ur4) then insert_cat("verbs with ambiguous radicals") end if radical_is_unambiguous_weak(ur1) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur1) .. " as first radical") end if radical_is_unambiguous_weak(ur2) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur2) .. " as second radical") end if radical_is_unambiguous_weak(ur3) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur3) .. " as third radical") end if ur4 and radical_is_unambiguous_weak(ur4) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur4) .. " as fourth radical") end end end if vform == "I" and not is_passive_only(base.passive) then for _, vowel_spec in ipairs(base.grouped_conj_vowels) do insert_ann("vowels", ("%s ~ %s"):format(table.concat(vowel_spec.pasts, "/"), table.concat(vowel_spec.nonpasts, "/"))) for _, past in ipairs(vowel_spec.pasts) do for _, nonpast in ipairs(vowel_spec.nonpasts) do if past == "-" or nonpast == "-" then error("Internal error: Saw form I past vowel %s and non-past vowel %s but - in place of vowel should have triggered an error earlier") end insert_cat(("form-I verbs with past vowel %s and non-past vowel %s"):format(past, nonpast)) --W: Add morphological pattern for form I based on vowels local form_i_with_vowels = ("I/%s~%s"):format(past, nonpast) local morph_pattern = FORM_TO_MORPHOLOGICAL_PATTERN[form_i_with_vowels] if morph_pattern then insert_ann("morph_pattern", morph_pattern) end end end end end for slot, name in pairs(slots_that_may_be_uncertain) do if base.slot_uncertain[slot] then -- An unspecified and non-defaulted verbal noun (form I) is considered uncertain rather than explicitly -- missing. Use <vn:-> to explicitly indicate the lack of verbal noun. Same for form-I stative active -- participles. insert_cat(("verbs with unknown or uncertain %ss"):format(name)) end end if base.irregular then insert_ann("irreg", "irregular") insert_cat("irregular verbs") end end -- Compute the categories to add the verb to, as well as the annotation to display in the conjugation title bar. We -- combine the code to do these functions as both categories and title bar contain similar information. local function compute_categories_and_annotation(alternant_multiword_spec) alternant_multiword_spec.categories = {} local ann = {} alternant_multiword_spec.annotation = ann ann.form = {} ann.morph_pattern = {} --W: arabic morphological pattern corresponding to the form ann.weakness = {} ann.vowels = {} ann.passive = nil ann.reduced = {} ann.irreg = {} ann.defective = {} local multiword_lemma = false for _, slot in ipairs(export.potential_lemma_slots) do if alternant_multiword_spec.forms[slot] then for _, formobj in ipairs(alternant_multiword_spec.forms[slot]) do if formobj.form:find(" ") then multiword_lemma = true break end end break end end local function insert_ann(anntype, value) m_table.insertIfNot(alternant_multiword_spec.annotation[anntype], value) end local function insert_cat(cat, also_when_multiword) -- Don't place multiword terms in categories like 'Arabic form-II verbs' to avoid spamming the categories with -- such terms. if also_when_multiword or not multiword_lemma then m_table.insertIfNot(alternant_multiword_spec.categories, "Arabic " .. cat) end end iut.map_word_specs(alternant_multiword_spec, function(base) add_categories_and_annotation(alternant_multiword_spec, base, multiword_lemma, insert_ann, insert_cat) end) for slot, name in pairs(slots_that_may_be_uncertain) do if alternant_multiword_spec.forms[slot] then for _, form in ipairs(alternant_multiword_spec.forms[slot]) do if form.uncertain then if form.form == "?" then insert_cat(("verbs with explicitly unknown %ss"):format(name)) else insert_cat(("verbs needing %s checked"):format(name)) end break end end end end if alternant_multiword_spec.has_active then if alternant_multiword_spec.has_passive and alternant_multiword_spec.has_non_impers_passive then insert_cat("verbs with full passive") ann.passive = "full passive" elseif alternant_multiword_spec.has_passive then insert_cat("verbs with impersonal passive") ann.passive = "impersonal passive" else insert_cat("verbs lacking passive forms") ann.passive = "no passive" end else if alternant_multiword_spec.has_non_impers_passive then insert_cat("passive verbs") insert_cat("verbs with full passive") ann.passive = "passive-only" else insert_cat("passive verbs") insert_cat("impersonal verbs") insert_cat("verbs with impersonal passive") ann.passive = "impersonal (passive-only)" end end if alternant_multiword_spec.passive_uncertain then insert_cat("verbs needing passive checked") ann.passive = ann.passive .. ' <abbr title="passive status uncertain">(?)</abbr>' end if alternant_multiword_spec.has_active and not alternant_multiword_spec.has_imp then insert_ann("defective", "no imperative") insert_cat("verbs lacking imperative forms") end if not alternant_multiword_spec.has_past then insert_ann("defective", "no past") insert_cat("verbs lacking past forms") end if not alternant_multiword_spec.has_nonpast then insert_ann("defective", "no non-past") insert_cat("verbs lacking non-past forms") end local ann_parts = {} local function insert_ann_part(part, conj) local val = table.concat(ann[part], conj or " or ") if val ~= "" and val ~= "regular" then table.insert(ann_parts, val) end end insert_ann_part("form") insert_ann_part("weakness") insert_ann_part("reduced") insert_ann_part("vowels") if ann.passive then table.insert(ann_parts, ann.passive) end insert_ann_part("irreg") insert_ann_part("defective", "، ") --W: Save morph_pattern before overwriting annotation with string alternant_multiword_spec.morph_pattern = ann.morph_pattern alternant_multiword_spec.annotation = table.concat(ann_parts, "، ") end local function show_forms(alternant_multiword_spec) local lemmas = {} for _, slot in ipairs(export.potential_lemma_slots) do if alternant_multiword_spec.forms[slot] then for _, formobj in ipairs(alternant_multiword_spec.forms[slot]) do table.insert(lemmas, formobj) end break end end alternant_multiword_spec.lemmas = lemmas -- save for later use in make_table() alternant_multiword_spec.vn = alternant_multiword_spec.forms.vn -- save for later use in make_table() -- Reconstruct the original verb spec without overrides for verbal nouns and participles, since those specific slots -- are ignored by {{ar-verb form}}. Compute this once beforehand; `transform_accel_obj` is called repeatedly on each -- form and we don't want to compute this repeatedly. local reconstructed_verb_spec = iut.reconstruct_original_spec(alternant_multiword_spec, { preprocess_angle_bracket_spec = function(spec) spec = spec:match("^<(.*)>$") assert(spec) local segments = put.parse_multi_delimiter_balanced_segment_run(spec, {{"[", "]"}, {"<", ">"}}) local dot_separated_groups = put.split_alternating_runs_and_strip_spaces(segments, "%.") -- Rejoin each dot-separated group into a single string, since we aren't actually going to do any parsing -- of bracket-bounded textual runs; then filter out overrides for verbal nouns and participles. local filtered_indicators = {} for _, dot_separated_group in ipairs(dot_separated_groups) do local indicator = table.concat(dot_separated_group) -- FIXME: Do we want to filter out any other indicators? if not (indicator:find("^vn:") or indicator:find("^[ap]p:")) then table.insert(filtered_indicators, indicator) end end return ("<%s>"):format(table.concat(filtered_indicators, ".")) end, }) -- If we're dealing with a single word, no alternants and a single verb form, use the auto-conjugation-fetching -- variant. local reconstructed_lemma, inside = reconstructed_verb_spec:match("^([^ <>()]+)(%b<>)$") if inside and alternant_multiword_spec.verb_forms[1] and not alternant_multiword_spec.verb_forms[2] then reconstructed_verb_spec = ("+%s<%s>"):format(reconstructed_lemma, alternant_multiword_spec.verb_forms[1]) end local function transform_accel_obj(slot, formobj, accel_obj) if not accel_obj then return accel_obj end if slot == "ap" or slot == "pp" or slot == "vn" then -- FIXME: [[Module:accel]] can't correctly handle more than one verb form for participles and verbal nouns accel_obj.form = slot .. "-" .. table.concat(alternant_multiword_spec.verb_forms, "،") else accel_obj.form = "verb-form-" .. reconstructed_verb_spec end return accel_obj end local function generate_link(data) local form = data.form local term = form.formval_for_link local alt = form.alt if term == "?" then term = nil alt = "?" end --W: Create simple link for verb forms without #Arabic anchor local display = alt or term local link if display then link = link_term_simple(display) else link = "" end link = link .. iut.get_footnote_text(form.footnotes, data.footnote_obj) --W: qualifier/label rendering removed along with the q/qq/l/ll modifiers; -- it required [[Module:pron qualifier]], absent on ar.wiktionary. return link end local props = { lang = lang, lemmas = lemmas, transform_accel_obj = transform_accel_obj, generate_link = generate_link, slot_list = alternant_multiword_spec.verb_slots, include_translit = false, } iut.show_forms(alternant_multiword_spec.forms, props) end --W: Tatweel and the invisible direction/joining marks survive copy-paste and -- MediaWiki keeps them in page titles, so strip them from both sides before -- comparing a |بؤرة= value against a generated form. local FOCUS_IGNORE = "[" .. u(0x0640) .. u(0x061C) .. u(0x200B) .. "-" .. u(0x200F) .. u(0x202A) .. "-" .. u(0x202E) .. u(0xFEFF) .. "]" local function normalize_for_focus(s) return (rsub(s, FOCUS_IGNORE, "")) end --W: Which slots hold `focus`? Must run before show_forms() turns the form -- lists into rendered wikitext. local function find_focus_slots(alternant_multiword_spec, focus) if not focus or focus == "" then return nil end local target = normalize_for_focus(focus) local slots = {} for slot, forms in pairs(alternant_multiword_spec.forms) do if type(forms) == "table" then for _, formobj in ipairs(forms) do if type(formobj) == "table" and formobj.form and normalize_for_focus(formobj.form) == target then slots[slot] = true break end end end end return slots end --W: 'yes' unless |موسع= asked for the table open. [[Template:inflection-table-top]] -- turns any non-empty |tall= into the mw-collapsed class. local function collapsed_arg(alternant_multiword_spec) if alternant_multiword_spec.expanded then return nil end return 'yes' end --W: Put the marker class on the cell itself rather than on a span around its -- contents, so the highlight covers the whole cell instead of hugging the word. -- Runs on the table wikitext below, before m_string_utilities.format() fills the -- {slot} placeholders in. Every form cell in either table has one of two shapes: -- -- | {past_1s} -> | class="avf-focus" | {past_1s} -- | rowspan="2" | {past_1s} -> | rowspan="2" class="avf-focus" | {past_1s} -- -- Styled by [[Template:ar-verb-form/style.css]] (.inflection-table td.avf-focus). local function mark_focus_cells(text, focus_slots) if not focus_slots then return text end return (rsub(text, "\n(|[^\n]*)", function(cell) local slot = rmatch(cell, "^|%s*{([%w_]+)}%s*$") if slot then if focus_slots[slot] then return '\n| class="avf-focus" | {' .. slot .. "}" end return nil end local attrs attrs, slot = rmatch(cell, "^|%s*(.-)%s*|%s*{([%w_]+)}%s*$") if slot and focus_slots[slot] then return "\n| " .. attrs .. ' class="avf-focus" | {' .. slot .. "}" end return nil end)) end ------------------------------------------------------------------------------- -- Functions to create inflection tables -- ------------------------------------------------------------------------------- -- Make the conjugation table. Called from export.show(). local function make_table(alternant_multiword_spec) local text = mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-top', args = { title = 'تصريف {title}', --W: |tall= is what puts mw-collapsed on the wrapper. |موسع= drops it. -- (Not `expanded and nil or 'yes'` — in Lua that is always 'yes'.) tall = collapsed_arg(alternant_multiword_spec), palette = 'green', category = 'conjugation', class = 'tr-alongside', -- temp hack to prevent extra line break } } text = text .. [=[ ! colspan="6" | <<الْمَصْدَر>> | colspan="7" | {vn} ]=] if alternant_multiword_spec.has_active then text = text .. [=[ |- ! colspan="6" | <<اِسْم الْفَاعِل>> | colspan="7" | {ap} ]=] end if alternant_multiword_spec.has_passive then text = text .. [=[ |- ! colspan="6" | <<اِسْم الْمَفْعُول>> | colspan="7" | {pp} ]=] end text = text .. [=[ |- ! colspan="999" class="separator" | ]=] if alternant_multiword_spec.has_active then text = text .. [=[ |- ! colspan="12" class="outer" | <<الْفِعْل الْمَعْلُوم>> |- ! colspan="2" | ! colspan="3" | <<الْمُفْرَد>> ! rowspan="12" class="separator" | ! colspan="2" | <<الْمُثَنَّى>> ! rowspan="12" class="separator" | ! colspan="3"| <<الْجَمْع>> |- ! colspan="2"| ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> |- ! rowspan="2" | <<الْمَاضِي>> ! class="secondary" | المذكر | rowspan="2" | {past_1s} | {past_2ms} | {past_3ms} | rowspan="2" | {past_2d} | {past_3md} | rowspan="2" | {past_1p} | {past_2mp} | {past_3mp} |- ! class="secondary" | المؤنث | {past_2fs} | {past_3fs} | {past_3fd} | {past_2fp} | {past_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَرْفُوع>> ! class="secondary" | المذكر | rowspan="2" | {ind_1s} | {ind_2ms} | {ind_3ms} | rowspan="2" | {ind_2d} | {ind_3md} | rowspan="2" | {ind_1p} | {ind_2mp} | {ind_3mp} |- ! class="secondary" | المؤنث | {ind_2fs} | {ind_3fs} | {ind_3fd} | {ind_2fp} | {ind_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَنْصُوب>> ! class="secondary" | المذكر | rowspan="2" | {sub_1s} | {sub_2ms} | {sub_3ms} | rowspan="2" | {sub_2d} | {sub_3md} | rowspan="2" | {sub_1p} | {sub_2mp} | {sub_3mp} |- ! class="secondary" | المؤنث | {sub_2fs} | {sub_3fs} | {sub_3fd} | {sub_2fp} | {sub_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَجْزُوم>> ! class="secondary" | المذكر | rowspan="2" | {juss_1s} | {juss_2ms} | {juss_3ms} | rowspan="2" | {juss_2d} | {juss_3md} | rowspan="2" | {juss_1p} | {juss_2mp} | {juss_3mp} |- ! class="secondary" | المؤنث | {juss_2fs} | {juss_3fs} | {juss_3fd} | {juss_2fp} | {juss_3fp} |- ! rowspan="2" | <<الْأَمْر>> ! class="secondary" | المذكر | rowspan="2" | | {imp_2ms} | rowspan="2" | | rowspan="2" | {imp_2d} | rowspan="2" | | rowspan="2" | | {imp_2mp} | rowspan="2" | |- ! class="secondary" | المؤنث | {imp_2fs} | {imp_2fp} ]=] end if alternant_multiword_spec.has_passive then text = text .. [=[ |- ! colspan="999" class="separator" | |- ! colspan="12" class="outer" | <<الْفِعْل الْمَجْهُول>> |- ! colspan="2" | ! colspan="3" | <<الْمُفْرَد>> ! rowspan="10" class="separator" | ! colspan="2" | <<الْمُثَنَّى>> ! rowspan="10" class="separator" | ! colspan="3" | <<الْجَمْع>> |- ! colspan="2" | ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> |- ! rowspan="2" | <<الْمَاضِي>> ! class="secondary" | المذكر | rowspan="2" | {past_pass_1s} | {past_pass_2ms} | {past_pass_3ms} | rowspan="2" | {past_pass_2d} | {past_pass_3md} | rowspan="2" | {past_pass_1p} | {past_pass_2mp} | {past_pass_3mp} |- ! class="secondary" | المؤنث | {past_pass_2fs} | {past_pass_3fs} | {past_pass_3fd} | {past_pass_2fp} | {past_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَرْفُوع>> ! class="secondary" | المذكر | rowspan="2" | {ind_pass_1s} | {ind_pass_2ms} | {ind_pass_3ms} | rowspan="2" | {ind_pass_2d} | {ind_pass_3md} | rowspan="2" | {ind_pass_1p} | {ind_pass_2mp} | {ind_pass_3mp} |- ! class="secondary" | المؤنث | {ind_pass_2fs} | {ind_pass_3fs} | {ind_pass_3fd} | {ind_pass_2fp} | {ind_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَنْصُوب>> ! class="secondary" | المذكر | rowspan="2" | {sub_pass_1s} | {sub_pass_2ms} | {sub_pass_3ms} | rowspan="2" | {sub_pass_2d} | {sub_pass_3md} | rowspan="2" | {sub_pass_1p} | {sub_pass_2mp} | {sub_pass_3mp} |- ! class="secondary" | المؤنث | {sub_pass_2fs} | {sub_pass_3fs} | {sub_pass_3fd} | {sub_pass_2fp} | {sub_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَجْزُوم>> ! class="secondary" | المذكر | rowspan="2" | {juss_pass_1s} | {juss_pass_2ms} | {juss_pass_3ms} | rowspan="2" | {juss_pass_2d} | {juss_pass_3md} | rowspan="2" | {juss_pass_1p} | {juss_pass_2mp} | {juss_pass_3mp} |- ! class="secondary" | المؤنث | {juss_pass_2fs} | {juss_pass_3fs} | {juss_pass_3fd} | {juss_pass_2fp} | {juss_pass_3fp} ]=] end text = text .. mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-bottom', args = { notes = '{footnote}', } } local forms = alternant_multiword_spec.forms if not alternant_multiword_spec.lemmas then forms.title = "—" else local linked_lemmas = {} for _, form in ipairs(alternant_multiword_spec.lemmas) do --W: Create simple link for lemmas without #Arabic anchor -- table.insert(linked_lemmas, link_term(form.form, "term")) table.insert(linked_lemmas, link_term_simple(form.form)) end forms.title = table.concat(linked_lemmas, "، ") end local ann_parts = {} if alternant_multiword_spec.annotation ~= "" then table.insert(ann_parts, alternant_multiword_spec.annotation) end if alternant_multiword_spec.vn then local linked_vns = {} for _, form in ipairs(alternant_multiword_spec.vn) do table.insert(linked_vns, link_term(form.form, "term")) end table.insert(ann_parts, (#linked_vns > 1 and "verbal nouns" or "verbal noun") .. " " .. table.concat(linked_vns, "، ")) end local annotation = table.concat(ann_parts, "، ") --W: use only arabic morphological pattern in title annotation local morph_patterns = alternant_multiword_spec.morph_pattern if morph_patterns and #morph_patterns > 0 then annotation = table.concat(morph_patterns, "، ") forms.title = forms.title .. " (" .. annotation .. ")" end -- Format the table. text = mark_focus_cells(text, alternant_multiword_spec.focus_slots) local tagged_table = rsub(text, "<<(.-)>>", tag_text) return m_string_utilities.format(tagged_table, forms) end -- Helper function to get root letters and verb classification for simplified table header -- Returns: root_display (e.g., "ح س ب"), verb_type (e.g., "فعل ثلاثي صحيح سالم") local function get_verb_info(alternant_multiword_spec) local root_display = "" local verb_type = "فعل" local rad1, rad2, rad3, rad4 -- Extract radicals from the base word spec using iut.map_word_specs -- This accesses the radicals directly from the parsed verb specification iut.map_word_specs(alternant_multiword_spec, function(base) -- Get radicals from the first conjugation vowel spec if base.conj_vowels and #base.conj_vowels > 0 then local vowel_spec = base.conj_vowels[1] -- Extract radicals, handling both table and string types local function extract_radical(rad) if type(rad) == "table" then -- If it's a table, get the first element or the string representation return rad[1] or rget(rad) else return rad end end rad1 = extract_radical(vowel_spec.rad1) rad2 = extract_radical(vowel_spec.rad2) rad3 = extract_radical(vowel_spec.rad3) rad4 = vowel_spec.rad4 and extract_radical(vowel_spec.rad4) end end) -- Get basic/derived classification from morphological pattern local basic_deriv if alternant_multiword_spec.morph_pattern and #alternant_multiword_spec.morph_pattern > 0 then local morph_pattern = alternant_multiword_spec.morph_pattern[1] basic_deriv = MORPHOLOGICAL_PATTERN_TO_BASIC_DERIV[morph_pattern] end -- Format root display with spaces between letters if rad1 and rad2 and rad3 then if rad4 then -- Quadrilateral root_display = rad1 .. " " .. rad2 .. " " .. rad3 .. " " .. rad4 verb_type = "فعل رباعي" if basic_deriv then verb_type = verb_type .. " " .. basic_deriv end else -- triliteral root_display = rad1 .. " " .. rad2 .. " " .. rad3 verb_type = "فعل ثلاثي" if basic_deriv then verb_type = verb_type .. " " .. basic_deriv if basic_deriv == "مُجرَّد" then --W: the classification concerns only triliteral basic verbs (الثلاثي المُجرَّد) local classification = export.classify_triliteral_verb(rad1, rad2, rad3) verb_type = verb_type .. " " .. classification end end end end return root_display, verb_type end -- Make simplified conjugation table for ar.wiktionary. Called from export.show(). -- This table shows only past, present indicative, and imperative forms. -- No passive voice, no participles, no verbal nouns. local function make_table_ar(alternant_multiword_spec) local text = mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-top', args = { title = 'تصريف {title}', --W: |tall= is what puts mw-collapsed on the wrapper. |موسع= drops it. -- (Not `expanded and nil or 'yes'` — in Lua that is always 'yes'.) tall = collapsed_arg(alternant_multiword_spec), palette = 'green', category = 'conjugation', class = 'tr-alongside', -- temp hack to prevent extra line break } } -- Get verb root and classification local root_display, verb_type = get_verb_info(alternant_multiword_spec) -- First row: Root and verb type spanning all columns text = text .. "! colspan=\"2\" | <<الجذر>>: " .. root_display .. "\n" text = text .. "| colspan=\"3\" | <<'''" .. verb_type .. "'''>>\n" text = text .. [=[ |- ! colspan="2" | <<الضمائر>> ! <<الماضي>> ! <<المضارع>> ! <<الأمر>> |- ! rowspan="2" | <<المتكلّم>> | <<أنا>> | {past_1s} | {ind_1s} | |- | <<نحن>> | {past_1p} | {ind_1p} | |- ! rowspan="5" | <<المخاطب>> | <<أنتَ>> | {past_2ms} | {ind_2ms} | {imp_2ms} |- | <<أنتِ>> | {past_2fs} | {ind_2fs} | {imp_2fs} |- | <<أنتما>> | {past_2d} | {ind_2d} | {imp_2d} |- | <<أنتم>> | {past_2mp} | {ind_2mp} | {imp_2mp} |- | <<أنتنّ>> | {past_2fp} | {ind_2fp} | {imp_2fp} |- ! rowspan="6" | <<الغائب>> | <<هو>> | {past_3ms} | {ind_3ms} | |- | <<هي>> | {past_3fs} | {ind_3fs} | |- | <<هما>> | {past_3md} | {ind_3md} | |- | <<هما>> | {past_3fd} | {ind_3fd} | |- | <<هم>> | {past_3mp} | {ind_3mp} | |- | <<هنّ>> | {past_3fp} | {ind_3fp} | ]=] text = text .. mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-bottom', args = { notes = '{footnote}', } } local forms = alternant_multiword_spec.forms if not alternant_multiword_spec.lemmas then forms.title = "—" else local linked_lemmas = {} for _, form in ipairs(alternant_multiword_spec.lemmas) do --W: Create simple link for lemmas without #Arabic anchor -- table.insert(linked_lemmas, link_term(form.form, "term")) table.insert(linked_lemmas, link_term_simple(form.form)) end forms.title = table.concat(linked_lemmas, "، ") end local ann_parts = {} if alternant_multiword_spec.annotation ~= "" then table.insert(ann_parts, alternant_multiword_spec.annotation) end local annotation = table.concat(ann_parts, "، ") --W: use only arabic morphological pattern in title annotation local morph_patterns = alternant_multiword_spec.morph_pattern if morph_patterns and #morph_patterns > 0 then annotation = table.concat(morph_patterns, "، ") forms.title = forms.title .. " (" .. annotation .. ")" end -- Format the table. text = mark_focus_cells(text, alternant_multiword_spec.focus_slots) local tagged_table = rsub(text, "<<(.-)>>", tag_text) return m_string_utilities.format(tagged_table, forms) end ------------------------------------------------------------------------------- -- Verb classification functions -- ------------------------------------------------------------------------------- --[=[ تصنيف الفعل الثلاثي المُجرَّد حسب نوعه Classify an Arabic triliteral basic verb root based on its three letters. @param r1 الحرف الأول (first root letter - الفاء) @param r2 الحرف الثاني (second root letter - العين) @param r3 الحرف الثالث (third root letter - اللام) @return str: نوع الفعل (verb type classification in Arabic) Examples: classify_triliteral_verb('ك', 'ت', 'ب') --> 'صحيح سالم' classify_triliteral_verb('أ', 'ت', 'ي') --> 'معتل ناقص يائي مهموز الفاء' classify_triliteral_verb('ق', 'ل', 'و') --> 'معتل ناقص واوي' ]=] function export.classify_triliteral_verb(r1, r2, r3) -- تعريف أحرف العلة والهمزة -- الحروف الضعيفة (أحرف العلة): واو وياء فقط local weak_letters = {[W] = true, [Y] = true} -- الهمزة: جميع أشكال الهمزة (ء، أ، إ، ؤ، ئ) local hamza_letters = { [HAMZA] = true, -- ء [HAMZA_ON_ALIF] = true, -- أ [HAMZA_UNDER_ALIF] = true, -- إ [HAMZA_ON_W] = true, -- ؤ [HAMZA_ON_Y] = true -- ئ } -- فحص كل حرف local is_hamza = {hamza_letters[r1] or false, hamza_letters[r2] or false, hamza_letters[r3] or false} local is_weak = {weak_letters[r1] or false, weak_letters[r2] or false, weak_letters[r3] or false} -- عدد أحرف العلة local weak_count = (is_weak[1] and 1 or 0) + (is_weak[2] and 1 or 0) + (is_weak[3] and 1 or 0) -- فحص التضعيف (الحرف الثاني = الثالث) local is_doubled = (r2 == r3) -- ═══════════════════════════════════════════════════════════ -- المُعتل: الفعل الذي يحتوي على حرف أو أكثر من أحرف العلة -- ═══════════════════════════════════════════════════════════ if weak_count > 0 then -- ───────────────────────────────────────────────────── -- لفيف: يحتوي على حرفين أو أكثر من أحرف العلة -- ───────────────────────────────────────────────────── if weak_count >= 2 then -- مقرون: حرفا العلة متجاوران if is_weak[2] and is_weak[3] then -- الثاني والثالث if is_hamza[1] then return "معتل لفيف مقرون مهموز الفاء" end return "معتل لفيف مقرون" elseif is_weak[1] and is_weak[2] then -- الأول والثاني return "معتل لفيف مقرون" -- مفروق: الأول والثالث حرفا علة elseif is_weak[1] and is_weak[3] then if is_hamza[2] then return "معتل لفيف مفروق مهموز العين" end return "معتل لفيف مفروق" end -- ───────────────────────────────────────────────────── -- مثال: الحرف الأول حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[1] then if r1 == W then -- واوي if is_hamza[3] then return "معتل مثال واوي مهموز اللام" elseif is_doubled then return "معتل مثال واوي مُضعَّف" elseif is_hamza[2] then return "معتل مثال واوي مهموز العين" end return "معتل مثال واوي" else -- يائي (ي) if is_doubled then return "معتل مثال يائي مُضعَّف" elseif is_hamza[2] then return "معتل مثال يائي مهموز العين" end return "معتل مثال يائي" end -- ───────────────────────────────────────────────────── -- أجوف: الحرف الثاني حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[2] then if r2 == W then -- واوي if is_hamza[3] then return "معتل أجوف واوي مهموز اللام" elseif is_hamza[1] then return "معتل أجوف واوي مهموز الفاء" end return "معتل أجوف واوي" else -- يائي (ي) if is_hamza[3] then return "معتل أجوف يائي مهموز اللام" elseif is_hamza[1] then return "معتل أجوف يائي مهموز الفاء" end return "معتل أجوف يائي" end -- ───────────────────────────────────────────────────── -- ناقص: الحرف الثالث حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[3] then if r3 == W then -- واوي if is_hamza[1] then return "معتل ناقص واوي مهموز الفاء" elseif is_hamza[2] then return "معتل ناقص واوي مهموز العين" end return "معتل ناقص واوي" else -- يائي (ي) if is_hamza[1] then return "معتل ناقص يائي مهموز الفاء" elseif is_hamza[2] then return "معتل ناقص يائي مهموز العين" end return "معتل ناقص يائي" end end -- ═══════════════════════════════════════════════════════════ -- الصحيح: الفعل الذي لا يحتوي على أحرف العلة -- ═══════════════════════════════════════════════════════════ else -- ───────────────────────────────────────────────────── -- مهموز: يحتوي على همزة -- ───────────────────────────────────────────────────── if is_hamza[1] or is_hamza[2] or is_hamza[3] then -- مهموز الفاء واللام: الأول والثالث همزة if is_hamza[1] and is_hamza[3] then return "صحيح مهموز الفاء واللام" -- مهموز الفاء elseif is_hamza[1] then if is_doubled then return "صحيح مُضعَّف مهموز الفاء" end return "صحيح مهموز الفاء" -- مهموز العين elseif is_hamza[2] then return "صحيح مهموز العين" -- مهموز اللام elseif is_hamza[3] then return "صحيح مهموز اللام" end -- ───────────────────────────────────────────────────── -- مُضعَّف: الحرف الثاني = الثالث -- ───────────────────────────────────────────────────── elseif is_doubled then return "صحيح مُضعَّف" -- ───────────────────────────────────────────────────── -- سالم: لا همزة، لا حرف علة، لا تضعيف -- ───────────────────────────────────────────────────── else return "صحيح سالم" end end return "غير معروف" end ------------------------------------------------------------------------------- -- External entry points -- ------------------------------------------------------------------------------- --W: Normalize root to use underscores as separators -- Accepts: "كتب", "ك_ت_ب", or "ك ت ب" -- Returns: "ك_ت_ب" local function normalize_root(root) if not root then return nil end -- Helper function to normalize a single character/radical local function normalize_char(char) -- Convert all hamza variants to canonical bare hamza if char == HAMZA_ON_ALIF or char == HAMZA_UNDER_ALIF or char == HAMZA_ON_W or char == HAMZA_ON_Y then return HAMZA -- Convert alif maqṣūra to yaa (ى → ي) elseif char == AMAQ then return Y -- Reject alif - it's not a root radical elseif char == ALIF then error("Root contains alif (ا) which is not a valid root radical. For weak verbs, use و (waw) or ي (yaa) instead.") end return char end -- Convert alif madda to hamza + alif for processing (will be caught as error below) root = root:gsub(AMAD, HAMZA .. ALIF) -- Replace spaces with underscores root = root:gsub(" ", "_") -- If no underscores yet, insert them between each character if not root:find("_") then -- Extract only Arabic letters (excluding diacritics) -- Arabic letter range: U+0621 to U+064A (basic letters) local chars = {} for char in mw.ustring.gmatch(root, "[ء-ي]") do table.insert(chars, normalize_char(char)) end root = table.concat(chars, "_") else -- Root already has underscores (e.g., "أ_م_ن"), normalize each radical local parts = mw.text.split(root, "_") for i, part in ipairs(parts) do parts[i] = normalize_char(part) end root = table.concat(parts, "_") end return root end --W: Construct the [1] parameter from root (جذر) and morphological pattern (وزن) -- Example: -- input: root="كتب", pattern="فعَل يفعُل" -- output: "ك_ت_ب<I/a~u>" -- `force_full_passive`, if true, appends ".pass" so the verb gets a full personal -- passive paradigm instead of the module's default impersonal-passive guess for -- vforms like I (past vowel i), V, VI, X and IIq (see |مبني للمجهول= on {{ar-conj}}). -- `reduced`, if true, appends ".reduced" so forms V/VI use the assimilated تاء -- التفعّل prefix (اِسَّمَّعَ) instead of the default تَ- prefix (تَسَمَّعَ) -- see -- |مدغم= on {{ar-conj}}. Not every root with an assimilable first radical is -- necessarily attested with the assimilated form, so this isn't inferred automatically -- from the root; it must be requested. local function construct_verb_spec(root, pattern, force_full_passive, reduced) if not root or not pattern then return nil end local normalized_root = normalize_root(root) local form_code = MORPHOLOGICAL_PATTERN_TO_FORM[pattern] if not form_code then error("Unknown morphological pattern (وزن صرفي): " .. pattern) end if force_full_passive then form_code = form_code .. ".pass" end if reduced then form_code = form_code .. ".reduced" end return normalized_root .. "<" .. form_code .. ">" end -- Wrapper function that accepts جذر (root) and وزن (pattern) as parameters -- Returns: verb_type (e.g., "فعل ثلاثي مُجرَّد صحيح سالم") -- Example: export.get_verb_type({["جذر"] = "كتب", ["وزن"] = "فعَل يفعُل"}) function export.get_verb_type(args) local verb_spec if args["جذر"] and args["وزن"] then verb_spec = construct_verb_spec(args["جذر"], args["وزن"]) elseif args[1] then verb_spec = args[1] else error("Must provide either جذر and وزن, or a verb specification in args[1]") end local alternant_multiword_spec = export.do_generate_forms({[1] = verb_spec}) local _, verb_type = get_verb_info(alternant_multiword_spec) return verb_type end -- Append two lists `l1` and `l2`, removing duplicates. If either is {nil}, just return the other. local function combine_lists(l1, l2) -- combine_footnotes() does exactly what we want. return iut.combine_footnotes(l1, l2) end local function combine_metadata(data) local src1 = data.form1 local src2 = data.form2 local dest = data.dest_form dest.uncertain = src1.uncertain or src2.uncertain if src1.genders and src2.genders and not m_table.deepEquals(src1.genders, src2.genders) then -- do nothing else dest.genders = src1.genders or src2.genders end if src1.pos and src2.pos and src1.pos ~= src2.pos then -- do nothing else dest.pos = src1.pos or src2.pos end -- Don't copy .alt, .gloss, .lit, .id, which describe a single term and don't extend to multiword terms. dest.q = combine_lists(src1.q, src2.q) dest.qq = combine_lists(src1.qq, src2.qq) dest.l = combine_lists(src1.l, src2.l) dest.ll = combine_lists(src1.ll, src2.ll) end -- Externally callable function to parse and conjugate a verb given user-specified arguments. -- Return value is WORD_SPEC, an object where the conjugated forms are in `WORD_SPEC.forms` -- for each slot. If there are no values for a slot, the slot key will be missing. The value -- for a given slot is a list of objects {form=FORM, footnotes=FOOTNOTES}. function export.do_generate_forms(args, source_template, headword_head) local PAGENAME = mw.loadData("Module:headword/data").pagename local function in_template_space() local nsText = mw.title.getCurrentTitle().nsText return nsText == "Template" or nsText == "قالب" end -- Determine the verb spec we're being asked to generate the conjugation of. This may be taken from the current page -- title or the value of |pagename=; but not when called from {{ar-verb form}}, where the page title is a -- non-lemma form. Note that the verb spec may omit the lemma; e.g. it may be "<II>". For this reason, we use the -- value of `pagename` computed here down below, when calling normalize_all_lemmas(). local pagename = source_template ~= "ar-verb form" and args.pagename or PAGENAME local head = headword_head or pagename --W: If both root (جذر) and morphological pattern (وزن) are provided, construct [1] if args["جذر"] and args["جذر"] ~= "" and args["وزن"] and args["وزن"] ~= "" then local constructed = construct_verb_spec(args["جذر"], args["وزن"], args["مبني للمجهول"], args["مدغم"]) if constructed and constructed ~= "" then args[1] = constructed end end local arg1 = args[1] if not arg1 then if (pagename == "ar-conj" or pagename == "ar-verb" or pagename == "ar-verb form" or pagename == "تصريف") and in_template_space() then arg1 = "كتب<I/a~u.pass>" else arg1 = "<>" end end -- When called from {{ar-verb form}}, determine the non-lemma form whose inflections we're being asked to -- determine. This normally comes from the page title or the value of |pagename=. local verb_form_of_form if source_template == "ar-verb form" then verb_form_of_form = args.pagename if not verb_form_of_form then if PAGENAME == "ar-verb form" and in_template_space() then verb_form_of_form = "كتبت" else verb_form_of_form = PAGENAME end end end local incorporated_headword_head_into_lemma = false if arg1:find("^<.*>$") then -- missing lemma if head:find(" ") then -- If multiword lemma, try to add arg spec after the first word. -- Try to preserve the brackets in the part after the verb, but don't do it -- if there aren't the same number of left and right brackets in the verb -- (which means the verb was linked as part of a larger expression). local first_word, post = rmatch(head, "^(.-)( .*)$") local left_brackets = rsub(first_word, "[^%[]", "") local right_brackets = rsub(first_word, "[^%]]", "") if #left_brackets == #right_brackets then arg1 = iut.remove_redundant_links(first_word) .. arg1 .. post incorporated_headword_head_into_lemma = true else -- Try again using the form without links. local linkless_head = m_links.remove_links(head) if linkless_head:find(" ") then first_word, post = rmatch(linkless_head, "^(.-)( .*)$") arg1 = first_word .. arg1 .. post else error("Unable to incorporate <...> spec into explicit head due to a multiword linked verb or " .. "unbalanced brackets; please include <> explicitly: " .. arg1) end end else -- Will be incorporated through `head` below in the call to normalize_all_lemmas(). incorporated_headword_head_into_lemma = true end end local parse_props = { parse_indicator_spec = parse_indicator_spec, angle_brackets_omittable = true, allow_blank_lemma = true, } local alternant_multiword_spec = iut.parse_inflected_text(arg1, parse_props) alternant_multiword_spec.pos = pos or "verbs" alternant_multiword_spec.args = args alternant_multiword_spec.source_template = source_template alternant_multiword_spec.verb_form_of_form = verb_form_of_form alternant_multiword_spec.incorporated_headword_head_into_lemma = incorporated_headword_head_into_lemma normalize_all_lemmas(alternant_multiword_spec, head) detect_all_indicator_specs(alternant_multiword_spec) local inflect_props = { lang = lang, slot_list = alternant_multiword_spec.verb_slots, inflect_word_spec = conjugate_verb, combine_metadata = combine_metadata, -- We add links around the generated verbal forms rather than allow the entire multiword -- expression to be a link, so ensure that user-specified links get included as well. include_user_specified_links = true, } iut.inflect_multiword_or_alternant_multiword_spec(alternant_multiword_spec, inflect_props) if debug_translit then for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do if form.translit then local full_form_translit = (lang:transliterate(m_links.remove_links(form.form))) if full_form_translit ~= form.translit then error(("Internal error: For slot '%s', form '%s' incremental translit '%s' not same as full translit '%s'"): format(slot, form.form, form.translit, full_form_translit)) end end form.form = iut.remove_redundant_links(form.form) end end end -- Remove redundant brackets around entire forms. for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do form.form = iut.remove_redundant_links(form.form) end end -- Apply NFC-style normalization: reorder shadda+vowel to vowel+shadda -- to match MediaWiki's automatic NFC normalization (which is applied on enwiktionary -- but not in local Lua environments). This reverses the postprocessing step. for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do form.form = apply_nfc_shadda(form.form) end end determine_slot_uncertainty_from_forms(alternant_multiword_spec) determine_verb_properties_from_forms(alternant_multiword_spec) compute_categories_and_annotation(alternant_multiword_spec) if args.json and source_template == "ar-conj" then -- There is a circular reference in `base.alternant_multiword_spec`, which points back to top level. iut.map_word_specs(alternant_multiword_spec, function(base) base.alternant_multiword_spec = nil end) return require("Module:JSON").toJSON(alternant_multiword_spec) end return alternant_multiword_spec end -- Entry point for {{ar-conj}}. Template-callable function to parse and conjugate a verb given -- user-specified arguments and generate a displayable table of the conjugated forms. function export.show(frame) local parent_args = frame:getParent() and frame:getParent().args or frame.args --W local params = { [1] = {}, ["noautolinktext"] = {type = "boolean"}, ["noautolinkverb"] = {type = "boolean"}, ["t"] = {}, -- for use by {{ar-verb form}}; otherwise ignored ["id"] = {}, -- for use by {{ar-verb form}}; otherwise ignored ["pagename"] = {}, -- for testing/documentation pages ["json"] = {type = "boolean"}, -- for bot use --W: additional parameters for ar.wiktionary.org ["جذر"] = {type = "string"}, --W: الجذر اللغوي, root ["وزن"] = {type = "string"}, --W: الوزن الصرفي, morphological pattern ["مبني للمجهول"] = {type = "boolean"}, --W: force a full personal passive paradigm, -- overriding the default impersonal-passive guess for vforms like I (past vowel -- i), V, VI, X and IIq ["مدغم"] = {type = "boolean"}, --W: for forms V/VI, assimilate the تاء التفعّل prefix -- into the first radical (اِسَّمَّعَ instead of تَسَمَّعَ) ["كامل"] = {type = "boolean"}, --W: show full table (default is simplified table) ["بؤرة"] = {type = "string"}, --W: highlight the cells holding this form ["موسع"] = {type = "boolean"}, --W: show the table expanded (default: collapsed) } local args = require("Module:parameters").process(parent_args, params) --W: If both root (جذر) and morphological pattern (وزن) are provided, construct [1] if args["جذر"] and args["جذر"] ~= "" and args["وزن"] and args["وزن"] ~= "" then local constructed = construct_verb_spec(args["جذر"], args["وزن"], args["مبني للمجهول"], args["مدغم"]) if constructed and constructed ~= "" then args[1] = constructed end end local alternant_multiword_spec = export.do_generate_forms(args, "ar-conj") if type(alternant_multiword_spec) == "string" then -- JSON return value return alternant_multiword_spec end --W: |بؤرة= marks the cells holding a given form, so that a conjugated-form -- page ([[Module:ar-verb-form]]) can show the reader where in the paradigm -- the word they looked up actually sits. The slots have to be found before -- show_forms(), which replaces each form list with rendered wikitext. alternant_multiword_spec.focus_slots = find_focus_slots(alternant_multiword_spec, args["بؤرة"]) --W: A table whose one relevant cell is highlighted is no use folded away, so -- [[Module:ar-verb-form]] asks for it open. Lemma pages keep it collapsed. alternant_multiword_spec.expanded = args["موسع"] or false show_forms(alternant_multiword_spec) --W: Choose table type: simplified (default) or full (comprehensive) local table_html if args["كامل"] then table_html = make_table(alternant_multiword_spec) else table_html = make_table_ar(alternant_multiword_spec) end --W: don't add categories --W: TODO: all lines with insert_cat() are to be removed. -- return table_html .. -- require("Module:utilities").format_categories(alternant_multiword_spec.categories, lang, nil, nil, force_cat) return table_html end function export.verb_forms(frame) local parargs = frame:getParent().args local params = { [1] = {}, [2] = {}, [3] = {}, [4] = {}, [5] = {}, pagename = {}, } for _, form in ipairs(allowed_vforms) do -- FIXME: We go up to 5 here. The code supports unlimited variants but it's unlikely we will ever see more than -- 2. for index = 1, 5 do local prefix = index == 1 and form or form .. index params[prefix .. "-pv"] = {} for _, extn in ipairs { "", "-vn", "-ap", "-pp" } do params[prefix .. extn] = {} params[prefix .. extn .. "-head"] = {} -- FIXME: No -tr? params[prefix .. extn .. "-gloss"] = {} end end end local args = require("Module:parameters").process(parargs, params) local i = 1 local past_vowel_re = "^[aui,]*$" local combined_root = nil if not args[i] or rfind(args[i], past_vowel_re) then combined_root = args.pagename or mw.loadData("Module:headword/data").pagename if not rfind(combined_root, "^([^ ]) ([^ ]) ([^ ])$") and not rfind(combined_root, "^([^ ]) ([^ ]) ([^ ]) ([^ ])$") then error("When inferring roots from page title, need three or four space-separated radicals: " .. combined_root) end elseif rfind(args[i], " ") then combined_root = args[i] i = i + 1 else local separate_roots = {} while args[i] and not rfind(args[i], past_vowel_re) do table.insert(separate_roots, args[i]) i = i + 1 end combined_root = table.concat(separate_roots, " ") end local past_vowel = args[i] i = i + 1 if past_vowel and not rfind(past_vowel, past_vowel_re) then error("Unrecognized past vowel, should be 'a', 'i', 'u', 'a,u', etc. or empty: " .. past_vowel) end -- Spaces interfere with parsing as a unit in [[Module:inflection utilities]], so replace with underscore. combined_root = combined_root:gsub(" ", "_") local split_root = rsplit(combined_root, "_") -- Map from verb forms (I, II, etc.) to a table of verb properties, -- which has entries e.g. for "verb" (either true to autogenerate the verb -- head, or an explicitly specified verb head using e.g. argument "I-head"), -- and for "verb-gloss" (which comes from e.g. the argument "I" or "I-gloss"), -- and for "vn" and "vn-gloss", "ap" and "ap-gloss", "pp" and "pp-gloss". local verb_properties = {} for _, form in ipairs(allowed_vforms) do local formpropslist = {} local derivs = {{"verb", ""}, {"vn", "-vn"}, {"ap", "-ap"}, {"pp", "-pp"}} local index = 1 while true do local formprops = {} local prefix = index == 1 and form or form .. index if prefix == "I" then formprops.pv = past_vowel end if args[prefix .. "-pv"] then formprops.pv = args[prefix .. "-pv"] end for _, deriv in ipairs(derivs) do local prop = deriv[1] local extn = deriv[2] if args[prefix .. extn] == "+" then formprops[prop] = true elseif args[prefix .. extn] == "-" then formprops[prop] = false elseif args[prefix .. extn] then formprops[prop] = true formprops[prop .. "-gloss"] = args[prefix .. extn] end if args[prefix .. extn .. "-head"] then if formprops[prop] == nil then formprops[prop] = true end formprops[prop] = args[prefix .. extn .. "-head"] end if args[prefix .. extn .. "-gloss"] then if formprops[prop] == nil then formprops[prop] = true end formprops[prop .. "-gloss"] = args[prefix .. extn .. "-gloss"] end end if formprops.verb then -- If a verb form specified, also turn on vn (unless form I, with -- unpredictable vn) and ap, and maybe pp, according to form, -- weakness and past vowel. But don't turn these on if there's -- an explicit on/off specification for them (e.g. I-pp=-). if form ~= "I" and formprops.vn == nil then formprops.vn = true end if formprops.ap == nil then formprops.ap = true end local weakness = weakness_from_radicals(form, split_root[1], split_root[2], split_root[3], split_root[4]) if formprops.pp == nil and not vform_probably_no_passive(form, weakness, rsplit(formprops.pv or "", ","), {}) then formprops.pp = true end if formprops.verb == true or formprops.vn == true or formprops.ap == true or formprops.pp == true then formprops.need_autogen = true end table.insert(formpropslist, formprops) index = index + 1 else break end end table.insert(verb_properties, {form, formpropslist}) end -- Go through and create the verb form derivations as necessary, when they haven't been explicitly given. for _, vplist in ipairs(verb_properties) do local vform = vplist[1] for _, props in ipairs(vplist[2]) do if props.need_autogen then local form_with_vowels if vform == "I" then local pv = props.pv if not pv then -- Make up likely past vowels based on weakness and actual radical. if split_root[3] == W then -- final-weak form_with_vowels = "I/a~u" elseif split_root[3] == Y then form_with_vowels = "I/a~i" elseif split_root[2] == W then --hollow form_with_vowels = "I/u~u" elseif split_root[2] == Y then form_with_vowels = "I/i~i" else -- most common; doesn't matter so much since we're not displaying the non-past form_with_vowels = "I/a~u" end else local pvs = rsplit(pv, ",") local vowel_sufs = {} for _, pv in ipairs(pvs) do local vowel_spec if pv == "a" then -- Make up likely past vowels based on weakness and actual radical. if split_root[3] == W then -- final-weak vowel_spec = "a~u" elseif split_root[3] == Y then vowel_spec = "a~i" elseif split_root[2] == W then --hollow vowel_spec = "a~u" elseif split_root[2] == Y then vowel_spec = "a~i" else -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "a~u" end elseif pv == "i" then -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "i~a" elseif pv == "u" then -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "u~u" else error(("Internal error: Bad past vowel '%s' in {{ar-verb forms}}"):format(pv)) end table.insert(vowel_sufs, vowel_spec) end form_with_vowels = "I/" .. table.concat(vowel_sufs, "/") end else form_with_vowels = vform end local angle_bracket_spec = ("%s<%s.pass>"):format(combined_root, form_with_vowels) local alternant_multiword_spec = export.do_generate_forms({angle_bracket_spec}, "ar-verb forms") local function format_forms(forms) if not forms then return "-" -- FIXME: Throw an error? end local formatted = {} for _, form in ipairs(forms) do if form.translit then table.insert(formatted, ("%s//%s"):format(form.form, form.translit)) else table.insert(formatted, form.form) end end return table.concat(formatted, "،") end if props.verb == true then props.verb = format_forms(alternant_multiword_spec.forms.past_3ms) end for _, deriv in ipairs({"vn", "ap", "pp"}) do if props[deriv] == true then props[deriv] = format_forms(alternant_multiword_spec.forms[deriv]) end end end end end -- Go through and output the result local formtextarr = {} for _, vplist in ipairs(verb_properties) do local form = vplist[1] for _, props in ipairs(vplist[2]) do local textarr = {} if props.verb then local text = "* '''[[Appendix:Arabic verbs#Form " .. form .. "|Form " .. form .. "]]''': " local linktext = {} local splitheads = rsplit(props.verb, "[,،]") for _, head in ipairs(splitheads) do table.insert(linktext, m_links.full_link({lang = lang, term = head, gloss = props["verb-gloss"]})) end text = text .. table.concat(linktext, "، ") table.insert(textarr, text) for _, derivengl in ipairs({{"vn", "Verbal noun"}, {"ap", "Active participle"}, {"pp", "Passive participle"}}) do local deriv = derivengl[1] local engl = derivengl[2] if props[deriv] then local text = "** " .. engl .. ": " local linktext = {} local splitheads = rsplit(props[deriv], "[,،]") for _, head in ipairs(splitheads) do local ar, translit = head:match("^(.*)//(.-)$") if not ar then ar = head end table.insert(linktext, m_links.full_link {lang = lang, term = ar, tr = translit, gloss = props[deriv .. "-gloss"]} ) end text = text .. table.concat(linktext, "، ") table.insert(textarr, text) end end table.insert(formtextarr, table.concat(textarr, "\n")) end end end return table.concat(formtextarr, "\n") end -- Infer radicals from lemma headword (i.e. 3rd masculine singular past) and verb form (I, II, etc.). Throw an error if -- headword is malformed. A given returned radical may be actually be a list of possible radicals, where the first one -- should be used if the user didn't explicitly give the radical. If the list contains a field `ambig = true`, the -- radical is considered ambiguous and should not be categorized. `is_reduced` indicates that the user specified -- `.reduced` to indicate that the verb form is reduced by assimilation and/or haplology (typically archaic Koranic -- forms such as اِدَّارَأَ instead of تَدَارَأَ; or اِسْطَاعَ instead of اِسْتِطَاعَ; etc. function export.infer_radicals(data) local headword, vform, passive, past_vowel, nonpast_vowel, is_reduced = data.headword, data.vform, data.passive, data.past_vowel, data.nonpast_vowel, data.is_reduced past_vowel = past_vowel or "-" nonpast_vowel = nonpast_vowel or "-" local function verify_vowel(vowel, param) --W: A past/non-past vowel may legitimately be a form object {form=..., footnotes=...} -- when the user attaches a footnote, e.g. <I/a[نادر]~u>, so compare the formval. -- Comparing the raw value crashed on Lua 5.1 ("bad argument #2 to 'format'"). local formval = rget(vowel) if formval ~= A and formval ~= I and formval ~= U and formval ~= "-" then error(("Internal error: Bad value for %s: %s (should be Arabic diacritic vowel or '-')"):format( param, dump(vowel))) end end verify_vowel(past_vowel, "past_vowel") verify_vowel(nonpast_vowel, "nonpast_vowel") local ch = {} local form_viii_assim, variant -- sub out alif-madda for easier processing headword = rsub(headword, AMAD, HAMZA .. ALIF) local function infer_err(msg, noann) local anns = {} local nohead, novform if noann == "nohead" then nohead = true elseif noann == "novform" then novform = true elseif noann == "nohead-vform" then nohead = true novform = true elseif noann then error(("Internal error: Unrecognized value for 'noann': %s"):format(dump(noann))) end if not nohead then table.insert(anns, ("headword=%s"):format(data.headword)) end if not novform then table.insert(anns, ("verb form=%s"):format(data.vform)) end anns = table.concat(anns, "، ") if anns ~= "" then anns = ": " .. anns end error(msg .. anns) end local len = ulen(headword) local expected_length -- extract the headword letters into an array for i = 1, len do table.insert(ch, usub(headword, i, i)) end -- check that the letter at the given index is the given string, or -- is one of the members of the given array local function check(index, must) local letter = ch[index] if type(must) == "string" then if not letter then infer_err("Letter " .. index .. " is nil") end if letter ~= must then infer_err(("For verb form %s, letter %s must be %s, not %s"):format(vform, index, must, letter), "novform") end elseif not m_table.contains(must, letter) then infer_err("For verb form " .. vform .. ", radical " .. index .. " must be one of " .. table.concat(must, " ") .. ", not " .. letter, "novform") end end -- Check that length of headword is within [min, max] local function check_len(min, max) if min and len < min then infer_err(("Not enough letters for verb form %s, expected at least %s"):format(vform, min), "novform") end if max and len > max then infer_err(("Too many letters for verb form %s, expected at most %s"):format(vform, max), "novform") end end -- A form I verb beginning with w- keeps the w in the non-past unless the non-past vowel is a kasra or a -- lowered kasra; see form_i_w_assimilated(), which the radicals path uses too. This used to be spelt out -- here as "i~a or u~u", which missed a~u (وَجَلَ يَوْجُلُ). local function form_I_w_non_assimilated() return not form_i_w_assimilated(past_vowel, nonpast_vowel) end -- Convert radicals to canonical form (handle various hamza varieties and check for misplaced alif or alif maqṣūra; -- legitimate cases of these letters are handled above). local function convert(rad, index) if type(rad) == "table" then for i, r in ipairs(rad) do rad[i] = convert(r, index) end return rad elseif rad == HAMZA_ON_ALIF or rad == HAMZA_UNDER_ALIF or rad == HAMZA_ON_W or rad == HAMZA_ON_Y then return HAMZA elseif rad == AMAQ then infer_err("Radical " .. index .. " must not be alif maqṣūra") elseif rad == ALIF then infer_err("Radical " .. index .. " must not be alif") else return rad end end local quadlit = vform:find("q$") -- find first radical, start of second/third radicals, check for -- required letters local radstart, rad1, rad2, rad3, rad4 local weakness if vform == "I" or vform == "II" then rad1 = ch[1] radstart = 2 elseif vform == "III" then rad1 = ch[1] check(2, {ALIF, W}) -- W occurs in passive-only verbs radstart = 3 elseif vform == "IV" then -- this would be alif-madda but we replaced it with hamza-alif above. if ch[1] == HAMZA and ch[2] == ALIF then rad1 = HAMZA else check(1, HAMZA_ON_ALIF) rad1 = ch[2] end radstart = 3 elseif vform == "V" then check(1, is_reduced and ALIF or T) rad1 = ch[2] radstart = 3 elseif vform == "VI" then check(1, is_reduced and ALIF or T) if ch[2] == AMAD then rad1 = HAMZA radstart = 3 else rad1 = ch[2] check(3, {ALIF, W}) -- W occurs in passive-only verbs radstart = 4 end elseif vform == "VII" then check(1, ALIF) if is_reduced then check(2, M) rad1 = M radstart = 3 else check(2, N) rad1 = ch[3] radstart = 4 end elseif vform == "VIII" then check(1, ALIF) rad1 = ch[2] if rad1 == "د" then rad1 = {"د", "ذ"} -- not considered ambiguous since it's usually د radstart = 3 form_viii_assim = "دّ" elseif rad1 == "ظ" and ch[3] == "ط" and len >= 5 then -- [[اظطلم]], variant of [[اظلم]] radstart = 4 form_viii_assim = "ظْط" elseif rad1 == "ذ" and ch[3] == "د" and len >= 5 then -- [[اذدكر]], variant of [[اذكر]] radstart = 4 form_viii_assim = "ذْد" elseif rad1 == T or rad1 == "ث" or rad1 == "ذ" or rad1 == "ط" or rad1 == "ظ" then radstart = 3 form_viii_assim = rad1 .. SH elseif rad1 == "ز" then check(3, "د") radstart = 4 form_viii_assim = "زْد" elseif rad1 == "ص" or rad1 == "ض" then check(3, "ط") radstart = 4 form_viii_assim = rad1 .. SK .. "ط" else check(3, T) radstart = 4 rad1 = convert(rad1, 1) form_viii_assim = rad1 .. SK .. "ت" end if rad1 == T then -- Radical is ambiguous, might be ت or و or ي but doesn't affect conjugation. Note that there are no -- form-VIII verbs with initial radical ي given in Hans Wehr but Lane mentions at least: -- - (page 2973) اِتَّأَسَ, with assimilation of the ي to ت, from root ي ء س; -- - (page 2975) اِتَّبَسَ non-past يَتَّبِسُ and alternative اِيتَبَسَ non-past يَاتَبِسُ from the root ي ب س; -- - (page 2976) اِتَّسَرَ non-past يَتَّسِرُ or alternatively يَأْتَسِرُ with hamza preserved from the root ي س ر. -- These alternative forms seem very rare and probably not worth worrying about, but if we want to handle -- them, we can do it when the time comes. rad1 = {T, W, Y, ambig = true} -- اِتَّخَذَ irregularly has hamza as the radical but assimilates like و if ch[3] == "خ" and ch[4] == "ذ" then rad1[4] = HAMZA end end elseif vform == "IX" then check(1, ALIF) rad1 = ch[2] radstart = 3 elseif vform == "X" then check(1, ALIF) check(2, S) if is_reduced then rad1 = ch[3] radstart = 4 else check(3, T) rad1 = ch[4] radstart = 5 end elseif vform == "Iq" then rad1 = ch[1] rad2 = ch[2] radstart = 3 elseif vform == "IIq" then check(1, T) rad1 = ch[2] rad2 = ch[3] radstart = 4 elseif vform == "IIIq" then check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) radstart = 5 elseif vform == "IVq" then check(1, ALIF) rad1 = ch[2] rad2 = ch[3] radstart = 4 elseif vform == "XI" then check_len(5, 5) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, ALIF) rad3 = ch[5] weakness = "sound" elseif vform == "XII" then check(1, ALIF) rad1 = ch[2] if ch[3] ~= ch[5] then infer_err("For verb form XII, letters 3 and 5 should be the same", "novform") end check(4, W) radstart = 5 elseif vform == "XIII" then check_len(5, 5) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, W) rad3 = ch[5] if rad3 == AMAQ then weakness = "final-weak" else weakness = "sound" end elseif vform == "XIV" then check_len(6, 6) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) rad3 = ch[5] if ch[6] == AMAQ then check_waw_ya(rad3) weakness = "final-weak" else if ch[5] ~= ch[6] then infer_err("For verb form XIV, letters 5 and 6 should be the same", "novform") end weakness = "sound" end elseif vform == "XV" then check_len(6, 6) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) rad3 = ch[5] if rad3 == Y then check(6, ALIF) else check(6, AMAQ) end weakness = "sound" else error("Internal error: Unrecognized verb form " .. vform) end -- Process the last two radicals. RADSTART is the index of the first of the two. If it's nil then all radicals have -- already been processed above, and we don't do anything. if radstart then -- There must (normally) be one or two letters left. if len == radstart then if vform == "I" and ch[len] == Y then -- short form حَيَّ weakness = "final-weak" rad2 = Y rad3 = Y variant = "short" elseif vform == "IV" and rad1 == "ر" and ch[len] == AMAQ then -- irregular verb أَرَى weakness = "final-weak" rad2 = HAMZA rad3 = Y elseif vform == "X" and rad1 == "ح" and ch[len] == AMAQ then -- irregular verb اِسْتَحَى weakness = "final-weak" rad2 = Y rad3 = Y variant = "short" else -- If one letter left, then it's a geminate verb. If the letter is alif or alif maqṣūra, it will trigger -- an error down the line. if vform_supports_geminate(vform) then weakness = "geminate" rad2 = ch[len] rad3 = ch[len] if vform == "III" or vform == "VI" then variant = "short" end else infer_err("Apparent geminate verb, but geminate verbs not allowed for this verb form") end end elseif quadlit then -- Process last two radicals of a quadriliteral verb form. rad3 = ch[radstart] rad4 = ch[radstart + 1] expected_length = radstart + 1 check_len(expected_length) if rad4 == AMAQ or rad4 == ALIF and rad3 == Y or rad4 == Y then -- rad4 can be Y in passive-only verbs. if vform_supports_final_weak(vform) then weakness = "final-weak" -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws in a row); it could be wāw or -- yāʾ, but doesn't affect the conjugation. rad4 = rad3 == W and {Y, W, ambig = true} or {W, Y, ambig = true} else infer_err("Last radical is " .. rad4 .. " but verb form " .. vform .. " doesn't support final-weak verbs", "novform") end else weakness = "sound" end else -- Process last two radicals of a triliteral verb form. rad2 = ch[radstart] rad3 = ch[radstart + 1] expected_length = radstart + 1 check_len(expected_length) if vform == "I" and (is_waw_ya(rad3) or rad3 == ALIF or rad3 == AMAQ) then local inferred_past_vowel, inferred_nonpast_vowel -- Check for final-weak form I verb. It can end in tall alif (rad3 = wāw) or alif maqṣūra (rad3 = yāʾ) -- or a wāw or yāʾ (with a past vowel of i or u, e.g. nasiya/yansā "forget" or with a passive-only -- verb). if rad1 == W and not form_I_w_non_assimilated() then weakness = "assimilated+final-weak" else weakness = "final-weak" end if rad3 == ALIF then rad3 = W inferred_past_vowel = A inferred_nonpast_vowel = U if is_passive_only(passive) then infer_err("Final-weak form-I passive verbs should end in yāʔ (ي), not tall alif (ا)", "novform") end elseif rad3 == AMAQ then rad3 = Y inferred_past_vowel = A inferred_nonpast_vowel = I if is_passive_only(passive) then infer_err("Final-weak form-I passive verbs should end in yāʔ (ي), not alif maqṣūra (ى)", "novform") end elseif rad1 == "ح" and rad2 == Y and rad3 == Y then -- Long variant حَيِيَ. inferred_past_vowel = I inferred_nonpast_vowel = A variant = "long" else if not is_passive_only(passive) then -- does a non-passive final-weak verb in -uwa ever happen? (YES: e.g. [[رجو]] "to be slack") inferred_past_vowel = rad3 == Y and I or U inferred_nonpast_vowel = A end -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws); it could be wāw or yāʾ, but -- doesn't affect the conjugation. rad3 = (rad1 == W or rad2 == W) and {Y, W, ambig = true} or {W, Y, ambig = true} -- ambiguous end if inferred_past_vowel then local raw_past_vowel = rget(past_vowel) local raw_nonpast_vowel = rget(nonpast_vowel) if raw_past_vowel ~= "-" then if raw_past_vowel ~= inferred_past_vowel then infer_err(("Final-weak form-I verb inferred past vowel %s, which disagrees with " .. "explicitly specified %s"):format(undia[inferred_past_vowel], undia[raw_past_vowel]), "novform") else -- in case of footnote in past_vowel inferred_past_vowel = past_vowel end end if raw_nonpast_vowel ~= "-" and raw_nonpast_vowel ~= A and inferred_nonpast_vowel == U then -- if inferred as I or A, the reality can be the reverse; form-I final-weak verbs with a~a and -- i~i exist, e.g. سَعَى/يَسْعَى, وَلِيَ/يَلِي. Weird verb [[صها]] (also written [[صهى]]) has non-past -- يصهى so we can't throw an error in this situation. if raw_nonpast_vowel ~= inferred_nonpast_vowel then infer_err(("Final-weak form-I verb inferred non-past vowel %s, which disagrees with " .. "explicitly specified %s"):format(undia[inferred_nonpast_vowel], undia[raw_nonpast_vowel]), "novform") else -- in case of footnote in nonpast_vowel inferred_nonpast_vowel = nonpast_vowel end end end if not is_passive_only(passive) then if rget(past_vowel) == "-" then past_vowel = inferred_past_vowel end if rget(nonpast_vowel) == "-" then nonpast_vowel = inferred_nonpast_vowel end end elseif vform == "IX" and is_waw_ya(rad3) and len == radstart + 2 and ch[len] == AMAQ then -- Final-weak form IX verbs like اِرْعَوَى "to desist, to repent, to see the light". weakness = "final-weak" expected_length = radstart + 2 elseif vform == "X" and rad1 == "ح" and rad2 == Y and rad3 == ALIF then -- Long variant اِسْتَحْيَا. weakness = "final-weak" rad3 = Y variant = "long" elseif rad3 == AMAQ or rad2 == Y and rad3 == ALIF or rad3 == Y then -- rad3 == Y happens in passive-only verbs. if vform_supports_final_weak(vform) then weakness = "final-weak" else infer_err("Last radical is " .. rad3 .. " but verb form doesn't support final-weak verbs") end -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws); it could be wāw or yāʾ, but -- doesn't affect the conjugation. rad3 = (rad1 == W or rad2 == W) and {Y, W, ambig = true} or {W, Y, ambig = true} elseif rad2 == ALIF then if vform_supports_hollow(vform) then weakness = "hollow" local function set_past_to_a() if req(past_vowel, A) then -- already set elseif req(past_vowel, "-") or req(past_vowel, rget(nonpast_vowel)) then past_vowel = A else infer_err(("Form I hollow verb with nonpast vowel set to '%s' must have past vowel set to 'a' or the same value, not %s"): format(undia[rget(nonpast_vowel)], undia[rget(past_vowel)]), "novform") end end if vform == "I" and req(nonpast_vowel, U) then rad2 = W set_past_to_a() elseif vform == "I" and req(nonpast_vowel, I) then rad2 = Y set_past_to_a() else if req(nonpast_vowel, A) and not req(past_vowel, I) then infer_err(("Form I hollow verb with nonpast vowel set to 'a' must have past vowel set to 'i', not %s"): format(undia[rget(past_vowel)]), "novform") end -- Ambiguous radical; could be wāw or yāʾ; if verb form I, it's critical to get this right, and -- the caller checks for this situation and throws an error if non-past vowel is "a" and second -- radical isn't explicitly given. rad2 = {W, Y, ambig = true, need_radical = true} end else infer_err("Second radical is alif but verb form doesn't support hollow verbs") end elseif vform == "I" and rad1 == W and not form_I_w_non_assimilated() then weakness = "assimilated" elseif rad2 == rad3 and (vform == "III" or vform == "VI") then weakness = "geminate" variant = "long" else weakness = "sound" end end if expected_length then check_len(expected_length, expected_length) end end rad1 = convert(rad1, 1) rad2 = convert(rad2, 2) rad3 = convert(rad3, 3) rad4 = convert(rad4, 4) if not weakness then error("Internal error: Returned weakness from infer_radicals() is nil") end return { weakness = weakness, rad1 = rad1, rad2 = rad2, rad3 = rad3, rad4 = rad4, past_vowel = past_vowel, nonpast_vowel = nonpast_vowel, form_viii_assim = form_viii_assim, variant = variant, } end -- bot interface to infer_radicals() function export.infer_radicals_json(frame) local iparams = { headword = {}, vform = {}, passive = {}, past_vowel = {}, nonpast_vowel = {}, is_reduced = {type = "boolean"}, } local iargs = require("Module:parameters").process(frame.args, iparams) return require("Module:JSON").toJSON(export.infer_radicals(iargs)) end -- Infer vocalization from participle headword (active or passive), verb form (I, II, etc.) and whether the headword is -- active or passive. Throw an error if headword is malformed. Returned radicals may contain Latin letters "t", "w" or "y" -- indicating ambiguous radicals guessed to be tāʾ, wāw or yāʾ respectively. function export.infer_participle_vocalization(headword, vform, weakness, is_active) local chars = {} local orig_headword = headword -- Sub out alif-madda for easier processing. headword = rsub(headword, AMAD, HAMZA .. ALIF) local len = ulen(headword) -- Extract the headword letters into an array. for i = 1, len do table.insert(chars, usub(headword, i, i)) end local function form_intro_error_msg() return ("For verb form %s %s%s participle %s, "):format(vform, orig_headword ~= headword and "normalized " or "", is_active and "active" or "passive", headword) end local function err(msg) error(form_intro_error_msg() .. msg, 1) end -- Check that length of headword is within [min, max]. local function check_len(min, max) if min and len < min then err(("expected at least %s letters but saw %s"):format(min, len)) elseif max and len > max then err(("expected at most %s letters but saw %s"):format(max, len)) end end -- Get the character at `ind`, making sure it exists. local function c(ind) check_len(ind) return chars[ind] end -- Check that the letter at the given index is the given string, or is one of the members of the given array local function check(index, must) local letter = chars[index] local function make_possible_values() if type(must) == "string" then return must else return list_to_text(must, nil, " or ") end end if not letter then err(("expected a letter (specifically %s) at position %s, but participle is too short"):format( make_possible_values(), index)) end local matches if type(must) == "string" then matches = letter == must else matches = m_table.contains(must, letter) end if not matches then err(("letter %s at index %s must be %s"):format(letter, index, make_possible_values())) end end local function check_weakness(values, allow_missing, invert_condition) local function make_possible_weaknesses() for i, val in ipairs(values) do values[i] = "'" .. val .. "'" end return list_to_text(values, nil, " or ") end if allow_missing and invert_condition then error("Internal error: Can't specify both allow_missing and invert_condition") end if not weakness then if allow_missing or invert_condition then return else err(("weakness is unspecified but must be %s"):format(make_possible_weaknesses())) end else local matches = m_table.contains(values, weakness) if invert_condition and matches then err(("weakness '%s' must not be %s"):format(weakness, make_possible_weaknesses())) elseif not invert_condition and not matches then err(("weakness '%s' must be %s"):format(weakness, make_possible_weaknesses())) end end end local vocalized local function handle_possibly_final_weak(sound_prefix, expected_length) check_len(expected_length, expected_length) if c(expected_length) == AMAQ then -- passive final-weak if is_active then err("participle in -ِى only allowed for passive participles") end check_weakness({"final-weak", "assimilated+final-weak"}, "allow missing") vocalized = sound_prefix .. AN .. AMAQ else -- all others behave as if sound check_weakness({"final-weak", "assimilated+final-weak"}, nil, "invert condition") vocalized = sound_prefix .. (is_active and I or A) .. c(expected_length) end end if not (vform == "I" and is_active) then -- all participles except verb form I active begin in م-. check(1, M) end if vform == "I" then if is_active then check(2, ALIF) local sound_prefix = c(1) .. AA .. c(3) if len == 3 then if c(3) == HAMZA then -- Either hollow with hamzated third radical, e.g. [[شاء]] active participle 'شَاءٍ', or final-weak -- with hamzated second radical, e.g. [[رأى]] active participle 'رَاءٍ'. Theoretically (?), also -- geminate with hamzated second/third radical, but I don't know if any such verbs exist. if weakness == "geminate" then vocalized = sound_prefix .. SH else check_weakness({"hollow", "final-weak"}, "allow missing") vocalized = sound_prefix .. IN end else check_weakness({"final-weak", "geminate"}) if weakness == "geminate" then vocalized = sound_prefix .. SH else vocalized = sound_prefix .. IN end end else check_len(4, 4) -- we will convert back to alif maqṣūra below as needed vocalized = sound_prefix .. I .. c(4) end else -- assimilated verbs: regular, e.g. مَوْزُون "weighed" -- geminate verbs: regular, e.g. مَبْلُول "moistened" -- third-hamzated verbs: مَبْرُوء -- hollow verbs: مَقُود "led, driven"; مَزِيد "added, increased" -- hollow first-hamzated verbs: مَئِيض "returned, reverted"; مَأْيُوس "despaired" (NOTE: formation is sound); -- مَأُود or مَؤُود "bent; depleted" -- hollow third-hamzated verbs: مَشِيء "willed, intended", مَضُوء "glittered?" -- final-weak: مَلْقِيّ "found, encountered"; مَصْغُوّ "inclined" -- hollow + final-weak: مَشْوِيّ "fried, grilled", مَهْوِيّ "loved" -- first-hamzated + hollow + final-weak: مَأْوِيّ "received hospitably" local sound_prefix = MA .. c(2) .. SK .. c(3) if len == 5 then -- sound, assimilated or geminate check(4, W) vocalized = sound_prefix .. UU .. c(5) else check_len(4, 4) if c(4) == W then -- final-weak third-wāw vocalized = sound_prefix .. U .. W .. SH elseif c(4) == Y then -- final-weak third-yāʾ vocalized = sound_prefix .. I .. Y .. SH else -- hollow check(3, {W, Y}) if c(3) == W then vocalized = MA .. c(2) .. UU .. c(4) else vocalized = MA .. c(2) .. II .. c(4) end end end end elseif vform == "II" or vform == "V" or vform == "XII" or vform == "XIII" or vform == "Iq" or vform == "IIq" or vform == "IIIq" then local sound_prefix, expected_length if vform == "II" then sound_prefix = MU .. c(2) .. A .. c(3) .. SH expected_length = 4 elseif vform == "V" then check(2, T) sound_prefix = MU .. T .. A .. c(3) .. A .. c(4) .. SH expected_length = 5 elseif vform == "XII" then -- e.g. [[احدودب]] "to be or become convex or humpbacked", مُحْدَوْدِب (active); -- [[اثنونى]] "to be bent; to be doubled up", مُثْنَوْنٍ (active) check(4, W) if c(3) ~= c(5) then err(("third letter %s should be the same as the fifth letter %s"):format(c(3), c(5))) end sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. W .. SK .. c(5) expected_length = 6 elseif vform == "XIII" then -- e.g. [[اخروط]] "to get entangled; to extend", مُخْرَوِّط (active), مُخْرَوَّط (passive) check(4, W) sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. W .. SH expected_length = 5 elseif vform == "Iq" then sound_prefix = MU .. c(2) .. A .. c(3) .. SK .. c(4) expected_length = 5 elseif vform == "IIq" then check(2, T) sound_prefix = MU .. T .. A .. c(3) .. A .. c(4) .. SK .. c(5) expected_length = 6 elseif vform == "IIIq" then -- e.g. [[اخرنطم]] "to be proud and angry" check(4, T) sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. N .. SK .. c(5) expected_length = 6 else error("Internal error: Unhandled verb form " .. vform) end if len == expected_length - 1 then -- active final-weak if not is_active then err(("length-%s participle only allowed for active participles"):format(len)) end check_weakness({"final-weak", "assimilated+final-weak"}, "allow missing") vocalized = sound_prefix .. IN else handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "III" or vform == "VI" then local sound_prefix, expected_length if vform == "VI" then check(2, T) check(4, ALIF) sound_prefix = MU .. T .. A .. c(3) .. AA .. c(5) expected_length = 6 else sound_prefix = MU .. c(2) .. AA .. c(4) expected_length = 5 end if len == expected_length - 1 then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. SH else vocalized = sound_prefix .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. SH end else handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "IV" or vform == "X" then -- form IV: -- sound: مُرْسِخ (active, "entrenching"), مُرْسَخ (passive, "entrenched") -- first-hamzated (like sound): مُؤْيِس (active, "causing to despair"), مُؤْيَس (passive, "caused to despair") -- final-weak: مُكْرٍ (active, "renting out"), مُكْرًى (passive, "rented out") -- assimilated: مُورِد (active, "transferring"), مُورَد (passive, "transferred"); same when first-Y, e.g. -- أَيْقَنَ "to be certain of": مُوقِن (active), مُوقَن (passive) -- assimilated + final-weak: مُورٍ (active, "setting fire, kindling"), مُورًى (passive, "set fire, kindled") -- geminate: مُمِدّ (active, "granting, helping"), مُمَدّ (passive, "granted, helped") -- hollow: مُزِيل (active, "eliminating"), مُزَال (passive, "eliminated") -- hollow + final-weak: مُعْيٍ (active, "tiring"), مُعْيًى (passive, "tired") local sound_prefix, expected_length if vform == "X" then check(2, S) check(3, T) sound_prefix = MU .. S .. SK .. T .. A .. c(4) expected_length = 6 else sound_prefix = MU .. c(2) expected_length = 4 end if len == expected_length and c(len - 1) == Y and c(len) ~= AMAQ then -- active hollow if not is_active then err("this shape only allowed for active participles") end check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. II .. c(len) elseif len == expected_length and c(len - 1) == ALIF then -- passive hollow if is_active then err("this shape only allowed for passive participles") end check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. AA .. c(len) elseif len == expected_length - 1 then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. I .. c(len) .. SH elseif vform == "IV" and c(2) == W then -- assimilated final-weak vocalized = sound_prefix .. c(len) .. IN else vocalized = sound_prefix .. SK .. c(len) .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. A .. c(len) .. SH end else if vform == "IV" and c(2) == W then -- assimilated, possibly final-weak sound_prefix = sound_prefix .. c(expected_length - 1) else sound_prefix = sound_prefix .. SK .. c(expected_length - 1) end handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "VII" or vform == "VIII" then -- form VII (passive participles are fairly rare but do exist): -- sound: مُنْكَتِب (active "subscribing"), مُنْكَتَب (passive "subscribed") -- geminate: مُنْضَمّ (both active "joining, containing" and passive "joined, contained") -- final-weak: مُنْطَلٍ (active "fooling (someone)"), مُنْطَلًى (passive "fooled") -- final-weak with medial wāw: مُنْطَوٍ (active "involving"), مُنْطَوًى (passive "involved") -- hollow: مُنْقَاد (both active "complying with" and passive "complied with") -- -- for form VIII, the same variants exist but things are complicated by assimilations involving the template T. -- sound third-hamzated no assimilation: مُبْتَدِئ (active "beginning"), مُبْتَدَأ (passive "begun") -- geminate no assimilation: مُبْتَزّ (both active "robbing" and passive "robbed") -- final-weak no assimilation: مُبْتَنٍ (active "building"), مُبْتَنًى (passive "built") -- final-weak with medial wāw no assimilation: مُحْتَوٍ (active "containing"), مُحْتَوًى (passive "contained") -- hollow no assimilation: مُخْتَار (both active "choosing" and passive "chosen") -- -- sound with total assimilation: مُتَّبِع (active "following"), مُتَّبَع (passive "followed") -- sound with total assimilation, assimilating wāw: مُتَّعِد (active "threatening"), مُتَّعَد (passive "threatened") -- sound with total assimilation, irregularly assimilating hamza: مُتَّخِذ (active "taking"), مُتَّخَذ (passive "taken") -- sound with total assimilation (to ḏāl, producing dāl): مُدَّخِر (active "reserving"), مُدَّخَر (passive "reserved") -- sound with total assimilation (to ḏāl): مُذَّكِر (active "remembering"), مُذَّكَر (passive "remembered") -- sound with total assimilation (to ṭāʔ): مُطَّرِح (active "discarding"), مُطَّرَح (passive "discarded") -- sound with total assimilation (to ẓāʔ): مُظَّلِم (active "tolerating"), مُظَّلَم (passive "tolerated") -- final-weak with total assimilation, assimilating wāw: مُتَّقٍ (active "guarding against"), مُتَّقًى (passive "guarded against") -- final-weak with total assimilation (to ṯāʔ): مُثَّنٍ (active "undulating"), مُثَّنًى (passive "undulated") -- final-weak with total assimilation (to dāl): مُدَّعٍ (active "claiming"), مُدَّعًى (passive "claimed") -- sound with partial assimilation (to zayn): مُزْدَهِر (active "thriving"), مُزْدَهَر (passive "thrived") -- sound with medial wāw with partial assimilation (to zayn): مُزْدَوِج (active "appearing twice") -- sound with partial assimilation (to ṣād): مُصْطَبِح (active "illuminating"), مُصْطَبَح (passive, "illuminated") -- sound with partial assimilation (to ḍād): مُضْطَرِب (active "to be disturbed"; no passive) -- geminate with partial assimilation (to ṣād): مُصْطَبّ (both active "effusing" and passive "effused") -- geminate with partial assimilation (to ḍād): مُضْطَرّ (both active "forcing" and passive "forced") -- final-weak with partial assimilation (to ṣād): مُصْطَلٍ (active "warming"), مُصْطَلًى (passive "warmed") -- hollow with partial assimilation (to zayn): مُزْدَاد (both active "increasing" and passive "increased") -- hollow with partial assimilation (to ṣad): مُصْطَاد (both active "hunting" and passive "hunted") local sound_prefix, sufind if vform == "VII" then check(2, N) sound_prefix = MU .. N .. SK .. c(3) sufind = 4 else local c2 = c(2) if c2 == T or c2 == "د" or c2 == "ث" or c2 == "ذ" or c2 == "ط" or c2 == "ظ" then -- full assimilation sound_prefix = MU .. c2 .. SH sufind = 3 else -- partial or no assimilation if c2 == "ز" then check(3, "د") elseif c2 == "ص" or c2 == "ض" then check(3, "ط") else check(3, T) end sound_prefix = MU .. c2 .. SK .. c(3) sufind = 4 end end if c(sufind) == ALIF then -- hollow, active or passive check_len(sufind + 1, sufind + 1) check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. AA .. c(sufind + 1) elseif len == sufind then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. A .. c(len) .. SH else vocalized = sound_prefix .. A .. c(len) .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. A .. c(len) .. SH end else sound_prefix = sound_prefix .. A .. c(sufind) handle_possibly_final_weak(sound_prefix, sufind + 1) end elseif vform == "IX" then check_len(4, 4) vocalized = MU .. c(2) .. SK .. c(3) .. A .. c(4) .. SH elseif vform == "IVq" then -- e.g. [[اذلعب]] "to scamper away", مُذْلَعِبّ (active), مُذْلَعَبّ (passive); -- [[اطمأن]] "to remain quietly; to be certain", مُطْمَئِنّ (active), مُطْمَأَنّ (passive) check_len(5, 5) local sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. c(4) if is_active then vocalized = sound_prefix .. I .. c(5) .. SH else vocalized = sound_prefix .. A .. c(5) .. SH end elseif vform == "XI" then check_len(5, 5) check(4, ALIF) vocalized = MU .. c(2) .. SK .. c(3) .. AA .. c(5) .. SH -- e.g. [[احمار]] "to turn red, to blush", مُحْمَارّ (active) elseif vform == "XIV" or vform == "XV" then -- FIXME: Implement. No examples in Wiktionary currently; need to look up in a grammar. error("Support for verb form " .. vform .. " not implemented yet") else error("Don't recognize verb form " .. vform) end vocalized = rsub(vocalized, HAMZA .. AA, AMAD) local reconstructed_headword = lang:makeEntryName(vocalized) if reconstructed_headword ~= orig_headword then error(("Internal error: Vocalized participle %s doesn't match original participle %s"):format( vocalized, orig_headword)) end -- Apply NFC-style normalization to match MediaWiki's automatic normalization return apply_nfc_shadda(vocalized) end function export.infer_participle_vocalization_json(frame) local iparams = { [1] = {required = true}, [2] = {required = true}, ["weakness"] = {}, ["passive"] = {type = "boolean"} } local iargs = require("Module:parameters").process(frame.args, iparams) return export.infer_participle_vocalization(iargs[1], iargs[2], iargs.weakness, not iargs.passive) end return export rncevvs4tewlz1ionuietua8gvhhfap 1158942 1158927 2026-09-10T05:22:35Z ForzaGreen 28665 both parked decisions: p154 (يَمَنَ) recorded corrected, 427 (اِهْبَيَّخَ) extracted after adding its نون assimilation; Template A 83/83, 381 paradigms / 13040 cells 1158942 Scribunto text/plain local export = {} --[=[ This module implements {{ar-conj}} and provides the underlying conjugation functions for {{ar-verb}} (whose actual formatting is done in [[Module:ar-headword]]). Author: User:Benwing, from an early version (2013-2014) by User:Atitarev, User:ZxxZxxZ. ]=] --[=[ TERMINOLOGY: -- "slot" = A particular combination of tense/mood/person/number/etc. Example slot names for verbs are "past_1s" (past tense first-person singular), "juss_pass_3fp" (non-past jussive passive third-person feminine plural) "ap" (active participle). Each slot is filled with zero or more forms. -- "form" = The conjugated Arabic form representing the value of a given slot. -- "lemma" = The dictionary form of a given Arabic term. For Arabic, normally the third person masculine singular past, although other forms may be used if this form is missing (e.g. in passive-only verbs or verbs lacking the past). ]=] --[=[ FIXME: 1. Finish unimplemented conjugation types. Only IX-final-weak left (extremely rare, possibly only one verb اِعْمَايَ (according to Haywood and Nahmad p. 244, who are very specific about the irregular occurrence of alif + yā instead of expected اِعْمَيَّ with doubled yā). Not in Hans Wehr. NOTE: Not true about this, cf. form IX اِرْعَوَى "to desist, to repent, to see the light". Also note form XII اِخْضَوْضَرَ = form IX اِخْضَرَّ "to be or become green". [DONE except for اِعْمَايَ] 2. Implement irregular verbs as special cases and recognize them, e.g. -- laysa "to not be"; only exists in the past tense, no non-past, no imperative, no participles, no passive, no verbal noun. Irregular alternation las-/lays-. [IMPLEMENTABLE USING OVERRIDES] -- istaḥā yastaḥī "be ashamed of" -- this is complex according to Hans Wehr because there are two verbs, regular istaḥyā yastaḥyī "to spare (someone)'s life" and irregular istaḥyā yastaḥyī "to be ashamed to face (someone)", which is irregular because it has the alternate irregular form istaḥā yastaḥī which only applies to this meaning. Currently we follow Haywood and Nahmad in saying that both varieties can be spelled istaḥyā/istaḥā/istaḥḥā, but we should instead use a variant= param similar to حَيَّ to distinguish the two possibilities, and maybe not include istaḥḥā. -- ʿayya/ʿayiya yaʿayyu/yaʿyā "to not find the right way, be incapable of, stammer, falter, fall ill". This appears to be a mixture of a geminate and final-weak verb. Unclear what the whole paradigm looks like. Do the consonant-ending parts in the past follow the final-weak paradigm? Is it the same in the non-past? Or can you conjugate the non-past fully as either geminate or final-weak? -- اِنْمَحَى inmaḥā or يمَّحَى immaḥā "to be effaced, obliterated; to disappear, vanish" has irregular assimilation of inm- to imm- as an alternative. inmalasa "to become smooth; to glide; to slip away; to escape" also has immalasa as an alternative. The only other form VII verbs in Hans Wehr beginning with -m- are inmalaḵa "to be pulled out, torn out, wrenched" and inmāʿa "to be melted, to melt, to dissolve", which are not listed with imm- alternatives, but might have them; if so, we should handle this generally. [DONE] -- يَرَعَ yaraʕa yariʕu "to be a coward, to be chickenhearted" as an alternative form of يَرِعَ yariʕa yayraʕu (as given in Wehr). [IMPLEMENTABLE USING OVERRIDES] 3. Implement individual override parameters for each paradigm part. See Module:fro-verb for an example of how to do this generally. Note that {{temp|ar-conj-I}} and other of the older templates already had such individual override params. [DONE] Irregular verbs already implemented: -- [ḥayya/ḥayiya yaḥyā "live" -- behaves like a normal final-weak verb (e.g. past first singular ḥayītu) except in the past-tense parts with vowel-initial endings (all the third person except for the third feminine plural). The normal singular and dual endings have -yiya- in them, which compresses to -yya-, with the normal endings the less preferred ones. In masculine third plural, expected ḥayū is replaced by ḥayyū by analogy to the -yy- parts, and the regular form is not given as an alternant in John Mace. Barron's 201 verbs appears to have the regular ḥayū as the part, however. Note also that final -yā appears with tall alif. This appears to be a spelling convention of Arabic, also applying in ḥayyā (form II, "to keep (someone) alive") and 'aḥyā (form IV, "to animate, revive, give birth to, give new life to").] -- implemented -- [ittaxadha yattaxidhu "take"] -- implemented -- [sa'ala yas'alu "ask" with alternative jussive/imperative yasal/sal] -- implemented -- [ra'ā yarā "see"] -- implemented -- ['arā yurī "show"] -- implemented -- ['akala ya'kulu "eat" with imperative kul] -- implemented -- ['axadha ya'xudhu "take" with imperative xudh] -- implemented -- ['amara ya'muru "order" with imperative mur] -- implemented --]=] local force_cat = false -- set to true for debugging -- if true, always maintain manual translit during processing, and compare against full translit at the end local debug_translit = false local lang = require("Module:languages").getByCode("ar") local m_links = require("Module:links") local m_string_utilities = require("Module:string utilities") local m_table = require("Module:table") local ar_utilities = require("Module:ar-utilities") local iut = require("Module:inflection utilities") local put = require("Module:parse utilities") local list_to_text = mw.text.listToText local rfind = m_string_utilities.find local rsubn = m_string_utilities.gsub local rmatch = m_string_utilities.match local rsplit = m_string_utilities.split local usub = m_string_utilities.sub local ulen = m_string_utilities.len local u = m_string_utilities.char local unpack = unpack or table.unpack -- Lua 5.2 compatibility local dump = mw.dumpObject -- Within this module, conjugations are the functions that do the actual -- conjugating by creating the parts of a basic verb. -- They are defined further down. local conjugations = {} -- hamza variants local HAMZA = u(0x0621) -- hamza on the line (stand-alone hamza) = ء local HAMZA_ON_ALIF = u(0x0623) local HAMZA_ON_W = u(0x0624) local HAMZA_UNDER_ALIF = u(0x0625) local HAMZA_ON_Y = u(0x0626) local HAMZA_ANY = "[" .. HAMZA .. HAMZA_ON_ALIF .. HAMZA_UNDER_ALIF .. HAMZA_ON_W .. HAMZA_ON_Y .. "]" local HAMZA_PH = u(0xFFF0) -- hamza placeholder local BAD = u(0xFFF1) local BORDER = u(0xFFF2) local KEEP_W_PH = u(0xFFF3) -- retained-wāw placeholder; see postprocess_term() local KEEP_HAMZA_PH = u(0xFFF4) -- quiescent-hamza placeholder; see postprocess_term() local WAW_SEAT_PH = u(0xFFF5) -- pre-seated hamza-on-wāw placeholder; see postprocess_term() local NO_MADDA_PH = u(0xFFF6) -- hamza held back from the madda rule; see postprocess_term() -- diacritics local A = u(0x064E) -- fatḥa local AN = u(0x064B) -- fatḥatān (fatḥa tanwīn) local U = u(0x064F) -- ḍamma local UN = u(0x064C) -- ḍammatān (ḍamma tanwīn) local I = u(0x0650) -- kasra local IN = u(0x064D) -- kasratān (kasra tanwīn) local SK = u(0x0652) -- sukūn = no vowel local SH = u(0x0651) -- šadda = gemination of consonants local DAGGER_ALIF = u(0x0670) local DIACRITIC_ANY_BUT_SH = "[" .. A .. I .. U .. AN .. IN .. UN .. SK .. DAGGER_ALIF .. "]" -- Pattern matching short vowels local AIU = "[" .. A .. I .. U .. "]" -- Pattern matching short vowels or sukūn local AIUSK = "[" .. A .. I .. U .. SK .. "]" -- Pattern matching any diacritics that may be on a consonant local DIACRITIC = SH .. "?" .. DIACRITIC_ANY_BUT_SH -- translit_patterns local vowels = "aeiouāēīōū" local NV = "[^" .. vowels .. "]" local dia = {a = A, i = I, u = U} local undia = {[A] = "a", [I] = "i", [U] = "u", ["-"] = "-"} -- various letters and signs local ALIF = u(0x0627) -- ʾalif = ا local AMAQ = u(0x0649) -- ʾalif maqṣūra = ى local AMAD = u(0x0622) -- ʾalif madda = آ local TAM = u(0x0629) -- tāʾ marbūṭa = ة local T = u(0x062A) -- tāʾ = ت local HYPHEN = u(0x0640) local N = u(0x0646) -- nūn = ن local W = u(0x0648) -- wāw = و local Y = u(0x064A) -- yāʾ = ي local S = "س" local M = "م" local LRM = u(0x200e) -- left-to-right mark -- common combinations local AH = A .. TAM local AT = A .. T local AA = A .. ALIF local AAMAQ = A .. AMAQ local AAH = AA .. TAM local AAT = AA .. T local II = I .. Y local UU = U .. W local AY = A .. Y local AW = A .. W local AYSK = AY .. SK local AWSK = AW .. SK local NA = N .. A local NI = N .. I local AAN = AA .. N local AANI = AA .. NI local AYNI = AYSK .. NI local AWNA = AWSK .. NA local AYNA = AYSK .. NA local AYAAT = AY .. AAT local UNU = "[" .. UN .. U .. "]" local MA = M .. A local MU = M .. U local TA = T .. A local TU = T .. U local _I = ALIF .. I local _U = ALIF .. U --W: Morphological pattern (الوزن الصرفي) to verb form mapping local MORPHOLOGICAL_PATTERN_TO_FORM = { ["فعَل يفعُل"] = "I/a~u", ["فعَل يفعِل"] = "I/a~i", ["فعَل يفعَل"] = "I/a~a", ["فعُل يفعُل"] = "I/u~u", ["فعِل يفعَل"] = "I/i~a", ["فعِل يفعِل"] = "I/i~i", ["فعّل"] = "II", ["فاعل"] = "III", ["أفعل"] = "IV", ["تفعّل"] = "V", ["تفاعل"] = "VI", ["انفعل"] = "VII", ["افتعل"] = "VIII", ["افعلّ"] = "IX", ["استفعل"] = "X", ["افعالّ"] = "XI", ["افعوعل"] = "XII", ["افعوّل"] = "XIII", ["فعلل"] = "Iq", ["تفعلل"] = "IIq", ["افعنلل"] = "IIIq", -- "XIV" should not be used ["افعللّ"] = "IVq", } --W: Create reverse mapping from form to morphological pattern local FORM_TO_MORPHOLOGICAL_PATTERN = {} for pattern, form in pairs(MORPHOLOGICAL_PATTERN_TO_FORM) do FORM_TO_MORPHOLOGICAL_PATTERN[form] = pattern end --W: local MORPHOLOGICAL_PATTERN_TO_BASIC_DERIV = { ["فعَل يفعُل"] = "مُجرَّد", ["فعَل يفعِل"] = "مُجرَّد", ["فعَل يفعَل"] = "مُجرَّد", ["فعُل يفعُل"] = "مُجرَّد", ["فعِل يفعَل"] = "مُجرَّد", ["فعِل يفعِل"] = "مُجرَّد", ["فعّل"] = "مزيد بحرف", ["فاعل"] = "مزيد بحرف", ["أفعل"] = "مزيد بحرف", ["تفعّل"] = "مزيد بحرفين", ["تفاعل"] = "مزيد بحرفين", ["انفعل"] = "مزيد بحرفين", ["افتعل"] = "مزيد بحرفين", ["افعلّ"] = "مزيد بحرفين", ["استفعل"] = "مزيد بثلاثة أحرف", ["افعالّ"] = "مزيد بثلاثة أحرف", ["افعوعل"] = "مزيد بثلاثة أحرف", ["افعوّل"] = "مزيد بثلاثة أحرف", ["فعلل"] = "مُجرَّد", ["تفعلل"] = "مزيد بحرف", ["افعنلل"] = "مزيد بحرفين", ["افعللّ"] = "مزيد بحرفين", } local translit_cache = { -- hamza variants [HAMZA] = "ʔ", [HAMZA_ON_ALIF] = "ʔ", [HAMZA_ON_W] = "ʔ", [HAMZA_UNDER_ALIF] = "ʔ", [HAMZA_ON_Y] = "ʔ", [HAMZA_PH] = "ʔ", -- diacritics [A] = "a", [AN] = "an", [U] = "u", [UN] = "un", [I] = "i", [IN] = "in", [SK] = "", [SH] = "*", -- handled specially [DAGGER_ALIF] = "ā", -- various letters and signs [""] = "", [ALIF] = BAD, -- we should never be transliterating ALIF by itself, as its translit in isolation is ambiguous [AMAQ] = BAD, [AMAD] = "ʔā", [TAM] = "", [T] = "t", [N] = "n", [W] = "w", [Y] = "y", [S] = "s", [M] = "m", [LRM] = "", -- common combinations [AH] = "a", [AT] = "at", [AA] = "ā", [AAMAQ] = "ā", [AAH] = "āh", [AAT] = "āt", [II] = "ī", [UU] = "ū", [AY] = "ay", [AW] = "aw", [AYSK] = "ay", [AWSK] = "aw", [NA] = "na", [NI] = "ni", [AAN] = "ān", [AANI] = "āni", [AYNI] = "ayni", [AWNA] = "awna", [AYNA] = "ayna", [AYAAT] = "ayāt", [MA] = "ma", [MU] = "mu", [TA] = "ta", [TU] = "tu", [_I] = "i", [_U] = "u", } local function transliterate(text) local cached = translit_cache[text] if cached then if cached == BAD then error(("Internal error: Unable to transliterate %s because explicitly marked as BAD"):format(text)) end return cached end local tr = (lang:transliterate(text)) if not tr then error(("Internal error: Unable to transliterate: %s"):format(text)) end translit_cache[text] = tr return tr end local all_person_number_list = { "1s", "2ms", "2fs", "3ms", "3fs", "2d", "3md", "3fd", "1p", "2mp", "2fp", "3mp", "3fp" } local function make_person_number_slot_accel_list(list) local slot_accel_list = {} return slot_accel_list end local imp_person_number_list = {} for _, pn in ipairs(all_person_number_list) do if pn:find("^2") then table.insert(imp_person_number_list, pn) end end local passive_types = m_table.listToSet { "pass", -- verb has both active and passive "ipass", -- verb is active with impersonal passive "nopass", -- verb is active-only "onlypass", -- verb is passive-only "onlypass-impers", -- verb itself is impersonal, meaning passive-only with impersonal passive } local indicator_flags = m_table.listToSet { "nopast", "no_nonpast", "noimp", "nocat", -- don't categorize or include annotations about this; useful in suppletive parts of verbs "reduced", -- verb has assimilation/reduction of initial coronals "altgem", -- form X with alternative past geminate forms with final-weak endings } export.potential_lemma_slots = {"past_3ms", "past_pass_3ms", "ind_3ms", "ind_pass_3ms", "imp_2ms"} export.unsettable_slots = {} for _, potential_lemma_slot in ipairs(export.potential_lemma_slots) do table.insert(export.unsettable_slots, potential_lemma_slot .. "_linked") end -- We don't set the active participle directly for form I because we don't want stative verbs (with past vowel i or u) -- to default to فَاعِل. Instead we set the special slot 'ap1' and later copy it to 'ap' for non-stative verbs. The user -- meanwhile can explicitly request the فَاعِل form for active participles for stative verbs using `ap:+`. table.insert(export.unsettable_slots, "ap1") -- primary default فَاعِل for form I active participles table.insert(export.unsettable_slots, "ap2") -- secondary default فَعِيل for form I active participles (stative I) table.insert(export.unsettable_slots, "ap3") -- secondary default فَعِل for form I active participles (stative II) table.insert(export.unsettable_slots, "apcd") -- secondary default أَفْعَل for form I active participles (color/defect) table.insert(export.unsettable_slots, "apan") -- secondary default فَعْلَان for form I active participles (in -ān) table.insert(export.unsettable_slots, "pp2") -- secondary default فَعِيل for form I passive participles (same as ap2) table.insert(export.unsettable_slots, "vn2") -- secondary default فِعَال for form III verbal nouns export.unsettable_slots_set = m_table.listToSet(export.unsettable_slots) local default_indicator_to_active_participle_slot = { ["+"] = "ap1", ["++"] = "ap2", ["+++"] = "ap3", ["+cd"] = "apcd", ["+an"] = "apan", } local slots_that_may_be_uncertain = { vn = "verbal noun", ap = "active participle", } -- Initialize all the slots for which we generate forms. local function add_slots(alternant_multiword_spec) alternant_multiword_spec.verb_slots = { {"ap", "act|part"}, {"pp", "pass|part"}, {"vn", "vnoun"}, } for _, unsettable_slot in ipairs(export.unsettable_slots) do table.insert(alternant_multiword_spec.verb_slots, {unsettable_slot, "-"}) end -- Add entries for a slot with person/number variants. -- `slot_prefix` is the prefix of the slot, typically specifying the tense/aspect. -- `tag_suffix` is a string listing the set of inflection tags to add after the person/number tags. -- `person_number_list` is a list of the person/number slot suffixes to add to `slot_prefix`. local function add_personal_slot(slot_prefix, tag_suffix, person_number_list) for _, persnum in ipairs(person_number_list) do local slot = slot_prefix .. "_" .. persnum local accel = persnum:gsub("(.)", "%1|") .. tag_suffix table.insert(alternant_multiword_spec.verb_slots, {slot, accel}) end end local tenses = { {"past", "past|%s"}, {"ind", "non-past|%s|ind"}, {"sub", "non-past|%s|sub"}, {"juss", "non-past|%s|juss"}, } for _, slot_accel in ipairs(tenses) do local slot, accel = unpack(slot_accel) for _, voice in ipairs {"act", "pass"} do add_personal_slot(voice == "act" and slot or slot .. "_pass", accel:format(voice), all_person_number_list) end end add_personal_slot("imp", "imp", imp_person_number_list) alternant_multiword_spec.verb_slots_map = {} for _, slot_accel in ipairs(alternant_multiword_spec.verb_slots) do local slot, accel = unpack(slot_accel) alternant_multiword_spec.verb_slots_map[slot] = accel end end local overridable_stems = {} local slot_override_param_mods = { footnote = { item_dest = "footnotes", store = "insert", }, alt = {}, t = { -- [[Module:links]] expects the gloss in "gloss". item_dest = "gloss", }, gloss = {}, g = { -- [[Module:links]] expects the genders in "g". `sublist = true` automatically splits on comma (optionally -- with surrounding whitespace). item_dest = "genders", sublist = true, }, pos = {}, lit = {}, id = {}, --W: The q/qq/l/ll (qualifier and label) modifiers were removed: rendering them -- needs [[Module:pron qualifier]], which does not exist on ar.wiktionary, so any -- use of them raised a script error. Accepting them here again requires importing -- that module first and restoring the block in generate_link() below. } local function generate_obj(formval, parse_err, prefix, is_slot_override) local val, uncertain = formval:match("^(.*)(%?)$") val = val or formval uncertain = not not uncertain local ar, translit = val:match("^(.*)//(.*)$") if not ar then ar = val end if ar == "" then if uncertain then ar = "?" else error(("Can't specify blank value for override for %s override '%s'"):format( is_slot_override and "slot" or "stem", prefix)) end end return {form = ar, translit = translit, uncertain = uncertain} end local function parse_inline_modifiers(comma_separated_group, parse_err, prefix, is_slot_override) local function this_generate_obj(formval, parse_err) return generate_obj(formval, parse_err, prefix, is_slot_override) end return put.parse_inline_modifiers_from_segments { group = comma_separated_group, props = { param_mods = slot_override_param_mods, parse_err = parse_err, generate_obj = this_generate_obj, pre_normalize_modifiers = function(data) local modtext = data.modtext modtext = modtext:match("^(%[.*%])$") if modtext then return ("<footnote:%s>"):format(modtext) end return data.modtext end, }, } end local function allow_multiple_values_for_override(comma_separated_groups, data, is_slot_override) local retvals = {} for _, comma_separated_group in ipairs(comma_separated_groups) do local retval if is_slot_override then --W: pass `prefix` and `is_slot_override` through; without them the error -- raised for a blank override value formatted a nil prefix, which is a -- crash on Lua 5.1 rather than the intended message. Reachable from e.g. -- <I/a~u.vn:[[كتابة]]>, where the [[...]] is consumed as a footnote and -- leaves the value empty. retval = parse_inline_modifiers(comma_separated_group, data.parse_err, data.prefix, is_slot_override) else retval = generate_obj(comma_separated_group[1], data.parse_err, data.prefix, is_slot_override) retval.footnotes = data.fetch_footnotes(comma_separated_group) end table.insert(retvals, retval) end for _, form in ipairs(retvals) do if form.form == "+" or default_indicator_to_active_participle_slot[form.form] then if form.form ~= "+" and default_indicator_to_active_participle_slot[form.form] and not is_slot_override then error(("Stem override '%s' cannot use %s to request a secondary default"):format( data.prefix, form.form)) end data.base.slot_override_uses_default[data.prefix] = true end end for _, form in ipairs(retvals) do if form.form == "-" then data.base.slot_explicitly_missing[data.prefix] = true break end end if data.base.slot_explicitly_missing[data.prefix] then for _, form in ipairs(retvals) do if form.form ~= "-" then data.parse_err(("For slot or stem '%s', saw both - and a value other than -, which isn't allowed"): format(data.prefix)) end end return nil end return retvals end local function simple_choice(choices) return function(separated_groups, data) if #separated_groups > 1 then data.parse_err("For spec '" .. data.prefix .. ":', only one value currently allowed") end if #separated_groups[1] > 1 then data.parse_err("For spec '" .. data.prefix .. ":', no footnotes currently allowed") end local choice = separated_groups[1][1] if not m_table.contains(choices, choice) then data.parse_err("For spec '" .. data.prefix .. ":', saw value '" .. choice .. "' but expected one of '" .. table.concat(choices, "،") .. "'") end return choice end end for _, overridable_stem in ipairs { "past", "past_v", "past_c", "past_pass", "past_pass_v", "past_pass_c", "nonpast", "nonpast_v", "nonpast_c", "nonpast_pass", "nonpast_pass_v", "nonpast_pass_c", "imp", "imp_v", "imp_c", } do overridable_stems[overridable_stem] = allow_multiple_values_for_override end overridable_stems.past_final_weak_vowel = simple_choice { "ay", "aw", "ī", "ū" } overridable_stems.past_pass_final_weak_vowel = simple_choice { "ay", "aw", "ī", "ū" } overridable_stems.nonpast_final_weak_vowel = simple_choice { "ā", "ī", "ū" } overridable_stems.nonpast_pass_final_weak_vowel = simple_choice { "ā", "ī", "ū" } ------------------------------------------------------------------------------- -- Utility functions -- ------------------------------------------------------------------------------- -- version of rsubn() that discards all but the first return value local function rsub(term, foo, bar) return (rsubn(term, foo, bar)) end -- version of rsubn() that returns a 2nd argument boolean indicating whether a substitution was made. local function rsubb(term, foo, bar) local retval, nsubs = rsubn(term, foo, bar) return retval, nsubs > 0 end -- Concatenate one or more strings or form objects. local function q(...) local not_all_strings = debug_translit local has_manual_translit = debug_translit for i = 1, select("#", ...) do local argt = select(i, ...) if not argt then error(("Internal error: Saw nil at index %s: %s"):format(i, dump({...}))) end if type(argt) ~= "string" then not_all_strings = true if argt.translit then has_manual_translit = true break end end end if not not_all_strings then -- just strings, concatenate directly return table.concat({...}) end local formvals = {} local translit = has_manual_translit and {} or nil local footnotes for i = 1, select("#", ...) do local argt = select(i, ...) if type(argt) == "string" then formvals[i] = argt if has_manual_translit then translit[i] = transliterate(argt) end else formvals[i] = argt.form if has_manual_translit then translit[i] = argt.translit or transliterate(argt.form) end footnotes = iut.combine_footnotes(footnotes, argt.footnotes) end end -- FIXME: Do we want to support other properties? return { form = table.concat(formvals), translit = has_manual_translit and table.concat(translit) or nil, footnotes = footnotes, } end -- Return the formval associated with `rad` (a radical or past/non-past vowel, either a string or form object). local function rget(rad) if type(rad) == "string" then return rad elseif type(rad) == "table" then return rad.form else error(("Internal error: Unexpected type for radical or past/non-past vowel: %s"):format(dump(rad))) end end export.rget = rget -- for use in [[Module:ar-headword]] -- Return the footnotes associated with `rad` (a radical or past/non-past vowel, either a string or form object). local function rget_footnotes(rad) if type(rad) == "string" then return nil elseif type(rad) == "table" then return rad.footnotes else error(("Internal error: Unexpected type for radical or past/non-past vowel: %s"):format(dump(rad))) end end -- Return true if the formval associated with `rad` (a radical or past/non-past vowel, either a string or form object) -- is `val`. local function req(rad, val) return rget(rad) == val end -- True if a form-I verb whose first radical is wāw elides that wāw in the non-past (يَعِدُ for -- *يَوْعِدُ) rather than keeping it (يَوْجَلُ). The elision belongs to the kasra non-past -- a~i -- وَعَدَ يَعِدُ, i~i وَرِثَ يَرِثُ -- and to the fatḥa non-past that is a throat-letter lowering of -- that same underlying kasra: a~a وَضَعَ يَضَعُ, وَهَبَ يَهَبُ. Everywhere else the wāw stays: -- i~a وَجِلَ يَوْجَلُ، وَنِيَ يَوْنَى (a fatḥa non-past that is the regular partner of a kasra past, -- not a lowered kasra) -- *~u وَجَلَ يَوْجُلُ، وَجُهَ يَوْجُهُ (a ḍamma non-past is never a lowered kasra) -- The two elided-anyway lexemes with a kasra past (وَسِعَ يَسَعُ, وَطِئَ يَطَأُ) are lexical and are -- what the explicit `I-assimilated` weakness exists to express -- see the comment on it in -- detect_indicator_spec(). -- -- The u-non-past half of this is attested directly: أنطوان الدحداح، معجم تصريف الأفعال العربية -- prints يَوْجُلُ/اُوْجُلْ for وَجَلَ (ص136) and says so in the page's own note -- "لا تُحذَفُ واوُ -- الفعلِ في المُضارعِ ولا تُقلَبُ في الأمرِ" -- and يَوْنَى/اِوْنَ for وَنِيَ (ص230). -- -- Returns true when the vowels aren't known, which keeps the historical default for callers -- that have radicals but no vowel pattern. local function form_i_w_assimilated(past_vowel, nonpast_vowel) if not past_vowel or not nonpast_vowel then return true end return req(nonpast_vowel, I) or req(nonpast_vowel, A) and req(past_vowel, A) end --W: Geminate form-I roots whose degemination vowel follows the مضارع by إتباع. -- When gemination breaks under a consonant-initial suffix, the vowel that surfaces on the -- second radical (مَدَدْتَ's second د) is normally the verb's own past vowel, which the وزن -- fixes at one value per measure: فتح for فعَل يفعُل (مَدَدْتَ، فَرَرْتَ، عَضَضْتَ), كسر for -- فعِل يفعَل (ظَلِلْتَ). Classical grammar also attests a minority إتباع variant on the -- فعَل يفعُل measure, where that vowel instead follows the مضارع's own ضمة (هَمُمْتَ). Which -- of the two a given verb takes is a fact about the lexeme, not about its وزن -- هَمَّ and -- مَدَّ share the measure and differ here -- so it cannot be derived from جذر + وزن and is -- listed, the same way SOUND_FORM_I_UU_ROOTS lists هَيُؤَ. Keyed rad1..rad2..rad3 and gated -- on the a~u pattern, so a listed root still degeminates normally under any other pattern. local GEMINATE_ITBAA_ROOTS = { ["همم"] = true, -- هَمَّ ـُ: أنطوان الدحداح، معجم تصريف الأفعال العربية، ص192 } -- The vowel that surfaces on the second radical of a geminate form-I verb once gemination -- breaks. `past_vowel` unless the root is an attested إتباع lexeme; see GEMINATE_ITBAA_ROOTS. local function geminate_degemination_vowel(rad1, rad2, rad3, past_vowel, nonpast_vowel) if past_vowel and nonpast_vowel and req(past_vowel, A) and req(nonpast_vowel, U) and GEMINATE_ITBAA_ROOTS[rad1 .. rad2 .. rad3] then return nonpast_vowel end return past_vowel end --W: Form-I final-weak roots that keep rad3 as a consonant in the 2fs non-past and -- imperative (تَسْهُوِينَ، اُسْهُوِي) instead of contracting it away (تَسْهِينَ، اِسْهِي). -- -- This was briefly gated on the فعُل يفعُل measure alone, then narrowed to a root table, -- then widened back to the measure -- and the middle step is why. The argument for the -- measure was always that the measure is what keeps rad3 consonantal in the first place -- (its past 3ms is سَهُوَ, not the سَها an a~u root like دَعَا gives), but at the time the -- corpus held exactly one فعُل يفعُل final-weak page, so there was no second paradigm to -- check it against and the rule it displaces -- contract before the 2fs -ī, as دَعَا does -- -- was the one the module had applied for years. Generalising an unverified reading over a -- whole class of live entries is the wrong direction to be wrong in, so it went back to -- listing سهو alone, keyed and gated exactly like SOUND_FORM_I_UU_ROOTS. -- -- Checkpoint 4 supplied the second page, and it answers the question. p222 (model 147) -- prints TWO فعُل يفعُل final-weak paradigms, حَلُوَ and سَرُوَ, and both retain rad3 in both -- eligible slots: تَحْلُوِينَ / اُحْلُوِي and تَسْرُوِينَ / اُسْرُوِي, the wāw carrying its kasra -- under it, all four read at 6x off the strip against the same columns' يَحْلُونَ / يَسْرُونَ. -- Two independent roots, one page, agreeing with سهو on p204 -- and سَرُوَ is precisely the -- root the narrowing had put back to تَسْرِينَ, so the cautious reading is the one the ink -- refutes. So the gate is the vowel pattern after all and the table is gone. local function final_weak_uu_retains_rad3(rad1, rad2, rad3, past_ending_vowel, nonpast_ending_vowel) return past_ending_vowel == "ū" and nonpast_ending_vowel == "ū" end -- Map `vow` (a past/non-past vowel, either a string or form object without translit) by passing the formval through -- `fn`. Don't call this on radicals because they may have manual translit and it isn't clear how to handle that. local function map_vowel(vow, fn) if type(vow) == "string" then return fn(vow) elseif type(vow) == "table" then return {form = fn(vow.form), footnotes = vow.footnotes} else error(("Internal error: Unexpected type for past/non-past vowel: %s"):format(dump(vow))) end end local function get_radicals_3(vowel_spec) return vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3, vowel_spec.past, vowel_spec.nonpast end local function get_radicals_4(vowel_spec) return vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3, vowel_spec.rad4 end local function is_final_weak(base, vowel_spec) return vowel_spec.weakness == "final-weak" or base.form == "XV" end --W: Create a simple link local function link_term_simple(text) return "[[" .. text .. "]]" end local function link_term(text, face, id) return m_links.full_link({lang = lang, term = text, tr = "-", id = id}, face) end local function tag_text(text, tag, class) return m_links.full_link({lang = lang, alt = text, tr = "-"}) end local function track(page) require("Module:debug/track")("ar-verb/" .. page) return true end local function track_if_ar_conj(base, page) if base.alternant_multiword_spec.source_template == "ar-conj" then require("Module:debug/track")("ar-verb/" .. page) end return true end local function reorder_shadda(word) -- shadda+short-vowel (including tanwīn vowels, i.e. -an -in -un) gets -- replaced with short-vowel+shadda during NFC normalisation, which -- MediaWiki does for all Unicode strings; however, it makes various -- processes inconvenient, so undo it. word = rsub(word, "(" .. DIACRITIC_ANY_BUT_SH .. ")" .. SH, SH .. "%1") return word end local function apply_nfc_shadda(word) -- Apply NFC-style normalization: reorder shadda+vowel to vowel+shadda -- to match MediaWiki's automatic NFC normalization. This reverses the -- internal shadda+vowel format used during processing. word = rsub(word, SH .. "(" .. DIACRITIC_ANY_BUT_SH .. ")", "%1" .. SH) return word end ------------------------------------------------------------------------------- -- Basic functions to inflect tenses -- ------------------------------------------------------------------------------- local function skip_slot(base, slot, allow_overrides) if base.slot_explicitly_missing[slot] then return true end if not allow_overrides and base.slot_overrides[slot] and not base.slot_override_uses_default[slot] then -- Skip any slots for which there are overrides, except those that request the default value using +, ++, etc. return true end if base.passive == "nopass" and (slot == "pp" or slot:find("_pass")) then return true elseif base.passive == "onlypass" and slot ~= "pp" and slot ~= "vn" and not slot:find("_pass") then return true elseif base.passive == "ipass" and slot:find("_pass") and not slot:find("3ms") then return true elseif base.passive == "onlypass-impers" and slot ~= "pp" and slot ~= "vn" and (not slot:find("_pass") or slot:find("_pass") and not slot:find("3ms")) then return true end if base.nopast and slot:find("^past_") then return true end if base.noimp and slot:find("^imp_") then return true end if base.no_nonpast and (slot:find("^ind_") or slot:find("^sub_") or slot:find("^juss")) then return true end return false end local function basic_combine_stem_ending(stem, ending) return stem .. ending end local function basic_combine_stem_ending_tr(stem, ending) return stem .. ending end -- Concatenate `prefixes`, `stems` and `endings` (any of which may be an abbreviate form list, i.e. strings, form -- objects or lists of strings or form objects) and store into `slot`. If a user-supplied override exists for the slot, -- nothing will happen unless `allow_overrides` is provided. local function add3(base, slot, prefixes, stems, endings, allow_overrides) if skip_slot(base, slot, allow_overrides) then return end -- Optimization since the prefixes are almost always single strings. if type(prefixes) == "string" then local function do_combine_stem_ending(stem, ending) return prefixes .. stem .. ending end local function do_combine_stem_ending_tr(stem, ending) return transliterate(prefixes) .. stem .. ending end iut.add_forms(base.forms, slot, stems, endings, do_combine_stem_ending, transliterate, do_combine_stem_ending_tr, base.form_footnotes) else iut.add_multiple_forms(base.forms, slot, {prefixes, stems, endings}, basic_combine_stem_ending, transliterate, basic_combine_stem_ending_tr, base.form_footnotes) end end -- Insert one or more forms in `form_or_forms` into `slot`. `form_or_forms` is an abbreviated form list (see comment at -- top of [[Module:inflection utilities]]). If a user-supplied override exists for the slot, nothing will happen unless -- `allow_overrides` is provided. BEWARE: One form object should never occur in two different slots, or twice in a given -- slot; if taking a form object from an existing slot, make sure to shallowCopy() it. local function insert_form_or_forms(base, slot, form_or_forms, allow_overrides, uncertain) if not skip_slot(base, slot, allow_overrides) then -- Some optimizations of the most common case of inserting a single string. if type(form_or_forms) == "string" and not base.form_footnotes then form_or_forms = {form = form_or_forms, uncertain = uncertain} iut.insert_form(base.forms, slot, form_or_forms) else local list = iut.convert_to_general_list_form(form_or_forms, base.form_footnotes) if uncertain then for _, formobj in ipairs(list) do formobj.uncertain = true end end iut.insert_forms(base.forms, slot, list) end end end -- Insert `string_or_form` into both the ap2 and pp2 slots, shallowCopying a form object to make sure no form objects -- occur in two slots. local function insert_ap2_pp2(base, string_or_form) insert_form_or_forms(base, "ap2", string_or_form) if type(string_or_form) == "table" then string_or_form = m_table.shallowCopy(string_or_form) end insert_form_or_forms(base, "pp2", string_or_form) end -- Convert `stemforms` (a string, a form object, or a list of strings and/or form objects) into "general form" (a list -- of form objects) and map `fn` over the list of objects. `fn` is passed two arguments (form value and translit) and -- should likewise return the new form value and translit. Footnotes will be preserved. FIXME: Preserve other metadata. local function map_general(stemforms, fn) return iut.map_forms(iut.convert_to_general_list_form(stemforms), fn) end -- Similar to map_general() except that `fn` should return a single value (one or more strings or form objects), instead -- of two values (form value and translit), and the resulting value(s) from all calls to `fn` will be flattened to -- construct the overall return value. Footnotes will be preserved. FIXME: Preserve other metadata. local function flatmap_general(stemforms, fn) return iut.flatmap_forms(iut.convert_to_general_list_form(stemforms), fn) end -- Given user-supplied stem overrides in `base`, construct any derived stem overrides (e.g. vowel-specific or -- consonant-specific variants), and truncate initial y-/ي- in any non-past overrides. local function construct_stems(base) local stems = base.stem_overrides stems.past_v = stems.past_v or stems.past stems.past_c = stems.past_c or stems.past stems.past_pass_v = stems.past_pass_v or stems.past_pass stems.past_pass_c = stems.past_pass_c or stems.past_pass stems.nonpast_v = stems.nonpast_v or stems.nonpast stems.nonpast_c = stems.nonpast_c or stems.nonpast stems.nonpast_pass_v = stems.nonpast_pass_v or stems.nonpast_pass stems.nonpast_pass_c = stems.nonpast_pass_c or stems.nonpast_pass stems.imp_v = stems.imp_v or stems.imp stems.imp_c = stems.imp_c or stems.imp local function truncate_nonpast_initial_cons(stem_type, form, translit) if form == "+" then return form, translit end if not form:find("^" .. Y) then error(("Form value %s for stem type '%s' should begin with ي"):format(form, stem_type)) end form = form:gsub("^" .. Y, "") if translit then if not translit:find("^y") then error(("Translit value %s for stem type '%s' should begin with y"):format(translit, stem_type)) end translit = translit:gsub("^y", "") end return form, translit end for _, nonpast_stem_type in ipairs { "nonpast_v", "nonpast_c", "nonpast_pass_v", "nonpast_pass_c" } do if stems[nonpast_stem_type] then stems[nonpast_stem_type] = map_general(stems[nonpast_stem_type], function(form, translit) return truncate_nonpast_initial_cons(nonpast_stem_type, form, translit) end) end end end -- Given user-specified overrides for stem `stemname`, return overrides with occurrences of + replaced by -- `default_stem`. If no overrides, return `default_stem`, or {} if no default. local function override_stem_if_needed(base, stemname, default_stem) local overrides = base.stem_overrides[stemname] if not overrides then return default_stem or {} end return map_general(overrides, function(form, translit) if form ~= "+" and default_indicator_to_active_participle_slot[form] then error(("Stem overrides cannot use secondary default indicators but saw %s in stem override '%s'"):format( form, stemname)) end if form == "+" then if translit then error(("Cannot supply manual translit along with + for stem override '%s'"):format(stemname)) end if not default_stem then error(("Cannot use + for stem override '%s' because no default is available"):format(stemname)) end if type(default_stem) ~= "string" then error(("Internal error: Default stem for '%s' is not a string: %s"):format(stemname, dump(default_stem))) end return default_stem end return form, translit end) end ------------------------------------------------------------------------------- -- Properties of different verbal forms -- ------------------------------------------------------------------------------- local allowed_vforms = {"I", "II", "III", "IV", "V", "VI", "VII", "VIII", "IX", "X", "XI", "XII", "XIII", "XIV", "XV", "Iq", "IIq", "IIIq", "IVq"} local allowed_vforms_set = m_table.listToSet(allowed_vforms) local allowed_vforms_with_weakness = m_table.shallowCopy(allowed_vforms) -- The user needs to be able to explicitly specify that a form-I verb (specifically one whose initial radical is و) is -- sound. Cf. wajiʕa yawjaʕu (not #yajaʕu) "to ache, to hurt". In general, i~a and u~u verbs whose initial radical is و -- seem to not assimilate the first radical; cf. وقح "to be shameless", variously waqaḥa~yaqiḥu, waquḥa~yawquḥu and -- waqiḥa~yawqaḥu, whereas a~i verbs (wafaḍa~yafiḍu "to rush"), i~i verbs (wafiqa~yafiqu "to be proper, to be suitable") -- and a~a verbs (waḍaʕa~yaḍaʕu "to set down, to place") do assimilate. But there are naturally exceptions, e.g. -- waṭiʔa~yaṭaʔu "to tread, to trample"; wasiʕa~yasaʕu "to be spacious; to be well-off"; waṯiʔa~yaṯaʔu "to get bruised, -- to be sprained". Also beware of waniya~yawnā "to be faint; to languish", which is sound in the first radical and -- final-weak in the last radical. Nonetheless, the regularity of the patterns mentioned above suggest we should provide -- them as defaults. -- Note that there are other cases of unexpectedly sound verbs, e.g. izdawaja~yazdawiju "to be in pairs", layisa~yalyasu -- "to be valiant, to be brave", ʔaḥwaja~yuḥwiju "to need", istahwana~yastahwinu "to consider easy", sawisa~yaswasu "to -- be or become moth-eaten or worm-eaten" (vs. sāsa~yasūsu "to govern, to rule" from the same radicals), ʕawira~yaʕwaru -- "to be one-eyed", istajwaba~yastajwibu "to interrogate", etc. But in these cases there is no need for explicit user -- specification as the lemma itself specifies the unexpected soundness. for _, form_with_weakness in ipairs { "I-sound", "I-assimilated", "none-sound", "none-hollow", "none-geminate", "none-final-weak" } do table.insert(allowed_vforms_with_weakness, form_with_weakness) end local allowed_vforms_with_weakness_set = m_table.listToSet(allowed_vforms_with_weakness) local function vform_supports_final_weak(vform) return vform ~= "XI" and vform ~= "XV" and vform ~= "IVq" end local function vform_supports_geminate(vform) return vform == "I" or vform == "III" or vform == "IV" or vform == "VI" or vform == "VII" or vform == "VIII" or vform == "X" end local function vform_supports_hollow(vform) return vform == "I" or vform == "IV" or vform == "VII" or vform == "VIII" or vform == "X" end local function vform_probably_impersonal_passive(vform, weakness, past_vowel, nonpast_vowel) return vform == "I" and req(past_vowel, I) or vform == "V" or vform == "VI" or vform == "X" or vform == "IIq" end local function vform_probably_full_passive(vform) return vform == "II" or vform == "III" or vform == "IV" or vform == "Iq" end local function vform_probably_no_passive(vform, weakness, past_vowel, nonpast_vowel) return vform == "I" and req(past_vowel, U) or vform == "VII" or vform == "IX" or vform == "XI" or vform == "XII" or vform == "XIII" or vform == "XIV" or vform == "XV" or vform == "IIIq" or vform == "IVq" end -- Active vforms II, III, IV, Iq use non-past prefixes in -u- instead of -a-. local function prefix_vowel_from_vform(vform) if vform == "II" or vform == "III" or vform == "IV" or vform == "Iq" then return "u" else return "a" end end -- True if the active non-past takes a-vocalization rather than i-vocalization in its last syllable. local function vform_nonpast_a_vowel(vform) return vform == "V" or vform == "VI" or vform == "XV" or vform == "IIq" end -- True if the `passive` spec indicates a passive-only verb. local function is_passive_only(passive) return passive == "onlypass" or passive == "onlypass-impers" end export.is_passive_only = is_passive_only -- for use in [[Module:ar-headword]] ------------------------------------------------------------------------------- -- Properties of specific sounds -- ------------------------------------------------------------------------------- -- Is radical wāw (و) or yāʾ (ي)? local function is_waw_ya(rad) return req(rad, W) or req(rad, Y) end -- Check that radical is wāw (و) or yāʾ (ي), error if not local function check_waw_ya(rad) if not is_waw_ya(rad) then error("Expecting weak radical: '" .. rget(rad) .. "' should be " .. W .. " or " .. Y) end end -- Form-I verb حيّ or حيي and form-X verb استحيا or استحى local function hayy_radicals(rad1, rad2, rad3) return req(rad1, "ح") and req(rad2, Y) and is_waw_ya(rad3) end -- FUCK ME HARD. "Lua error at line 1514: main function has more than 200 local variables". local function create_conjugations() ------------------------------------------------------------------------------- -- Radicals associated with various irregular verbs -- ------------------------------------------------------------------------------- -- Form-I verb أخذ or form-VIII verb اتخذ local function axadh_radicals(rad1, rad2, rad3) return req(rad1, HAMZA) and req(rad2, "خ") and req(rad3, "ذ") end -- Form-I verb whose imperative has a reduced form: أكل and أخذ and أمر. Return "shortonly" if only -- short-form imperatives exist (أكل and أخذ) or "shortlong" if long-form imperatives also exist (أمر); -- they are used after a clitic like فَ and وَ. local function reduced_imperative_verb(rad1, rad2, rad3) return axadh_radicals(rad1, rad2, rad3) and "shortonly" or req(rad1, HAMZA) and req(rad2, "ك") and req(rad3, "ل") and "shortonly" or req(rad1, HAMZA) and req(rad2, "م") and req(rad3, "ر") and "shortlong" end -- Form-I verb رأى and form-IV verb أرى local function raa_radicals(rad1, rad2, rad3) return req(rad1, "ر") and req(rad2, HAMZA) and is_waw_ya(rad3) end -- Form-I verb سأل local function saal_radicals(rad1, rad2, rad3) return req(rad1, "س") and req(rad2, HAMZA) and req(rad3, "ل") end -- Form-I verb كان local function kaan_radicals(rad1, rad2, rad3) return req(rad1, "ك") and req(rad2, W) and req(rad3, N) end ------------------------------------------------------------------------------- -- Sets of past endings -- ------------------------------------------------------------------------------- -- The 13 endings of the sound/hollow/geminate past tense. local past_endings = { -- singular SK .. TU, SK .. TA, SK .. "تِ", A, A .. "تْ", --dual SK .. "تُمَا", AA, A .. "تَا", -- plural SK .. "نَا", SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this SK .. "تُن" .. SH .. A, UU .. ALIF, SK .. "نَ" } -- Make endings for final-weak past in -aytu or -awtu. AYAW is AY or AW as appropriate. Note that AA and AW are -- global variables. local function make_past_endings_ay_aw(ayaw, third_sg_masc) return { -- singular ayaw .. SK .. TU, ayaw .. SK .. TA, ayaw .. SK .. "تِ", third_sg_masc, A .. "تْ", --dual ayaw .. SK .. "تُمَا", ayaw .. AA, A .. "تَا", -- plural ayaw .. SK .. "نَا", ayaw .. SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this ayaw .. SK .. "تُن" .. SH .. A, AW .. SK .. ALIF, ayaw .. SK .. "نَ" } end -- past final-weak -aytu endings local past_endings_ay = make_past_endings_ay_aw(AY, AAMAQ) -- past final-weak -awtu endings local past_endings_aw = make_past_endings_ay_aw(AW, AA) -- used for alternative endings for form-X geminate verbs like اِسْتَمَرَّ local past_endings_ay_12_person_only = { -- singular AY .. SK .. TU, AY .. SK .. TA, AY .. SK .. "تِ", {}, {}, --dual AY .. SK .. "تُمَا", {}, {}, -- plural AY .. SK .. "نَا", AY .. SK .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this AY .. SK .. "تُن" .. SH .. A, {}, {}, } -- Make endings for final-weak past in -ītu or -ūtu. IIUU is ī or ū as appropriate. Note that AA and UU are global -- variables. local function make_past_endings_ii_uu(iiuu) return { -- singular iiuu .. TU, iiuu .. TA, iiuu .. "تِ", iiuu .. A, iiuu .. A .. "تْ", --dual iiuu .. "تُمَا", iiuu .. AA, iiuu .. A .. "تَا", -- plural iiuu .. "نَا", iiuu .. "تُمْ", -- shadda + vowel diacritic ends up in the wrong order due to Unicode -- bug, so keep them separate to avoid this iiuu .. "تُن" .. SH .. A, UU .. ALIF, iiuu .. "نَ" } end -- past final-weak -ītu endings local past_endings_ii = make_past_endings_ii_uu(II) -- past final-weak -ūtu endings local past_endings_uu = make_past_endings_ii_uu(UU) ------------------------------------------------------------------------------- -- Sets of non-past prefixes and endings -- ------------------------------------------------------------------------------- local nonpast_prefix_consonants = { -- singular HAMZA, T, T, Y, T, -- dual T, Y, T, -- plural N, T, T, Y, Y } -- There are only five distinct endings in all non-past verbs. Make any set of non-past endings given these five -- distinct endings. local function make_nonpast_endings(null, fem, dual, pl, fempl) return { -- singular null, null, fem, null, null, -- dual dual, dual, dual, -- plural null, pl, fempl, pl, fempl } end -- endings for non-past indicative local ind_endings = make_nonpast_endings( U, II .. NA, AANI, UU .. NA, SK .. NA ) -- Make the endings for non-past subjunctive/jussive, given the vowel diacritic used in "null" endings -- (1s/2ms/3ms/3fs/1p). local function make_sub_juss_endings(dia_null) return make_nonpast_endings( dia_null, II, AA, UU .. ALIF, SK .. NA ) end -- endings for non-past subjunctive local sub_endings = make_sub_juss_endings(A) -- endings for non-past jussive local juss_endings = make_sub_juss_endings(SK) -- endings for alternative geminate non-past jussive in -a; same as subjunctive local juss_endings_alt_a = sub_endings -- endings for alternative geminate non-past jussive in -i local juss_endings_alt_i = make_sub_juss_endings(I) -- Endings for final-weak non-past indicative in -ā. Note that AY, AW and AAMAQ are global variables. local ind_endings_aa = make_nonpast_endings( AAMAQ, AYSK .. NA, AY .. AANI, AWSK .. NA, AYSK .. NA ) -- Make endings for final-weak non-past indicative in -ī or -ū; IIUU is ī or ū as appropriate. Note that II and UU -- are global variables. local function make_ind_endings_ii_uu(iiuu) return make_nonpast_endings( iiuu, II .. NA, iiuu .. AANI, UU .. NA, iiuu .. NA ) end -- endings for final-weak non-past indicative in -ī local ind_endings_ii = make_ind_endings_ii_uu(II) -- endings for final-weak non-past indicative in -ū local ind_endings_uu = make_ind_endings_ii_uu(UU) -- Endings for final-weak non-past subjunctive in -ā. Note that AY, AW, ALIF, AAMAQ are global variables. local sub_endings_aa = make_nonpast_endings( AAMAQ, AYSK, AY .. AA, AWSK .. ALIF, AYSK .. NA ) -- Make endings for final-weak non-past subjunctive in -ī or -ū. IIUU is ī or ū as appropriate. Note that AA, II, -- UU, ALIF are global variables. local function make_sub_endings_ii_uu(iiuu) return make_nonpast_endings( iiuu .. A, II, iiuu .. AA, UU .. ALIF, iiuu .. NA ) end -- endings for final-weak non-past subjunctive in -ī local sub_endings_ii = make_sub_endings_ii_uu(II) -- endings for final-weak non-past subjunctive in -ū local sub_endings_uu = make_sub_endings_ii_uu(UU) -- endings for final-weak non-past jussive in -ā local juss_endings_aa = make_nonpast_endings( A, AYSK, AY .. AA, AWSK .. ALIF, AYSK .. NA ) -- Make endings for final-weak non-past jussive in -ī or -ū. IU is short i or u, IIUU is long ī or ū as appropriate. -- Note that AA, II, UU, ALIF are global variables. local function make_juss_endings_ii_uu(iu, iiuu) return make_nonpast_endings( iu, II, iiuu .. AA, UU .. ALIF, iiuu .. NA ) end -- endings for final-weak non-past jussive in -ī local juss_endings_ii = make_juss_endings_ii_uu(I, II) -- endings for final-weak non-past jussive in -ū local juss_endings_uu = make_juss_endings_ii_uu(U, UU) ------------------------------------------------------------------------------- -- Sets of imperative endings -- ------------------------------------------------------------------------------- -- Extract the second person jussive endings to get corresponding imperative endings. local function imperative_endings_from_jussive(endings) return {endings[2], endings[3], endings[6], endings[10], endings[11]} end -- normal imperative endings local imp_endings = imperative_endings_from_jussive(juss_endings) -- alternative geminate imperative endings in -a local imp_endings_alt_a = imperative_endings_from_jussive(juss_endings_alt_a) -- alternative geminate imperative endings in -i local imp_endings_alt_i = imperative_endings_from_jussive(juss_endings_alt_i) -- final-weak imperative endings in -ā local imp_endings_aa = imperative_endings_from_jussive(juss_endings_aa) -- final-weak imperative endings in -ī local imp_endings_ii = imperative_endings_from_jussive(juss_endings_ii) -- final-weak imperative endings in -ū local imp_endings_uu = imperative_endings_from_jussive(juss_endings_uu) ------------------------------------------------------------------------------- -- Basic functions to inflect tenses -- ------------------------------------------------------------------------------- -- Add to `base` the inflections for the tense indicated by `tense` (the prefix in the slot names, e.g. 'past' -- or 'juss_pass'), formed by combining the `prefixes`, `stems` and `endings`. Each of `prefixes`, `stems` and -- `endings` is either a sequence of 5 (for the imperative) or 13 (for other tenses) abbreviated form lists (each of -- which is either a string, a form object, or a list of strings and/or form objects; see -- [[Module:inflection utilities]] for more info). Alternatively, any of `prefixes`, `stems` or `endings` can be a -- single-element list containing an abbreviated form list, with an additional key `all_same` set to true, or (as a -- special case) a single string; in the latter cases, the same value is used for all 5 or 13 slots. If existing -- inflections already exist, they will be added to, not overridden. `pnums` is the list of person/number slot name -- suffixes, which must match up with the elements in `prefixes`, `stems` and `endings` (i.e. 5 for imperative, 13 -- otherwise). local function inflect_tense_1(base, tense, prefixes, stems, endings, pnums) if not prefixes or not stems or not endings then return end local function verify_affixes(affixname, affixes) local function interr(msg) error(("Internal error: For tense '%s', '%s' %s: %s"):format(tense, affixname, msg, dump(affixes))) end if type(affixes) == "string" then -- do nothing elseif type(affixes) ~= "table" then interr("is not a table or string") elseif affixes.all_same then if #affixes ~= 1 then interr(("with all_same = true should have length 1 but has length %s"):format(#affixes)) end else if #affixes ~= #pnums then interr(("should have length %s but has length %s"):format(#pnums, #affixes)) end end end verify_affixes("prefixes", prefixes) verify_affixes("stems", stems) verify_affixes("endings", endings) local function get_affix(affixes, i) if type(affixes) == "string" then return affixes elseif affixes.all_same then return affixes[1] else return affixes[i] end end for i, pnum in ipairs(pnums) do local prefix = get_affix(prefixes, i) local stem = get_affix(stems, i) local ending = get_affix(endings, i) local slot = tense .. "_" .. pnum add3(base, slot, prefix, stem, ending) end end -- Add to `base` the inflections for the tense indicated by `tense` (the prefix in the slot names, e.g. 'past' -- or 'juss_pass'), formed by combining the `prefixes`, `stems` and `endings`. This is a simple wrapper around -- inflect_tense_1() that applies to all tenses other than the imperative; see inflect_tense_1() for more -- information about the parameters. local function inflect_tense(base, tense, prefixes, stems, endings) inflect_tense_1(base, tense, prefixes, stems, endings, all_person_number_list) end -- Like inflect_tense() but for the imperative, which has only five parts instead of 13 and no prefixes. local function inflect_tense_imp(base, stems, endings) inflect_tense_1(base, "imp", "", stems, endings, imp_person_number_list) end ------------------------------------------------------------------------------- -- Functions to inflect the past tense -- ------------------------------------------------------------------------------- -- Generate past verbs using specified vowel and consonant stems; works for sound, assimilated, hollow, and geminate -- verbs, active and passive. local function past_2stem_conj(base, tense, v_stem, c_stem, footnote_12) local passive = tense:find("_pass") and "_pass" or "" -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "past" .. passive .. "_v", v_stem) local c_stem_12 = c_stem if footnote_12 then c_stem_12 = iut.combine_form_and_footnotes(c_stem_12, footnote_12) end c_stem_12 = override_stem_if_needed(base, "past" .. passive .. "_c", c_stem_12) local c_stem_3 = override_stem_if_needed(base, "past" .. passive .. "_c", c_stem) inflect_tense(base, tense, "", { -- singular c_stem_12, c_stem_12, c_stem_12, v_stem, v_stem, --dual c_stem_12, v_stem, v_stem, -- plural c_stem_12, c_stem_12, c_stem_12, v_stem, c_stem_3 }, past_endings) end -- Generate past verbs using single specified stem; works for sound and assimilated verbs, active and passive. local function past_1stem_conj(base, tense, stem) past_2stem_conj(base, tense, stem, stem) end ------------------------------------------------------------------------------- -- Functions to inflect non-past tenses -- ------------------------------------------------------------------------------- -- Generate non-past conjugation, with two stems, for vowel-initial and consonant-initial endings, respectively. -- Useful for active and passive; for all forms; for all weaknesses (sound, assimilated, hollow, final-weak and -- geminate) and for all types of non-past (indicative, subjunctive, jussive) except for the imperative. (There is a -- separate wrapper function below for geminate jussives because they have three alternants.) Both stems may be the -- same, e.g. for sound verbs. -- `prefix_vowel` will be either "a" or "u". `endings` should be an array of 13 items. If `endings` is nil or -- omitted, infer the endings from the tense. If `jussive` is true, or `endings` is nil and `tense` indicatives -- jussive, use the jussive pattern of vowel/consonant stems (different from the normal ones). local function nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, endings, jussive) local passive = tense:find("_pass") and "_pass" or "" -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "nonpast" .. passive .. "_v", v_stem and q(dia[prefix_vowel], v_stem) or nil) c_stem = override_stem_if_needed(base, "nonpast" .. passive .. "_c", c_stem and q(dia[prefix_vowel], c_stem) or nil) if not endings then if tense:find("^ind") then endings = ind_endings elseif tense:find("^sub") then endings = sub_endings elseif tense:find("^juss") then jussive = true endings = juss_endings else error("Internal error: Unrecognized tense '" .. tense .."'") end end if not jussive then inflect_tense(base, tense, nonpast_prefix_consonants, { -- singular v_stem, v_stem, v_stem, v_stem, v_stem, --dual v_stem, v_stem, v_stem, -- plural v_stem, v_stem, c_stem, v_stem, c_stem }, endings) else inflect_tense(base, tense, nonpast_prefix_consonants, { -- singular -- 'adlul, tadlul, tadullī, yadlul, tadlul c_stem, c_stem, v_stem, c_stem, c_stem, --dual -- tadullā, yadullā, tadullā v_stem, v_stem, v_stem, -- plural -- nadlul, tadullū, tadlulna, yadullū, yadlulna c_stem, v_stem, c_stem, v_stem, c_stem }, endings) end end -- Generate non-past conjugation with one stem (no distinct stems for vowel-initial and consonant-initial endings). -- See nonpast_2stem_conj(). local function nonpast_1stem_conj(base, tense, prefix_vowel, stem, endings, jussive) nonpast_2stem_conj(base, tense, prefix_vowel, stem, stem, endings, jussive) end -- Generate active/passive jussive geminative. There are three alternants, two with terminations -a and -i and one -- in a null termination with a distinct pattern of vowel/consonant stem usage. See nonpast_2stem_conj() for a -- description of the arguments. local function jussive_gem_conj(base, tense, prefix_vowel, v_stem, c_stem) -- alternative in -a nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings_alt_a) -- alternative in -i nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings_alt_i) -- alternative in -null; requires different combination of v_stem and -- c_stem since the null endings require the c_stem (e.g. "tadlul" here) -- whereas the corresponding endings above in -a or -i require the v_stem -- (e.g. "tadulla, tadulli" above) nonpast_2stem_conj(base, tense, prefix_vowel, v_stem, c_stem, juss_endings, "jussive") end ------------------------------------------------------------------------------- -- Functions to inflect the imperative -- ------------------------------------------------------------------------------- -- Generate imperative conjugation, with two stems, for vowel-initial and consonant-initial endings, respectively. -- Useful for all forms, and for all weaknesses other than final-weak. Note that the two stems may be the same -- (specifically for sound and assimilated verbs). If `endings` is nil or omitted, use `imp_endings`. If `alt_gem` -- is specified, use the pattern of vowel and consonant stems appropriate for the alternative geminate imperatives -- that use a null ending of -a or -i instead of an empty ending. local function make_2stem_imperative(base, v_stem, c_stem, endings, alt_gem) endings = endings or imp_endings -- Override stems with user-specified stems if available. v_stem = override_stem_if_needed(base, "imp_v", v_stem) c_stem = override_stem_if_needed(base, "imp_c", c_stem) if alt_gem then inflect_tense_imp(base, {v_stem, v_stem, v_stem, v_stem, c_stem}, endings) else inflect_tense_imp(base, {c_stem, v_stem, v_stem, v_stem, c_stem}, endings) end end -- Generate imperative parts for sound or assimilated verbs. local function make_1stem_imperative(base, stem) make_2stem_imperative(base, stem, stem) end -- Generate imperative parts for geminate verbs form I (also IV, VII, VIII, X). local function make_gem_imperative(base, v_stem, c_stem) make_2stem_imperative(base, v_stem, c_stem, imp_endings_alt_a, "alt gem") make_2stem_imperative(base, v_stem, c_stem, imp_endings_alt_i, "alt gem") make_2stem_imperative(base, v_stem, c_stem) end ------------------------------------------------------------------------------- -- Functions to inflect entire verbs -- ------------------------------------------------------------------------------- -- Generate finite parts of a sound verb (also works for assimilated verbs) from five stems (past and non-past, -- active and passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u"). local function make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel) past_1stem_conj(base, "past", past_stem) past_1stem_conj(base, "past_pass", past_pass_stem) nonpast_1stem_conj(base, "ind", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "sub", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "juss", prefix_vowel, nonpast_stem) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem) make_1stem_imperative(base, imp_stem) end local function past_final_weak_endings_from_vowel(vowel) if vowel == "ay" then return past_endings_ay elseif vowel == "aw" then return past_endings_aw elseif vowel == "ī" then return past_endings_ii elseif vowel == "ū" then return past_endings_uu elseif not vowel then return nil else error(("Internal error: Unrecognized past final-weak vowel spec '%s'"):format(vowel)) end end local function nonpast_final_weak_endings_from_vowel(vowel) if vowel == "ā" then return ind_endings_aa, sub_endings_aa, juss_endings_aa, imp_endings_aa elseif vowel == "ī" then return ind_endings_ii, sub_endings_ii, juss_endings_ii, imp_endings_ii elseif vowel == "ū" then return ind_endings_uu, sub_endings_uu, juss_endings_uu, imp_endings_uu elseif not vowel then return nil else error(("Internal error: Unrecognized non-past final-weak vowel spec '%s'"):format(vowel)) end end -- Generate finite parts of a final-weak verb from five stems (past and non-past, active and passive, plus -- imperative), the past active ending vowel (ay, aw, ī or ū), the non-past active ending vowel (ā, ī or ū) and the -- prefix vowel in the active non-past (a or u). All five imperative persons share `imp_stem`, so the -- hamzat wasl carries the verb's own vowel class throughout -- اُدْعُ and اُدْعِي alike. The 2fs used to -- take a kasra one of its own (finding #5, now #11): its ending is -ī regardless of vowel class, so the -- stem vowel surfaces as a kasra there, and #5 read that as pulling the hamza with it. The book prints -- the ḍamma -- اُدْعِي (p.141), اُعْلِي and اُغْزِي (p159) -- so the surfacing kasra is عارض and leaves the -- hamza alone, exactly as the ḍamma of اِمْشُوا leaves it alone in the mirror case. -- -- `retain_rad3_2fs`, if true, keeps rad3 as a consonant before that -ī instead of letting it -- contract away: تَسْهُوِينَ / اُسْهُوِي rather than تَسْهِينَ / اِسْهِي. See README "Findings" #8 -- and final_weak_uu_retains_rad3(). local function make_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, past_ending_vowel, nonpast_ending_vowel, prefix_vowel, retain_rad3_2fs) past_stem = override_stem_if_needed(base, "past", past_stem) past_pass_stem = override_stem_if_needed(base, "past_pass", past_pass_stem) -- Don't call override_stem_if_needed() here for non-past stems; it's called in nonpast_2stem_conj(). imp_stem = override_stem_if_needed(base, "imp", imp_stem) -- + not supported for ending vowel overrides past_ending_vowel = base.stem_overrides.past_final_weak_vowel or past_ending_vowel local past_pass_ending_vowel = base.stem_overrides.past_pass_final_weak_vowel or "ī" nonpast_ending_vowel = base.stem_overrides.nonpast_final_weak_vowel or nonpast_ending_vowel local nonpast_pass_ending_vowel = base.stem_overrides.nonpast_pass_final_weak_vowel or "ā" local past_endings = past_final_weak_endings_from_vowel(past_ending_vowel) local past_pass_endings = past_final_weak_endings_from_vowel(past_pass_ending_vowel) local ind_endings, sub_endings, juss_endings, imp_endings = nonpast_final_weak_endings_from_vowel(nonpast_ending_vowel) local ind_pass_endings, sub_pass_endings, juss_pass_endings = nonpast_final_weak_endings_from_vowel(nonpast_pass_ending_vowel) if retain_rad3_2fs then -- rad3 surfaces as a consonant carrying the class's own ḍamma, then the regular -ī: -- سْهـ + ُوِينَ, not سْهـ + ِينَ. The 2fs sits at index 3 of a make_nonpast_endings() -- table (`fem`) and at index 2 of the five imperative endings that -- imperative_endings_from_jussive() extracts from it. Copied rather than mutated -- -- these ending tables are module-level constants shared by every final-weak verb. -- Written out inline, not via a helper: create_conjugations() is at Lua's 60-upvalue -- ceiling, and one more captured local stops the module loading at all. local function set_2fs(endings, index, ending) local out = {} for i, v in ipairs(endings) do out[i] = v end out[index] = ending return out end ind_endings = set_2fs(ind_endings, 3, UU .. II .. NA) sub_endings = set_2fs(sub_endings, 3, UU .. II) juss_endings = set_2fs(juss_endings, 3, UU .. II) imp_endings = set_2fs(imp_endings, 2, UU .. II) end inflect_tense(base, "past", "", {past_stem, all_same = 1}, past_endings) inflect_tense(base, "past_pass", "", {past_pass_stem, all_same = 1}, past_pass_endings) nonpast_1stem_conj(base, "ind", prefix_vowel, nonpast_stem, ind_endings) nonpast_1stem_conj(base, "sub", prefix_vowel, nonpast_stem, sub_endings) nonpast_1stem_conj(base, "juss", prefix_vowel, nonpast_stem, juss_endings) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem, ind_pass_endings) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem, sub_pass_endings) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem, juss_pass_endings) inflect_tense_imp(base, {imp_stem, all_same = 1}, imp_endings) end -- Generate finite parts of an augmented (form II+) final-weak verb from five stems (past and non-past, active and -- passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u") and a flag indicating if it -- behaves like a form V/VI verb in taking non-past endings in -ā instead of -ī. local function make_augmented_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel, form56) make_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, "ay", form56 and "ā" or "ī", prefix_vowel) end -- Generate finite parts of an augmented (form II+) sound or final-weak verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (active past stem minus last syllable (= -al or -ā)); -- * `nonpast_stem_base` (non-past stem minus last syllable (= -al/-il or -ā/-ī); -- * `past_pass_stem_base` (passive past stem minus last syllable (= -il or -ī)); -- * `vn` (verbal noun). local function make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) insert_form_or_forms(base, "vn", vn) local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local final_weak = is_final_weak(base, vowel_spec) local prefix_vowel = prefix_vowel_from_vform(base.verb_form) local form56 = vform_nonpast_a_vowel(base.verb_form) local a_base_suffix = final_weak and "" or q(A, lastrad) local i_base_suffix = final_weak and "" or q(I, lastrad) -- past and non-past stems, active and passive local past_stem = q(past_stem_base, a_base_suffix) -- In forms 5 and 6, non-past has /a/ as last stem vowel in the non-past -- in both active and passive, but /i/ in the active participle and /a/ -- in the passive participle. Elsewhere, consistent /i/ in active non-past -- and participle, consistent /a/ in passive non-past and participle. -- Hence, forms 5 and 6 differ only in the non-past active (but not -- active participle), so we have to split the finite non-past stem and -- active participle stem. local nonpast_stem = q(nonpast_stem_base, form56 and a_base_suffix or i_base_suffix) local ap_stem = q(nonpast_stem_base, i_base_suffix) local past_pass_stem = q(past_pass_stem_base, i_base_suffix) local nonpast_pass_stem = q(nonpast_stem_base, a_base_suffix) -- imperative stem local imp_stem = q(past_stem_base, form56 and a_base_suffix or i_base_suffix) -- make parts if final_weak then make_augmented_final_weak_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel, form56) else make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, prefix_vowel) end -- active and passive participle if final_weak then insert_form_or_forms(base, "ap", q(MU, ap_stem, IN)) insert_form_or_forms(base, "pp", q(MU, nonpast_pass_stem, AN, AMAQ)) else insert_form_or_forms(base, "ap", q(MU, ap_stem)) insert_form_or_forms(base, "pp", q(MU, nonpast_pass_stem)) end end -- Generate finite parts of a hollow or geminate verb from ten stems (vowel and consonant stems for each of past and -- non-past, active and passive, plus imperative) plus the prefix vowel in the active non-past ("a" or "u"), plus a -- flag indicating if we are a geminate verb. local function make_hollow_geminate_verb(base, geminate, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel, altgem_note) past_2stem_conj(base, "past", past_v_stem, past_c_stem, altgem_note) past_2stem_conj(base, "past_pass", past_pass_v_stem, past_pass_c_stem) nonpast_2stem_conj(base, "ind", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "sub", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "ind_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) nonpast_2stem_conj(base, "sub_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) if geminate then jussive_gem_conj(base, "juss", prefix_vowel, nonpast_v_stem, nonpast_c_stem) jussive_gem_conj(base, "juss_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) make_gem_imperative(base, imp_v_stem, imp_c_stem) else nonpast_2stem_conj(base, "juss", prefix_vowel, nonpast_v_stem, nonpast_c_stem) nonpast_2stem_conj(base, "juss_pass", "u", nonpast_pass_v_stem, nonpast_pass_c_stem) make_2stem_imperative(base, imp_v_stem, imp_c_stem) end end -- Generate finite parts of an augmented (form II+) hollow verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (invariable part of active past stem); -- * `nonpast_stem_base` (invariable part of nonpast stem); -- * `past_pass_stem_base` (invariable part of passive past stem); -- * `vn` (verbal noun). local function make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) insert_form_or_forms(base, "vn", vn) local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local form410 = base.verb_form == "IV" or base.verb_form == "X" local prefix_vowel = prefix_vowel_from_vform(base.verb_form) local a_base_suffix_v, a_base_suffix_c local i_base_suffix_v, i_base_suffix_c a_base_suffix_v = q(AA, lastrad) -- 'af-āl-a, inf-āl-a a_base_suffix_c = q(A, lastrad) -- 'af-al-tu, inf-al-tu i_base_suffix_v = q(II, lastrad) -- 'uf-īl-a, unf-īl-a i_base_suffix_c = q(I, lastrad) -- 'uf-il-tu, unf-il-tu -- past and non-past stems, active and passive, for vowel-initial and -- consonant-initial endings local past_v_stem = q(past_stem_base, a_base_suffix_v) local past_c_stem = q(past_stem_base, a_base_suffix_c) -- yu-f-īl-u, ya-staf-īl-u but yanf-āl-u, yaft-āl-u local nonpast_v_stem = q(nonpast_stem_base, form410 and i_base_suffix_v or a_base_suffix_v) local nonpast_c_stem = q(nonpast_stem_base, form410 and i_base_suffix_c or a_base_suffix_c) local past_pass_v_stem = q(past_pass_stem_base, i_base_suffix_v) local past_pass_c_stem = q(past_pass_stem_base, i_base_suffix_c) local nonpast_pass_v_stem = q(nonpast_stem_base, a_base_suffix_v) local nonpast_pass_c_stem = q(nonpast_stem_base, a_base_suffix_c) -- imperative stem local imp_v_stem = q(past_stem_base, form410 and i_base_suffix_v or a_base_suffix_v) local imp_c_stem = q(past_stem_base, form410 and i_base_suffix_c or a_base_suffix_c) -- make parts make_hollow_geminate_verb(base, false, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel) -- active participle insert_form_or_forms(base, "ap", q(MU, nonpast_v_stem)) -- passive participle insert_form_or_forms(base, "pp", q(MU, nonpast_pass_v_stem)) end -- Generate finite parts of an augmented (form II+) geminate verb, given: -- * `base` (conjugation data structure); -- * `vowel_spec` (radicals, weakness); -- * `past_stem_base` (invariable part of active past stem; this and the stem bases below will end with a consonant -- for forms IV, X, IVq, and a short vowel for the others); -- * `nonpast_stem_base` (invariable part of nonpast stem); -- * `past_pass_stem_base` (invariable part of passive past stem); -- * `vn` (verbal noun); -- * `altgem_note` (footnote to add to active past 1/2-person forms, when alternative forms are supplied [form X]). local function make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn, altgem_note) insert_form_or_forms(base, "vn", vn) local vform = base.verb_form local lastrad = base.quadlit and vowel_spec.rad4 or vowel_spec.rad3 local prefix_vowel = prefix_vowel_from_vform(vform) local a_base_suffix_v, a_base_suffix_c local i_base_suffix_v, i_base_suffix_c if vform == "IV" or vform == "X" or vform == "IVq" then a_base_suffix_v = q(A, lastrad, SH) -- 'af-all a_base_suffix_c = q(SK, lastrad, A, lastrad) -- 'af-lal i_base_suffix_v = q(I, lastrad, SH) -- yuf-ill i_base_suffix_c = q(SK, lastrad, I, lastrad) -- yuf-lil else a_base_suffix_v = q(lastrad, SH) -- fā-ll, infa-ll a_base_suffix_c = q(lastrad, A, lastrad) -- fā-lal, infa-lal i_base_suffix_v = q(lastrad, SH) -- yufā-ll, yanfa-ll i_base_suffix_c = q(lastrad, I, lastrad) -- yufā-lil, yanfa-lil end -- past and non-past stems, active and passive, for vowel-initial and -- consonant-initial endings local past_v_stem = q(past_stem_base, a_base_suffix_v) local past_c_stem = q(past_stem_base, a_base_suffix_c) local nonpast_v_stem = q(nonpast_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_v or i_base_suffix_v) local nonpast_c_stem = q(nonpast_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_c or i_base_suffix_c) -- NOTE: Formerly had a comment that "vform III and VI passive past do not have contracted parts, only -- uncontracted parts, which are added separately by those functions". This is based on Mace -- "Arabic Verbs and Essential Grammar" (1999) entry 63 (continued), which shows passive ḥūjija but no ḥūjja; -- but that is apparently a mistake, as (1) verb tables in other books do show contracted passive parts for -- these forms; (2) there is no mention of such an exception on p. 99, which explains how geminate ("doubled") -- verbs work (on the contrary, it says "The contracted and uncontracted pairs (see above) are found all -- over Forms III and VI of the doubled verbs"). local past_pass_v_stem = q(past_pass_stem_base, i_base_suffix_v) local past_pass_c_stem = q(past_pass_stem_base, i_base_suffix_c) local nonpast_pass_v_stem = q(nonpast_stem_base, a_base_suffix_v) local nonpast_pass_c_stem = q(nonpast_stem_base, a_base_suffix_c) -- imperative stem local imp_v_stem = q(past_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_v or i_base_suffix_v) local imp_c_stem = q(past_stem_base, vform_nonpast_a_vowel(vform) and a_base_suffix_c or i_base_suffix_c) -- make parts make_hollow_geminate_verb(base, "geminate", past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, prefix_vowel, altgem_note) -- active participle insert_form_or_forms(base, "ap", q(MU, nonpast_v_stem)) -- passive participle insert_form_or_forms(base, "pp", q(MU, nonpast_pass_v_stem)) end ------------------------------------------------------------------------------- -- Conjugation functions for specific conjugation types -- ------------------------------------------------------------------------------- local function form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1) local imp_vowel = map_vowel(nonpast_vowel, function(vow) if vow == A or vow == I then return I elseif vow == U then return U elseif not skip_slot(base, "imp_2ms") then error(("Internal error: Non-past vowel %s isn't a, i, or u, should have been caught earlier"):format( dump(nonpast_vowel))) else -- Passive-only; imperative won't ever be displayed so it doesn't matter. return I end end) -- Mace ("Arabic Verbs and Essentials of Grammar" p. 63: [https://archive.org/details/arabicverbsessen00john/page/62/mode/2up]) -- claims that initial hamza is assimilated/elided into a long vowel in the form-I imperative, but apparently -- this isn't corrrect. local vowel_on_alif = map_vowel(imp_vowel, function(vow) return ALIF .. vow end) return q(vowel_on_alif, rad1, SK) end -- Implement form-I sound or assimilated verb. ASSIMILATED is true for assimilated verbs. local function make_form_i_sound_assimilated_verb(base, vowel_spec, assimilated) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied -- past and non-past stems, active and passive local past_stem = q(rad1, A, rad2, past_vowel, rad3) local nonpast_stem = assimilated and q(rad2, nonpast_vowel, rad3) or q(rad1, SK, rad2, nonpast_vowel, rad3) local past_pass_stem = q(rad1, U, rad2, I, rad3) local nonpast_pass_stem = q(rad1, SK, rad2, A, rad3) -- imperative stem -- check for irregular verb with reduced imperative (أَخَذَ or أَكَلَ or أَمَرَ) local reducedimp = reduced_imperative_verb(rad1, rad2, rad3) if reducedimp then base.irregular = true end local imp_stem_suffix = q(rad2, nonpast_vowel, rad3) local long_imp_stem_base = form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1) local short_imp_stem_base = "" local imp_stem = q((assimilated or reducedimp) and "" or long_imp_stem_base, imp_stem_suffix) -- A form-I mithal that keeps its wāw spells its imperative on a hamzat wasl, -- اِوْجَلْ / اِوْنَ. Tell postprocess_term() not to reduce that اِوْ to اِي -- see #10 -- there for why the substitution list can't decide this for itself. if not assimilated and req(rad1, W) then base.orth.keep_initial_w = true end -- make parts make_sound_verb(base, past_stem, past_pass_stem, nonpast_stem, nonpast_pass_stem, imp_stem, "a") if reducedimp == "shortlong" then make_1stem_imperative(base, iut.combine_form_and_footnotes(q(long_imp_stem_base, imp_stem_suffix), mw.getCurrentFrame():preprocess("[used especially with a clitic such as {{m|ar|فَ}} or {{m|ar|وَ}}]"))) end -- Check for irregular verb سَأَلَ with alternative jussive and imperative. Calling this after make_sound_verb() -- adds additional entries to the paradigm parts. if saal_radicals(rad1, rad2, rad3) then base.irregular = true nonpast_1stem_conj(base, "juss", "a", "سَل") nonpast_1stem_conj(base, "juss_pass", "u", "سَل") make_1stem_imperative(base, "سَل") end -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, I, rad3)) -- Insert alternative active participle (stative type I) فَعِيل. Since not all verbs have this, we require that -- verbs that do have it specify it explicitly; a shortcut ++ is provided to make this easier (e.g. <ap:++> to -- indicate that the alternative form should be used for the active participle, <ap:+,++> to indicate that both -- forms can be used, and <ap:-> to indicate that there is no active participle). The same form is used for -- secondary default passive participle. insert_ap2_pp2(base, q(rad1, A, rad2, II, rad3)) -- Active participle, stative type II فَعِل (+++). insert_form_or_forms(base, "ap3", q(rad1, A, rad2, I, rad3)) -- Active participle, color/defect أَفْعَل (+cd). insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, A, rad3)) -- Active participle, -ān فَعْلَان (+an). insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, UU, rad3)) end conjugations["I-sound"] = function(base, vowel_spec) make_form_i_sound_assimilated_verb(base, vowel_spec, false) end conjugations["none-sound"] = function(base, vowel_spec) -- All default stems are nil. make_sound_verb(base) end conjugations["none-hollow"] = function(base, vowel_spec) -- All default stems are nil. make_hollow_geminate_verb(base, false) end conjugations["none-geminate"] = function(base, vowel_spec) -- All default stems are nil. make_hollow_geminate_verb(base, "geminate") end conjugations["none-final-weak"] = function(base, vowel_spec) -- All default stems are nil. make_final_weak_verb(base) end conjugations["I-assimilated"] = function(base, vowel_spec) make_form_i_sound_assimilated_verb(base, vowel_spec, "assimilated") end local function make_form_i_hayy_verb(base, vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied base.irregular = true -- past and non-past stems, active and passive, and imperative stem local past_c_stem = "حَيِي" local past_v_stem_long = past_c_stem local past_v_stem_short = "حَيّ" -- 3mp of the long variant (حَيِيَ) is حَيُوا, not the geminated حَيُّوا of the short -- variant (حَيَّ) -- rad2 (ي) here is a plain radical, not doubled; see README -- "Open findings" #4. Needs its own stem, distinct from both the long and short ones. local past_v_stem_3mp = "حَي" local past_pass_c_stem = "حُيِي" local past_pass_v_stem_long = past_pass_c_stem local past_pass_v_stem_short = "حُيّ" local past_pass_v_stem_3mp = "حُي" local nonpast_stem = "حْي" local nonpast_pass_stem = nonpast_stem local imp_stem = _I .. nonpast_stem -- make parts past_2stem_conj(base, "past", {}, past_c_stem) past_2stem_conj(base, "past_pass", {}, past_pass_c_stem) local variant = vowel_spec.variant or "both" if variant == "short" or variant == "both" then past_2stem_conj(base, "past", past_v_stem_short, {}) past_2stem_conj(base, "past_pass", past_pass_v_stem_short, {}) end function inflect_long_variant(tense, long_stem, mp_stem) inflect_tense_1(base, tense, "", {long_stem, long_stem, long_stem, long_stem, mp_stem}, {past_endings[4], past_endings[5], past_endings[7], past_endings[8], past_endings[12]}, {"3ms", "3fs", "3md", "3fd", "3mp"}) end if variant == "long" or variant == "both" then inflect_long_variant("past", past_v_stem_long, past_v_stem_3mp) inflect_long_variant("past_pass", past_pass_v_stem_long, past_pass_v_stem_3mp) end nonpast_1stem_conj(base, "ind", "a", nonpast_stem, ind_endings_aa) nonpast_1stem_conj(base, "sub", "a", nonpast_stem, sub_endings_aa) nonpast_1stem_conj(base, "juss", "a", nonpast_stem, juss_endings_aa) nonpast_1stem_conj(base, "ind_pass", "u", nonpast_pass_stem, ind_endings_aa) nonpast_1stem_conj(base, "sub_pass", "u", nonpast_pass_stem, sub_endings_aa) nonpast_1stem_conj(base, "juss_pass", "u", nonpast_pass_stem, juss_endings_aa) inflect_tense_imp(base, {imp_stem, all_same = 1}, imp_endings_aa) -- active and passive participles apparently do not exist for this verb end -- Implement form-I final-weak assimilated+final-weak verb. ASSIMILATED is true for assimilated verbs. local function make_form_i_final_weak_verb(base, vowel_spec, assimilated) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- حَيَّ or حَيِيَ is weird enough that we handle it as a separate function. if hayy_radicals(rad1, rad2, rad3) then make_form_i_hayy_verb(base, vowel_spec) return end -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- Past and non-past stems, active and passive, and imperative stem. local past_stem = q(rad1, A, rad2) local past_pass_stem = q(rad1, U, rad2) local nonpast_stem, nonpast_pass_stem, imp_stem if raa_radicals(rad1, rad2, rad3) then base.irregular = true nonpast_stem = rad1 nonpast_pass_stem = rad1 imp_stem = rad1 else nonpast_pass_stem = q(rad1, SK, rad2) if assimilated then nonpast_stem = rad2 imp_stem = rad2 else nonpast_stem = nonpast_pass_stem imp_stem = q(form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1), rad2) end end -- A form-I mithal that keeps its wāw spells its imperative on a hamzat wasl, -- اِوْجَلْ / اِوْنَ. Tell postprocess_term() not to reduce that اِوْ to اِي -- see #10 -- there for why the substitution list can't decide this for itself. if not assimilated and req(rad1, W) then base.orth.keep_initial_w = true end -- Make parts. local past_ending_vowel = req(rad3, Y) and req(past_vowel, A) and "ay" or req(rad3, W) and req(past_vowel, A) and "aw" or req(past_vowel, I) and "ī" or "ū" -- Try to preserve footnotes attached to the third radical and/or past and/or non-past vowels. local past_footnotes = iut.combine_footnotes(rget_footnotes(rad3), rget_footnotes(past_vowel)) local nonpast_ending_vowel = req(nonpast_vowel, A) and "ā" or req(nonpast_vowel, I) and "ī" or "ū" local nonpast_footnotes = iut.combine_footnotes(rget_footnotes(rad3), rget_footnotes(nonpast_vowel)) make_final_weak_verb(base, iut.combine_form_and_footnotes(past_stem, past_footnotes), iut.combine_form_and_footnotes(past_pass_stem, past_footnotes), iut.combine_form_and_footnotes(nonpast_stem, nonpast_footnotes), iut.combine_form_and_footnotes(nonpast_pass_stem, nonpast_footnotes), iut.combine_form_and_footnotes(imp_stem, nonpast_footnotes), past_ending_vowel, nonpast_ending_vowel, "a", final_weak_uu_retains_rad3(rad1, rad2, rad3, past_ending_vowel, nonpast_ending_vowel)) -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, IN)) -- Active participle, stative type I فَعِيّ (++). FIXME: Is this correct when rad3 is W? insert_ap2_pp2(base, q(rad1, A, rad2, II, SH)) -- Active participle, stative type II فَعٍ (+++). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "ap3", q(rad1, A, rad2, IN)) -- Active participle, color/defect أَفْعَى (+cd). insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, AAMAQ)) -- Active participle, -ān فَعْيَان or فَعْوَان (+an). -- FIXME: Any examples of this for both rad3 = W and y to verify it's correct? insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, req(rad3, Y) and II or UU, SH)) end conjugations["I-final-weak"] = function(base, vowel_spec) make_form_i_final_weak_verb(base, vowel_spec, false) end conjugations["I-assimilated+final-weak"] = function(base, vowel_spec) make_form_i_final_weak_verb(base, vowel_spec, "assimilated") end conjugations["I-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- In some sense, hollow vowels i~i and u~u are more "correct" than a~i and a~u, but the latter follow the -- pattern of other form-I verbs, so we map i~i to a~i and u~u to a~u in infer_radicals(). Now however we have -- to undo this to get the actual past vowel based on the non-past vowel. if req(past_vowel, A) then past_vowel = map_vowel(past_vowel, function(vow) return req(nonpast_vowel, A) and I or rget(nonpast_vowel) end) end local lengthened_nonpast = map_vowel(nonpast_vowel, function(vow) return vow == U and UU or vow == I and II or AA end) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- active past stems - vowel (v) and consonant (c) local past_v_stem = q(rad1, AA, rad3) local past_c_stem = q(rad1, past_vowel, rad3) -- active non-past stems - vowel (v) and consonant (c) local nonpast_v_stem = q(rad1, lengthened_nonpast, rad3) local nonpast_c_stem = q(rad1, nonpast_vowel, rad3) -- passive past stems - vowel (v) and consonant (c) -- 'ufīla always; the contracted stem takes whichever of i/u the active past_c_stem -- *isn't* -- xāfa/xiftu (i) -> passive xufta (u), qāla/qultu (u) -> passive qilta (i). -- Contraction drops rad2, so this vowel is the only thing distinguishing active from -- passive in these forms, hence the two roots' behaviour is opposite, not independent. -- Confirmed on two printed pages (خافَ p.224 "خُفْتَ", قالَ p.138 "قِلْتَ"); previously -- hardcoded to u from خافَ alone, which is only half the pattern -- see README. local past_pass_v_stem = q(rad1, II, rad3) local past_pass_c_stem = q(rad1, map_vowel(past_vowel, function(vow) return vow == I and U or I end), rad3) -- passive non-past stems - vowel (v) and consonant (c) -- yufāla/yufalna -- stem is built differently but conjugation is identical to sound verbs local nonpast_pass_v_stem = q(rad1, AA, rad3) local nonpast_pass_c_stem = q(rad1, A, rad3) -- imperative stem local imp_v_stem = nonpast_v_stem local imp_c_stem = nonpast_c_stem -- make parts make_hollow_geminate_verb(base, false, past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, "a") if kaan_radicals(rad1, rad2, rad3) then local endings = make_nonpast_endings(U, {}, {}, {}, {}) inflect_tense(base, "juss", nonpast_prefix_consonants, q(A, rad1), endings) base.irregular = true end -- Active participle. insert_form_or_forms(base, "ap1", req(rad3, HAMZA) and q(rad1, AA, HAMZA, IN) or q(rad1, AA, HAMZA, I, rad3)) -- Active participle, stative type I فَيِّد (++). FIXME: Any examples of this to verify it's correct? insert_ap2_pp2(base, q(rad1, A, Y, SH, I, rad3)) -- Active participle, stative type II فَيِد (+++). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "ap3", q(rad1, A, Y, I, rad3)) -- Active participle, color/defect أَفّيَد or أَفّوَد (+cd). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, SK, rad2, A, rad3)) -- Active participle, -ān فَيْدَان or فَوْدَان (+an). Example: جَاعَ "to be hungry", act part جَوْعَان insert_form_or_forms(base, "apan", q(rad1, A, rad2, SK, rad3, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, req(rad2, Y) and II or UU, rad3)) end conjugations["I-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3, past_vowel, nonpast_vowel = get_radicals_3(vowel_spec) -- Verbal nouns (maṣādir) for form I are unpredictable and have to be supplied. -- active past stems - vowel (v) and consonant (c) -- The contracted stem's vowel is the past vowel for most roots (مَدَدْتَ) but follows the -- مضارع for the إتباع lexemes (هَمُمْتَ); see geminate_degemination_vowel(). local past_v_stem = q(rad1, A, rad2, SH) local past_c_stem = q(rad1, A, rad2, geminate_degemination_vowel(rad1, rad2, rad3, past_vowel, nonpast_vowel), rad2) -- active non-past stems - vowel (v) and consonant (c) local nonpast_v_stem = q(rad1, nonpast_vowel, rad2, SH) local nonpast_c_stem = q(rad1, SK, rad2, nonpast_vowel, rad2) -- passive past stems - vowel (v) and consonant (c) -- dulla/dulilta local past_pass_v_stem = q(rad1, U, rad2, SH) local past_pass_c_stem = q(rad1, U, rad2, I, rad2) -- passive non-past stems - vowel (v) and consonant (c) --yudallu/yudlalna -- stem is built differently but conjugation is identical to sound verbs local nonpast_pass_v_stem = q(rad1, A, rad2, SH) local nonpast_pass_c_stem = q(rad1, SK, rad2, A, rad2) -- imperative stem local imp_v_stem = q(rad1, nonpast_vowel, rad2, SH) local imp_c_stem = q(form_i_imp_stem_through_rad1(base, nonpast_vowel, rad1), rad2, nonpast_vowel, rad2) -- make parts make_hollow_geminate_verb(base, "geminate", past_v_stem, past_c_stem, past_pass_v_stem, past_pass_c_stem, nonpast_v_stem, nonpast_c_stem, nonpast_pass_v_stem, nonpast_pass_c_stem, imp_v_stem, imp_c_stem, "a") -- Active participle. insert_form_or_forms(base, "ap1", q(rad1, AA, rad2, SH)) -- Active participle, stative type I فَعِيع (++). FIXME: Any examples of this to verify it's correct? insert_ap2_pp2(base, q(rad1, A, rad2, II, rad2)) -- Active participle, stative type II فَعّ (+++). Example: بَرَّ "to be pious", active participle بَرّ insert_form_or_forms(base, "ap3", q(rad1, A, rad2, SH)) -- Active participle, color/defect أَفَعّ (+cd). -- Example: لَصَّ "to be thievish, to steal repeatedly", active participle أَلَصّ. insert_form_or_forms(base, "apcd", q(HAMZA, A, rad1, A, rad2, SH)) -- Active participle, -ān فَعَّان (+an). FIXME: Any examples of this to verify it's correct? insert_form_or_forms(base, "apan", q(rad1, A, rad2, SH, AAN)) -- Passive participle. insert_form_or_forms(base, "pp", q(MA, rad1, SK, rad2, UU, rad2)) end -- Return the ta- (active, past and non-past) and tu- (passive past) prefixes for a form II/III/V/VI verb. -- Form V and VI verbs normally use ta- and tu-, but reduced (base.reduced) verbs use different prefixes. Form II -- and III verbs have no prefix. local function form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vform = base.verb_form if vform == "V" or vform == "VI" then if base.reduced then -- To simplify the code, we generate two rad1's with a sukūn between them, which is cleaned up in -- postprocessing. return q(_I, rad1, SK), q(rad1, SK), q(_U, rad1, SK) else return TA, TA, TU end else return "", "", "" end end -- Make form II or V sound or final-weak verb. local function make_form_ii_v_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vn = vform == "V" and q(ta_past_prefix, rad1, A, rad2, SH, final_weak and IN or q(U, rad3)) or q(TA, rad1, SK, rad2, II, final_weak and AH or rad3) -- various stem bases local past_stem_base = q(ta_past_prefix, rad1, A, rad2, SH) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, A, rad2, SH) local past_pass_stem_base = q(tu_past_prefix, rad1, U, rad2, SH) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["II-sound"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["II-final-weak"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end local function make_form_iii_alt_vn(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) -- Insert alternative verbal noun فِعَال. Since not all verbs have this, we require that verbs that do have it -- specify it explicitly; a shortcut ++ is provided to make this easier (e.g. <vn:+,++> to indicate that -- both the normal verbal noun مُفَاعَلَة and secondary verbal noun فِعَال are available). insert_form_or_forms(base, "vn2", q(rad1, I, rad2, AA, final_weak and HAMZA or rad3)) end -- Make form III or VI sound or final-weak verb. local function make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) local vn = vform == "VI" and q(ta_past_prefix, rad1, AA, rad2, final_weak and IN or q(U, rad3)) or q(MU, rad1, AA, rad2, final_weak and AAH or q(A, rad3, AH)) -- various stem bases local past_stem_base = q(ta_past_prefix, rad1, AA, rad2) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, AA, rad2) local past_pass_stem_base = q(tu_past_prefix, rad1, UU, rad2) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) if vform == "III" then make_form_iii_alt_vn(base, vowel_spec) end end conjugations["III-sound"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["III-final-weak"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end -- Make form III or VI geminate verb. local function make_form_iii_vi_geminate_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vform = base.verb_form local ta_past_prefix, ta_nonpast_prefix, tu_past_prefix = form_ii_iii_v_vi_ta_tu_prefix(base, rad1) -- Alternative verbal noun فِعَال will be inserted when we add sound parts below. local vn = vform == "VI" and q(ta_past_prefix, rad1, AA, rad2, SH) or q(MU, rad1, AA, rad2, SH, AH) -- Various stem bases. local past_stem_base = q(ta_past_prefix, rad1, AA) local nonpast_stem_base = q(ta_nonpast_prefix, rad1, AA) local past_pass_stem_base = q(tu_past_prefix, rad1, UU) -- Make parts. local variant = vowel_spec.variant or "short" if variant == "short" or variant == "both" then make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Also add alternative sound (non-compressed) parts. This will lead to some duplicate entries, but they are -- removed during addition. if variant == "long" or variant == "both" then make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) elseif vform == "III" then -- Still need to add the alternative form-III verbal noun. make_form_iii_alt_vn(base, vowel_spec) end end conjugations["III-geminate"] = function(base, vowel_spec) make_form_iii_vi_geminate_verb(base, vowel_spec) end -- Make form IV sound or final-weak verb. local function make_form_iv_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) -- core of stem base, minus stem prefixes local stem_core -- check for irregular verb أَرَى local is_raa = raa_radicals(rad1, rad2, rad3) if is_raa then base.irregular = true stem_core = rad1 else stem_core = q(rad1, SK, rad2) end -- verbal noun local vn = is_raa and q(HAMZA, I, stem_core, AA, HAMZA, AH) or q(HAMZA, I, stem_core, AA, final_weak and HAMZA or rad3) -- various stem bases local past_stem_base = q(HAMZA, A, stem_core) local nonpast_stem_base = stem_core local past_pass_stem_base = q(HAMZA, U, stem_core) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IV-sound"] = function(base, vowel_spec) make_form_iv_sound_final_weak_verb(base, vowel_spec) end conjugations["IV-final-weak"] = function(base, vowel_spec) make_form_iv_sound_final_weak_verb(base, vowel_spec) end conjugations["IV-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- verbal noun local vn = q(HAMZA, I, rad1, AA, rad3, AH) -- various stem bases local past_stem_base = q(HAMZA, A, rad1) local nonpast_stem_base = rad1 local past_pass_stem_base = q(HAMZA, U, rad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IV-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(HAMZA, I, rad1, SK, rad2, AA, rad2) -- various stem bases local past_stem_base = q(HAMZA, A, rad1) local nonpast_stem_base = rad1 local past_pass_stem_base = q(HAMZA, U, rad1) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["V-sound"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["V-final-weak"] = function(base, vowel_spec) make_form_ii_v_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-sound"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-final-weak"] = function(base, vowel_spec) make_form_iii_vi_sound_final_weak_verb(base, vowel_spec) end conjugations["VI-geminate"] = function(base, vowel_spec) make_form_iii_vi_geminate_verb(base, vowel_spec) end -- Make a verbal noun of the general form that applies to forms VII and above. RAD12 is the first consonant cluster -- (after initial اِ) and RAD34 is the second consonant cluster. RAD5 is the final consonant. local function high_form_verbal_noun(rad12, rad34, rad5) return q(_I, rad12, I, rad34, AA, rad5) end -- Populate a sound or final-weak verb for any of the various high-numbered augmented forms (form VII and up) that -- have up to 5 consonants in two clusters in the stem and the same pattern of vowels between. Some of these -- consonants in certain verb parts are w's, which leads to apparent anomalies in certain stems of these parts, but -- these anomalies are handled automatically in postprocessing, where we resolve sequences of iwC -> īC, uwC -> ūC, -- w + sukūn + w -> w + shadda. -- RAD12 is the first consonant cluster (after initial اِ) and RAD34 is the second consonant cluster. RAD5 is the -- final consonant. local function make_high_form_sound_final_weak_verb(base, vowel_spec, rad12, rad34, rad5) local final_weak = is_final_weak(base, vowel_spec) local vn = high_form_verbal_noun(rad12, rad34, final_weak and HAMZA or rad5) -- various stem bases local nonpast_stem_base = q(rad12, A, rad34) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad12, U, rad34) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end local function form_vii_nrad1(base, rad1) if base.reduced then if not req(rad1, M) then error(("Internal error: Form VII first radical %s is not م but .reduced specified; should have been caught earlier"): format(rget(rad1))) end return M .. SH else return q("نْ", rad1) end end -- Make form VII sound or final-weak verb. local function make_form_vii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high_form_sound_final_weak_verb(base, vowel_spec, form_vii_nrad1(base, rad1), rad2, rad3) end conjugations["VII-sound"] = function(base, vowel_spec) make_form_vii_sound_final_weak_verb(base, vowel_spec) end conjugations["VII-final-weak"] = function(base, vowel_spec) make_form_vii_sound_final_weak_verb(base, vowel_spec) end conjugations["VII-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local nrad1 = form_vii_nrad1(base, rad1) local vn = high_form_verbal_noun(nrad1, Y, rad3) -- various stem bases local nonpast_stem_base = nrad1 local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nrad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["VII-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local nrad1 = form_vii_nrad1(base, rad1) local vn = high_form_verbal_noun(nrad1, rad2, rad2) -- various stem bases local nonpast_stem_base = q(nrad1, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nrad1, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Return Form VIII verbal noun. local function form_viii_verbal_noun(base, vowel_spec, rad1, rad2, rad3) local final_weak = is_final_weak(base, vowel_spec) rad3 = final_weak and HAMZA or rad3 return {high_form_verbal_noun(vowel_spec.form_viii_assim, rad2, rad3)} end -- Make form VIII sound or final-weak verb. local function make_form_viii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- check for irregular verb اِتَّخَذَ if axadh_radicals(rad1, rad2, rad3) then base.irregular = true rad1 = T end make_high_form_sound_final_weak_verb(base, vowel_spec, vowel_spec.form_viii_assim, rad2, rad3) end conjugations["VIII-sound"] = function(base, vowel_spec) make_form_viii_sound_final_weak_verb(base, vowel_spec) end conjugations["VIII-final-weak"] = function(base, vowel_spec) make_form_viii_sound_final_weak_verb(base, vowel_spec) end conjugations["VIII-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = form_viii_verbal_noun(base, vowel_spec, rad1, Y, rad3) -- various stem bases local nonpast_stem_base = vowel_spec.form_viii_assim local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, nonpast_stem_base) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["VIII-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = form_viii_verbal_noun(base, vowel_spec, rad1, rad2, rad2) -- various stem bases local nonpast_stem_base = q(vowel_spec.form_viii_assim, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, vowel_spec.form_viii_assim, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IX-sound"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(_I, rad1, SK, rad2, I, rad3, AA, rad3) -- various stem bases local nonpast_stem_base = q(rad1, SK, rad2, A) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad1, SK, rad2, U) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["IX-final-weak"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high_form_sound_final_weak_verb(base, vowel_spec, q(rad1, SK, rad2), rad3, rad3) end -- Populate a sound or final-weak verb for any of the various high-numbered -- augmented forms that have 5 consonants in the stem and the same pattern of -- vowels. Some of these consonants in certain verb parts are w's, which leads to -- apparent anomalies in certain stems of these parts, but these anomalies -- are handled automatically in postprocessing, where we resolve sequences of -- iwC -> īC, uwC -> ūC, w + sukūn + w -> w + shadda. local function make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, rad3, rad4, rad5) make_high_form_sound_final_weak_verb(base, vowel_spec, q(rad1, SK, rad2), q(rad3, SK, rad4), rad5) end -- Make form X sound or final-weak verb. local function make_form_x_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) -- check for irregular verb اِسْتَحْيَا (also اِسْتَحَى) local is_hayy = hayy_radicals(rad1, rad2, rad3) local variant = vowel_spec.variant or "both" if not is_hayy or variant == "long" or variant == "both" then make_high5_form_sound_final_weak_verb(base, vowel_spec, S, T, rad1, rad2, rad3) end if is_hayy and (variant == "short" or variant == "both") then base.irregular = true -- Add alternative entries to the verbal paradigms. Any duplicates are removed during addition. make_high_form_sound_final_weak_verb(base, vowel_spec, S .. SK .. T, rad1, rad3) end end conjugations["X-sound"] = function(base, vowel_spec) make_form_x_sound_final_weak_verb(base, vowel_spec) end conjugations["X-final-weak"] = function(base, vowel_spec) make_form_x_sound_final_weak_verb(base, vowel_spec) end conjugations["X-hollow"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(base.reduced and "اِسْ" or "اِسْتِ", rad1, AA, rad3, AH) -- various stem bases local past_stem_base = q(base.reduced and "اِسْ" or "اِسْتَ", rad1) local nonpast_stem_base = q(base.reduced and "سْ" or "سْتَ", rad1) local past_pass_stem_base = q(base.reduced and "اُسْ" or "اُسْتُ", rad1) -- make parts make_augmented_hollow_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["X-geminate"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q("اِسْتِ", rad1, SK, rad2, AA, rad2) -- various stem bases local past_stem_base = q("اِسْتَ", rad1) local nonpast_stem_base = q("سْتَ", rad1) local past_pass_stem_base = q("اُسْتُ", rad1) -- make parts if base.altgem then inflect_tense(base, "past", "", {q(past_stem_base, A, rad2, SH), all_same = 1}, past_endings_ay_12_person_only) end make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn, base.altgem and "[uncommon]" or nil) end conjugations["XI-sound"] = function(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local vn = q(_I, rad1, SK, rad2, II, rad3, AA, rad3) -- various stem bases local nonpast_stem_base = q(rad1, SK, rad2, AA) local past_stem_base = q(_I, nonpast_stem_base) local past_pass_stem_base = q(_U, rad1, SK, rad2, UU) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Probably no form XI final-weak, since already geminate in form; would behave as XI-sound. -- Make form XII sound or final-weak verb. local function make_form_xii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, W, rad2, rad3) end conjugations["XII-sound"] = function(base, vowel_spec) make_form_xii_sound_final_weak_verb(base, vowel_spec) end conjugations["XII-final-weak"] = function(base, vowel_spec) make_form_xii_sound_final_weak_verb(base, vowel_spec) end -- Make form XIII sound or final-weak verb. local function make_form_xiii_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, W, W, rad3) end conjugations["XIII-sound"] = function(base, vowel_spec) make_form_xiii_sound_final_weak_verb(base, vowel_spec) end conjugations["XIII-final-weak"] = function(base, vowel_spec) make_form_xiii_sound_final_weak_verb(base, vowel_spec) end -- Make a form XIV or XV sound or final-weak verb. Last radical appears twice (if`anlala / yaf`anlilu) so if it were -- w or y you'd get if`anwā / yaf`anwī or if`anyā / yaf`anyī, i.e. unlike for most augmented verbs, the identity of -- the radical matters. local function make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3 = get_radicals_3(vowel_spec) local lastrad = base.verb_form == "XV" and Y or rad3 make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, N, rad3, lastrad) end conjugations["XIV-sound"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end conjugations["XIV-final-weak"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end conjugations["XV-sound"] = function(base, vowel_spec) make_form_xiv_xv_sound_final_weak_verb(base, vowel_spec) end -- Probably no form XV final-weak, since already final-weak in form; would behave as XV-sound. -- Make form Iq or IIq sound or final-weak verb. local function make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local vform = base.verb_form local vn = vform == "IIq" and q(TA, rad1, A, rad2, SK, rad3, (final_weak and IN or q(U, rad4))) or q(rad1, A, rad2, SK, rad3, (final_weak and AAH or q(A, rad4, AH))) local ta_pref = vform == "IIq" and TA or "" local tu_pref = vform == "IIq" and TU or "" -- various stem bases local past_stem_base = q(ta_pref, rad1, A, rad2, SK, rad3) local nonpast_stem_base = past_stem_base local past_pass_stem_base = q(tu_pref, rad1, U, rad2, SK, rad3) -- make parts make_augmented_sound_final_weak_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end conjugations["Iq-sound"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["Iq-final-weak"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIq-sound"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIq-final-weak"] = function(base, vowel_spec) make_form_iq_iiq_sound_final_weak_verb(base, vowel_spec) end -- The نون of اِفْعَنْلَلَ, or the إدغام that replaces it. Where ل1 is weak the reference -- prints the نون assimilated into it: model 427 (p.415) sets اِهْبَنْيَخَ as اِهْبَيَّخَ, -- and its own footnote gives the root as أصْلُهُ هَبْيَخَ, so the ي is ل1 and the نون is -- the wazn's. Emitting ل1 twice with a sukūn between them is how reduced forms V/VI -- write their تاء التفعّل too -- postprocess_subs' "same letter separated by sukūn" -- rule turns the pair into the šadda, and the مضارع and أمر come out اِهْبَيِّخْ / -- يَهْبَيِّخُ, which is what the page prints. -- -- Requested with |مدغم=, never inferred from the radical -- the same call forms V/VI, -- VII and X already make. p.415 is the only افعنلل paradigm in the corpus with a weak -- ل1 (the other five, اِفْعَنْلَلَ اِحْوَنْصَلَ اِحْرَنْجَمَ اِفْرَنْقَعَ اِسْلَنْطَأَ -- اِطْلَنْفَأَ, are all sound there), and one attestation is not enough to make every -- root of the shape assimilate for every caller on ar.wiktionary. local function form_iiiq_n_augment(base, rad3) if not base.reduced then return N end if not (req(rad3, W) or req(rad3, Y)) then error(("Form IIIq .reduced assimilates the نون into ل1, which must be و or ي, but saw %s"): format(rget(rad3))) end return rad3 end -- Make form IIIq sound or final-weak verb. local function make_form_iiiq_sound_final_weak_verb(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) make_high5_form_sound_final_weak_verb(base, vowel_spec, rad1, rad2, form_iiiq_n_augment(base, rad3), rad3, rad4) end conjugations["IIIq-sound"] = function(base, vowel_spec) make_form_iiiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IIIq-final-weak"] = function(base, vowel_spec) make_form_iiiq_sound_final_weak_verb(base, vowel_spec) end conjugations["IVq-sound"] = function(base, vowel_spec) local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local vn = q(_I, rad1, SK, rad2, I, rad3, SK, rad4, AA, rad4) -- various stem bases local past_stem_base = q(_I, rad1, SK, rad2, A, rad3) local nonpast_stem_base = q(rad1, SK, rad2, A, rad3) local past_pass_stem_base = q(_U, rad1, SK, rad2, U, rad3) -- make parts make_augmented_geminate_verb(base, vowel_spec, past_stem_base, nonpast_stem_base, past_pass_stem_base, vn) end -- Probably no form IVq final-weak, since already geminate in form; would behave as IVq-sound. end create_conjugations() ------------------------------------------------------------------------------- -- Guts of main conjugation function -- ------------------------------------------------------------------------------- -- Given form, weakness and radicals, check to make sure the radicals present are allowable for the weakness. Hamzas on -- alif/wāw/yāʾ seats are never allowed (should always appear as hamza-on-the-line), and various weaknesses have various -- strictures on allowable consonants. local function check_radicals(form, weakness, rad1, rad2, rad3, rad4) local function hamza_check(index, rad) if rad == HAMZA_ON_ALIF or rad == HAMZA_UNDER_ALIF or rad == HAMZA_ON_W or rad == HAMZA_ON_Y then error("Radical " .. index .. " is " .. rad .. " but should be ء (hamza on the line)") end end local function check_waw_ya(index, rad) if not is_waw_ya(rad) then error("Radical " .. index .. " is " .. rad .. " but should be و or ي") end end local function check_not_waw_ya(index, rad) if is_waw_ya(rad) then error("In a sound verb, radical " .. index .. " should not be و or ي") end end hamza_check(rad1) hamza_check(rad2) hamza_check(rad3) hamza_check(rad4) if weakness == "assimilated" or weakness == "assimilated+final-weak" then if rad1 ~= W then error("Radical 1 is " .. rad1 .. " but should be و") end -- don't check that non-assimilated form I verbs don't have wāw as their -- first radical because some form-I verbs exist where a first-radical wāw -- behaves as sound, e.g. wajuha yawjuhu "to be distinguished". end if weakness == "final-weak" or weakness == "assimilated+final-weak" then if rad4 then check_waw_ya(4, rad4) else check_waw_ya(3, rad3) end elseif vform_supports_final_weak(form) then -- non-final-weak verbs cannot have weak final radical if there's a corresponding -- final-weak verb category. I think this is safe. We may have problems with -- ḥayya/ḥayiya yaḥyā if we treat it as a geminate verb. if rad4 then check_not_waw_ya(4, rad4) else check_not_waw_ya(3, rad3) end end if weakness == "hollow" then check_waw_ya(2, rad2) -- don't check that non-hollow verbs in forms that support hollow verbs -- don't have wāw or yāʾ as their second radical because some verbs exist -- where a middle-radical wāw/yāʾ behaves as sound, e.g. form-VIII izdawaja -- "to be in pairs". end if weakness == "geminate" then if rad4 then error("Internal error: No geminate quadrilaterals, should not be seen") end if rad2 ~= rad3 then error("Weakness is geminate; radical 3 is " .. rad3 .. " but should be same as radical 2 " .. rad2) end elseif vform_supports_geminate(form) then -- non-geminate verbs cannot have second and third radical same if there's -- a corresponding geminate verb category. I think this is safe. We -- don't fuss over double wāw or double yāʾ because this could legitimately -- be a final-weak verb with middle wāw/yāʾ, treated as sound. if rad4 then error("Internal error: No quadrilaterals should support geminate verbs") end if rad2 == rad3 and not is_waw_ya(rad2) then error("Weakness is '" .. weakness .. "'; radical 2 and 3 are same at " .. rad2 .. " but should not be; consider making weakness 'geminate'") end end end -- array of substitutions; each element is a 2-entry array FROM, TO; do it -- this way so the concatenations only get evaluated once local postprocess_subs = { -- reorder short-vowel + shadda -> shadda + short-vowel for easier processing {"(" .. AIU .. ")" .. SH, SH .. "%1"}, ----------same letter separated by sukūn should instead use shadda--------- ------------happens e.g. in kun-nā "we were".----------------- {"(.)" .. SK .. "%1", "%1" .. SH}, ---------------------------- assimilated verbs ---------------------------- -- iw, iy -> ī (assimilated verbs) {I .. W .. SK, II}, {I .. Y .. SK, II}, -- uw, uy -> ū (assimilated verbs) {U .. W .. SK, UU}, {U .. Y .. SK, UU}, -------------- final -yā uses tall alif not alif maqṣūra ------------------ {"(" .. Y .. SH .. "?" .. A .. ")" .. AMAQ, "%1" .. ALIF}, ----------------------- handle hamza assimilation ------------------------- -- initial hamza + short-vowel + hamza + sukūn -> hamza + long vowel {HAMZA .. A .. HAMZA .. SK, HAMZA .. A .. ALIF}, {HAMZA .. I .. HAMZA .. SK, HAMZA .. I .. Y}, {HAMZA .. U .. HAMZA .. SK, HAMZA .. U .. W} } local postprocess_tr_subs = { {"ī([" .. vowels .. "y*])", "iy%1"}, {"ū([" .. vowels .. "w*])", "uw%1"}, {"(.)%*", "%1%1"}, -- implement shadda ---------------------------- assimilated verbs ---------------------------- -- iw, iy -> ī (assimilated verbs) {"iw([^" .. vowels .. "w])", "ī%1"}, {"iy([^" .. vowels .. "y])", "ī%1"}, -- uw, uy -> ū (assimilated verbs) {"uw([^" .. vowels .. "w])", "ū%1"}, {"uy([^" .. vowels .. "y])", "ū%1"}, ----------------------- handle hamza assimilation ------------------------- -- initial hamza + short-vowel + hamza + sukūn -> hamza + long vowel {"ʔaʔ(" .. NV .. ")", "ʔā%1"}, {"ʔiʔ(" .. NV .. ")", "ʔī%1"}, {"ʔuʔ(" .. NV .. ")", "ʔū%1"}, } -- Post-process verb parts to eliminate phonological anomalies. Many of the changes, particularly the tricky ones, -- involve converting hamza to have the proper seat. The rules for this are complicated and are documented on the -- [[w:Hamza]] Wikipedia page. In some cases there are alternatives allowed, and we handle them below by returning -- multiple possibilities. -- -- ORTHOGRAPHIC EXEMPTIONS (`orth`, set on `base` by the code that knows the fact) -- -- postprocess_subs above and ar_utilities.process_hamza apply general Arabic orthography -- to a finished word. A handful of the reference's verb spellings are decisions neither -- pass can make, because what settles them is not in the string: that a hamza is a -- *radical*, that a short vowel is the stem's rather than iʿrāb, that a verb is form VIII -- rather than form I. Each of the three has both halves in the corpus. Form I أَتَى and -- form VIII اِئْتَمَنَ reach this function's 1s مضارع as the same ءَ + ءْ + ت, and the book -- prints آتِي for the one and أَأْتَمِنُ for the other (#14). شَأَى's imperative اُشْؤُ and -- جَرُؤَ's past both end in hamza + a short vowel, and only the first of the two vowels -- belongs to the stem (#12). -- -- module/ar-utilities.lua is byte-identical to module/vendor/ar-utilities.lua, and its -- other callers are on-wiki nominal templates that cannot be regression-tested from this -- repo, so none of this is fixed by editing it. Each case is instead handled on the -- caller's side of the shared passes, in one of three ways: -- -- * park the sequence on a placeholder so a shared rule cannot see it, and restore it -- afterwards -- #10's original mechanism, now also #14's; -- * add a spelling beside whatever the shared code returns, because the reference -- attests both and this module already carries alternatives elsewhere -- #16, #17; or -- * rewrite the sequence, either before the shared pass so its rule never matches -- (#12's past clause writes the ā as an ألف مقصورة, which the madda rule does not -- look for) or after it, where the shared code has made a decision the caller's fact -- overrides (#12's seating clause, #18's). -- -- Adding one is three things: a key in `base.orth` set by the builder that knows the -- fact, a clause here naming the finding, and the XFAIL entries in tests/run.lua deleted -- in the same commit. -- -- Three of the clauses below are unconditional, and deliberately so: they repair a -- spelling the shared code produces that Arabic does not use (#18), restore one it -- discards (#17), or extend one of its rules to a seat it never reaches (#15). None of -- the three needs a fact from the caller -- only a place to live that is not the shared -- file. local function postprocess_term(term, orth) if term == "?" then return "?" end orth = orth or {} -- Phase 1: the reductions in postprocess_subs, with the parked exemptions held out -- across them. `park_w` is a parameter rather than read from `orth` because #16 runs -- this phase both ways -- see below. local function reduce(t, park_w) -- Add BORDER at text boundaries. t = BORDER .. t .. BORDER -- #10: a word-initial hamzat wasl + wāw + sukūn is how a form-I mithal that keeps -- its wāw spells its imperative (اِوْنَ، اِوْجَلْ). The iw -> ī reduction is right -- everywhere else and cannot simply be dropped -- it is what turns the form VIII -- mithal past اِوْتَعَدَ into اِيتَعَدَ, in the same word-initial position, so position -- alone doesn't separate the two cases. if park_w then t = rsub(t, BORDER .. ALIF .. I .. W .. SK, BORDER .. ALIF .. I .. KEEP_W_PH .. SK) end -- #14: the reference spells the 1s مضارع of a hamza-initial form VIII analytically, -- أَأْتَمِنُ -- the prefix hamza, then the quiescent radical hamza on its own seat -- -- where the أَ + أْ -> آ contraction below and the alif-madda rule in process_hamza -- between them return آتَمِنُ. Form I of a hamza-initial root arrives here with the -- identical sequence and *is* printed contracted (آكُلُ، آخُذُ، آتِي), so the string -- cannot decide this and the flag does. if orth.analytic_hamza then t = rsub(t, BORDER .. HAMZA .. A .. HAMZA .. SK, BORDER .. HAMZA .. A .. KEEP_HAMZA_PH .. SK) end -- Do the main post-processing, based on the pattern substitutions in postprocess_subs. for _, sub in ipairs(postprocess_subs) do t = rsub(t, sub[1], sub[2]) end -- #12, the past half: a form-I final-weak verb on a wāw rad3 builds its past 3ms ā -- as a real alif (دَعَا), and after a hamza rad2 that puts hamza-on-alif + fatḥa + -- alif in front of process_hamza's madda rule (أَا -> آ), which swallows the seat: -- شَأَى comes out شَآ. The rule itself is right -- it is what gives قَرَأَ its dual قَرَآ -- -- and it keys on ALIF, not on AMAQ, which is why the module already gets رَأَى and -- نَأَى right: a yāʾ rad3 never produces the second alif. So the ā is written as an -- ألف مقصورة here instead, which sidesteps the madda rule rather than suppressing -- it. Only the past 3ms of an -aw verb reaches this sequence; every derived form -- already ends its past 3ms in AMAQ. if orth.hamza_rad2_final_weak then t = rsub(t, HAMZA .. A .. ALIF .. BORDER, HAMZA .. A .. AMAQ .. BORDER) end t = t:gsub(KEEP_W_PH, W) -- Restored as a plain hamza, not a seat: the seating pass below still has to place it, -- and gets it right (a quiescent medial hamza after a fatḥa takes the alif). t = t:gsub(KEEP_HAMZA_PH, HAMZA) return (t:gsub(BORDER, "")) end -- Phase 2: hamza seats. Returns the list of spellings for one reduced term. local function seat(t) if not rfind(t, HAMZA) then return { t } end -- #17: after a long ū, process_hamza treats و + ؤُ + و as having no acceptable -- original spelling and returns only its two alternatives (ئُو، ءُو) -- while the -- same function keeps the original beside the alternative after a long ā, which is -- what gives سَاؤُوا. The reference prints the wāw seat after a long ū throughout -- (يَسُوؤُونَ، تَسُوؤُونَ، سُوؤُوا, against its own تَسُوئِينَ / سُوئِي for the kasra: the -- seat follows the hamza's own vowel). So the discarded spelling is added back -- here, after theirs rather than instead of them -- the same order the long-ā -- branch of that function already uses. local waw_seat if rfind(t, W .. HAMZA .. U .. W) then local parked = t:gsub(W .. HAMZA .. U .. W, W .. WAW_SEAT_PH .. U .. W) waw_seat = ar_utilities.process_hamza(parked:gsub(HAMZA, HAMZA_PH))[1] waw_seat = waw_seat:gsub(WAW_SEAT_PH, HAMZA_ON_W) end local out = ar_utilities.process_hamza(t:gsub(HAMZA, HAMZA_PH)) if waw_seat then out[#out + 1] = waw_seat end -- #12, the imperative/jussive half: apocopation strands the radical hamza at the -- end of the word, where process_hamza takes its seat from the *preceding* letter's -- vowel and, finding a sukūn, drops it on the line -- اُشْءُ, أَفْءِ. That rule is right -- for a hamza whose own vowel is iʿrāb (جُزْء), and cannot be made position-sensitive -- inside the shared file: the fatḥa of جَرُؤَ is the ماضٍ ending and the ḍamma before -- it rightly wins the seat. Here the short vowel left after rad3 goes is the stem's -- own, which is the fact this flag carries. The reference seats it: اُشْؤُ (p160, شأو, -- form I) and أَفْإِ (p300, فأو, form IV) -- two roots, two measures, two pages, and -- the two vowels between them. -- -- The seat is the wāw for a ḍamma and the alif otherwise; #18's substitution below -- then writes the alif's hamza underneath when the vowel is a kasra, which is how -- the alif branch comes out as إ. That is not a third rule invented here: the same -- إ is what p368's اِرْتَإِ and p182's imperative إ already print, and what the module -- already produces for both, having reached the alif seat by its own route (a -- preceding fatḥa, and word-initial position). Only p300 needed the seat itself. -- -- Gated on a preceding sukūn, which is the whole of what the two pages attest. The -- hamza after a *long* vowel is left alone (وأي's passive jussive تُوءَ), where the -- shared code also leaves it on the line and no page here prints the slot; and the -- pattern wants the vowel directly after the hamza, so a šadda'd seat (p271's رَأِّ) -- is not matched, the same gate #18 gets for free. if orth.hamza_rad2_final_weak then for i, v in ipairs(out) do v = rsub(BORDER .. v .. BORDER, SK .. HAMZA .. "(" .. AIU .. ")" .. BORDER, function(vowel) return SK .. (vowel == U and HAMZA_ON_W or HAMZA_ON_ALIF) .. vowel .. BORDER end) out[i] = (v:gsub(BORDER, "")) end end -- #15: a hamza left on the line, carrying a fatḥa, with an alif after it, is -- printed as a madda -- but not when that would end the word directly after an -- alif. The reference sets يَداآنِ / تَداآنِ (p201, دوء), يَهاآنِ / تَهاآنِ (p242, -- هوء) and يَسُوآنِ / تَسُوآنِ (p157, سوء) for the مضارع duals, and سُوآ for p157's -- imperative dual, while the same three verbs print their *word-final* duals -- داءا / هاءا / سَاءَا with the hamza on the line -- so the exception is carried by -- the preceding alif, not by the position alone: سُوآ is word-final too and does -- contract, after a long ū. Ten cells, three roots, three pages, and the blocked -- case attested on all three. process_hamza applies this rule only to a hamza it -- has seated on an alif (which is what gives تَدَرْبَآ); after a long vowel it -- leaves the seat on the line and the rule never fires, so the extension is here. -- -- Confined to a hamza after a *long vowel*, which is the whole of what those three -- pages attest. Written without that gate it also contracts the فَعْلان participle -- سَوْءَان to سَوْآن -- a hamza after the sukūn of a diphthong, not after a long vowel, -- in a slot no fixture covers and no page here prints. The mater carries no sukūn -- in a long vowel as this module builds one, so requiring the two to be adjacent -- is the gate. for i, v in ipairs(out) do v = BORDER .. v .. BORDER v = rsub(v, ALIF .. HAMZA .. A .. ALIF .. BORDER, ALIF .. NO_MADDA_PH .. A .. ALIF .. BORDER) v = rsub(v, "([" .. AMAD .. ALIF .. W .. Y .. "])" .. HAMZA .. A .. ALIF, "%1" .. AMAD) v = v:gsub(NO_MADDA_PH, HAMZA) out[i] = (v:gsub(BORDER, "")) end -- #18: a hamza seated on an alif and carrying a bare kasra is written *below* the -- alif. The seat is chosen from the preceding letter's vowel, which is right when -- the hamza's own vowel is iʿrāb (مَلْجَأٍ) and is why the rule lives in the shared -- code; in a verb there is no iʿrāb, and the kasra that follows an apocopated -- radical hamza is the stem's own -- so اِرْتَأَى's jussive is اِرْتَإِ. -- -- Not when the alif also carries a šadda, and the corpus is what says so: رَأَّى -- (p271, form II of the *same* root رأو as p368's اِرْتَأَى) prints its apocopated -- imperative رَأِّ, hamza above, where the doubled radical puts a šadda on the seat. -- Written unconditionally first, this rule broke that cell and nothing else in 379 -- paradigms, which is both the counter-example and its own control: one root, two -- forms, two pages, and the presence of the šadda the only difference between them. -- No test for the šadda is needed: postprocess_subs has already reordered it in -- front of the vowel, so أ + šadda + kasra simply isn't the sequence below. for i, v in ipairs(out) do out[i] = v:gsub(HAMZA_ON_ALIF .. I, HAMZA_UNDER_ALIF .. I) end return out end local out = seat(reduce(term, orth.keep_initial_w)) -- #16: a retained-wāw مثال reduces اِوْ to اِيْ in the imperative on p240 (اِيْجَعْ, all -- five cells) and does not on p247 (اِوْنَ) -- the book carries both spellings and -- grammar allows both, so the module carries both too, the way it already carries -- سَائُوا | سَاؤُوا. #10's park is still right; what was missing is the second form -- beside the first. Widening loose() instead would retire the suite's only check on -- the mithal-wāwī / mithal-yāʾī imperative distinction, which is why this is here. if orth.keep_initial_w then for _, reduced in ipairs(seat(reduce(term, false))) do local seen = false for _, v in ipairs(out) do if v == reduced then seen = true break end end if not seen then out[#out + 1] = reduced end end end if #out == 1 then return out[1] end return out end local function postprocess_translit(translit) if translit == "?" then return "?" end -- Add BORDER at text boundaries. translit = BORDER .. translit .. BORDER -- Do the main post-processing, based on the pattern substitutions in postprocess_tr_subs. for _, sub in ipairs(postprocess_tr_subs) do translit = rsub(translit, sub[1], sub[2]) end translit = translit:gsub(BORDER, "") return translit end -- #13: نقل الحركة + إدغام in a لفيف مقرون verb outside form I. Where the paradigm brings -- two vowelled yāʾs together on a *quiescent* radical, the reference moves the first -- yāʾ's vowel back onto that radical and geminates the pair: أَحْيَيَا -> أَحَيَّا, -- يُحْيِيَانِ / تُحْيِيَانِ -> يُحِيَّانِ / تُحِيَّانِ, أَحْيِيَا -> أَحِيَّا (p302, form IV on حيي), -- and the same five slots of اِسْتَحْيَا (p389, form X on the same root). -- -- Only the نقل is done here. The إدغام is postprocess_subs' own "same letter separated -- by a sukūn -> šadda" rule, which already puts the šadda in front of the vowel the way -- the rest of the module writes it -- so this has to run *before* postprocess_forms(), -- and does, from conjugate_verb(). -- -- The gate is that sukūn, and the reference's pages calibrate it from both sides: -- * a *long* vowel in front of the first yāʾ can host nothing, and عايا (p287, form -- III on عيي) prints يُعايِيانِ، تُعايِيانِ and عايِيا with both yāʾs. The pattern below -- does not match them because there is no consonant + sukūn there to match; -- * the second yāʾ has to carry a short vowel of its own. The control is in the same -- columns of the pages that do contract: يُحْيِي (3ms), يُحْيِينَ (3fp) and تُحْيِينَ -- (2fs), where it is the mater of a long ī, are printed uncontracted, and the -- pattern does not match those either. -- Between them the two conditions pick out exactly the five slots of p302 and p389 that -- contract, against the twenty-odd cells of the same two paradigms that do not. -- -- Form I is excluded, and by attestation rather than by phonology: حَيِيَ (p229, 017_159) -- and عَيِيَ (p250, 238_159) put the identical consonant + sukūn + yāʾ sequence in their -- مضارع and imperative duals and both print them uncontracted -- يَحْيَيانِ، اِحْيَيا، -- يَعْيَيانِ، اِعْيَيا. Form I has its own version of this alternation, حَيَّ beside حَيِيَ, -- and it is a different mechanism: the vowel is dropped rather than moved, which is why -- make_form_i_hayy_verb() spells it out as a stem pair instead. -- -- Added beside the uncontracted form and never instead of it: both spellings are -- current, and form X already carries the two the module builds from its own stems -- (اِسْتَحْيَيَا | اِسْتَحَيَا), so this is a third entry in a slot that already had a choice. local NAQL_FROM = "(.)" .. SK .. Y .. "(" .. AIU .. ")" .. Y .. "(" .. AIU .. ")" local NAQL_TO = "%1%2" .. Y .. SK .. Y .. "%3" local function add_naql_idgham_forms(base) -- "none" is a user-supplied paradigm with no verb form to reason from; form I is -- excluded on the evidence above. if base.verb_form == "I" or base.verb_form == "none" then return end -- Collected first and inserted afterwards: inserting into base.forms while iterating -- over it would let a contracted form be re-read as a source in the same pass. local additions for slot, forms in pairs(base.forms) do for _, form in ipairs(forms) do local naql, changed = rsubb(form.form, NAQL_FROM, NAQL_TO) if changed then additions = additions or {} additions[#additions + 1] = {slot, {form = naql, footnotes = form.footnotes}} end end end if additions then for _, addition in ipairs(additions) do iut.insert_form(base.forms, addition[1], addition[2]) end end end local function postprocess_forms(base) local converted_values = {} for slot, forms in pairs(base.forms) do local need_dedup = false for i, form in ipairs(forms) do local term = postprocess_term(form.form, base.orth) local translit = form.translit and postprocess_translit(form.translit) or nil if term ~= form.form or translit ~= form.translit then need_dedup = true end converted_values[i] = {term, translit} end if need_dedup then local temp_dedup = {} for i = 1, #forms do local new_term, new_translit = unpack(converted_values[i]) if type(new_term) == "table" then for _, nt in ipairs(new_term) do local new_formobj = { form = nt, translit = new_translit, footnotes = forms[i].footnotes, } iut.insert_form(temp_dedup, "temp", new_formobj) end else local new_formobj = { form = new_term, translit = new_translit, footnotes = forms[i].footnotes, } iut.insert_form(temp_dedup, "temp", new_formobj) end end base.forms[slot] = temp_dedup.temp end end end local function process_slot_overrides(base) for slot, forms in pairs(base.slot_overrides) do local existing_values = base.forms[slot] base.forms[slot] = nil for _, form in ipairs(forms) do -- + in active participle for form I requests slot ap1 if form.form == "+" and (base.verb_form ~= "I" or slot ~= "ap") then if not existing_values then error(("Slot '%s' requested the default value but no such value available"):format(slot)) end -- We maintain an invariant that no two slots share a form object (although they may share the footnote -- lists inside the form objects). However, there is no need to copy the form objects here because there -- is a one-to-one correspondence between slots and slot overrides, i.e. you can't have a default value -- go into two slots. insert_form_or_forms(base, slot, existing_values, "allow overrides", form.uncertain) elseif default_indicator_to_active_participle_slot[form.form] then if form.form == "++" then if slot ~= "vn" and slot ~= "ap" and slot ~= "pp" then error(("Secondary default value request '++' only applicable to verbal nouns and pariciples, but found in slot '%s'"): format(slot)) end else if slot ~= "ap" then error(("Secondary default value request '%s' only applicable to active pariciples, but found in slot '%s'"): format(form.form, slot)) end end local secondary_default_slot = slot == "vn" and "vn2" or slot == "pp" and "pp2" or default_indicator_to_active_participle_slot[form.form] local existing_values = base.forms[secondary_default_slot] if not existing_values then error(("Slot '%s' requested a secondary default value using '%s' but no such value available"): format(slot, form.form)) end -- See comment above about the lack of need to copy the form objects. insert_form_or_forms(base, slot, existing_values, "allow overrides", form.uncertain) -- To make sure there aren't shared form objects. base.forms[secondary_default_slot] = nil else insert_form_or_forms(base, slot, form, "allow overrides", form.uncertain) end end end -- Now, for non-stative form-I verbs, fill the active participle slot from ap1 unless it should be missing (e.g. -- passive-only or user specified 'ap:-'). if base.verb_form == "I" and not base.forms.ap and base.forms.ap1 and not skip_slot(base, "ap") then local saw_non_stative = false for _, vowel_spec in ipairs(base.conj_vowels) do if req(vowel_spec.past, A) then saw_non_stative = true break end end if saw_non_stative then base.forms.ap = base.forms.ap1 -- To make sure there aren't shared form objects. base.forms.ap1 = nil end end end local function handle_lemma_linked(base) -- Compute linked versions of potential lemma slots, for use in {{ar-verb}}. We substitute the original lemma -- (before removing links) for forms that are the same as the lemma, if the original lemma has links. for _, slot in ipairs(export.potential_lemma_slots) do if base.forms[slot] then insert_form_or_forms(base, slot .. "_linked", iut.map_forms(base.forms[slot], function(form) if form == base.lemma and rfind(base.linked_lemma, "%[%[") then return base.linked_lemma else return form end end)) end end end -- Process specs given by the user using 'addnote[SLOTSPEC][FOOTNOTE][FOOTNOTE][...]'. local function process_addnote_specs(base) for _, spec in ipairs(base.addnote_specs) do for _, slot_spec in ipairs(spec.slot_specs) do slot_spec = "^" .. slot_spec .. "$" for slot, forms in pairs(base.forms) do if rfind(slot, slot_spec) then -- To save on memory, side-effect the existing forms. for _, form in ipairs(forms) do form.footnotes = iut.combine_footnotes(form.footnotes, spec.footnotes) end end end end end end local function add_missing_links_to_forms(base) -- Any forms without links should get them now. Redundant ones will be stripped later. for slot, forms in pairs(base.forms) do for _, form in ipairs(forms) do if not form.form:find("%[%[") then form.form = "[[" .. form.form .. "]]" end end end end local function conjugate_verb(base) construct_stems(base) -- Orthographic exemptions the shared hamza/reduction passes can't decide for -- themselves; see the block comment on postprocess_term(). Most are set by the -- stem builder that knows the fact, so this table has to exist before dispatch. base.orth = base.orth or {} for _, vowel_spec in ipairs(base.conj_vowels) do -- Reconstruct conjugation type from verb form and (possibly inferred) weakness. conj_type = base.verb_form .. "-" .. vowel_spec.weakness -- #14: a hamza-initial form VIII spells its 1s مضارع analytically (أَأْتَمِنُ), where -- form I of a hamza-initial root contracts to a madda (آتِي) from the identical -- sequence. Declared here rather than in the four VIII builders because it holds -- for all of them and depends on nothing else they compute. if base.verb_form == "VIII" and vowel_spec.rad1 and rfind(rget(vowel_spec.rad1) or "", HAMZA_ANY) then base.orth.analytic_hamza = true end -- #12: a final-weak verb whose rad2 is a hamza. Two of the shared passes read that -- radical as an ordinary word-final hamza and get it wrong in the two slots where -- rad3 stops being a consonant -- see the two clauses in postprocess_term(). Like -- #14 this is declared here rather than in a builder: it holds for every form the -- root can take (شأو, فأو, رأو, وأي in the corpus, forms I, II, IV and VIII) and -- depends on nothing else the builders compute. if rfind(vowel_spec.weakness or "", "final%-weak") and vowel_spec.rad2 and rfind(rget(vowel_spec.rad2) or "", HAMZA_ANY) then base.orth.hamza_rad2_final_weak = true end -- Check that the conjugation type is recognized. if not conjugations[conj_type] then error("Unknown conjugation type '" .. conj_type .. "'") end -- The way the conjugation functions work is they always add entries to the appropriate parts of the paradigm -- (each of which is an array), rather than setting the values. This makes it possible to call more than one -- conjugation function and essentially get a paradigm of the "either A or B" kind. Doing this may insert -- duplicate entries into a particular paradigm part, but this is not a problem because we check for duplicate -- entries when adding them, and don't insert in that case. conjugations[conj_type](base, vowel_spec) end -- #13: before postprocess_forms(), which supplies the إدغام half of the alternation. add_naql_idgham_forms(base) postprocess_forms(base) process_slot_overrides(base) -- This should happen before add_missing_links_to_forms() so that the comparison `form == base.lemma` in -- handle_lemma_linked() works correctly and compares unlinked forms to unlinked forms. handle_lemma_linked(base) process_addnote_specs(base) if not base.alternant_multiword_spec.args.noautolinkverb then add_missing_links_to_forms(base) end end local function parse_indicator_spec(angle_bracket_spec) -- Store the original angle bracket spec so we can reconstruct the overall conj spec with the lemma(s) in them. local base = { angle_bracket_spec = angle_bracket_spec, conj_vowels = {}, root_consonants = {}, user_stem_overrides = {}, user_slot_overrides = {}, slot_explicitly_missing = {}, slot_uncertain = {}, slot_override_uses_default = {}, addnote_specs = {}, } local function parse_err(msg) error(msg .. ": " .. angle_bracket_spec) end local function fetch_footnotes(separated_group) local footnotes for j = 2, #separated_group - 1, 2 do if separated_group[j + 1] ~= "" then parse_err("Extraneous text after bracketed footnotes: '" .. table.concat(separated_group) .. "'") end if not footnotes then footnotes = {} end table.insert(footnotes, separated_group[j]) end return footnotes end local inside = angle_bracket_spec:match("^<(.*)>$") assert(inside) local segments = put.parse_multi_delimiter_balanced_segment_run(inside, {{"[", "]"}, {"<", ">"}}) local dot_separated_groups = put.split_alternating_runs_and_strip_spaces(segments, "%.") -- The first dot-separated element must specify the verb form, e.g. IV or IIq. If the form is I, it needs to include -- the the past and non-past vowels, e.g. I/a~u for kataba ~ yaktubu. More than one vowel can be given, -- comma-separated, and more than one past~non-past pair can be given, slash-separated, e.g. I/a,u~u/i~a for form I -- كمل, which can be conjugated as kamala/kamula ~ yakmulu or kamila ~ yakmalu. An individual vowel spec must be one -- of a, i or u and in general (a) at least one past~non-past pair most be given, and (b) both past and non-past -- vowels must be given even though sometimes the vowel can be determined from the unvocalized form. An exception is -- passive-only verbs, where the vowels can't in general be determined (except indirectly in some cases by looking -- at an associated non-passive verb); in that case, the vowel~vowel spec can left out. local slash_separated_groups = put.split_alternating_runs_and_strip_spaces(dot_separated_groups[1], "/") local form_spec = slash_separated_groups[1] base.form_footnotes = fetch_footnotes(form_spec) if form_spec[1] == "" then parse_err("Missing verb form") end if not allowed_vforms_with_weakness_set[form_spec[1]] then parse_err(("Unrecognized verb form '%s', should be one of %s"):format( form_spec[1], list_to_text(allowed_vforms, nil, " or "))) end if form_spec[1]:find("%-") then base.verb_form, base.explicit_weakness = form_spec[1]:match("^(.-)%-(.*)$") else base.verb_form = form_spec[1] end if #slash_separated_groups > 1 then if base.verb_form ~= "I" then parse_err(("Past~non-past vowels can only be specified when verb form is I, but saw form '%s'"):format( base.verb_form)) end for i = 2, #slash_separated_groups do local slash_separated_group = slash_separated_groups[i] local tilde_separated_groups = put.split_alternating_runs_and_strip_spaces(slash_separated_group, "~") if #tilde_separated_groups ~= 2 then parse_err(("Expected two tilde-separated vowel specs: %s"):format(table.concat(slash_separated_group))) end local function parse_conj_vowels(tilde_separated_group, vtype) local conj_vowel_objects = {} local comma_separated_groups = put.split_alternating_runs_and_strip_spaces(tilde_separated_group, ",") for _, comma_separated_group in ipairs(comma_separated_groups) do local conj_vowel = comma_separated_group[1] if conj_vowel ~= "a" and conj_vowel ~= "i" and conj_vowel ~= "u" then parse_err(("Expected %s conjugation vowel '%s' to be one of a, i or u in %s"):format( vtype, conj_vowel, table.concat(slash_separated_group))) end conj_vowel = dia[conj_vowel] local conj_vowel_footnotes = fetch_footnotes(comma_separated_group) -- Try to use strings when possible as it makes q() significantly more efficient. if conj_vowel_footnotes then table.insert(conj_vowel_objects, {form = conj_vowel, footnotes = conj_vowel_footnotes}) else table.insert(conj_vowel_objects, conj_vowel) end end return conj_vowel_objects end local conj_vowel_spec = { past = parse_conj_vowels(tilde_separated_groups[1], "past"), nonpast = parse_conj_vowels(tilde_separated_groups[2], "non-past"), } table.insert(base.conj_vowels, conj_vowel_spec) end end for i = 2, #dot_separated_groups do local dot_separated_group = dot_separated_groups[i] local first_element = dot_separated_group[1] if first_element == "addnote" then local spec_and_footnotes = fetch_footnotes(dot_separated_group) if #spec_and_footnotes < 2 then parse_err("Spec with 'addnote' should be of the form 'addnote[SLOTSPEC][FOOTNOTE][FOOTNOTE][...]'") end local slot_spec = table.remove(spec_and_footnotes, 1) local slot_spec_inside = rmatch(slot_spec, "^%[(.*)%]$") if not slot_spec_inside then parse_err("Internal error: slot_spec " .. slot_spec .. " should be surrounded with brackets") end local slot_specs = rsplit(slot_spec_inside, ",") -- FIXME: Here, [[Module:it-verb]] called strip_spaces(). Generally we don't do this. Should we? table.insert(base.addnote_specs, {slot_specs = slot_specs, footnotes = spec_and_footnotes}) elseif first_element:find("^var:") then if #dot_separated_group > 1 then parse_err(("Can't attach footnotes to 'var:' spec '%s'"):format(first_element)) end base.variant = first_element:match("^var:(.*)$") elseif first_element:find("^I+V?:") then local root_cons, root_cons_value = first_element:match("^(I+V?):(.*)$") local root_index if root_cons == "I" then root_index = 1 elseif root_cons == "II" then root_index = 2 elseif root_cons == "III" then root_index = 3 elseif root_cons == "IV" then root_index = 4 if not base.verb_form:find("q$") then parse_err(("Can't specify root consonant IV for non-quadriliteral verb form '%s': %s"):format( base.verb_form, first_element)) end end local cons, translit = root_cons_value:match("^(.*)//(.*)$") if not cons then cons = root_cons_value end local root_footnotes = fetch_footnotes(dot_separated_group) if not translit and not root_footnotes then base.root_consonants[root_index] = cons else base.root_consonants[root_index] = {form = cons, translit = translit, footnotes = root_footnotes} end elseif first_element:find("^[a-z][a-z0-9_]*:") then local slot_or_stem, remainder = first_element:match("^(.-):(.*)$") dot_separated_group[1] = remainder local comma_separated_groups = put.split_alternating_runs_and_strip_spaces(dot_separated_group, "[,،]") if overridable_stems[slot_or_stem] then if base.user_stem_overrides[slot_or_stem] then parse_err("Overridable stem '" .. slot_or_stem .. "' specified twice") end base.user_stem_overrides[slot_or_stem] = overridable_stems[slot_or_stem](comma_separated_groups, {prefix = slot_or_stem, base = base, parse_err = parse_err, fetch_footnotes = fetch_footnotes}) else -- assume a form override; we validate further later when the possible slots are available if base.user_slot_overrides[slot_or_stem] then parse_err("Form override '" .. slot_or_stem .. "' specified twice") end base.user_slot_overrides[slot_or_stem] = allow_multiple_values_for_override(comma_separated_groups, {prefix = slot_or_stem, base = base, parse_err = parse_err, fetch_footnotes = fetch_footnotes}, "is form override") end elseif indicator_flags[first_element] then if #dot_separated_group > 1 then parse_err("No footnotes allowed with '" .. first_element .. "' spec") end if base[first_element] then parse_err("Spec '" .. first_element .. "' specified twice") end base[first_element] = true else local passive, uncertain = first_element:match("^(.*)(%?)$") passive = passive or first_element uncertain = not not uncertain if passive_types[passive] then if #dot_separated_group > 1 then parse_err("No footnotes allowed with '" .. passive .. "' spec") end if base.passive then parse_err("Value for passive type specified twice") end base.passive = passive base.passive_uncertain = uncertain else parse_err("Unrecognized spec '" .. first_element .. "'") end end end return base end -- Normalize all lemmas, substituting the pagename for blank lemmas and adding links to multiword lemmas. local function normalize_all_lemmas(alternant_multiword_spec, head) -- (1) Add links to all before and after text. Remember the original text so we can reconstruct the verb spec later. if not alternant_multiword_spec.args.noautolinktext then iut.add_links_to_before_and_after_text(alternant_multiword_spec, "remember original") end -- (2) Remove any links from the lemma, but remember the original form so we can use it below in the 'lemma_linked' -- form. iut.map_word_specs(alternant_multiword_spec, function(base) if base.lemma == "" then base.lemma = head end base.user_specified_lemma = base.lemma base.lemma = m_links.remove_links(base.lemma) base.user_specified_verb = base.lemma base.verb = base.user_specified_verb local linked_lemma if alternant_multiword_spec.args.noautolinkverb or base.user_specified_lemma:find("%[%[") then linked_lemma = base.user_specified_lemma else -- Add links to the lemma so the user doesn't specifically need to, since we preserve -- links in multiword lemmas and include links in non-lemma forms rather than allowing -- the entire form to be a link. linked_lemma = iut.add_links(base.user_specified_lemma) end base.linked_lemma = linked_lemma end) end --W: Lexical exceptions — yāʾ-middle roots whose Form IV (أفعل) keeps the yāʾ SOUND -- instead of undergoing the regular إعلال to a hollow verb (أَفْيَلَ, not أَفَالَ). -- These are attested verbs (not the صفة مشبهة like أَخْيَف/خَيْفاء); each is sourced below. -- Only Form IV is affected; forms VII/VIII/X of these roots remain hollow. -- Keyed by rad1..rad2..rad3. local SOUND_FORM_IV_ROOTS = { ["خيل"] = true, -- أَخْيَلَ: أخيلت السماءُ تهيأت للمطر — المعجم الوسيط ج1 ص266 (وبه أَخالت أيضًا) ["غيل"] = true, -- أَغْيَلَ: أغيلته فهي مُغْيِل، سقته الغَيْل — لسان العرب ج11 ص111 (وبه أغالت أيضًا) ["حيج"] = true, -- أَحْيَجَ: أحيجت الأرضُ، على خلاف القياس — تاج العروس ج5 ص500 ["حين"] = true, -- أَحْيَنَ: أحين القومُ آن لهم أن يبلغوا ما أملوه — المعجم الوسيط ج1 ص212 ["خيف"] = true, -- أَخْيَفَ: أخيف الحاجُّ نزلوا خيف منى — كتاب الأفعال لابن القوطية ص35 ["ريف"] = true, -- أَرْيَفَ: أريفت الأرضُ إريافًا (خصبت) — المحيط في اللغة ج10 ص256 ["زين"] = true, -- أَزْيَنَ: أزينت الأرضُ، حسّنها عشبها — مقاييس اللغة ج3 ص41 } --W: Lexical exceptions — hollow roots in the Form I فَعُلَ pattern (past~non-past u~u) -- that keep the middle radical SOUND instead of contracting to a long vowel (هَيُؤَ / -- يَهْيُؤُ, not هَاءَ / يَهُوءُ). This is not the general rule for u~u hollow verbs -- -- طَالَ / يَطُولُ (ط-و-ل) contracts as normal -- so each entry here is a specific -- attested exception, not a phonological class. -- Keyed by rad1..rad2..rad3. -- Keyed with bare hamza (ء) for rad3, matching normalize_root()'s canonicalization of root radicals -- (HAMZA_ON_ALIF etc. all fold to HAMZA) -- the raw root "هيأ" normalizes to radicals ه, ي, ء. local SOUND_FORM_I_UU_ROOTS = { ["هي" .. HAMZA] = true, -- هَيُؤَ ـُ: أنطوان الدحداح، معجم تصريف الأفعال العربية، مكتبة لبنان، ط2، 1995، ص203 } -- Determine weakness from radicals. Used when root given in place of lemma (e.g. for {{ar-verb forms}}). -- `past_vowel`/`nonpast_vowel`, if given, are the past/non-past vowel of a form-I verb (a string or -- footnote-bearing form object, per `req()`); they let SOUND_FORM_I_UU_ROOTS single out the specific -- vowel pattern under which a hollow root is attested as sound, without affecting its other patterns -- (e.g. a root could still contract normally under a different past~non-past pairing). local function weakness_from_radicals(form, rad1, rad2, rad3, rad4, past_vowel, nonpast_vowel) local weakness = nil local quadlit = form:find("q$") -- If weakness unspecified, derive from radicals. if not quadlit then if is_waw_ya(rad3) and rad1 == W and form == "I" then weakness = form_i_w_assimilated(past_vowel, nonpast_vowel) and "assimilated+final-weak" or "final-weak" elseif is_waw_ya(rad3) and vform_supports_final_weak(form) then weakness = "final-weak" elseif rad2 == rad3 and vform_supports_geminate(form) then weakness = "geminate" elseif is_waw_ya(rad2) and vform_supports_hollow(form) then --W: Both wāw (و) and yāʾ (ي) in the middle position create hollow verbs, in form I -- (قَالَ from ق-و-ل, بَاعَ from ب-ي-ع) and in the augmented forms IV, VII, VIII, X -- (أَتَاسَ from ت-ي-س, أَبَانَ from ب-ي-ن, أَطَاحَ from ط-ي-ح — معجم متن اللغة). if form == "IV" and rad2 == Y and SOUND_FORM_IV_ROOTS[rad1 .. rad2 .. rad3] then -- Lexical exceptions whose Form IV keeps the yāʾ sound (أَفْيَلَ); see SOUND_FORM_IV_ROOTS. weakness = "sound" elseif form == "I" and past_vowel and nonpast_vowel and req(past_vowel, U) and req(nonpast_vowel, U) and SOUND_FORM_I_UU_ROOTS[rad1 .. rad2 .. rad3] then -- Lexical exceptions whose Form I فَعُلَ keeps the middle radical sound (هَيُؤَ); see -- SOUND_FORM_I_UU_ROOTS. Gated to u~u specifically since that's the attested pattern. weakness = "sound" else weakness = "hollow" end elseif rad1 == W and form == "I" then -- Vowel-conditioned, like the lemma path below: a wāw-initial form-I verb only -- elides its wāw under a kasra (or lowered-kasra) non-past. Before this, the two -- inference paths in this module disagreed -- infer_radicals() kept the wāw for -- i~a and u~u while this one dropped it for every vowel pattern. weakness = form_i_w_assimilated(past_vowel, nonpast_vowel) and "assimilated" or "sound" else weakness = "sound" end else if is_waw_ya(rad4) then weakness = "final-weak" else weakness = "sound" end end return weakness end -- Join the infixed tāʔ (ت) to the first radical in form VIII verbs. This may cause assimilation of the tāʔ to the -- radical or in some cases the radical to the tāʔ. Used when a root is supplied instead of a lemma (which already has -- the appropriate assimilation in it). local function form_viii_join_ta(rad) if rad == W or rad == Y or rad == "ت" then return "تّ" elseif rad == "د" then return "دّ" elseif rad == "ث" then return "ثّ" elseif rad == "ذ" then return "ذّ" elseif rad == "ز" then return "زْد" elseif rad == "ص" then return "صْط" elseif rad == "ض" then return "ضْط" elseif rad == "ط" then return "طّ" elseif rad == "ظ" then return "ظّ" else return rad .. SK .. "ت" end end local function detect_indicator_spec(base) base.forms = {} base.stem_overrides = {} base.slot_overrides = {} if not base.conj_vowels[1] then -- These may be converted to inferred vowels. If not, we throw an error if form I and not passive-only. base.conj_vowels = {{ past = "-", nonpast = "-", }} else -- If multiple vowels specified for a given vowel type (e.g. a,u~u), expand so that each spec in local expansion = {} for _, spec in ipairs(base.conj_vowels) do for _, past in ipairs(spec.past) do for _, nonpast in ipairs(spec.nonpast) do table.insert(expansion, {past = past, nonpast = nonpast}) end end end base.conj_vowels = expansion end local vform = base.verb_form -- check for quadriliteral form (Iq, IIq, IIIq, IVq) base.quadlit = not not vform:find("q$") -- Infer radicals as necessary. We infer a separate set of radicals for each past~non-past vowel combination because -- they may be different (particularly with form-I hollow verbs). for _, vowel_spec in ipairs(base.conj_vowels) do -- NOTE: rad1, rad2, etc. refer to user-specified radicals, which are formobj tables that optionally specify an -- explicit manual translit, whereas ir1, ir2, etc. refer to inferred radicals, which are either strings or -- lists of possible radicals. local rads = base.root_consonants local rad1, rad2, rad3, rad4 = rads[1], rads[2], rads[3], rads[4] -- Default any unspecified radicals to radicals determined from the headword. The returned radicals may be -- lists of possible radicals, where the first radical should be chosen if the user didn't explicitly specify a -- radical but all are allowed. If `ambig = true` is set in the table, the radical is considered ambiguous and -- categories won't be created for weak radicals. local weakness, ir1, ir2, ir3, ir4 if vform ~= "none" then ir1, ir2, ir3 = rmatch(base.lemma, "^([^_])_([^_])_([^_])$") if not ir1 then ir1, ir2, ir3, ir4 = rmatch(base.lemma, "^([^_])_([^_])_([^_])_([^_])$") end if ir1 then -- root given instead of lemma weakness = weakness_from_radicals(vform, ir1, ir2, ir3, ir4, vowel_spec.past, vowel_spec.nonpast) if vform == "VIII" then vowel_spec.form_viii_assim = form_viii_join_ta(ir1) end else local ret = export.infer_radicals { headword = base.lemma, vform = vform, passive = base.passive, past_vowel = vowel_spec.past, nonpast_vowel = vowel_spec.nonpast, is_reduced = base.reduced, } weakness, ir1, ir2, ir3, ir4 = ret.weakness, ret.rad1, ret.rad2, ret.rad3, ret.rad4 vowel_spec.form_viii_assim = ret.form_viii_assim vowel_spec.past = ret.past_vowel vowel_spec.nonpast = ret.nonpast_vowel vowel_spec.variant = base.variant or ret.variant end end -- For most ambiguous radicals, the choice of radical doesn't matter because it doesn't affect the conjugation -- one way or another. For form I hollow verbs, however, it definitely does. In fact, the choice of radical is -- critical even beyond the past and non-past vowels because it affects the form of the passive participle. So, -- check for this and signal an error if the radical could not be inferred and is not given explicitly. if vform == "I" and type(ir2) == "table" and ir2.need_radical and not rad2 then error("Unable to guess middle radical of hollow form I verb; need to specify radical explicitly") end if vform == "I" and not is_passive_only(base.passive) and ( rget(vowel_spec.past) == "-" or rget(vowel_spec.nonpast) == "-") then error("Form I verb that isn't passive-only or final-weak must have past~non-past vowels specified") end -- Convert ambiguous radicals. local function regularize_inferred_radical(rad) if type(rad) == "table" then if rad.ambig then return {form = rad[1], ambig = true} else return rad[1] end else return rad end end -- Return the appropriate radical at index `index` (1 through 4), based either on the user-specified radical -- `user_radical` or (if unspecified) `inferred_radical`, inferred from the unvocalized lemma. Two values are -- returned, the "regularized" version of the radical (where ambiguous inferred radicals are converted to their -- most likely actual radical) and the non-regularized version. The returned values are form objects rather than -- strings. local function fetch_radical(user_radical, inferred_radical, index) if not user_radical then return regularize_inferred_radical(inferred_radical), inferred_radical else local rad_formval = rget(user_radical) if type(inferred_radical) == "table" then local allowed_radical_set = m_table.listToSet(inferred_radical) if not allowed_radical_set[rad_formval] then error(("For lemma %s, radical %s ambiguously inferred as %s but user radical incompatibly given as %s"): format(base.lemma, index, list_to_text(inferred_radical, nil, " or "), rad_formval)) end elseif rad_formval ~= inferred_radical then error(("For lemma %s, radical %s inferred as %s but user radical incompatibly given as %s"): format(base.lemma, index, inferred_radical, rad_formval)) end return user_radical, user_radical end end if vform ~= "none" then vowel_spec.rad1, vowel_spec.unreg_rad1 = fetch_radical(rad1, ir1, 1) vowel_spec.rad2, vowel_spec.unreg_rad2 = fetch_radical(rad2, ir2, 2) vowel_spec.rad3, vowel_spec.unreg_rad3 = fetch_radical(rad3, ir3, 3) if base.quadlit then vowel_spec.rad4, vowel_spec.unreg_rad4 = fetch_radical(rad4, ir4, 4) end end if vform == "I" then -- If explicit weakness given using 'I-sound' or 'I-assimilated', we may need to adjust the inferred weakness. if base.explicit_weakness == "sound" then if weakness == "assimilated" then weakness = "sound" elseif weakness == "assimilated+final-weak" then -- Verbs like waniya~yawnā "to be faint; to languish" (although the defaults should handle this -- correctly) weakness = "final-weak" else error(("Can't specify form 'I-sound' when inferred weakness is '%s' for lemma %s"):format( weakness, base.lemma)) end elseif base.explicit_weakness == "assimilated" then if weakness == "sound" then -- i~a verbs like waṭiʔa~yaṭaʔu "to tread, to trample"; wasiʕa~yasaʕu "to be spacious; to be well-off"; -- waṯiʔa~yaṯaʔu "to get bruised, to be sprained", which would default to sound. weakness = "assimilated" elseif weakness == "final-weak" then -- For completeness; not clear if any verbs occur where this is needed. (There are plenty of -- assimilated+final-weak verbs but the defaults should take care of them.) weakness = "assimilated+final-weak" else error(("Can't specify form 'I-assimilated' when inferred weakness is '%s' for lemma %s"):format( weakness, base.lemma)) end elseif base.explicit_weakness then error(("Internal error: Unrecognized value '%s' for base.explicit_weakness"):format(base.explicit_weakness)) end elseif vform == "none" then weakness = base.explicit_weakness elseif base.explicit_weakness then error(("Internal error: Explicit weakness should not be specifiable except with forms I and none, but saw explicit weakness '%s' with verb form '%s'"): format(base.explicit_weakness, vform)) end vowel_spec.weakness = weakness if vform ~= "none" then -- Error if radicals are wrong given the weakness. More likely to happen if the weakness is explicitly given -- rather than inferred. Will also happen if certain incorrect letters are included as radicals e.g. hamza on -- top of various letters, alif maqṣūra, tā' marbūṭa. check_radicals(vform, weakness, rget(vowel_spec.rad1), rget(vowel_spec.rad2), rget(vowel_spec.rad3), base.quadlit and rget(vowel_spec.rad4) or nil) end -- Check the variant value. local form_iii_vi_geminate = (vform == "III" or vform == "VI") and rget(vowel_spec.rad2) == rget(vowel_spec.rad3) and not req(vowel_spec.rad2, Y) local hayy_i_x = hayy_radicals(vowel_spec.rad1, vowel_spec.rad2, vowel_spec.rad3) and (vform == "I" or vform == "X") if form_iii_vi_geminate or hayy_i_x then if vowel_spec.variant and vowel_spec.variant ~= "long" and vowel_spec.variant ~= "short" and vowel_spec.variant ~= "both" then error(("For form-III/VI geminate verb or form-I/X verb with ح-ي-ي radicals, saw unrecognized 'var:%s' value; should be 'var:long', 'var:short' or 'var:both'"):format( vowel_spec.variant)) end elseif vowel_spec.variant then error(("Variant value 'var:%s' not allowed in this context"):format(vowel_spec.variant)) end end -- If form I, regroup expanded vowels for display purposes. if vform == "I" then local group_by_past = {} for _, vowel_spec in ipairs(base.conj_vowels) do m_table.insertIfNot(group_by_past, { past = undia[rget(vowel_spec.past)], nonpasts = {undia[rget(vowel_spec.nonpast)]}, }, { key = function(obj) return obj.past end, combine = function(obj1, obj2) for _, nonpast in ipairs(obj2.nonpasts) do m_table.insertIfNot(obj1.nonpasts, nonpast) end end, }) end local group_by_nonpast = {} for _, vowel_spec in ipairs(group_by_past) do m_table.insertIfNot(group_by_nonpast, { pasts = {vowel_spec.past}, nonpasts = vowel_spec.nonpasts, }, { key = function(obj) return obj.nonpasts end, combine = function(obj1, obj2) for _, past in ipairs(obj2.pasts) do m_table.insertIfNot(obj1.pasts, past) end end, }) end base.grouped_conj_vowels = group_by_nonpast end -- Set value of passive. If not specified, default is yes for forms II, III, IV and Iq; no but uncertainly for -- forms VII, IX, XI - XV and IIIq - IVq, as well as form I with past vowel u; impersonal but uncertainly for form -- V, VI, X and IIq, as well as form I with past vowel i; and yes but uncertainly for the remainder (form I with -- past vowel only a and form VIII). if not base.passive then base.passive_defaulted = true -- Temporary tracking for defaulted passives by verb form, weakness and (for form I) past/non-past vowels. track_if_ar_conj(base, "passive-defaulted/" .. vform) for _, vowel_spec in ipairs(base.conj_vowels) do track_if_ar_conj(base, "passive-defaulted/" .. vform.. "/" .. vowel_spec.weakness) if vform == "I" then --W: use rget(); the vowel is a form object when the user attached a footnote -- (e.g. <I/a[نادر]~u>), and undia[<table>] is nil, which crashes format() on Lua 5.1. local past_nonpast = ("%s~%s"):format(undia[rget(vowel_spec.past)], undia[rget(vowel_spec.nonpast)]) track_if_ar_conj(base, "passive-defaulted/I/" .. past_nonpast) track_if_ar_conj(base, "passive-defaulted/I/" .. vowel_spec.weakness .. "/" .. past_nonpast) end end if vform_probably_full_passive(vform) then base.passive = "pass" else base.passive_uncertain = true for _, vowel_spec in ipairs(base.conj_vowels) do if vform_probably_no_passive(vform, vowel_spec.weakness, vowel_spec.past, vowel_spec.nonpast) then base.passive = "nopass" break elseif vform_probably_impersonal_passive(vform, vowel_spec.weakness, vowel_spec.past, vowel_spec.nonpast) then base.passive = "ipass" break end end base.passive = base.passive or "pass" end end -- NOTE: Currently there are no built-in stems or form overrides for Arabic; this code is inherited from -- [[Module:ca-verb]], where such things do exist, and is kept for generality in case we decide in the future to -- implement such things. -- Override built-in verb stems and overrides with user-specified ones. for stem, values in pairs(base.user_stem_overrides) do base.stem_overrides[stem] = values end for slot, values in pairs(base.user_slot_overrides) do if not base.alternant_multiword_spec.verb_slots_map[slot] then error("Unrecognized override slot '" .. slot .. "': " .. base.angle_bracket_spec) end if export.unsettable_slots_set[slot] then error("Slot '" .. slot .. "' cannot be set using an override: " .. base.angle_bracket_spec) end if skip_slot(base, slot, "allow overrides") then error("Override slot '" .. slot .. "' would be skipped based on the passive, 'noimp' and/or 'no_nonpast' settings: " .. base.angle_bracket_spec) end base.slot_overrides[slot] = values end if base.verb_form == "none-final-weak" then for _, stem_type in ipairs { "past", "past_pass", "nonpast", "nonpast_pass" } do if base.stem_overrides[stem_type .. "_c"] or base.stem_overrides[stem_type .. "_v"] then error(("Specify past stem for verb type 'none-final-weak' using '%s:...' not '%s_c:...' or '%s_v:...'"): format(stem_type, stem_type, stem_type)) end end for _, stem_type in ipairs { "past", "nonpast" } do if base.stem_overrides[stem_type] or not base.stem_overrides[stem_type .. "_final_weak_vowel"] then error(("For verb type 'none-final-weak', if '%s:...' specified, so must '%s_final_weak_vowel:...'"): format(stem_type, stem_type)) end end end end local function detect_all_indicator_specs(alternant_multiword_spec) add_slots(alternant_multiword_spec) alternant_multiword_spec.verb_forms = {} -- This means at least one individual base had the slot marked as explicitly missing. Another base (e.g. when -- there are multiple alternants) might have a value for the slot. In practice, we only respect this when there are -- no overall values in the slot and `slot_uncertain` isn't set; in this case, we display "no ..." for the slot -- instead of simply not displaying anything for the slot. alternant_multiword_spec.slot_explicitly_missing = {} -- This means at least one individual base had no values for the slot and the slot marked as explicitly uncertain. -- Note that this is different from a value being present but marked as uncertain (e.g. if an override was given -- with a ? after it); this causes the form object for the value to have `uncertain = true` set. If there are no -- overall values in the slot and `slot_uncertain` is set, we display this in the headword. alternant_multiword_spec.slot_uncertain = {} iut.map_word_specs(alternant_multiword_spec, function(base) -- So arguments, etc. can be accessed. WARNING: Creates circular reference. base.alternant_multiword_spec = alternant_multiword_spec detect_indicator_spec(base) if not base.nocat then m_table.insertIfNot(alternant_multiword_spec.verb_forms, base.verb_form) end if base.passive then alternant_multiword_spec.passive = base.passive end if base.passive_uncertain then alternant_multiword_spec.passive_uncertain = true end for slot, _ in pairs(base.slot_explicitly_missing) do alternant_multiword_spec.slot_explicitly_missing[slot] = true end end) end local function determine_slot_uncertainty_from_forms(alternant_multiword_spec) iut.map_word_specs(alternant_multiword_spec, function(base) -- If no verbal noun and verb form is not 'none' (manually-specified stems) — which currently only happens for -- form I — and the verbal noun wasn't explicitly indicated as missing using <vn:->, we assume it's just -- unknown/unspecified rather than missing. Same with active participles. for uncertain_slot, _ in pairs(slots_that_may_be_uncertain) do if not base.forms[uncertain_slot] and vform ~= "none" and not skip_slot(base, uncertain_slot) then base.slot_uncertain[uncertain_slot] = true end end -- Propagate slot uncertainty up. Currently only the verbal noun can have this set but we write the code -- generally. for slot, _ in pairs(base.slot_uncertain) do alternant_multiword_spec.slot_uncertain[slot] = true end end) -- If slot is uncertain and has no value, explicitly set its value to "?". for uncertain_slot, _ in pairs(slots_that_may_be_uncertain) do if not alternant_multiword_spec.forms[uncertain_slot] and alternant_multiword_spec.slot_uncertain[uncertain_slot] then alternant_multiword_spec.forms[uncertain_slot] = {{form = "?"}} end end end -- Determine certain properties of the verb from the overall forms, such as whether the verb is active-only or -- passive-only, is impersonal, lacks an imperative, etc. local function determine_verb_properties_from_forms(alternant_multiword_spec) alternant_multiword_spec.has_active = false alternant_multiword_spec.has_passive = false alternant_multiword_spec.has_non_impers_active = false alternant_multiword_spec.has_non_impers_passive = false alternant_multiword_spec.has_imp = false alternant_multiword_spec.has_past = false alternant_multiword_spec.has_nonpast = false for slot, _ in pairs(alternant_multiword_spec.forms) do if slot == "ap" or slot:find("[123]") and not slot:find("_pass") then alternant_multiword_spec.has_active = true end if slot == "pp" or slot:find("[123]") and slot:find("_pass") then alternant_multiword_spec.has_passive = true end if slot:find("[123]") and not slot:find("pass_[123]") and not slot:find("3ms") then alternant_multiword_spec.has_non_impers_active = true end if slot:find("pass_[123]") and not slot:find("3ms") then alternant_multiword_spec.has_non_impers_passive = true end if slot:find("^imp_") then alternant_multiword_spec.has_imp = true end if slot:find("^past_") then alternant_multiword_spec.has_past = true end if slot:find("^ind_") or slot:find("^sub_") or slot:find("^juss_") then alternant_multiword_spec.has_nonpast = true end end end local function add_categories_and_annotation(alternant_multiword_spec, base, multiword_lemma, insert_ann, insert_cat) -- Useful e.g. in constructing suppletive verbs out of parts. For a verb like جاء or أتى whose imperative comes -- from the unrelated verb تعالى, we don't want the latter verb showing up in categories or annotations. if base.nocat then return end local vform = base.verb_form if vform ~= "none" then insert_ann("form", vform) insert_cat("form-" .. vform .. " verbs") --W: Add morphological pattern corresponding to the form local morph_pattern = FORM_TO_MORPHOLOGICAL_PATTERN[vform] if morph_pattern then insert_ann("morph_pattern", morph_pattern) end end if base.reduced then insert_ann("reduced", "reduced") if vform ~= "none" then insert_cat("form-" .. vform .. " reduced verbs") end end if base.quadlit then insert_cat("verbs with quadriliteral roots") end if base.passive_defaulted then insert_cat("verbs with defaulted passive") end for _, vowel_spec in ipairs(base.conj_vowels) do local rad1, rad2, rad3, rad4 = get_radicals_4(vowel_spec) local final_weak = is_final_weak(base, vowel_spec) local weakness = vowel_spec.weakness -- We have to distinguish weakness by form and weakness by conjugation. Weakness by form merely indicates the -- presence of weak letters in certain positions in the radicals. Weakness by conjugation is related to how the -- verbs are conjugated. For example, form-II verbs that are "hollow by form" (middle radical is wāw or yāʾ) are -- conjugated as sound verbs. Another example: form-I verbs with initial wāw are "assimilated by form" and most -- are assimilated by conjugation as well, but a few are sound by conjugation, e.g. wajuha yawjuhu "to be -- distinguished" (rather than wajuha yajuhu); similarly for some hollow-by-form verbs in various forms, e.g. -- form VIII izdawaja yazdawiju "to be in pairs" (rather than izdāja yazdāju). Categories referring to weakness -- always refer to weakness by conjugation; weakness by form is distinguished only by categories such as -- [[:Category:Arabic form-III verbs with و as second radical]]. insert_ann("weakness", weakness) if vform ~= "none" then insert_cat(("%s form-%s verbs"):format(weakness, vform)) end local function radical_is_ambiguous(rad) return type(rad) == "table" and rad.ambig end local function radical_is_unambiguous_weak(rad) return not radical_is_ambiguous(rad) and (is_waw_ya(rad) or req(rad, HAMZA)) end if vform ~= "none" then local ur1, ur2, ur3, ur4 = vowel_spec.unreg_rad1, vowel_spec.unreg_rad2, vowel_spec.unreg_rad3, vowel_spec.unreg_rad4 -- Create headword categories based on the radicals. Do the following before -- converting the Latin radicals into Arabic ones so we distinguish -- between ambiguous and non-ambiguous radicals. if radical_is_ambiguous(ur1) or radical_is_ambiguous(ur2) or radical_is_ambiguous(ur3) or ur4 and radical_is_ambiguous(ur4) then insert_cat("verbs with ambiguous radicals") end if radical_is_unambiguous_weak(ur1) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur1) .. " as first radical") end if radical_is_unambiguous_weak(ur2) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur2) .. " as second radical") end if radical_is_unambiguous_weak(ur3) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur3) .. " as third radical") end if ur4 and radical_is_unambiguous_weak(ur4) then insert_cat("form-" .. vform .. " verbs with " .. rget(ur4) .. " as fourth radical") end end end if vform == "I" and not is_passive_only(base.passive) then for _, vowel_spec in ipairs(base.grouped_conj_vowels) do insert_ann("vowels", ("%s ~ %s"):format(table.concat(vowel_spec.pasts, "/"), table.concat(vowel_spec.nonpasts, "/"))) for _, past in ipairs(vowel_spec.pasts) do for _, nonpast in ipairs(vowel_spec.nonpasts) do if past == "-" or nonpast == "-" then error("Internal error: Saw form I past vowel %s and non-past vowel %s but - in place of vowel should have triggered an error earlier") end insert_cat(("form-I verbs with past vowel %s and non-past vowel %s"):format(past, nonpast)) --W: Add morphological pattern for form I based on vowels local form_i_with_vowels = ("I/%s~%s"):format(past, nonpast) local morph_pattern = FORM_TO_MORPHOLOGICAL_PATTERN[form_i_with_vowels] if morph_pattern then insert_ann("morph_pattern", morph_pattern) end end end end end for slot, name in pairs(slots_that_may_be_uncertain) do if base.slot_uncertain[slot] then -- An unspecified and non-defaulted verbal noun (form I) is considered uncertain rather than explicitly -- missing. Use <vn:-> to explicitly indicate the lack of verbal noun. Same for form-I stative active -- participles. insert_cat(("verbs with unknown or uncertain %ss"):format(name)) end end if base.irregular then insert_ann("irreg", "irregular") insert_cat("irregular verbs") end end -- Compute the categories to add the verb to, as well as the annotation to display in the conjugation title bar. We -- combine the code to do these functions as both categories and title bar contain similar information. local function compute_categories_and_annotation(alternant_multiword_spec) alternant_multiword_spec.categories = {} local ann = {} alternant_multiword_spec.annotation = ann ann.form = {} ann.morph_pattern = {} --W: arabic morphological pattern corresponding to the form ann.weakness = {} ann.vowels = {} ann.passive = nil ann.reduced = {} ann.irreg = {} ann.defective = {} local multiword_lemma = false for _, slot in ipairs(export.potential_lemma_slots) do if alternant_multiword_spec.forms[slot] then for _, formobj in ipairs(alternant_multiword_spec.forms[slot]) do if formobj.form:find(" ") then multiword_lemma = true break end end break end end local function insert_ann(anntype, value) m_table.insertIfNot(alternant_multiword_spec.annotation[anntype], value) end local function insert_cat(cat, also_when_multiword) -- Don't place multiword terms in categories like 'Arabic form-II verbs' to avoid spamming the categories with -- such terms. if also_when_multiword or not multiword_lemma then m_table.insertIfNot(alternant_multiword_spec.categories, "Arabic " .. cat) end end iut.map_word_specs(alternant_multiword_spec, function(base) add_categories_and_annotation(alternant_multiword_spec, base, multiword_lemma, insert_ann, insert_cat) end) for slot, name in pairs(slots_that_may_be_uncertain) do if alternant_multiword_spec.forms[slot] then for _, form in ipairs(alternant_multiword_spec.forms[slot]) do if form.uncertain then if form.form == "?" then insert_cat(("verbs with explicitly unknown %ss"):format(name)) else insert_cat(("verbs needing %s checked"):format(name)) end break end end end end if alternant_multiword_spec.has_active then if alternant_multiword_spec.has_passive and alternant_multiword_spec.has_non_impers_passive then insert_cat("verbs with full passive") ann.passive = "full passive" elseif alternant_multiword_spec.has_passive then insert_cat("verbs with impersonal passive") ann.passive = "impersonal passive" else insert_cat("verbs lacking passive forms") ann.passive = "no passive" end else if alternant_multiword_spec.has_non_impers_passive then insert_cat("passive verbs") insert_cat("verbs with full passive") ann.passive = "passive-only" else insert_cat("passive verbs") insert_cat("impersonal verbs") insert_cat("verbs with impersonal passive") ann.passive = "impersonal (passive-only)" end end if alternant_multiword_spec.passive_uncertain then insert_cat("verbs needing passive checked") ann.passive = ann.passive .. ' <abbr title="passive status uncertain">(?)</abbr>' end if alternant_multiword_spec.has_active and not alternant_multiword_spec.has_imp then insert_ann("defective", "no imperative") insert_cat("verbs lacking imperative forms") end if not alternant_multiword_spec.has_past then insert_ann("defective", "no past") insert_cat("verbs lacking past forms") end if not alternant_multiword_spec.has_nonpast then insert_ann("defective", "no non-past") insert_cat("verbs lacking non-past forms") end local ann_parts = {} local function insert_ann_part(part, conj) local val = table.concat(ann[part], conj or " or ") if val ~= "" and val ~= "regular" then table.insert(ann_parts, val) end end insert_ann_part("form") insert_ann_part("weakness") insert_ann_part("reduced") insert_ann_part("vowels") if ann.passive then table.insert(ann_parts, ann.passive) end insert_ann_part("irreg") insert_ann_part("defective", "، ") --W: Save morph_pattern before overwriting annotation with string alternant_multiword_spec.morph_pattern = ann.morph_pattern alternant_multiword_spec.annotation = table.concat(ann_parts, "، ") end local function show_forms(alternant_multiword_spec) local lemmas = {} for _, slot in ipairs(export.potential_lemma_slots) do if alternant_multiword_spec.forms[slot] then for _, formobj in ipairs(alternant_multiword_spec.forms[slot]) do table.insert(lemmas, formobj) end break end end alternant_multiword_spec.lemmas = lemmas -- save for later use in make_table() alternant_multiword_spec.vn = alternant_multiword_spec.forms.vn -- save for later use in make_table() -- Reconstruct the original verb spec without overrides for verbal nouns and participles, since those specific slots -- are ignored by {{ar-verb form}}. Compute this once beforehand; `transform_accel_obj` is called repeatedly on each -- form and we don't want to compute this repeatedly. local reconstructed_verb_spec = iut.reconstruct_original_spec(alternant_multiword_spec, { preprocess_angle_bracket_spec = function(spec) spec = spec:match("^<(.*)>$") assert(spec) local segments = put.parse_multi_delimiter_balanced_segment_run(spec, {{"[", "]"}, {"<", ">"}}) local dot_separated_groups = put.split_alternating_runs_and_strip_spaces(segments, "%.") -- Rejoin each dot-separated group into a single string, since we aren't actually going to do any parsing -- of bracket-bounded textual runs; then filter out overrides for verbal nouns and participles. local filtered_indicators = {} for _, dot_separated_group in ipairs(dot_separated_groups) do local indicator = table.concat(dot_separated_group) -- FIXME: Do we want to filter out any other indicators? if not (indicator:find("^vn:") or indicator:find("^[ap]p:")) then table.insert(filtered_indicators, indicator) end end return ("<%s>"):format(table.concat(filtered_indicators, ".")) end, }) -- If we're dealing with a single word, no alternants and a single verb form, use the auto-conjugation-fetching -- variant. local reconstructed_lemma, inside = reconstructed_verb_spec:match("^([^ <>()]+)(%b<>)$") if inside and alternant_multiword_spec.verb_forms[1] and not alternant_multiword_spec.verb_forms[2] then reconstructed_verb_spec = ("+%s<%s>"):format(reconstructed_lemma, alternant_multiword_spec.verb_forms[1]) end local function transform_accel_obj(slot, formobj, accel_obj) if not accel_obj then return accel_obj end if slot == "ap" or slot == "pp" or slot == "vn" then -- FIXME: [[Module:accel]] can't correctly handle more than one verb form for participles and verbal nouns accel_obj.form = slot .. "-" .. table.concat(alternant_multiword_spec.verb_forms, "،") else accel_obj.form = "verb-form-" .. reconstructed_verb_spec end return accel_obj end local function generate_link(data) local form = data.form local term = form.formval_for_link local alt = form.alt if term == "?" then term = nil alt = "?" end --W: Create simple link for verb forms without #Arabic anchor local display = alt or term local link if display then link = link_term_simple(display) else link = "" end link = link .. iut.get_footnote_text(form.footnotes, data.footnote_obj) --W: qualifier/label rendering removed along with the q/qq/l/ll modifiers; -- it required [[Module:pron qualifier]], absent on ar.wiktionary. return link end local props = { lang = lang, lemmas = lemmas, transform_accel_obj = transform_accel_obj, generate_link = generate_link, slot_list = alternant_multiword_spec.verb_slots, include_translit = false, } iut.show_forms(alternant_multiword_spec.forms, props) end --W: Tatweel and the invisible direction/joining marks survive copy-paste and -- MediaWiki keeps them in page titles, so strip them from both sides before -- comparing a |بؤرة= value against a generated form. local FOCUS_IGNORE = "[" .. u(0x0640) .. u(0x061C) .. u(0x200B) .. "-" .. u(0x200F) .. u(0x202A) .. "-" .. u(0x202E) .. u(0xFEFF) .. "]" local function normalize_for_focus(s) return (rsub(s, FOCUS_IGNORE, "")) end --W: Which slots hold `focus`? Must run before show_forms() turns the form -- lists into rendered wikitext. local function find_focus_slots(alternant_multiword_spec, focus) if not focus or focus == "" then return nil end local target = normalize_for_focus(focus) local slots = {} for slot, forms in pairs(alternant_multiword_spec.forms) do if type(forms) == "table" then for _, formobj in ipairs(forms) do if type(formobj) == "table" and formobj.form and normalize_for_focus(formobj.form) == target then slots[slot] = true break end end end end return slots end --W: 'yes' unless |موسع= asked for the table open. [[Template:inflection-table-top]] -- turns any non-empty |tall= into the mw-collapsed class. local function collapsed_arg(alternant_multiword_spec) if alternant_multiword_spec.expanded then return nil end return 'yes' end --W: Put the marker class on the cell itself rather than on a span around its -- contents, so the highlight covers the whole cell instead of hugging the word. -- Runs on the table wikitext below, before m_string_utilities.format() fills the -- {slot} placeholders in. Every form cell in either table has one of two shapes: -- -- | {past_1s} -> | class="avf-focus" | {past_1s} -- | rowspan="2" | {past_1s} -> | rowspan="2" class="avf-focus" | {past_1s} -- -- Styled by [[Template:ar-verb-form/style.css]] (.inflection-table td.avf-focus). local function mark_focus_cells(text, focus_slots) if not focus_slots then return text end return (rsub(text, "\n(|[^\n]*)", function(cell) local slot = rmatch(cell, "^|%s*{([%w_]+)}%s*$") if slot then if focus_slots[slot] then return '\n| class="avf-focus" | {' .. slot .. "}" end return nil end local attrs attrs, slot = rmatch(cell, "^|%s*(.-)%s*|%s*{([%w_]+)}%s*$") if slot and focus_slots[slot] then return "\n| " .. attrs .. ' class="avf-focus" | {' .. slot .. "}" end return nil end)) end ------------------------------------------------------------------------------- -- Functions to create inflection tables -- ------------------------------------------------------------------------------- -- Make the conjugation table. Called from export.show(). local function make_table(alternant_multiword_spec) local text = mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-top', args = { title = 'تصريف {title}', --W: |tall= is what puts mw-collapsed on the wrapper. |موسع= drops it. -- (Not `expanded and nil or 'yes'` — in Lua that is always 'yes'.) tall = collapsed_arg(alternant_multiword_spec), palette = 'green', category = 'conjugation', class = 'tr-alongside', -- temp hack to prevent extra line break } } text = text .. [=[ ! colspan="6" | <<الْمَصْدَر>> | colspan="7" | {vn} ]=] if alternant_multiword_spec.has_active then text = text .. [=[ |- ! colspan="6" | <<اِسْم الْفَاعِل>> | colspan="7" | {ap} ]=] end if alternant_multiword_spec.has_passive then text = text .. [=[ |- ! colspan="6" | <<اِسْم الْمَفْعُول>> | colspan="7" | {pp} ]=] end text = text .. [=[ |- ! colspan="999" class="separator" | ]=] if alternant_multiword_spec.has_active then text = text .. [=[ |- ! colspan="12" class="outer" | <<الْفِعْل الْمَعْلُوم>> |- ! colspan="2" | ! colspan="3" | <<الْمُفْرَد>> ! rowspan="12" class="separator" | ! colspan="2" | <<الْمُثَنَّى>> ! rowspan="12" class="separator" | ! colspan="3"| <<الْجَمْع>> |- ! colspan="2"| ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> |- ! rowspan="2" | <<الْمَاضِي>> ! class="secondary" | المذكر | rowspan="2" | {past_1s} | {past_2ms} | {past_3ms} | rowspan="2" | {past_2d} | {past_3md} | rowspan="2" | {past_1p} | {past_2mp} | {past_3mp} |- ! class="secondary" | المؤنث | {past_2fs} | {past_3fs} | {past_3fd} | {past_2fp} | {past_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَرْفُوع>> ! class="secondary" | المذكر | rowspan="2" | {ind_1s} | {ind_2ms} | {ind_3ms} | rowspan="2" | {ind_2d} | {ind_3md} | rowspan="2" | {ind_1p} | {ind_2mp} | {ind_3mp} |- ! class="secondary" | المؤنث | {ind_2fs} | {ind_3fs} | {ind_3fd} | {ind_2fp} | {ind_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَنْصُوب>> ! class="secondary" | المذكر | rowspan="2" | {sub_1s} | {sub_2ms} | {sub_3ms} | rowspan="2" | {sub_2d} | {sub_3md} | rowspan="2" | {sub_1p} | {sub_2mp} | {sub_3mp} |- ! class="secondary" | المؤنث | {sub_2fs} | {sub_3fs} | {sub_3fd} | {sub_2fp} | {sub_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَجْزُوم>> ! class="secondary" | المذكر | rowspan="2" | {juss_1s} | {juss_2ms} | {juss_3ms} | rowspan="2" | {juss_2d} | {juss_3md} | rowspan="2" | {juss_1p} | {juss_2mp} | {juss_3mp} |- ! class="secondary" | المؤنث | {juss_2fs} | {juss_3fs} | {juss_3fd} | {juss_2fp} | {juss_3fp} |- ! rowspan="2" | <<الْأَمْر>> ! class="secondary" | المذكر | rowspan="2" | | {imp_2ms} | rowspan="2" | | rowspan="2" | {imp_2d} | rowspan="2" | | rowspan="2" | | {imp_2mp} | rowspan="2" | |- ! class="secondary" | المؤنث | {imp_2fs} | {imp_2fp} ]=] end if alternant_multiword_spec.has_passive then text = text .. [=[ |- ! colspan="999" class="separator" | |- ! colspan="12" class="outer" | <<الْفِعْل الْمَجْهُول>> |- ! colspan="2" | ! colspan="3" | <<الْمُفْرَد>> ! rowspan="10" class="separator" | ! colspan="2" | <<الْمُثَنَّى>> ! rowspan="10" class="separator" | ! colspan="3" | <<الْجَمْع>> |- ! colspan="2" | ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> ! <<الْمُتَكَلِّم>> ! <<الْمُخَاطَب>> ! <<الْغَائِب>> |- ! rowspan="2" | <<الْمَاضِي>> ! class="secondary" | المذكر | rowspan="2" | {past_pass_1s} | {past_pass_2ms} | {past_pass_3ms} | rowspan="2" | {past_pass_2d} | {past_pass_3md} | rowspan="2" | {past_pass_1p} | {past_pass_2mp} | {past_pass_3mp} |- ! class="secondary" | المؤنث | {past_pass_2fs} | {past_pass_3fs} | {past_pass_3fd} | {past_pass_2fp} | {past_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَرْفُوع>> ! class="secondary" | المذكر | rowspan="2" | {ind_pass_1s} | {ind_pass_2ms} | {ind_pass_3ms} | rowspan="2" | {ind_pass_2d} | {ind_pass_3md} | rowspan="2" | {ind_pass_1p} | {ind_pass_2mp} | {ind_pass_3mp} |- ! class="secondary" | المؤنث | {ind_pass_2fs} | {ind_pass_3fs} | {ind_pass_3fd} | {ind_pass_2fp} | {ind_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَنْصُوب>> ! class="secondary" | المذكر | rowspan="2" | {sub_pass_1s} | {sub_pass_2ms} | {sub_pass_3ms} | rowspan="2" | {sub_pass_2d} | {sub_pass_3md} | rowspan="2" | {sub_pass_1p} | {sub_pass_2mp} | {sub_pass_3mp} |- ! class="secondary" | المؤنث | {sub_pass_2fs} | {sub_pass_3fs} | {sub_pass_3fd} | {sub_pass_2fp} | {sub_pass_3fp} |- ! rowspan="2" | <<الْمُضَارِع الْمَجْزُوم>> ! class="secondary" | المذكر | rowspan="2" | {juss_pass_1s} | {juss_pass_2ms} | {juss_pass_3ms} | rowspan="2" | {juss_pass_2d} | {juss_pass_3md} | rowspan="2" | {juss_pass_1p} | {juss_pass_2mp} | {juss_pass_3mp} |- ! class="secondary" | المؤنث | {juss_pass_2fs} | {juss_pass_3fs} | {juss_pass_3fd} | {juss_pass_2fp} | {juss_pass_3fp} ]=] end text = text .. mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-bottom', args = { notes = '{footnote}', } } local forms = alternant_multiword_spec.forms if not alternant_multiword_spec.lemmas then forms.title = "—" else local linked_lemmas = {} for _, form in ipairs(alternant_multiword_spec.lemmas) do --W: Create simple link for lemmas without #Arabic anchor -- table.insert(linked_lemmas, link_term(form.form, "term")) table.insert(linked_lemmas, link_term_simple(form.form)) end forms.title = table.concat(linked_lemmas, "، ") end local ann_parts = {} if alternant_multiword_spec.annotation ~= "" then table.insert(ann_parts, alternant_multiword_spec.annotation) end if alternant_multiword_spec.vn then local linked_vns = {} for _, form in ipairs(alternant_multiword_spec.vn) do table.insert(linked_vns, link_term(form.form, "term")) end table.insert(ann_parts, (#linked_vns > 1 and "verbal nouns" or "verbal noun") .. " " .. table.concat(linked_vns, "، ")) end local annotation = table.concat(ann_parts, "، ") --W: use only arabic morphological pattern in title annotation local morph_patterns = alternant_multiword_spec.morph_pattern if morph_patterns and #morph_patterns > 0 then annotation = table.concat(morph_patterns, "، ") forms.title = forms.title .. " (" .. annotation .. ")" end -- Format the table. text = mark_focus_cells(text, alternant_multiword_spec.focus_slots) local tagged_table = rsub(text, "<<(.-)>>", tag_text) return m_string_utilities.format(tagged_table, forms) end -- Helper function to get root letters and verb classification for simplified table header -- Returns: root_display (e.g., "ح س ب"), verb_type (e.g., "فعل ثلاثي صحيح سالم") local function get_verb_info(alternant_multiword_spec) local root_display = "" local verb_type = "فعل" local rad1, rad2, rad3, rad4 -- Extract radicals from the base word spec using iut.map_word_specs -- This accesses the radicals directly from the parsed verb specification iut.map_word_specs(alternant_multiword_spec, function(base) -- Get radicals from the first conjugation vowel spec if base.conj_vowels and #base.conj_vowels > 0 then local vowel_spec = base.conj_vowels[1] -- Extract radicals, handling both table and string types local function extract_radical(rad) if type(rad) == "table" then -- If it's a table, get the first element or the string representation return rad[1] or rget(rad) else return rad end end rad1 = extract_radical(vowel_spec.rad1) rad2 = extract_radical(vowel_spec.rad2) rad3 = extract_radical(vowel_spec.rad3) rad4 = vowel_spec.rad4 and extract_radical(vowel_spec.rad4) end end) -- Get basic/derived classification from morphological pattern local basic_deriv if alternant_multiword_spec.morph_pattern and #alternant_multiword_spec.morph_pattern > 0 then local morph_pattern = alternant_multiword_spec.morph_pattern[1] basic_deriv = MORPHOLOGICAL_PATTERN_TO_BASIC_DERIV[morph_pattern] end -- Format root display with spaces between letters if rad1 and rad2 and rad3 then if rad4 then -- Quadrilateral root_display = rad1 .. " " .. rad2 .. " " .. rad3 .. " " .. rad4 verb_type = "فعل رباعي" if basic_deriv then verb_type = verb_type .. " " .. basic_deriv end else -- triliteral root_display = rad1 .. " " .. rad2 .. " " .. rad3 verb_type = "فعل ثلاثي" if basic_deriv then verb_type = verb_type .. " " .. basic_deriv if basic_deriv == "مُجرَّد" then --W: the classification concerns only triliteral basic verbs (الثلاثي المُجرَّد) local classification = export.classify_triliteral_verb(rad1, rad2, rad3) verb_type = verb_type .. " " .. classification end end end end return root_display, verb_type end -- Make simplified conjugation table for ar.wiktionary. Called from export.show(). -- This table shows only past, present indicative, and imperative forms. -- No passive voice, no participles, no verbal nouns. local function make_table_ar(alternant_multiword_spec) local text = mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-top', args = { title = 'تصريف {title}', --W: |tall= is what puts mw-collapsed on the wrapper. |موسع= drops it. -- (Not `expanded and nil or 'yes'` — in Lua that is always 'yes'.) tall = collapsed_arg(alternant_multiword_spec), palette = 'green', category = 'conjugation', class = 'tr-alongside', -- temp hack to prevent extra line break } } -- Get verb root and classification local root_display, verb_type = get_verb_info(alternant_multiword_spec) -- First row: Root and verb type spanning all columns text = text .. "! colspan=\"2\" | <<الجذر>>: " .. root_display .. "\n" text = text .. "| colspan=\"3\" | <<'''" .. verb_type .. "'''>>\n" text = text .. [=[ |- ! colspan="2" | <<الضمائر>> ! <<الماضي>> ! <<المضارع>> ! <<الأمر>> |- ! rowspan="2" | <<المتكلّم>> | <<أنا>> | {past_1s} | {ind_1s} | |- | <<نحن>> | {past_1p} | {ind_1p} | |- ! rowspan="5" | <<المخاطب>> | <<أنتَ>> | {past_2ms} | {ind_2ms} | {imp_2ms} |- | <<أنتِ>> | {past_2fs} | {ind_2fs} | {imp_2fs} |- | <<أنتما>> | {past_2d} | {ind_2d} | {imp_2d} |- | <<أنتم>> | {past_2mp} | {ind_2mp} | {imp_2mp} |- | <<أنتنّ>> | {past_2fp} | {ind_2fp} | {imp_2fp} |- ! rowspan="6" | <<الغائب>> | <<هو>> | {past_3ms} | {ind_3ms} | |- | <<هي>> | {past_3fs} | {ind_3fs} | |- | <<هما>> | {past_3md} | {ind_3md} | |- | <<هما>> | {past_3fd} | {ind_3fd} | |- | <<هم>> | {past_3mp} | {ind_3mp} | |- | <<هنّ>> | {past_3fp} | {ind_3fp} | ]=] text = text .. mw.getCurrentFrame():expandTemplate{ title = 'inflection-table-bottom', args = { notes = '{footnote}', } } local forms = alternant_multiword_spec.forms if not alternant_multiword_spec.lemmas then forms.title = "—" else local linked_lemmas = {} for _, form in ipairs(alternant_multiword_spec.lemmas) do --W: Create simple link for lemmas without #Arabic anchor -- table.insert(linked_lemmas, link_term(form.form, "term")) table.insert(linked_lemmas, link_term_simple(form.form)) end forms.title = table.concat(linked_lemmas, "، ") end local ann_parts = {} if alternant_multiword_spec.annotation ~= "" then table.insert(ann_parts, alternant_multiword_spec.annotation) end local annotation = table.concat(ann_parts, "، ") --W: use only arabic morphological pattern in title annotation local morph_patterns = alternant_multiword_spec.morph_pattern if morph_patterns and #morph_patterns > 0 then annotation = table.concat(morph_patterns, "، ") forms.title = forms.title .. " (" .. annotation .. ")" end -- Format the table. text = mark_focus_cells(text, alternant_multiword_spec.focus_slots) local tagged_table = rsub(text, "<<(.-)>>", tag_text) return m_string_utilities.format(tagged_table, forms) end ------------------------------------------------------------------------------- -- Verb classification functions -- ------------------------------------------------------------------------------- --[=[ تصنيف الفعل الثلاثي المُجرَّد حسب نوعه Classify an Arabic triliteral basic verb root based on its three letters. @param r1 الحرف الأول (first root letter - الفاء) @param r2 الحرف الثاني (second root letter - العين) @param r3 الحرف الثالث (third root letter - اللام) @return str: نوع الفعل (verb type classification in Arabic) Examples: classify_triliteral_verb('ك', 'ت', 'ب') --> 'صحيح سالم' classify_triliteral_verb('أ', 'ت', 'ي') --> 'معتل ناقص يائي مهموز الفاء' classify_triliteral_verb('ق', 'ل', 'و') --> 'معتل ناقص واوي' ]=] function export.classify_triliteral_verb(r1, r2, r3) -- تعريف أحرف العلة والهمزة -- الحروف الضعيفة (أحرف العلة): واو وياء فقط local weak_letters = {[W] = true, [Y] = true} -- الهمزة: جميع أشكال الهمزة (ء، أ، إ، ؤ، ئ) local hamza_letters = { [HAMZA] = true, -- ء [HAMZA_ON_ALIF] = true, -- أ [HAMZA_UNDER_ALIF] = true, -- إ [HAMZA_ON_W] = true, -- ؤ [HAMZA_ON_Y] = true -- ئ } -- فحص كل حرف local is_hamza = {hamza_letters[r1] or false, hamza_letters[r2] or false, hamza_letters[r3] or false} local is_weak = {weak_letters[r1] or false, weak_letters[r2] or false, weak_letters[r3] or false} -- عدد أحرف العلة local weak_count = (is_weak[1] and 1 or 0) + (is_weak[2] and 1 or 0) + (is_weak[3] and 1 or 0) -- فحص التضعيف (الحرف الثاني = الثالث) local is_doubled = (r2 == r3) -- ═══════════════════════════════════════════════════════════ -- المُعتل: الفعل الذي يحتوي على حرف أو أكثر من أحرف العلة -- ═══════════════════════════════════════════════════════════ if weak_count > 0 then -- ───────────────────────────────────────────────────── -- لفيف: يحتوي على حرفين أو أكثر من أحرف العلة -- ───────────────────────────────────────────────────── if weak_count >= 2 then -- مقرون: حرفا العلة متجاوران if is_weak[2] and is_weak[3] then -- الثاني والثالث if is_hamza[1] then return "معتل لفيف مقرون مهموز الفاء" end return "معتل لفيف مقرون" elseif is_weak[1] and is_weak[2] then -- الأول والثاني return "معتل لفيف مقرون" -- مفروق: الأول والثالث حرفا علة elseif is_weak[1] and is_weak[3] then if is_hamza[2] then return "معتل لفيف مفروق مهموز العين" end return "معتل لفيف مفروق" end -- ───────────────────────────────────────────────────── -- مثال: الحرف الأول حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[1] then if r1 == W then -- واوي if is_hamza[3] then return "معتل مثال واوي مهموز اللام" elseif is_doubled then return "معتل مثال واوي مُضعَّف" elseif is_hamza[2] then return "معتل مثال واوي مهموز العين" end return "معتل مثال واوي" else -- يائي (ي) if is_doubled then return "معتل مثال يائي مُضعَّف" elseif is_hamza[2] then return "معتل مثال يائي مهموز العين" end return "معتل مثال يائي" end -- ───────────────────────────────────────────────────── -- أجوف: الحرف الثاني حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[2] then if r2 == W then -- واوي if is_hamza[3] then return "معتل أجوف واوي مهموز اللام" elseif is_hamza[1] then return "معتل أجوف واوي مهموز الفاء" end return "معتل أجوف واوي" else -- يائي (ي) if is_hamza[3] then return "معتل أجوف يائي مهموز اللام" elseif is_hamza[1] then return "معتل أجوف يائي مهموز الفاء" end return "معتل أجوف يائي" end -- ───────────────────────────────────────────────────── -- ناقص: الحرف الثالث حرف علة -- ───────────────────────────────────────────────────── elseif is_weak[3] then if r3 == W then -- واوي if is_hamza[1] then return "معتل ناقص واوي مهموز الفاء" elseif is_hamza[2] then return "معتل ناقص واوي مهموز العين" end return "معتل ناقص واوي" else -- يائي (ي) if is_hamza[1] then return "معتل ناقص يائي مهموز الفاء" elseif is_hamza[2] then return "معتل ناقص يائي مهموز العين" end return "معتل ناقص يائي" end end -- ═══════════════════════════════════════════════════════════ -- الصحيح: الفعل الذي لا يحتوي على أحرف العلة -- ═══════════════════════════════════════════════════════════ else -- ───────────────────────────────────────────────────── -- مهموز: يحتوي على همزة -- ───────────────────────────────────────────────────── if is_hamza[1] or is_hamza[2] or is_hamza[3] then -- مهموز الفاء واللام: الأول والثالث همزة if is_hamza[1] and is_hamza[3] then return "صحيح مهموز الفاء واللام" -- مهموز الفاء elseif is_hamza[1] then if is_doubled then return "صحيح مُضعَّف مهموز الفاء" end return "صحيح مهموز الفاء" -- مهموز العين elseif is_hamza[2] then return "صحيح مهموز العين" -- مهموز اللام elseif is_hamza[3] then return "صحيح مهموز اللام" end -- ───────────────────────────────────────────────────── -- مُضعَّف: الحرف الثاني = الثالث -- ───────────────────────────────────────────────────── elseif is_doubled then return "صحيح مُضعَّف" -- ───────────────────────────────────────────────────── -- سالم: لا همزة، لا حرف علة، لا تضعيف -- ───────────────────────────────────────────────────── else return "صحيح سالم" end end return "غير معروف" end ------------------------------------------------------------------------------- -- External entry points -- ------------------------------------------------------------------------------- --W: Normalize root to use underscores as separators -- Accepts: "كتب", "ك_ت_ب", or "ك ت ب" -- Returns: "ك_ت_ب" local function normalize_root(root) if not root then return nil end -- Helper function to normalize a single character/radical local function normalize_char(char) -- Convert all hamza variants to canonical bare hamza if char == HAMZA_ON_ALIF or char == HAMZA_UNDER_ALIF or char == HAMZA_ON_W or char == HAMZA_ON_Y then return HAMZA -- Convert alif maqṣūra to yaa (ى → ي) elseif char == AMAQ then return Y -- Reject alif - it's not a root radical elseif char == ALIF then error("Root contains alif (ا) which is not a valid root radical. For weak verbs, use و (waw) or ي (yaa) instead.") end return char end -- Convert alif madda to hamza + alif for processing (will be caught as error below) root = root:gsub(AMAD, HAMZA .. ALIF) -- Replace spaces with underscores root = root:gsub(" ", "_") -- If no underscores yet, insert them between each character if not root:find("_") then -- Extract only Arabic letters (excluding diacritics) -- Arabic letter range: U+0621 to U+064A (basic letters) local chars = {} for char in mw.ustring.gmatch(root, "[ء-ي]") do table.insert(chars, normalize_char(char)) end root = table.concat(chars, "_") else -- Root already has underscores (e.g., "أ_م_ن"), normalize each radical local parts = mw.text.split(root, "_") for i, part in ipairs(parts) do parts[i] = normalize_char(part) end root = table.concat(parts, "_") end return root end --W: Construct the [1] parameter from root (جذر) and morphological pattern (وزن) -- Example: -- input: root="كتب", pattern="فعَل يفعُل" -- output: "ك_ت_ب<I/a~u>" -- `force_full_passive`, if true, appends ".pass" so the verb gets a full personal -- passive paradigm instead of the module's default impersonal-passive guess for -- vforms like I (past vowel i), V, VI, X and IIq (see |مبني للمجهول= on {{ar-conj}}). -- `reduced`, if true, appends ".reduced" so an affix consonant assimilates into the -- radical beside it: forms V/VI use the assimilated تاء التفعّل prefix (اِسَّمَّعَ) -- instead of the default تَ- prefix (تَسَمَّعَ), and form IIIq assimilates اِفْعَنْلَلَ's -- نون into a weak ل1 (اِهْبَيَّخَ for اِهْبَنْيَخَ) -- see |مدغم= on {{ar-conj}}. Not every -- root with an assimilable radical is necessarily attested with the assimilated form, so -- this isn't inferred automatically from the root; it must be requested. local function construct_verb_spec(root, pattern, force_full_passive, reduced) if not root or not pattern then return nil end local normalized_root = normalize_root(root) local form_code = MORPHOLOGICAL_PATTERN_TO_FORM[pattern] if not form_code then error("Unknown morphological pattern (وزن صرفي): " .. pattern) end if force_full_passive then form_code = form_code .. ".pass" end if reduced then form_code = form_code .. ".reduced" end return normalized_root .. "<" .. form_code .. ">" end -- Wrapper function that accepts جذر (root) and وزن (pattern) as parameters -- Returns: verb_type (e.g., "فعل ثلاثي مُجرَّد صحيح سالم") -- Example: export.get_verb_type({["جذر"] = "كتب", ["وزن"] = "فعَل يفعُل"}) function export.get_verb_type(args) local verb_spec if args["جذر"] and args["وزن"] then verb_spec = construct_verb_spec(args["جذر"], args["وزن"]) elseif args[1] then verb_spec = args[1] else error("Must provide either جذر and وزن, or a verb specification in args[1]") end local alternant_multiword_spec = export.do_generate_forms({[1] = verb_spec}) local _, verb_type = get_verb_info(alternant_multiword_spec) return verb_type end -- Append two lists `l1` and `l2`, removing duplicates. If either is {nil}, just return the other. local function combine_lists(l1, l2) -- combine_footnotes() does exactly what we want. return iut.combine_footnotes(l1, l2) end local function combine_metadata(data) local src1 = data.form1 local src2 = data.form2 local dest = data.dest_form dest.uncertain = src1.uncertain or src2.uncertain if src1.genders and src2.genders and not m_table.deepEquals(src1.genders, src2.genders) then -- do nothing else dest.genders = src1.genders or src2.genders end if src1.pos and src2.pos and src1.pos ~= src2.pos then -- do nothing else dest.pos = src1.pos or src2.pos end -- Don't copy .alt, .gloss, .lit, .id, which describe a single term and don't extend to multiword terms. dest.q = combine_lists(src1.q, src2.q) dest.qq = combine_lists(src1.qq, src2.qq) dest.l = combine_lists(src1.l, src2.l) dest.ll = combine_lists(src1.ll, src2.ll) end -- Externally callable function to parse and conjugate a verb given user-specified arguments. -- Return value is WORD_SPEC, an object where the conjugated forms are in `WORD_SPEC.forms` -- for each slot. If there are no values for a slot, the slot key will be missing. The value -- for a given slot is a list of objects {form=FORM, footnotes=FOOTNOTES}. function export.do_generate_forms(args, source_template, headword_head) local PAGENAME = mw.loadData("Module:headword/data").pagename local function in_template_space() local nsText = mw.title.getCurrentTitle().nsText return nsText == "Template" or nsText == "قالب" end -- Determine the verb spec we're being asked to generate the conjugation of. This may be taken from the current page -- title or the value of |pagename=; but not when called from {{ar-verb form}}, where the page title is a -- non-lemma form. Note that the verb spec may omit the lemma; e.g. it may be "<II>". For this reason, we use the -- value of `pagename` computed here down below, when calling normalize_all_lemmas(). local pagename = source_template ~= "ar-verb form" and args.pagename or PAGENAME local head = headword_head or pagename --W: If both root (جذر) and morphological pattern (وزن) are provided, construct [1] if args["جذر"] and args["جذر"] ~= "" and args["وزن"] and args["وزن"] ~= "" then local constructed = construct_verb_spec(args["جذر"], args["وزن"], args["مبني للمجهول"], args["مدغم"]) if constructed and constructed ~= "" then args[1] = constructed end end local arg1 = args[1] if not arg1 then if (pagename == "ar-conj" or pagename == "ar-verb" or pagename == "ar-verb form" or pagename == "تصريف") and in_template_space() then arg1 = "كتب<I/a~u.pass>" else arg1 = "<>" end end -- When called from {{ar-verb form}}, determine the non-lemma form whose inflections we're being asked to -- determine. This normally comes from the page title or the value of |pagename=. local verb_form_of_form if source_template == "ar-verb form" then verb_form_of_form = args.pagename if not verb_form_of_form then if PAGENAME == "ar-verb form" and in_template_space() then verb_form_of_form = "كتبت" else verb_form_of_form = PAGENAME end end end local incorporated_headword_head_into_lemma = false if arg1:find("^<.*>$") then -- missing lemma if head:find(" ") then -- If multiword lemma, try to add arg spec after the first word. -- Try to preserve the brackets in the part after the verb, but don't do it -- if there aren't the same number of left and right brackets in the verb -- (which means the verb was linked as part of a larger expression). local first_word, post = rmatch(head, "^(.-)( .*)$") local left_brackets = rsub(first_word, "[^%[]", "") local right_brackets = rsub(first_word, "[^%]]", "") if #left_brackets == #right_brackets then arg1 = iut.remove_redundant_links(first_word) .. arg1 .. post incorporated_headword_head_into_lemma = true else -- Try again using the form without links. local linkless_head = m_links.remove_links(head) if linkless_head:find(" ") then first_word, post = rmatch(linkless_head, "^(.-)( .*)$") arg1 = first_word .. arg1 .. post else error("Unable to incorporate <...> spec into explicit head due to a multiword linked verb or " .. "unbalanced brackets; please include <> explicitly: " .. arg1) end end else -- Will be incorporated through `head` below in the call to normalize_all_lemmas(). incorporated_headword_head_into_lemma = true end end local parse_props = { parse_indicator_spec = parse_indicator_spec, angle_brackets_omittable = true, allow_blank_lemma = true, } local alternant_multiword_spec = iut.parse_inflected_text(arg1, parse_props) alternant_multiword_spec.pos = pos or "verbs" alternant_multiword_spec.args = args alternant_multiword_spec.source_template = source_template alternant_multiword_spec.verb_form_of_form = verb_form_of_form alternant_multiword_spec.incorporated_headword_head_into_lemma = incorporated_headword_head_into_lemma normalize_all_lemmas(alternant_multiword_spec, head) detect_all_indicator_specs(alternant_multiword_spec) local inflect_props = { lang = lang, slot_list = alternant_multiword_spec.verb_slots, inflect_word_spec = conjugate_verb, combine_metadata = combine_metadata, -- We add links around the generated verbal forms rather than allow the entire multiword -- expression to be a link, so ensure that user-specified links get included as well. include_user_specified_links = true, } iut.inflect_multiword_or_alternant_multiword_spec(alternant_multiword_spec, inflect_props) if debug_translit then for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do if form.translit then local full_form_translit = (lang:transliterate(m_links.remove_links(form.form))) if full_form_translit ~= form.translit then error(("Internal error: For slot '%s', form '%s' incremental translit '%s' not same as full translit '%s'"): format(slot, form.form, form.translit, full_form_translit)) end end form.form = iut.remove_redundant_links(form.form) end end end -- Remove redundant brackets around entire forms. for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do form.form = iut.remove_redundant_links(form.form) end end -- Apply NFC-style normalization: reorder shadda+vowel to vowel+shadda -- to match MediaWiki's automatic NFC normalization (which is applied on enwiktionary -- but not in local Lua environments). This reverses the postprocessing step. for slot, forms in pairs(alternant_multiword_spec.forms) do for _, form in ipairs(forms) do form.form = apply_nfc_shadda(form.form) end end determine_slot_uncertainty_from_forms(alternant_multiword_spec) determine_verb_properties_from_forms(alternant_multiword_spec) compute_categories_and_annotation(alternant_multiword_spec) if args.json and source_template == "ar-conj" then -- There is a circular reference in `base.alternant_multiword_spec`, which points back to top level. iut.map_word_specs(alternant_multiword_spec, function(base) base.alternant_multiword_spec = nil end) return require("Module:JSON").toJSON(alternant_multiword_spec) end return alternant_multiword_spec end -- Entry point for {{ar-conj}}. Template-callable function to parse and conjugate a verb given -- user-specified arguments and generate a displayable table of the conjugated forms. function export.show(frame) local parent_args = frame:getParent() and frame:getParent().args or frame.args --W local params = { [1] = {}, ["noautolinktext"] = {type = "boolean"}, ["noautolinkverb"] = {type = "boolean"}, ["t"] = {}, -- for use by {{ar-verb form}}; otherwise ignored ["id"] = {}, -- for use by {{ar-verb form}}; otherwise ignored ["pagename"] = {}, -- for testing/documentation pages ["json"] = {type = "boolean"}, -- for bot use --W: additional parameters for ar.wiktionary.org ["جذر"] = {type = "string"}, --W: الجذر اللغوي, root ["وزن"] = {type = "string"}, --W: الوزن الصرفي, morphological pattern ["مبني للمجهول"] = {type = "boolean"}, --W: force a full personal passive paradigm, -- overriding the default impersonal-passive guess for vforms like I (past vowel -- i), V, VI, X and IIq ["مدغم"] = {type = "boolean"}, --W: assimilate the affix into the radical beside it: -- forms V/VI's تاء التفعّل into the first radical (اِسَّمَّعَ instead of -- تَسَمَّعَ), or form IIIq's نون into a weak ل1 (اِهْبَيَّخَ for اِهْبَنْيَخَ) ["كامل"] = {type = "boolean"}, --W: show full table (default is simplified table) ["بؤرة"] = {type = "string"}, --W: highlight the cells holding this form ["موسع"] = {type = "boolean"}, --W: show the table expanded (default: collapsed) } local args = require("Module:parameters").process(parent_args, params) --W: If both root (جذر) and morphological pattern (وزن) are provided, construct [1] if args["جذر"] and args["جذر"] ~= "" and args["وزن"] and args["وزن"] ~= "" then local constructed = construct_verb_spec(args["جذر"], args["وزن"], args["مبني للمجهول"], args["مدغم"]) if constructed and constructed ~= "" then args[1] = constructed end end local alternant_multiword_spec = export.do_generate_forms(args, "ar-conj") if type(alternant_multiword_spec) == "string" then -- JSON return value return alternant_multiword_spec end --W: |بؤرة= marks the cells holding a given form, so that a conjugated-form -- page ([[Module:ar-verb-form]]) can show the reader where in the paradigm -- the word they looked up actually sits. The slots have to be found before -- show_forms(), which replaces each form list with rendered wikitext. alternant_multiword_spec.focus_slots = find_focus_slots(alternant_multiword_spec, args["بؤرة"]) --W: A table whose one relevant cell is highlighted is no use folded away, so -- [[Module:ar-verb-form]] asks for it open. Lemma pages keep it collapsed. alternant_multiword_spec.expanded = args["موسع"] or false show_forms(alternant_multiword_spec) --W: Choose table type: simplified (default) or full (comprehensive) local table_html if args["كامل"] then table_html = make_table(alternant_multiword_spec) else table_html = make_table_ar(alternant_multiword_spec) end --W: don't add categories --W: TODO: all lines with insert_cat() are to be removed. -- return table_html .. -- require("Module:utilities").format_categories(alternant_multiword_spec.categories, lang, nil, nil, force_cat) return table_html end function export.verb_forms(frame) local parargs = frame:getParent().args local params = { [1] = {}, [2] = {}, [3] = {}, [4] = {}, [5] = {}, pagename = {}, } for _, form in ipairs(allowed_vforms) do -- FIXME: We go up to 5 here. The code supports unlimited variants but it's unlikely we will ever see more than -- 2. for index = 1, 5 do local prefix = index == 1 and form or form .. index params[prefix .. "-pv"] = {} for _, extn in ipairs { "", "-vn", "-ap", "-pp" } do params[prefix .. extn] = {} params[prefix .. extn .. "-head"] = {} -- FIXME: No -tr? params[prefix .. extn .. "-gloss"] = {} end end end local args = require("Module:parameters").process(parargs, params) local i = 1 local past_vowel_re = "^[aui,]*$" local combined_root = nil if not args[i] or rfind(args[i], past_vowel_re) then combined_root = args.pagename or mw.loadData("Module:headword/data").pagename if not rfind(combined_root, "^([^ ]) ([^ ]) ([^ ])$") and not rfind(combined_root, "^([^ ]) ([^ ]) ([^ ]) ([^ ])$") then error("When inferring roots from page title, need three or four space-separated radicals: " .. combined_root) end elseif rfind(args[i], " ") then combined_root = args[i] i = i + 1 else local separate_roots = {} while args[i] and not rfind(args[i], past_vowel_re) do table.insert(separate_roots, args[i]) i = i + 1 end combined_root = table.concat(separate_roots, " ") end local past_vowel = args[i] i = i + 1 if past_vowel and not rfind(past_vowel, past_vowel_re) then error("Unrecognized past vowel, should be 'a', 'i', 'u', 'a,u', etc. or empty: " .. past_vowel) end -- Spaces interfere with parsing as a unit in [[Module:inflection utilities]], so replace with underscore. combined_root = combined_root:gsub(" ", "_") local split_root = rsplit(combined_root, "_") -- Map from verb forms (I, II, etc.) to a table of verb properties, -- which has entries e.g. for "verb" (either true to autogenerate the verb -- head, or an explicitly specified verb head using e.g. argument "I-head"), -- and for "verb-gloss" (which comes from e.g. the argument "I" or "I-gloss"), -- and for "vn" and "vn-gloss", "ap" and "ap-gloss", "pp" and "pp-gloss". local verb_properties = {} for _, form in ipairs(allowed_vforms) do local formpropslist = {} local derivs = {{"verb", ""}, {"vn", "-vn"}, {"ap", "-ap"}, {"pp", "-pp"}} local index = 1 while true do local formprops = {} local prefix = index == 1 and form or form .. index if prefix == "I" then formprops.pv = past_vowel end if args[prefix .. "-pv"] then formprops.pv = args[prefix .. "-pv"] end for _, deriv in ipairs(derivs) do local prop = deriv[1] local extn = deriv[2] if args[prefix .. extn] == "+" then formprops[prop] = true elseif args[prefix .. extn] == "-" then formprops[prop] = false elseif args[prefix .. extn] then formprops[prop] = true formprops[prop .. "-gloss"] = args[prefix .. extn] end if args[prefix .. extn .. "-head"] then if formprops[prop] == nil then formprops[prop] = true end formprops[prop] = args[prefix .. extn .. "-head"] end if args[prefix .. extn .. "-gloss"] then if formprops[prop] == nil then formprops[prop] = true end formprops[prop .. "-gloss"] = args[prefix .. extn .. "-gloss"] end end if formprops.verb then -- If a verb form specified, also turn on vn (unless form I, with -- unpredictable vn) and ap, and maybe pp, according to form, -- weakness and past vowel. But don't turn these on if there's -- an explicit on/off specification for them (e.g. I-pp=-). if form ~= "I" and formprops.vn == nil then formprops.vn = true end if formprops.ap == nil then formprops.ap = true end local weakness = weakness_from_radicals(form, split_root[1], split_root[2], split_root[3], split_root[4]) if formprops.pp == nil and not vform_probably_no_passive(form, weakness, rsplit(formprops.pv or "", ","), {}) then formprops.pp = true end if formprops.verb == true or formprops.vn == true or formprops.ap == true or formprops.pp == true then formprops.need_autogen = true end table.insert(formpropslist, formprops) index = index + 1 else break end end table.insert(verb_properties, {form, formpropslist}) end -- Go through and create the verb form derivations as necessary, when they haven't been explicitly given. for _, vplist in ipairs(verb_properties) do local vform = vplist[1] for _, props in ipairs(vplist[2]) do if props.need_autogen then local form_with_vowels if vform == "I" then local pv = props.pv if not pv then -- Make up likely past vowels based on weakness and actual radical. if split_root[3] == W then -- final-weak form_with_vowels = "I/a~u" elseif split_root[3] == Y then form_with_vowels = "I/a~i" elseif split_root[2] == W then --hollow form_with_vowels = "I/u~u" elseif split_root[2] == Y then form_with_vowels = "I/i~i" else -- most common; doesn't matter so much since we're not displaying the non-past form_with_vowels = "I/a~u" end else local pvs = rsplit(pv, ",") local vowel_sufs = {} for _, pv in ipairs(pvs) do local vowel_spec if pv == "a" then -- Make up likely past vowels based on weakness and actual radical. if split_root[3] == W then -- final-weak vowel_spec = "a~u" elseif split_root[3] == Y then vowel_spec = "a~i" elseif split_root[2] == W then --hollow vowel_spec = "a~u" elseif split_root[2] == Y then vowel_spec = "a~i" else -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "a~u" end elseif pv == "i" then -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "i~a" elseif pv == "u" then -- most common; doesn't matter so much since we're not displaying the non-past vowel_spec = "u~u" else error(("Internal error: Bad past vowel '%s' in {{ar-verb forms}}"):format(pv)) end table.insert(vowel_sufs, vowel_spec) end form_with_vowels = "I/" .. table.concat(vowel_sufs, "/") end else form_with_vowels = vform end local angle_bracket_spec = ("%s<%s.pass>"):format(combined_root, form_with_vowels) local alternant_multiword_spec = export.do_generate_forms({angle_bracket_spec}, "ar-verb forms") local function format_forms(forms) if not forms then return "-" -- FIXME: Throw an error? end local formatted = {} for _, form in ipairs(forms) do if form.translit then table.insert(formatted, ("%s//%s"):format(form.form, form.translit)) else table.insert(formatted, form.form) end end return table.concat(formatted, "،") end if props.verb == true then props.verb = format_forms(alternant_multiword_spec.forms.past_3ms) end for _, deriv in ipairs({"vn", "ap", "pp"}) do if props[deriv] == true then props[deriv] = format_forms(alternant_multiword_spec.forms[deriv]) end end end end end -- Go through and output the result local formtextarr = {} for _, vplist in ipairs(verb_properties) do local form = vplist[1] for _, props in ipairs(vplist[2]) do local textarr = {} if props.verb then local text = "* '''[[Appendix:Arabic verbs#Form " .. form .. "|Form " .. form .. "]]''': " local linktext = {} local splitheads = rsplit(props.verb, "[,،]") for _, head in ipairs(splitheads) do table.insert(linktext, m_links.full_link({lang = lang, term = head, gloss = props["verb-gloss"]})) end text = text .. table.concat(linktext, "، ") table.insert(textarr, text) for _, derivengl in ipairs({{"vn", "Verbal noun"}, {"ap", "Active participle"}, {"pp", "Passive participle"}}) do local deriv = derivengl[1] local engl = derivengl[2] if props[deriv] then local text = "** " .. engl .. ": " local linktext = {} local splitheads = rsplit(props[deriv], "[,،]") for _, head in ipairs(splitheads) do local ar, translit = head:match("^(.*)//(.-)$") if not ar then ar = head end table.insert(linktext, m_links.full_link {lang = lang, term = ar, tr = translit, gloss = props[deriv .. "-gloss"]} ) end text = text .. table.concat(linktext, "، ") table.insert(textarr, text) end end table.insert(formtextarr, table.concat(textarr, "\n")) end end end return table.concat(formtextarr, "\n") end -- Infer radicals from lemma headword (i.e. 3rd masculine singular past) and verb form (I, II, etc.). Throw an error if -- headword is malformed. A given returned radical may be actually be a list of possible radicals, where the first one -- should be used if the user didn't explicitly give the radical. If the list contains a field `ambig = true`, the -- radical is considered ambiguous and should not be categorized. `is_reduced` indicates that the user specified -- `.reduced` to indicate that the verb form is reduced by assimilation and/or haplology (typically archaic Koranic -- forms such as اِدَّارَأَ instead of تَدَارَأَ; or اِسْطَاعَ instead of اِسْتِطَاعَ; etc. function export.infer_radicals(data) local headword, vform, passive, past_vowel, nonpast_vowel, is_reduced = data.headword, data.vform, data.passive, data.past_vowel, data.nonpast_vowel, data.is_reduced past_vowel = past_vowel or "-" nonpast_vowel = nonpast_vowel or "-" local function verify_vowel(vowel, param) --W: A past/non-past vowel may legitimately be a form object {form=..., footnotes=...} -- when the user attaches a footnote, e.g. <I/a[نادر]~u>, so compare the formval. -- Comparing the raw value crashed on Lua 5.1 ("bad argument #2 to 'format'"). local formval = rget(vowel) if formval ~= A and formval ~= I and formval ~= U and formval ~= "-" then error(("Internal error: Bad value for %s: %s (should be Arabic diacritic vowel or '-')"):format( param, dump(vowel))) end end verify_vowel(past_vowel, "past_vowel") verify_vowel(nonpast_vowel, "nonpast_vowel") local ch = {} local form_viii_assim, variant -- sub out alif-madda for easier processing headword = rsub(headword, AMAD, HAMZA .. ALIF) local function infer_err(msg, noann) local anns = {} local nohead, novform if noann == "nohead" then nohead = true elseif noann == "novform" then novform = true elseif noann == "nohead-vform" then nohead = true novform = true elseif noann then error(("Internal error: Unrecognized value for 'noann': %s"):format(dump(noann))) end if not nohead then table.insert(anns, ("headword=%s"):format(data.headword)) end if not novform then table.insert(anns, ("verb form=%s"):format(data.vform)) end anns = table.concat(anns, "، ") if anns ~= "" then anns = ": " .. anns end error(msg .. anns) end local len = ulen(headword) local expected_length -- extract the headword letters into an array for i = 1, len do table.insert(ch, usub(headword, i, i)) end -- check that the letter at the given index is the given string, or -- is one of the members of the given array local function check(index, must) local letter = ch[index] if type(must) == "string" then if not letter then infer_err("Letter " .. index .. " is nil") end if letter ~= must then infer_err(("For verb form %s, letter %s must be %s, not %s"):format(vform, index, must, letter), "novform") end elseif not m_table.contains(must, letter) then infer_err("For verb form " .. vform .. ", radical " .. index .. " must be one of " .. table.concat(must, " ") .. ", not " .. letter, "novform") end end -- Check that length of headword is within [min, max] local function check_len(min, max) if min and len < min then infer_err(("Not enough letters for verb form %s, expected at least %s"):format(vform, min), "novform") end if max and len > max then infer_err(("Too many letters for verb form %s, expected at most %s"):format(vform, max), "novform") end end -- A form I verb beginning with w- keeps the w in the non-past unless the non-past vowel is a kasra or a -- lowered kasra; see form_i_w_assimilated(), which the radicals path uses too. This used to be spelt out -- here as "i~a or u~u", which missed a~u (وَجَلَ يَوْجُلُ). local function form_I_w_non_assimilated() return not form_i_w_assimilated(past_vowel, nonpast_vowel) end -- Convert radicals to canonical form (handle various hamza varieties and check for misplaced alif or alif maqṣūra; -- legitimate cases of these letters are handled above). local function convert(rad, index) if type(rad) == "table" then for i, r in ipairs(rad) do rad[i] = convert(r, index) end return rad elseif rad == HAMZA_ON_ALIF or rad == HAMZA_UNDER_ALIF or rad == HAMZA_ON_W or rad == HAMZA_ON_Y then return HAMZA elseif rad == AMAQ then infer_err("Radical " .. index .. " must not be alif maqṣūra") elseif rad == ALIF then infer_err("Radical " .. index .. " must not be alif") else return rad end end local quadlit = vform:find("q$") -- find first radical, start of second/third radicals, check for -- required letters local radstart, rad1, rad2, rad3, rad4 local weakness if vform == "I" or vform == "II" then rad1 = ch[1] radstart = 2 elseif vform == "III" then rad1 = ch[1] check(2, {ALIF, W}) -- W occurs in passive-only verbs radstart = 3 elseif vform == "IV" then -- this would be alif-madda but we replaced it with hamza-alif above. if ch[1] == HAMZA and ch[2] == ALIF then rad1 = HAMZA else check(1, HAMZA_ON_ALIF) rad1 = ch[2] end radstart = 3 elseif vform == "V" then check(1, is_reduced and ALIF or T) rad1 = ch[2] radstart = 3 elseif vform == "VI" then check(1, is_reduced and ALIF or T) if ch[2] == AMAD then rad1 = HAMZA radstart = 3 else rad1 = ch[2] check(3, {ALIF, W}) -- W occurs in passive-only verbs radstart = 4 end elseif vform == "VII" then check(1, ALIF) if is_reduced then check(2, M) rad1 = M radstart = 3 else check(2, N) rad1 = ch[3] radstart = 4 end elseif vform == "VIII" then check(1, ALIF) rad1 = ch[2] if rad1 == "د" then rad1 = {"د", "ذ"} -- not considered ambiguous since it's usually د radstart = 3 form_viii_assim = "دّ" elseif rad1 == "ظ" and ch[3] == "ط" and len >= 5 then -- [[اظطلم]], variant of [[اظلم]] radstart = 4 form_viii_assim = "ظْط" elseif rad1 == "ذ" and ch[3] == "د" and len >= 5 then -- [[اذدكر]], variant of [[اذكر]] radstart = 4 form_viii_assim = "ذْد" elseif rad1 == T or rad1 == "ث" or rad1 == "ذ" or rad1 == "ط" or rad1 == "ظ" then radstart = 3 form_viii_assim = rad1 .. SH elseif rad1 == "ز" then check(3, "د") radstart = 4 form_viii_assim = "زْد" elseif rad1 == "ص" or rad1 == "ض" then check(3, "ط") radstart = 4 form_viii_assim = rad1 .. SK .. "ط" else check(3, T) radstart = 4 rad1 = convert(rad1, 1) form_viii_assim = rad1 .. SK .. "ت" end if rad1 == T then -- Radical is ambiguous, might be ت or و or ي but doesn't affect conjugation. Note that there are no -- form-VIII verbs with initial radical ي given in Hans Wehr but Lane mentions at least: -- - (page 2973) اِتَّأَسَ, with assimilation of the ي to ت, from root ي ء س; -- - (page 2975) اِتَّبَسَ non-past يَتَّبِسُ and alternative اِيتَبَسَ non-past يَاتَبِسُ from the root ي ب س; -- - (page 2976) اِتَّسَرَ non-past يَتَّسِرُ or alternatively يَأْتَسِرُ with hamza preserved from the root ي س ر. -- These alternative forms seem very rare and probably not worth worrying about, but if we want to handle -- them, we can do it when the time comes. rad1 = {T, W, Y, ambig = true} -- اِتَّخَذَ irregularly has hamza as the radical but assimilates like و if ch[3] == "خ" and ch[4] == "ذ" then rad1[4] = HAMZA end end elseif vform == "IX" then check(1, ALIF) rad1 = ch[2] radstart = 3 elseif vform == "X" then check(1, ALIF) check(2, S) if is_reduced then rad1 = ch[3] radstart = 4 else check(3, T) rad1 = ch[4] radstart = 5 end elseif vform == "Iq" then rad1 = ch[1] rad2 = ch[2] radstart = 3 elseif vform == "IIq" then check(1, T) rad1 = ch[2] rad2 = ch[3] radstart = 4 elseif vform == "IIIq" then check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) radstart = 5 elseif vform == "IVq" then check(1, ALIF) rad1 = ch[2] rad2 = ch[3] radstart = 4 elseif vform == "XI" then check_len(5, 5) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, ALIF) rad3 = ch[5] weakness = "sound" elseif vform == "XII" then check(1, ALIF) rad1 = ch[2] if ch[3] ~= ch[5] then infer_err("For verb form XII, letters 3 and 5 should be the same", "novform") end check(4, W) radstart = 5 elseif vform == "XIII" then check_len(5, 5) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, W) rad3 = ch[5] if rad3 == AMAQ then weakness = "final-weak" else weakness = "sound" end elseif vform == "XIV" then check_len(6, 6) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) rad3 = ch[5] if ch[6] == AMAQ then check_waw_ya(rad3) weakness = "final-weak" else if ch[5] ~= ch[6] then infer_err("For verb form XIV, letters 5 and 6 should be the same", "novform") end weakness = "sound" end elseif vform == "XV" then check_len(6, 6) check(1, ALIF) rad1 = ch[2] rad2 = ch[3] check(4, N) rad3 = ch[5] if rad3 == Y then check(6, ALIF) else check(6, AMAQ) end weakness = "sound" else error("Internal error: Unrecognized verb form " .. vform) end -- Process the last two radicals. RADSTART is the index of the first of the two. If it's nil then all radicals have -- already been processed above, and we don't do anything. if radstart then -- There must (normally) be one or two letters left. if len == radstart then if vform == "I" and ch[len] == Y then -- short form حَيَّ weakness = "final-weak" rad2 = Y rad3 = Y variant = "short" elseif vform == "IV" and rad1 == "ر" and ch[len] == AMAQ then -- irregular verb أَرَى weakness = "final-weak" rad2 = HAMZA rad3 = Y elseif vform == "X" and rad1 == "ح" and ch[len] == AMAQ then -- irregular verb اِسْتَحَى weakness = "final-weak" rad2 = Y rad3 = Y variant = "short" else -- If one letter left, then it's a geminate verb. If the letter is alif or alif maqṣūra, it will trigger -- an error down the line. if vform_supports_geminate(vform) then weakness = "geminate" rad2 = ch[len] rad3 = ch[len] if vform == "III" or vform == "VI" then variant = "short" end else infer_err("Apparent geminate verb, but geminate verbs not allowed for this verb form") end end elseif quadlit then -- Process last two radicals of a quadriliteral verb form. rad3 = ch[radstart] rad4 = ch[radstart + 1] expected_length = radstart + 1 check_len(expected_length) if rad4 == AMAQ or rad4 == ALIF and rad3 == Y or rad4 == Y then -- rad4 can be Y in passive-only verbs. if vform_supports_final_weak(vform) then weakness = "final-weak" -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws in a row); it could be wāw or -- yāʾ, but doesn't affect the conjugation. rad4 = rad3 == W and {Y, W, ambig = true} or {W, Y, ambig = true} else infer_err("Last radical is " .. rad4 .. " but verb form " .. vform .. " doesn't support final-weak verbs", "novform") end else weakness = "sound" end else -- Process last two radicals of a triliteral verb form. rad2 = ch[radstart] rad3 = ch[radstart + 1] expected_length = radstart + 1 check_len(expected_length) if vform == "I" and (is_waw_ya(rad3) or rad3 == ALIF or rad3 == AMAQ) then local inferred_past_vowel, inferred_nonpast_vowel -- Check for final-weak form I verb. It can end in tall alif (rad3 = wāw) or alif maqṣūra (rad3 = yāʾ) -- or a wāw or yāʾ (with a past vowel of i or u, e.g. nasiya/yansā "forget" or with a passive-only -- verb). if rad1 == W and not form_I_w_non_assimilated() then weakness = "assimilated+final-weak" else weakness = "final-weak" end if rad3 == ALIF then rad3 = W inferred_past_vowel = A inferred_nonpast_vowel = U if is_passive_only(passive) then infer_err("Final-weak form-I passive verbs should end in yāʔ (ي), not tall alif (ا)", "novform") end elseif rad3 == AMAQ then rad3 = Y inferred_past_vowel = A inferred_nonpast_vowel = I if is_passive_only(passive) then infer_err("Final-weak form-I passive verbs should end in yāʔ (ي), not alif maqṣūra (ى)", "novform") end elseif rad1 == "ح" and rad2 == Y and rad3 == Y then -- Long variant حَيِيَ. inferred_past_vowel = I inferred_nonpast_vowel = A variant = "long" else if not is_passive_only(passive) then -- does a non-passive final-weak verb in -uwa ever happen? (YES: e.g. [[رجو]] "to be slack") inferred_past_vowel = rad3 == Y and I or U inferred_nonpast_vowel = A end -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws); it could be wāw or yāʾ, but -- doesn't affect the conjugation. rad3 = (rad1 == W or rad2 == W) and {Y, W, ambig = true} or {W, Y, ambig = true} -- ambiguous end if inferred_past_vowel then local raw_past_vowel = rget(past_vowel) local raw_nonpast_vowel = rget(nonpast_vowel) if raw_past_vowel ~= "-" then if raw_past_vowel ~= inferred_past_vowel then infer_err(("Final-weak form-I verb inferred past vowel %s, which disagrees with " .. "explicitly specified %s"):format(undia[inferred_past_vowel], undia[raw_past_vowel]), "novform") else -- in case of footnote in past_vowel inferred_past_vowel = past_vowel end end if raw_nonpast_vowel ~= "-" and raw_nonpast_vowel ~= A and inferred_nonpast_vowel == U then -- if inferred as I or A, the reality can be the reverse; form-I final-weak verbs with a~a and -- i~i exist, e.g. سَعَى/يَسْعَى, وَلِيَ/يَلِي. Weird verb [[صها]] (also written [[صهى]]) has non-past -- يصهى so we can't throw an error in this situation. if raw_nonpast_vowel ~= inferred_nonpast_vowel then infer_err(("Final-weak form-I verb inferred non-past vowel %s, which disagrees with " .. "explicitly specified %s"):format(undia[inferred_nonpast_vowel], undia[raw_nonpast_vowel]), "novform") else -- in case of footnote in nonpast_vowel inferred_nonpast_vowel = nonpast_vowel end end end if not is_passive_only(passive) then if rget(past_vowel) == "-" then past_vowel = inferred_past_vowel end if rget(nonpast_vowel) == "-" then nonpast_vowel = inferred_nonpast_vowel end end elseif vform == "IX" and is_waw_ya(rad3) and len == radstart + 2 and ch[len] == AMAQ then -- Final-weak form IX verbs like اِرْعَوَى "to desist, to repent, to see the light". weakness = "final-weak" expected_length = radstart + 2 elseif vform == "X" and rad1 == "ح" and rad2 == Y and rad3 == ALIF then -- Long variant اِسْتَحْيَا. weakness = "final-weak" rad3 = Y variant = "long" elseif rad3 == AMAQ or rad2 == Y and rad3 == ALIF or rad3 == Y then -- rad3 == Y happens in passive-only verbs. if vform_supports_final_weak(vform) then weakness = "final-weak" else infer_err("Last radical is " .. rad3 .. " but verb form doesn't support final-weak verbs") end -- Ambiguous radical; randomly pick wāw as radical (but avoid two wāws); it could be wāw or yāʾ, but -- doesn't affect the conjugation. rad3 = (rad1 == W or rad2 == W) and {Y, W, ambig = true} or {W, Y, ambig = true} elseif rad2 == ALIF then if vform_supports_hollow(vform) then weakness = "hollow" local function set_past_to_a() if req(past_vowel, A) then -- already set elseif req(past_vowel, "-") or req(past_vowel, rget(nonpast_vowel)) then past_vowel = A else infer_err(("Form I hollow verb with nonpast vowel set to '%s' must have past vowel set to 'a' or the same value, not %s"): format(undia[rget(nonpast_vowel)], undia[rget(past_vowel)]), "novform") end end if vform == "I" and req(nonpast_vowel, U) then rad2 = W set_past_to_a() elseif vform == "I" and req(nonpast_vowel, I) then rad2 = Y set_past_to_a() else if req(nonpast_vowel, A) and not req(past_vowel, I) then infer_err(("Form I hollow verb with nonpast vowel set to 'a' must have past vowel set to 'i', not %s"): format(undia[rget(past_vowel)]), "novform") end -- Ambiguous radical; could be wāw or yāʾ; if verb form I, it's critical to get this right, and -- the caller checks for this situation and throws an error if non-past vowel is "a" and second -- radical isn't explicitly given. rad2 = {W, Y, ambig = true, need_radical = true} end else infer_err("Second radical is alif but verb form doesn't support hollow verbs") end elseif vform == "I" and rad1 == W and not form_I_w_non_assimilated() then weakness = "assimilated" elseif rad2 == rad3 and (vform == "III" or vform == "VI") then weakness = "geminate" variant = "long" else weakness = "sound" end end if expected_length then check_len(expected_length, expected_length) end end rad1 = convert(rad1, 1) rad2 = convert(rad2, 2) rad3 = convert(rad3, 3) rad4 = convert(rad4, 4) if not weakness then error("Internal error: Returned weakness from infer_radicals() is nil") end return { weakness = weakness, rad1 = rad1, rad2 = rad2, rad3 = rad3, rad4 = rad4, past_vowel = past_vowel, nonpast_vowel = nonpast_vowel, form_viii_assim = form_viii_assim, variant = variant, } end -- bot interface to infer_radicals() function export.infer_radicals_json(frame) local iparams = { headword = {}, vform = {}, passive = {}, past_vowel = {}, nonpast_vowel = {}, is_reduced = {type = "boolean"}, } local iargs = require("Module:parameters").process(frame.args, iparams) return require("Module:JSON").toJSON(export.infer_radicals(iargs)) end -- Infer vocalization from participle headword (active or passive), verb form (I, II, etc.) and whether the headword is -- active or passive. Throw an error if headword is malformed. Returned radicals may contain Latin letters "t", "w" or "y" -- indicating ambiguous radicals guessed to be tāʾ, wāw or yāʾ respectively. function export.infer_participle_vocalization(headword, vform, weakness, is_active) local chars = {} local orig_headword = headword -- Sub out alif-madda for easier processing. headword = rsub(headword, AMAD, HAMZA .. ALIF) local len = ulen(headword) -- Extract the headword letters into an array. for i = 1, len do table.insert(chars, usub(headword, i, i)) end local function form_intro_error_msg() return ("For verb form %s %s%s participle %s, "):format(vform, orig_headword ~= headword and "normalized " or "", is_active and "active" or "passive", headword) end local function err(msg) error(form_intro_error_msg() .. msg, 1) end -- Check that length of headword is within [min, max]. local function check_len(min, max) if min and len < min then err(("expected at least %s letters but saw %s"):format(min, len)) elseif max and len > max then err(("expected at most %s letters but saw %s"):format(max, len)) end end -- Get the character at `ind`, making sure it exists. local function c(ind) check_len(ind) return chars[ind] end -- Check that the letter at the given index is the given string, or is one of the members of the given array local function check(index, must) local letter = chars[index] local function make_possible_values() if type(must) == "string" then return must else return list_to_text(must, nil, " or ") end end if not letter then err(("expected a letter (specifically %s) at position %s, but participle is too short"):format( make_possible_values(), index)) end local matches if type(must) == "string" then matches = letter == must else matches = m_table.contains(must, letter) end if not matches then err(("letter %s at index %s must be %s"):format(letter, index, make_possible_values())) end end local function check_weakness(values, allow_missing, invert_condition) local function make_possible_weaknesses() for i, val in ipairs(values) do values[i] = "'" .. val .. "'" end return list_to_text(values, nil, " or ") end if allow_missing and invert_condition then error("Internal error: Can't specify both allow_missing and invert_condition") end if not weakness then if allow_missing or invert_condition then return else err(("weakness is unspecified but must be %s"):format(make_possible_weaknesses())) end else local matches = m_table.contains(values, weakness) if invert_condition and matches then err(("weakness '%s' must not be %s"):format(weakness, make_possible_weaknesses())) elseif not invert_condition and not matches then err(("weakness '%s' must be %s"):format(weakness, make_possible_weaknesses())) end end end local vocalized local function handle_possibly_final_weak(sound_prefix, expected_length) check_len(expected_length, expected_length) if c(expected_length) == AMAQ then -- passive final-weak if is_active then err("participle in -ِى only allowed for passive participles") end check_weakness({"final-weak", "assimilated+final-weak"}, "allow missing") vocalized = sound_prefix .. AN .. AMAQ else -- all others behave as if sound check_weakness({"final-weak", "assimilated+final-weak"}, nil, "invert condition") vocalized = sound_prefix .. (is_active and I or A) .. c(expected_length) end end if not (vform == "I" and is_active) then -- all participles except verb form I active begin in م-. check(1, M) end if vform == "I" then if is_active then check(2, ALIF) local sound_prefix = c(1) .. AA .. c(3) if len == 3 then if c(3) == HAMZA then -- Either hollow with hamzated third radical, e.g. [[شاء]] active participle 'شَاءٍ', or final-weak -- with hamzated second radical, e.g. [[رأى]] active participle 'رَاءٍ'. Theoretically (?), also -- geminate with hamzated second/third radical, but I don't know if any such verbs exist. if weakness == "geminate" then vocalized = sound_prefix .. SH else check_weakness({"hollow", "final-weak"}, "allow missing") vocalized = sound_prefix .. IN end else check_weakness({"final-weak", "geminate"}) if weakness == "geminate" then vocalized = sound_prefix .. SH else vocalized = sound_prefix .. IN end end else check_len(4, 4) -- we will convert back to alif maqṣūra below as needed vocalized = sound_prefix .. I .. c(4) end else -- assimilated verbs: regular, e.g. مَوْزُون "weighed" -- geminate verbs: regular, e.g. مَبْلُول "moistened" -- third-hamzated verbs: مَبْرُوء -- hollow verbs: مَقُود "led, driven"; مَزِيد "added, increased" -- hollow first-hamzated verbs: مَئِيض "returned, reverted"; مَأْيُوس "despaired" (NOTE: formation is sound); -- مَأُود or مَؤُود "bent; depleted" -- hollow third-hamzated verbs: مَشِيء "willed, intended", مَضُوء "glittered?" -- final-weak: مَلْقِيّ "found, encountered"; مَصْغُوّ "inclined" -- hollow + final-weak: مَشْوِيّ "fried, grilled", مَهْوِيّ "loved" -- first-hamzated + hollow + final-weak: مَأْوِيّ "received hospitably" local sound_prefix = MA .. c(2) .. SK .. c(3) if len == 5 then -- sound, assimilated or geminate check(4, W) vocalized = sound_prefix .. UU .. c(5) else check_len(4, 4) if c(4) == W then -- final-weak third-wāw vocalized = sound_prefix .. U .. W .. SH elseif c(4) == Y then -- final-weak third-yāʾ vocalized = sound_prefix .. I .. Y .. SH else -- hollow check(3, {W, Y}) if c(3) == W then vocalized = MA .. c(2) .. UU .. c(4) else vocalized = MA .. c(2) .. II .. c(4) end end end end elseif vform == "II" or vform == "V" or vform == "XII" or vform == "XIII" or vform == "Iq" or vform == "IIq" or vform == "IIIq" then local sound_prefix, expected_length if vform == "II" then sound_prefix = MU .. c(2) .. A .. c(3) .. SH expected_length = 4 elseif vform == "V" then check(2, T) sound_prefix = MU .. T .. A .. c(3) .. A .. c(4) .. SH expected_length = 5 elseif vform == "XII" then -- e.g. [[احدودب]] "to be or become convex or humpbacked", مُحْدَوْدِب (active); -- [[اثنونى]] "to be bent; to be doubled up", مُثْنَوْنٍ (active) check(4, W) if c(3) ~= c(5) then err(("third letter %s should be the same as the fifth letter %s"):format(c(3), c(5))) end sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. W .. SK .. c(5) expected_length = 6 elseif vform == "XIII" then -- e.g. [[اخروط]] "to get entangled; to extend", مُخْرَوِّط (active), مُخْرَوَّط (passive) check(4, W) sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. W .. SH expected_length = 5 elseif vform == "Iq" then sound_prefix = MU .. c(2) .. A .. c(3) .. SK .. c(4) expected_length = 5 elseif vform == "IIq" then check(2, T) sound_prefix = MU .. T .. A .. c(3) .. A .. c(4) .. SK .. c(5) expected_length = 6 elseif vform == "IIIq" then -- e.g. [[اخرنطم]] "to be proud and angry" check(4, T) sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. N .. SK .. c(5) expected_length = 6 else error("Internal error: Unhandled verb form " .. vform) end if len == expected_length - 1 then -- active final-weak if not is_active then err(("length-%s participle only allowed for active participles"):format(len)) end check_weakness({"final-weak", "assimilated+final-weak"}, "allow missing") vocalized = sound_prefix .. IN else handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "III" or vform == "VI" then local sound_prefix, expected_length if vform == "VI" then check(2, T) check(4, ALIF) sound_prefix = MU .. T .. A .. c(3) .. AA .. c(5) expected_length = 6 else sound_prefix = MU .. c(2) .. AA .. c(4) expected_length = 5 end if len == expected_length - 1 then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. SH else vocalized = sound_prefix .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. SH end else handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "IV" or vform == "X" then -- form IV: -- sound: مُرْسِخ (active, "entrenching"), مُرْسَخ (passive, "entrenched") -- first-hamzated (like sound): مُؤْيِس (active, "causing to despair"), مُؤْيَس (passive, "caused to despair") -- final-weak: مُكْرٍ (active, "renting out"), مُكْرًى (passive, "rented out") -- assimilated: مُورِد (active, "transferring"), مُورَد (passive, "transferred"); same when first-Y, e.g. -- أَيْقَنَ "to be certain of": مُوقِن (active), مُوقَن (passive) -- assimilated + final-weak: مُورٍ (active, "setting fire, kindling"), مُورًى (passive, "set fire, kindled") -- geminate: مُمِدّ (active, "granting, helping"), مُمَدّ (passive, "granted, helped") -- hollow: مُزِيل (active, "eliminating"), مُزَال (passive, "eliminated") -- hollow + final-weak: مُعْيٍ (active, "tiring"), مُعْيًى (passive, "tired") local sound_prefix, expected_length if vform == "X" then check(2, S) check(3, T) sound_prefix = MU .. S .. SK .. T .. A .. c(4) expected_length = 6 else sound_prefix = MU .. c(2) expected_length = 4 end if len == expected_length and c(len - 1) == Y and c(len) ~= AMAQ then -- active hollow if not is_active then err("this shape only allowed for active participles") end check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. II .. c(len) elseif len == expected_length and c(len - 1) == ALIF then -- passive hollow if is_active then err("this shape only allowed for passive participles") end check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. AA .. c(len) elseif len == expected_length - 1 then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. I .. c(len) .. SH elseif vform == "IV" and c(2) == W then -- assimilated final-weak vocalized = sound_prefix .. c(len) .. IN else vocalized = sound_prefix .. SK .. c(len) .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. A .. c(len) .. SH end else if vform == "IV" and c(2) == W then -- assimilated, possibly final-weak sound_prefix = sound_prefix .. c(expected_length - 1) else sound_prefix = sound_prefix .. SK .. c(expected_length - 1) end handle_possibly_final_weak(sound_prefix, expected_length) end elseif vform == "VII" or vform == "VIII" then -- form VII (passive participles are fairly rare but do exist): -- sound: مُنْكَتِب (active "subscribing"), مُنْكَتَب (passive "subscribed") -- geminate: مُنْضَمّ (both active "joining, containing" and passive "joined, contained") -- final-weak: مُنْطَلٍ (active "fooling (someone)"), مُنْطَلًى (passive "fooled") -- final-weak with medial wāw: مُنْطَوٍ (active "involving"), مُنْطَوًى (passive "involved") -- hollow: مُنْقَاد (both active "complying with" and passive "complied with") -- -- for form VIII, the same variants exist but things are complicated by assimilations involving the template T. -- sound third-hamzated no assimilation: مُبْتَدِئ (active "beginning"), مُبْتَدَأ (passive "begun") -- geminate no assimilation: مُبْتَزّ (both active "robbing" and passive "robbed") -- final-weak no assimilation: مُبْتَنٍ (active "building"), مُبْتَنًى (passive "built") -- final-weak with medial wāw no assimilation: مُحْتَوٍ (active "containing"), مُحْتَوًى (passive "contained") -- hollow no assimilation: مُخْتَار (both active "choosing" and passive "chosen") -- -- sound with total assimilation: مُتَّبِع (active "following"), مُتَّبَع (passive "followed") -- sound with total assimilation, assimilating wāw: مُتَّعِد (active "threatening"), مُتَّعَد (passive "threatened") -- sound with total assimilation, irregularly assimilating hamza: مُتَّخِذ (active "taking"), مُتَّخَذ (passive "taken") -- sound with total assimilation (to ḏāl, producing dāl): مُدَّخِر (active "reserving"), مُدَّخَر (passive "reserved") -- sound with total assimilation (to ḏāl): مُذَّكِر (active "remembering"), مُذَّكَر (passive "remembered") -- sound with total assimilation (to ṭāʔ): مُطَّرِح (active "discarding"), مُطَّرَح (passive "discarded") -- sound with total assimilation (to ẓāʔ): مُظَّلِم (active "tolerating"), مُظَّلَم (passive "tolerated") -- final-weak with total assimilation, assimilating wāw: مُتَّقٍ (active "guarding against"), مُتَّقًى (passive "guarded against") -- final-weak with total assimilation (to ṯāʔ): مُثَّنٍ (active "undulating"), مُثَّنًى (passive "undulated") -- final-weak with total assimilation (to dāl): مُدَّعٍ (active "claiming"), مُدَّعًى (passive "claimed") -- sound with partial assimilation (to zayn): مُزْدَهِر (active "thriving"), مُزْدَهَر (passive "thrived") -- sound with medial wāw with partial assimilation (to zayn): مُزْدَوِج (active "appearing twice") -- sound with partial assimilation (to ṣād): مُصْطَبِح (active "illuminating"), مُصْطَبَح (passive, "illuminated") -- sound with partial assimilation (to ḍād): مُضْطَرِب (active "to be disturbed"; no passive) -- geminate with partial assimilation (to ṣād): مُصْطَبّ (both active "effusing" and passive "effused") -- geminate with partial assimilation (to ḍād): مُضْطَرّ (both active "forcing" and passive "forced") -- final-weak with partial assimilation (to ṣād): مُصْطَلٍ (active "warming"), مُصْطَلًى (passive "warmed") -- hollow with partial assimilation (to zayn): مُزْدَاد (both active "increasing" and passive "increased") -- hollow with partial assimilation (to ṣad): مُصْطَاد (both active "hunting" and passive "hunted") local sound_prefix, sufind if vform == "VII" then check(2, N) sound_prefix = MU .. N .. SK .. c(3) sufind = 4 else local c2 = c(2) if c2 == T or c2 == "د" or c2 == "ث" or c2 == "ذ" or c2 == "ط" or c2 == "ظ" then -- full assimilation sound_prefix = MU .. c2 .. SH sufind = 3 else -- partial or no assimilation if c2 == "ز" then check(3, "د") elseif c2 == "ص" or c2 == "ض" then check(3, "ط") else check(3, T) end sound_prefix = MU .. c2 .. SK .. c(3) sufind = 4 end end if c(sufind) == ALIF then -- hollow, active or passive check_len(sufind + 1, sufind + 1) check_weakness({"hollow"}, "allow missing") vocalized = sound_prefix .. AA .. c(sufind + 1) elseif len == sufind then -- active final-weak or active or passive geminate if is_active then check_weakness({"geminate", "final-weak", "assimilated+final-weak"}) if weakness == "geminate" then vocalized = sound_prefix .. A .. c(len) .. SH else vocalized = sound_prefix .. A .. c(len) .. IN end else check_weakness({"geminate"}, "allow missing") vocalized = sound_prefix .. A .. c(len) .. SH end else sound_prefix = sound_prefix .. A .. c(sufind) handle_possibly_final_weak(sound_prefix, sufind + 1) end elseif vform == "IX" then check_len(4, 4) vocalized = MU .. c(2) .. SK .. c(3) .. A .. c(4) .. SH elseif vform == "IVq" then -- e.g. [[اذلعب]] "to scamper away", مُذْلَعِبّ (active), مُذْلَعَبّ (passive); -- [[اطمأن]] "to remain quietly; to be certain", مُطْمَئِنّ (active), مُطْمَأَنّ (passive) check_len(5, 5) local sound_prefix = MU .. c(2) .. SK .. c(3) .. A .. c(4) if is_active then vocalized = sound_prefix .. I .. c(5) .. SH else vocalized = sound_prefix .. A .. c(5) .. SH end elseif vform == "XI" then check_len(5, 5) check(4, ALIF) vocalized = MU .. c(2) .. SK .. c(3) .. AA .. c(5) .. SH -- e.g. [[احمار]] "to turn red, to blush", مُحْمَارّ (active) elseif vform == "XIV" or vform == "XV" then -- FIXME: Implement. No examples in Wiktionary currently; need to look up in a grammar. error("Support for verb form " .. vform .. " not implemented yet") else error("Don't recognize verb form " .. vform) end vocalized = rsub(vocalized, HAMZA .. AA, AMAD) local reconstructed_headword = lang:makeEntryName(vocalized) if reconstructed_headword ~= orig_headword then error(("Internal error: Vocalized participle %s doesn't match original participle %s"):format( vocalized, orig_headword)) end -- Apply NFC-style normalization to match MediaWiki's automatic normalization return apply_nfc_shadda(vocalized) end function export.infer_participle_vocalization_json(frame) local iparams = { [1] = {required = true}, [2] = {required = true}, ["weakness"] = {}, ["passive"] = {type = "boolean"} } local iargs = require("Module:parameters").process(frame.args, iparams) return export.infer_participle_vocalization(iargs[1], iargs[2], iargs.weakness, not iargs.passive) end return export dzrsvuy05bapy9zw7pld88uamn6no4q 4 216488 1158941 1124541 2026-09-10T04:50:12Z Alexis Jazz 28437 Updating gadget usage statistics from [[Special:GadgetUsage]] ([[phab:T121049]]) 1158941 wikitext text/x-wiki {{#ifexist:Project:GUS2Wiki/top|{{/top}}|This page provides a historical record of [[Special:GadgetUsage]] through its page history. To get the data in CSV format, see wikitext. To customize this message or add categories, create [[/top]].}} البيانات التالية مخزنة، وكان آخر تحديث لها في 2026-09-07T05:05:32Z. العدد الأقصى للنتائج المخزنة هو {{PLURAL:5000||نتيجة واحدة|نتيجتان|5000 نتائج|5000 نتيجة}}. {| class="sortable wikitable" ! الإضافة !! data-sort-type="number" | عدد المستخدمين !! data-sort-type="number" | مستخدمين نشطين |- |AjaxEdit || 30 || 2 |- |BiDiEditing || 72 || 1 |- |Cat-a-lot || 10 || 1 |- |HotCat || 87 || 7 |- |JSL || 14 || 0 |- |Linkscount || 18 || 2 |- |MobileCategories || 14 || 1 |- |Numeral converter || 1 || 0 |- |Numeral-converter-toggle || 5 || 1 |- |ShortLink || 60 || 1 |- |UTCLiveClock || 84 || 1 |- |VE-wditem-input-loader || 11 || 1 |- |XTools-ArticleInfo || 2 || 0 |- |auto-cite-wikidata-loader || data-sort-value="Infinity" | افتراضي || data-sort-value="Infinity" | افتراضي |- |autocomplete || 27 || 1 |- |mobile-sidebar || 2 || 0 |- |newsectionbottom || 11 || 0 |- |searchlang || 22 || 0 |- |speed-deletion || 6 || 2 |- |userinfo-without-deletedcontribs || 22 || 3 |- |wikEd || 44 || 2 |} * [[خاص:GadgetUsage]] * [[m:Meta:GUS2Wiki/Script|GUS2Wiki]] <!-- data in CSV format: AjaxEdit,30,2 BiDiEditing,72,1 Cat-a-lot,10,1 HotCat,87,7 JSL,14,0 Linkscount,18,2 MobileCategories,14,1 Numeral converter,1,0 Numeral-converter-toggle,5,1 ShortLink,60,1 UTCLiveClock,84,1 VE-wditem-input-loader,11,1 XTools-ArticleInfo,2,0 auto-cite-wikidata-loader,default,default autocomplete,27,1 mobile-sidebar,2,0 newsectionbottom,11,0 searchlang,22,0 speed-deletion,6,2 userinfo-without-deletedcontribs,22,3 wikEd,44,2 --> dxwlr5mjhl9850ycgi5k6j6yfj0mihr 828 221771 1158962 972700 2026-09-10T10:59:24Z حبيشان 31560 تحديث من ويكيبيديا العربية 1158962 Scribunto text/plain local aliases = { width = {'العرض', 'عرض'}, title = {'عنوان', 'العنوان'}, shaer = {'الشاعر', 'شاعر'}, bahar = {'البحر', 'بحر'}, nass = {1,'النص', 'نص'}, shaterbyshater = {'شطر شطر'}, fasil = {'فاصل','separator','fasil'}, free = {'حر', 'شعر حر'}, } local string = mw.ustring local getArgs = require('Module:Arguments with aliases').getArgs local function validString(s) if(type(s) == 'string') then s = mw.text.trim(s) if( s ~= '' ) then return s end end return nil end local function abyat(frame) local A = getArgs(frame, {aliases=aliases, art_defautls=mw.loadData("Module:Art pref/data")["أبيات"]}) local mashtoor = false local header = '' local free = A.free and (A.free ~= "0") local fasil = A.fasil and string.gsub(A.fasil, '[%%%^%$%(%)%[%]%.%*%+%-%?]', '%%%1') or "\\\\" local fasil_s,fasil_e = string.find(mw.text.trim(A.nass),fasil) local sdrdiv = '<div class="abyat-sdr">' if not fasil_s and not A[2] then mashtoor=true sdrdiv = '<div>' end local nass= "" local ret = '<div class="abyat-wrapper' .. (mashtoor and ' mashtoor' or '') .. (free and ' free' or '') .. (A.shaterbyshater and ' abyat-shaterbyshater' or '') .. '" style="width:' .. (A.width or 'fit-content') .. ';">' if A.title then header = header .. "<center>'''" .. A.title .. "'''</center>" end if A.shaer then header = header .. "<center>" .. A.shaer .. "</center>" end if A.bahar then header = header .. "<center>''[من " .. A.bahar .. "]''</center>" end -- نمط متوافق مع أساليب القوالب القديمة if A[2] then local shater_num =1 local aa,bb aa = validString(A[shater_num]) bb = validString(A[shater_num+1]) while aa ~= nil or bb ~= nil do if aa~= nil and bb ~= nil then ret=ret .. '<div class="abyat-sdr">' ..aa .. "</div>" ret=ret .. '<div class="abyat-ajz">' ..bb .. "</div>\n" else ret=ret .. '<div class="abyat-single_bayt">' ..(aa or bb) .. "</div>\n" end shater_num=shater_num+2 aa = validString(A[shater_num]) bb = validString(A[shater_num+1]) end else -- نمط أبيات nass = mw.text.trim(A.nass) if fasil_s == 1 then ret = ret .. '<div class="abyat-single_bayt">' nass = string.sub(nass,fasil_e+1) else ret = ret .. sdrdiv end nass=string.gsub(nass,"\n"..fasil,'</div><div class="abyat-single_bayt">') nass=string.gsub(nass,"\n([^|])",'</div>' .. sdrdiv ..'%1') nass=string.gsub(nass, fasil,'</div><div class="abyat-ajz">') end return header .. ret .. nass .. "</div></div>" end return {abyat=abyat} bznhqqvt94hsxr7dyj30qxhua8iwnyn 10 231222 1158953 1097323 2026-09-10T10:41:54Z حبيشان 31560 1158953 wikitext text/x-wiki * {{استشهاد بويكي بيانات/ملعب|Q115727209|مرجع=أبو السعود|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات|Q115727209|مرجع=أبو السعود}} * {{استشهاد بويكي بيانات/ملعب|Q115641715|مرجع=الشربيني1868}} * {{استشهاد بويكي بيانات|Q115641715|مرجع=الشربيني1868|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات/ملعب|Q115683910|مرجع=القرطبي1935}} * {{استشهاد بويكي بيانات|Q115683910|مرجع=القرطبي1935}} * {{استشهاد بويكي بيانات/ملعب|Q115728353|مرجع=السيوطي1974}} * {{استشهاد بويكي بيانات|Q115728353|مرجع=السيوطي1974}} * {{استشهاد بويكي بيانات/ملعب|Q115732199|مرجع=السيوطي1978}} * {{استشهاد بويكي بيانات|Q115732199|مرجع=السيوطي1978}} * {{استشهاد بويكي بيانات/ملعب|Q115641917|مرجع=القاضي1981}} * {{استشهاد بويكي بيانات|Q115641917|مرجع=القاضي1981}} * {{استشهاد بويكي بيانات/ملعب|Q115641984|مرجع=ابن عاشور1984}} * {{استشهاد بويكي بيانات|Q115641984|مرجع=ابن عاشور1984}} * {{استشهاد بويكي بيانات/ملعب|Q115683837}} * {{استشهاد بويكي بيانات|Q115683837}} * {{استشهاد بويكي بيانات/ملعب|Q115782668|مرجع=النحاس1985|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات|Q115782668|مرجع=النحاس1985}} * {{استشهاد بويكي بيانات/ملعب|Q115746114|مرجع=القحطاني1989}} * {{استشهاد بويكي بيانات|Q115746114|مرجع=القحطاني1989}} * {{استشهاد بويكي بيانات/ملعب|Q19443162}} * {{استشهاد بويكي بيانات|Q19443162}} * {{استشهاد بويكي بيانات/ملعب|Q115640897|مرجع=درويش1992}} * {{استشهاد بويكي بيانات|Q115640897|مرجع=درويش1992}} * {{استشهاد بويكي بيانات/ملعب|Q115754123|مرجع=الواحدي1992}} * {{استشهاد بويكي بيانات|Q115754123|مرجع=الواحدي1992}} * {{استشهاد بويكي بيانات/ملعب|Q115651867|مرجع=الداني1994}} * {{استشهاد بويكي بيانات|Q115651867|مرجع=الداني1994}} * {{استشهاد بويكي بيانات/ملعب|Q115756452|مرجع=عطية1995}} * {{استشهاد بويكي بيانات|Q115756452|مرجع=عطية1995}} * {{استشهاد بويكي بيانات/ملعب|Q115701544|مرجع=ابن كثير1998}} * {{استشهاد بويكي بيانات|Q115701544|مرجع=ابن كثير1998}} * {{استشهاد بويكي بيانات/ملعب|Q115730072|مرجع=ابن عادل1998}} * {{استشهاد بويكي بيانات|Q115730072|مرجع=ابن عادل1998}} * {{استشهاد بويكي بيانات/ملعب|Q115730115|مرجع=الرازي2000}} * {{استشهاد بويكي بيانات|Q115730115|مرجع=الرازي2000}} * {{استشهاد بويكي بيانات/ملعب|Q97221368|مرجع=الطبري2001}} * {{استشهاد بويكي بيانات|Q97221368|مرجع=الطبري2001}} * {{استشهاد بويكي بيانات/ملعب|Q115484580|مرجع=أحمد2001}} * {{استشهاد بويكي بيانات|Q115484580|مرجع=أحمد2001}} * {{استشهاد بويكي بيانات/ملعب|Q115683684}} * {{استشهاد بويكي بيانات|Q115683684}} * {{استشهاد بويكي بيانات/ملعب|Q115640158|مرجع=الدعاس2004}} * {{استشهاد بويكي بيانات|Q115640158|مرجع=الدعاس2004}} * {{استشهاد بويكي بيانات/ملعب|Q115683765|مرجع=سلامة2006}} * {{استشهاد بويكي بيانات|Q115683765|مرجع=سلامة2006}} * {{استشهاد بويكي بيانات/ملعب|Q115701920|مرجع=ابن القيم2007}} * {{استشهاد بويكي بيانات|Q115701920|مرجع=ابن القيم2007}} * {{استشهاد بويكي بيانات/ملعب|Q115743500}} * {{استشهاد بويكي بيانات|Q115743500}} * {{استشهاد بويكي بيانات/ملعب|Q115615370|مرجع=الشايع2011}} * {{استشهاد بويكي بيانات|Q115615370|مرجع=الشايع2011}} * {{استشهاد بويكي بيانات/ملعب|Q140501357}} * {{استشهاد بويكي بيانات|Q140501357}} * {{استشهاد بويكي بيانات/ملعب|Q136804563|ج=2|ص=442}} * {{استشهاد بويكي بيانات|Q136804563|ج=2|ص=442}} * {{استشهاد بويكي بيانات/ملعب|Q115796840|ج=2|ص=100}} * {{استشهاد بويكي بيانات|Q115796840|ج=2|ص=100}} * {{استشهاد بويكي بيانات/ملعب|Q114810954|ص=100}} * {{استشهاد بويكي بيانات|Q114810954|ص=100}} <div dir="ltr"> * {{استشهاد بويكي بيانات/ملعب|Q115917529|مرجع=Dyson1999}} * {{استشهاد بويكي بيانات|Q115917529|مرجع=Dyson1999}} * {{استشهاد بويكي بيانات/ملعب|Q118316084|مرجع=Javvin2005}} * {{استشهاد بويكي بيانات|Q118316084|مرجع=Javvin2005}} * {{استشهاد بويكي بيانات/ملعب|Q115917346|مرجع=Kozierok2005}} * {{استشهاد بويكي بيانات|Q115917346|مرجع=Kozierok2005}} * {{استشهاد بويكي بيانات/ملعب|Q115922106|مرجع=Fall2012}} * {{استشهاد بويكي بيانات|Q115922106|مرجع=Fall2012}} * {{استشهاد بويكي بيانات/ملعب|Q115919723|مرجع=Odom2013}} * {{استشهاد بويكي بيانات|Q115919723|مرجع=Odom2013}} * {{استشهاد بويكي بيانات/ملعب|Q9065143}} </div> bno8i4lhh380o6b88fmd2ryvjws8pr1 1158970 1158953 2026-09-10T11:58:02Z حبيشان 31560 1158970 wikitext text/x-wiki {{تفضيات مقالة|استشهاد بويكي بيانات:سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات/ملعب|Q115727209|مرجع=أبو السعود|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات|Q115727209|مرجع=أبو السعود}} * {{استشهاد بويكي بيانات/ملعب|Q115641715|مرجع=الشربيني1868}} * {{استشهاد بويكي بيانات|Q115641715|مرجع=الشربيني1868|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات/ملعب|Q115683910|مرجع=القرطبي1935}} * {{استشهاد بويكي بيانات|Q115683910|مرجع=القرطبي1935}} * {{استشهاد بويكي بيانات/ملعب|Q115728353|مرجع=السيوطي1974}} * {{استشهاد بويكي بيانات|Q115728353|مرجع=السيوطي1974}} * {{استشهاد بويكي بيانات/ملعب|Q115732199|مرجع=السيوطي1978}} * {{استشهاد بويكي بيانات|Q115732199|مرجع=السيوطي1978}} * {{استشهاد بويكي بيانات/ملعب|Q115641917|مرجع=القاضي1981}} * {{استشهاد بويكي بيانات|Q115641917|مرجع=القاضي1981}} * {{استشهاد بويكي بيانات/ملعب|Q115641984|مرجع=ابن عاشور1984}} * {{استشهاد بويكي بيانات|Q115641984|مرجع=ابن عاشور1984}} * {{استشهاد بويكي بيانات/ملعب|Q115683837}} * {{استشهاد بويكي بيانات|Q115683837}} * {{استشهاد بويكي بيانات/ملعب|Q115782668|مرجع=النحاس1985|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات|Q115782668|مرجع=النحاس1985}} * {{استشهاد بويكي بيانات/ملعب|Q115746114|مرجع=القحطاني1989}} * {{استشهاد بويكي بيانات|Q115746114|مرجع=القحطاني1989}} * {{استشهاد بويكي بيانات/ملعب|Q19443162}} * {{استشهاد بويكي بيانات|Q19443162}} * {{استشهاد بويكي بيانات/ملعب|Q115640897|مرجع=درويش1992}} * {{استشهاد بويكي بيانات|Q115640897|مرجع=درويش1992}} * {{استشهاد بويكي بيانات/ملعب|Q115754123|مرجع=الواحدي1992}} * {{استشهاد بويكي بيانات|Q115754123|مرجع=الواحدي1992}} * {{استشهاد بويكي بيانات/ملعب|Q115651867|مرجع=الداني1994}} * {{استشهاد بويكي بيانات|Q115651867|مرجع=الداني1994}} * {{استشهاد بويكي بيانات/ملعب|Q115756452|مرجع=عطية1995}} * {{استشهاد بويكي بيانات|Q115756452|مرجع=عطية1995}} * {{استشهاد بويكي بيانات/ملعب|Q115701544|مرجع=ابن كثير1998}} * {{استشهاد بويكي بيانات|Q115701544|مرجع=ابن كثير1998}} * {{استشهاد بويكي بيانات/ملعب|Q115730072|مرجع=ابن عادل1998}} * {{استشهاد بويكي بيانات|Q115730072|مرجع=ابن عادل1998}} * {{استشهاد بويكي بيانات/ملعب|Q115730115|مرجع=الرازي2000}} * {{استشهاد بويكي بيانات|Q115730115|مرجع=الرازي2000}} * {{استشهاد بويكي بيانات/ملعب|Q97221368|مرجع=الطبري2001}} * {{استشهاد بويكي بيانات|Q97221368|مرجع=الطبري2001}} * {{استشهاد بويكي بيانات/ملعب|Q115484580|مرجع=أحمد2001}} * {{استشهاد بويكي بيانات|Q115484580|مرجع=أحمد2001}} * {{استشهاد بويكي بيانات/ملعب|Q115683684}} * {{استشهاد بويكي بيانات|Q115683684}} * {{استشهاد بويكي بيانات/ملعب|Q115640158|مرجع=الدعاس2004}} * {{استشهاد بويكي بيانات|Q115640158|مرجع=الدعاس2004}} * {{استشهاد بويكي بيانات/ملعب|Q115683765|مرجع=سلامة2006}} * {{استشهاد بويكي بيانات|Q115683765|مرجع=سلامة2006}} * {{استشهاد بويكي بيانات/ملعب|Q115701920|مرجع=ابن القيم2007}} * {{استشهاد بويكي بيانات|Q115701920|مرجع=ابن القيم2007}} * {{استشهاد بويكي بيانات/ملعب|Q115743500}} * {{استشهاد بويكي بيانات|Q115743500}} * {{استشهاد بويكي بيانات/ملعب|Q115615370|مرجع=الشايع2011}} * {{استشهاد بويكي بيانات|Q115615370|مرجع=الشايع2011}} * {{استشهاد بويكي بيانات/ملعب|Q140501357}} * {{استشهاد بويكي بيانات|Q140501357}} * {{استشهاد بويكي بيانات/ملعب|Q136804563|ج=2|ص=442}} * {{استشهاد بويكي بيانات|Q136804563|ج=2|ص=442}} * {{استشهاد بويكي بيانات/ملعب|Q115796840|ج=2|ص=100}} * {{استشهاد بويكي بيانات|Q115796840|ج=2|ص=100}} * {{استشهاد بويكي بيانات/ملعب|Q114810954|ص=100}} * {{استشهاد بويكي بيانات|Q114810954|ص=100}} <div dir="ltr"> * {{استشهاد بويكي بيانات/ملعب|Q115917529|مرجع=Dyson1999}} * {{استشهاد بويكي بيانات|Q115917529|مرجع=Dyson1999}} * {{استشهاد بويكي بيانات/ملعب|Q118316084|مرجع=Javvin2005}} * {{استشهاد بويكي بيانات|Q118316084|مرجع=Javvin2005}} * {{استشهاد بويكي بيانات/ملعب|Q115917346|مرجع=Kozierok2005}} * {{استشهاد بويكي بيانات|Q115917346|مرجع=Kozierok2005}} * {{استشهاد بويكي بيانات/ملعب|Q115922106|مرجع=Fall2012}} * {{استشهاد بويكي بيانات|Q115922106|مرجع=Fall2012}} * {{استشهاد بويكي بيانات/ملعب|Q115919723|مرجع=Odom2013}} * {{استشهاد بويكي بيانات|Q115919723|مرجع=Odom2013}} * {{استشهاد بويكي بيانات/ملعب|Q9065143}} </div> 6f4yltel6vj425ctuemzk9iz9ctuyfe 828 231288 1158948 1097948 2026-09-10T10:25:10Z حبيشان 31560 1158948 Scribunto text/plain require ('strict'); --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- each of these counts against the Lua upvalue limit ]] local validation; -- functions in Module:Citation/CS1/Date_validation local utilities; -- functions in Module:Citation/CS1/Utilities local z = {}; -- table of tables in Module:Citation/CS1/Utilities local identifiers; -- functions and tables in Module:Citation/CS1/Identifiers local metadata; -- functions in Module:Citation/CS1/COinS local cfg = {}; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist --[[------------------< P A G E S C O P E V A R I A B L E S >--------------- declare variables here that have page-wide scope that are not brought in from other modules; that are created here and used here ]] local added_deprecated_cat; -- boolean flag so that the category is added only once local added_vanc_errs; -- boolean flag so we only emit one Vancouver error / category local added_generic_name_errs; -- boolean flag so we only emit one generic name error / category and stop testing names once an error is encountered local added_numeric_name_errs; -- boolean flag so we only emit one numeric name error / category and stop testing names once an error is encountered local added_numeric_name_maint; -- boolean flag so we only emit one numeric name maint category and stop testing names once a category has been emitted --#to be deleted as source local Frame; -- holds the module's frame table local is_preview_mode; -- true when article is in preview mode; false when using 'Preview page with this template' (previewing the module) local is_sandbox; -- true when using sandbox modules to render citation local url_was_archive_today; -- boolean flag set to true when has_archive_today_url() has unset |url= because it held a deprecated archive.today url --#Added local wikidatacite = false; --[[--------------------------< F I R S T _ S E T >------------------------------------------------------------ Locates and returns the first set value in a table of values where the order established in the table, left-to-right (or top-to-bottom), is the order in which the values are evaluated. Returns nil if none are set. This version replaces the original 'for _, val in pairs do' and a similar version that used ipairs. With the pairs version the order of evaluation could not be guaranteed. With the ipairs version, a nil value would terminate the for-loop before it reached the actual end of the list. ]] local function first_set (list, count) local i = 1; while i <= count do -- loop through all items in list if utilities.is_set( list[i] ) then return list[i]; -- return the first set list member end i = i + 1; -- point to next end end --[[--------------------------< A D D _ V A N C _ E R R O R >---------------------------------------------------- Adds a single Vancouver system error message to the template's output regardless of how many error actually exist. To prevent duplication, added_vanc_errs is nil until an error message is emitted. added_vanc_errs is a Boolean declared in page scope variables above ]] local function add_vanc_error (source, position) if added_vanc_errs then return end added_vanc_errs = true; -- note that we've added this category utilities.set_message ('err_vancouver', {source, position}); end --[[--------------------------< I S _ S C H E M E >------------------------------------------------------------ does this thing that purports to be a URI scheme seem to be a valid scheme? The scheme is checked to see if it is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says: Scheme names consist of a sequence of characters beginning with a letter and followed by any combination of letters, digits, plus ("+"), period ("."), or hyphen ("-"). returns true if it does, else false ]] local function is_scheme (scheme) return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern end --[=[-------------------------< I S _ D O M A I N _ N A M E >-------------------------------------------------- Does this thing that purports to be a domain name seem to be a valid domain name? Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5 BNF defined here: https://tools.ietf.org/html/rfc4234 Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15; see also [[Single-letter second-level domain]] list of TLDs: https://www.iana.org/domains/root/db RFC 952 (modified by RFC 1123) requires the first and last character of a hostname to be a letter or a digit. Between the first and last characters the name may use letters, digits, and the hyphen. Also allowed are IPv4 addresses. IPv6 not supported domain is expected to be stripped of any path so that the last character in the last character of the TLD. tld is two or more alpha characters. Any preceding '//' (from splitting a URL with a scheme) will be stripped here. Perhaps not necessary but retained in case it is necessary for IPv4 dot decimal. There are several tests: the first character of the whole domain name including subdomains must be a letter or a digit internationalized domain name (ASCII characters with .xn-- ASCII Compatible Encoding (ACE) prefix xn-- in the TLD) see https://tools.ietf.org/html/rfc3490 single-letter/digit second-level domains in the .org, .cash, and .today TLDs q, x, and z SL domains in the .com TLD i and q SL domains in the .net TLD single-letter SL domains in the ccTLDs (where the ccTLD is two letters) two-character SL domains in gTLDs (where the gTLD is two or more letters) three-plus-character SL domains in gTLDs (where the gTLD is two or more letters) IPv4 dot-decimal address format; TLD not allowed returns true if domain appears to be a proper name and TLD or IPv4 address, else false ]=] local function is_domain_name (domain) if not domain then return false; -- if not set, abandon end domain = domain:gsub ('^//', ''); -- strip '//' from domain name if present; done here so we only have to do it once if not domain:match ('^[%w]') then -- first character must be letter or digit return false; end if domain:match ('^%a+:') then -- hack to detect things that look like s:Page:Title where Page: is namespace at Wikisource return false; end local patterns = { -- patterns that look like URLs '%f[%w][%w][%w%-]+[%w]%.%a%a+$', -- three or more character hostname.hostname or hostname.tld '%f[%w][%w][%w%-]+[%w]%.xn%-%-[%w]+$', -- internationalized domain name with ACE prefix '%f[%a][qxz]%.com$', -- assigned one character .com hostname (x.com times out 2015-12-10) '%f[%a][iq]%.net$', -- assigned one character .net hostname (q.net registered but not active 2015-12-10) '%f[%w][%w]%.%a%a$', -- one character hostname and ccTLD (2 chars) '%f[%w][%w][%w]%.%a%a+$', -- two character hostname and TLD '^%d%d?%d?%.%d%d?%d?%.%d%d?%d?%.%d%d?%d?', -- IPv4 address '[%a%d]+%:?' -- IPv6 address } for _, pattern in ipairs (patterns) do -- loop through the patterns list if domain:match (pattern) then return true; -- if a match then we think that this thing that purports to be a URL is a URL end end for _, d in ipairs (cfg.single_letter_2nd_lvl_domains_t) do -- look for single letter second level domain names for these top level domains if domain:match ('%f[%w][%w]%.' .. d) then return true end end return false; -- no matches, we don't know what this thing is end --[[--------------------------< I S _ U R L >------------------------------------------------------------------ returns true if the scheme and domain parts of a URL appear to be a valid URL; else false. This function is the last step in the validation process. This function is separate because there are cases that are not covered by split_url(), for example is_parameter_ext_wikilink() which is looking for bracketted external wikilinks. ]] local function is_url (scheme, domain) if utilities.is_set (scheme) then -- if scheme is set check it and domain return is_scheme (scheme) and is_domain_name (domain); else return is_domain_name (domain); -- scheme not set when URL is protocol-relative end end --[[--------------------------< S P L I T _ U R L >------------------------------------------------------------ Split a URL into a scheme, authority indicator, and domain. First remove Fully Qualified Domain Name terminator (a dot following TLD) (if any) and any path(/), query(?) or fragment(#). If protocol-relative URL, return nil scheme and domain else return nil for both scheme and domain. When not protocol-relative, get scheme, authority indicator, and domain. If there is an authority indicator (one or more '/' characters immediately following the scheme's colon), make sure that there are only 2. Any URL that does not have news: scheme must have authority indicator (//). TODO: are there other common schemes like news: that don't use authority indicator? Strip off any port and path; ]] local function split_url (url_str) local scheme, authority, domain; url_str = url_str:gsub ('([%a%d])%.?[/%?#].*$', '%1'); -- strip FQDN terminator and path(/), query(?), fragment (#) (the capture prevents false replacement of '//') if url_str:match ('^//%S*') then -- if there is what appears to be a protocol-relative URL domain = url_str:match ('^//(%S*)') elseif url_str:match ('%S-:/*%S+') then -- if there is what appears to be a scheme, optional authority indicator, and domain name scheme, authority, domain = url_str:match ('(%S-:)(/*)(%S+)'); -- extract the scheme, authority indicator, and domain portions if utilities.is_set (authority) then authority = authority:gsub ('//', '', 1); -- replace place 1 pair of '/' with nothing; if utilities.is_set(authority) then -- if anything left (1 or 3+ '/' where authority should be) then return scheme; -- return scheme only making domain nil which will cause an error message end else if not scheme:match ('^news:') then -- except for news:..., MediaWiki won't link URLs that do not have authority indicator; TODO: a better way to do this test? return scheme; -- return scheme only making domain nil which will cause an error message end end domain = domain:gsub ('(%a):%d+', '%1'); -- strip port number if present end return scheme, domain; end --[[--------------------------< L I N K _ P A R A M _ O K >--------------------------------------------------- checks the content of |title-link=, |series-link=, |author-link=, etc. for properly formatted content: no wikilinks, no URLs Link parameters are to hold the title of a Wikipedia article, so none of the WP:TITLESPECIALCHARACTERS are allowed: # < > [ ] | { } _ except the underscore which is used as a space in wiki URLs and # which is used for section links returns false when the value contains any of these characters. When there are no illegal characters, this function returns TRUE if value DOES NOT appear to be a valid URL (the |<param>-link= parameter is ok); else false when value appears to be a valid URL (the |<param>-link= parameter is NOT ok). ]] local function link_param_ok (value) local scheme, domain; if value:find ('[<>%[%]|{}]') then -- if any prohibited characters return false; end scheme, domain = split_url (value); -- get scheme or nil and domain or nil from URL; return not is_url (scheme, domain); -- return true if value DOES NOT appear to be a valid URL end --[[--------------------------< L I N K _ T I T L E _ O K >--------------------------------------------------- Use link_param_ok() to validate |<param>-link= value and its matching |<title>= value. |<title>= may be wiki-linked but not when |<param>-link= has a value. This function emits an error message when that condition exists check <link> for inter-language interwiki-link prefix. prefix must be a MediaWiki-recognized language code and must begin with a colon. ]] local function link_title_ok (link, lorig, title, torig) local orig; if utilities.is_set (link) then -- don't bother if <param>-link doesn't have a value if not link_param_ok (link) then -- check |<param>-link= markup orig = lorig; -- identify the failing link parameter elseif title:find ('%[%[') then -- check |title= for wikilink markup orig = torig; -- identify the failing |title= parameter elseif link:match ('^%a+:') then -- if the link is what looks like an interwiki local prefix = link:match ('^(%a+):'):lower(); -- get the interwiki prefix if cfg.inter_wiki_map[prefix] then -- if prefix is in the map, must have preceding colon orig = lorig; -- flag as error end end end if utilities.is_set (orig) then link = ''; -- unset utilities.set_message ('err_bad_paramlink', orig); -- URL or wikilink in |title= with |title-link=; end return link; -- link if ok, empty string else end --[[--------------------------< C H E C K _ U R L >------------------------------------------------------------ Determines whether a URL string appears to be valid. First we test for space characters. If any are found, return false. Then split the URL into scheme and domain portions, or for protocol-relative (//example.com) URLs, just the domain. Use is_url() to validate the two portions of the URL. If both are valid, or for protocol-relative if domain is valid, return true, else false. Because it is different from a standard URL, and because this module used external_link() to make external links that work for standard and news: links, we validate newsgroup names here. The specification for a newsgroup name is at https://tools.ietf.org/html/rfc5536#section-3.1.4 ]] local function check_url( url_str ) if nil == url_str:match ("^%S+$") then -- if there are any spaces in |url=value it can't be a proper URL return false; end local scheme, domain; scheme, domain = split_url (url_str); -- get scheme or nil and domain or nil from URL; if 'news:' == scheme then -- special case for newsgroups return domain:match('^[%a%d%+%-_]+%.[%a%d%+%-_%.]*[%a%d%+%-_]$'); end return is_url (scheme, domain); -- return true if value appears to be a valid URL end --[=[-------------------------< I S _ P A R A M E T E R _ E X T _ W I K I L I N K >---------------------------- Return true if a parameter value has a string that begins and ends with square brackets [ and ] and the first non-space characters following the opening bracket appear to be a URL. The test will also find external wikilinks that use protocol-relative URLs. Also finds bare URLs. The frontier pattern prevents a match on interwiki-links which are similar to scheme:path URLs. The tests that find bracketed URLs are required because the parameters that call this test (currently |title=, |chapter=, |work=, and |publisher=) may have wikilinks and there are articles or redirects like '//Hus' so, while uncommon, |title=[[//Hus]] is possible as might be [[en://Hus]]. ]=] local function is_parameter_ext_wikilink (value) local scheme, domain; if value:match ('%f[%[]%[%a%S*:%S+.*%]') then -- if ext. wikilink with scheme and domain: [xxxx://yyyyy.zzz] scheme, domain = split_url (value:match ('%f[%[]%[(%a%S*:%S+).*%]')); elseif value:match ('%f[%[]%[//%S+.*%]') then -- if protocol-relative ext. wikilink: [//yyyyy.zzz] scheme, domain = split_url (value:match ('%f[%[]%[(//%S+).*%]')); elseif value:match ('%a%S*:%S+') then -- if bare URL with scheme; may have leading or trailing plain text scheme, domain = split_url (value:match ('(%a%S*:%S+)')); elseif value:match ('^//%S+') or value:match ('%s//%S+') then -- if protocol-relative bare URL: //yyyyy.zzz; authority indicator (//) must be be preceded nothing or by whitespace scheme, domain = split_url (value:match ('(//%S+)')); -- what is left should be the domain else return false; -- didn't find anything that is obviously a URL end return is_url (scheme, domain); -- return true if value appears to be a valid URL end --[[--------------------------< I S _ A R C H I V E _ T O D A Y _ U R L >-------------------------------------- archive.today (and .ph, .is, .md, .li, .fo, .vn) are deprecated. See Wikipedia:Requests_for_comment/Archive.is_RFC_5 (Special:Permalink/1339352976) <val> is a string that may or may not be a url ]] local function is_archive_today_url (val) if is_parameter_ext_wikilink (val) then -- is <val> a url? local tld = val:match ('archive%.(%l+)'); -- get the tld return tld and cfg.archive_today_tlds_t[tld]; -- return true if <tld> is one of the archive.today tlds; nil else end end --[[--------------------------< H A S _ A R C H I V E _ T O D A Y _ U R L >------------------------------------ blank any parameter that holds one of the archive.today tlds (.today, .ph, .is, .md, .li, .fo, .vn) and also blank parameters that are dependent on the offending parameter. input <cite_args_t> is a table of all of this template's parameters This function modifies <cite_args_t> when |url= or its alias |URL= hold an archive.today url, this function sets the flag <url_was_archive_today> which is used later to suppress the missing url error message from {{citation}} when it has |website=<summat> or from the templates {{cite web}}, {{cite podcast}}, {{cite mailing list}} which all require a url. returns nothing ]] local function has_archive_today_url (cite_args_t) local unset_params_t = {}; -- to hold a sequence of parameters to unset from cite_args_t for p, v in pairs (cite_args_t) do -- look at each parameter in the template if is_archive_today_url (v:lower()) then -- if it has a parameter with an archive.today url table.insert (unset_params_t, p); -- add the url-holding parameter name to our list of parameters to unset (may or may not be a legitimate url parameter) if 'url' == p:lower() then -- if this parameter is |url= or |URL= url_was_archive_today = true; -- set a flag so that cite web / podcast / mailinglist don't emit a missing url error message end if cfg.dependencies_t[p] then -- if this url-holding parameter (<p>) has dependencies for _, dependent in ipairs (cfg.dependencies_t[p]) do -- loop through the url-holding parameter's dependency sequence table.insert (unset_params_t, dependent); -- add dependent parameter to the list to be unset end end end end if 0 < #unset_params_t then -- when there are parameters to unset if not is_preview_mode then -- when in preview mode do not suppress any parameters for _, unset_param in ipairs (unset_params_t) do -- loop through the sequence cite_args_t[unset_param] = nil; -- and unset the url-holding parameter and its dependent parameters to avoid parameter-requires-other-parameter errors end end utilities.set_message ('maint_deprecated_archive'); -- add a maintenance message end end --[[-------------------------< C H E C K _ F O R _ U R L >----------------------------------------------------- loop through a list of parameters and their values. Look at the value and if it has an external link, emit an error message. ]] local function check_for_url (parameter_list, error_list) for k, v in pairs (parameter_list) do -- for each parameter in the list if is_parameter_ext_wikilink (v) then -- look at the value; if there is a URL add an error message table.insert (error_list, utilities.wrap_style ('parameter', k)); end end end --[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------ Escape sequences for content that will be used for URL descriptions ]] local function safe_for_url( str ) if str:match( "%[%[.-%]%]" ) ~= nil then utilities.set_message ('err_wikilink_in_url', {}); end return str:gsub( '[%[%]\n]', { ['['] = '&#91;', [']'] = '&#93;', ['\n'] = ' ' } ); end --[[--------------------------< E X T E R N A L _ L I N K >---------------------------------------------------- Format an external link with error checking ]] local function external_link (URL, label, source, access) local err_msg = ''; local domain; local path; local base_url; if not utilities.is_set (label) then label = URL; if utilities.is_set (source) then --#Modified utilities.set_message (wikidatacite and 'err_wikidata_bare_url_missing_title' or 'err_bare_url_missing_title', {utilities.wrap_style ('parameter', source)}); else error (cfg.messages["bare_url_no_origin"]); -- programmer error; valid parameter name does not have matching meta-parameter end end if not check_url (URL) then utilities.set_message ('err_bad_url', {utilities.wrap_style ('parameter', source)}); end domain, path = URL:match ('^([/%.%-%+:%a%d]+)([/%?#].*)$'); -- split the URL into scheme plus domain and path if path then -- if there is a path portion path = path:gsub ('[%[%]]', {['['] = '%5b', [']'] = '%5d'}); -- replace '[' and ']' with their percent-encoded values URL = table.concat ({domain, path}); -- and reassemble end base_url = table.concat ({ "[", URL, " ", safe_for_url (label), "]" }); -- assemble a wiki-markup URL if utilities.is_set (access) then -- access level (subscription, registration, limited) base_url = utilities.substitute (cfg.presentation['ext-link-access-signal'], {cfg.presentation[access].class, cfg.presentation[access].title, base_url}); -- add the appropriate icon end return base_url; end --[[--------------------------< D E P R E C A T E D _ P A R A M E T E R >-------------------------------------- Categorize and emit an error message when the citation contains one or more deprecated parameters. The function includes the offending parameter name to the error message. Only one error message is emitted regardless of the number of deprecated parameters in the citation. added_deprecated_cat is a Boolean declared in page scope variables above ]] local function deprecated_parameter(name) if not added_deprecated_cat then added_deprecated_cat = true; -- note that we've added this category utilities.set_message ('err_deprecated_params', {name}); -- add error message end end --[=[-------------------------< K E R N _ Q U O T E S >-------------------------------------------------------- Apply kerning to open the space between the quote mark provided by the module and a leading or trailing quote mark contained in a |title= or |chapter= parameter's value. This function will positive kern either single or double quotes: "'Unkerned title with leading and trailing single quote marks'" " 'Kerned title with leading and trailing single quote marks' " (in real life the kerning isn't as wide as this example) Double single quotes (italic or bold wiki-markup) are not kerned. Replaces Unicode quote marks in plain text or in the label portion of a [[L|D]] style wikilink with typewriter quote marks regardless of the need for kerning. Unicode quote marks are not replaced in simple [[D]] wikilinks. Call this function for chapter titles, for website titles, etc.; not for book titles. ]=] local function kern_quotes (str) local cap = ''; local wl_type, label, link; wl_type, label, link = utilities.is_wikilink (str); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]] if 1 == wl_type then -- [[D]] simple wikilink with or without quote marks if mw.ustring.match (str, '%[%[[\"“”\'‘’].+[\"“”\'‘’]%]%]') then -- leading and trailing quote marks str = utilities.substitute (cfg.presentation['kern-left'], str); str = utilities.substitute (cfg.presentation['kern-right'], str); elseif mw.ustring.match (str, '%[%[[\"“”\'‘’].+%]%]') then -- leading quote marks str = utilities.substitute (cfg.presentation['kern-left'], str); elseif mw.ustring.match (str, '%[%[.+[\"“”\'‘’]%]%]') then -- trailing quote marks str = utilities.substitute (cfg.presentation['kern-right'], str); end else -- plain text or [[L|D]]; text in label variable label = mw.ustring.gsub (label, '[“”]', '\"'); -- replace “” (U+201C & U+201D) with " (typewriter double quote mark) label = mw.ustring.gsub (label, '[‘’]', '\''); -- replace ‘’ (U+2018 & U+2019) with ' (typewriter single quote mark) cap = mw.ustring.match (label, "^([\"\'][^\'].+)"); -- match leading double or single quote but not doubled single quotes (italic markup) if utilities.is_set (cap) then label = utilities.substitute (cfg.presentation['kern-left'], cap); end cap = mw.ustring.match (label, "^(.+[^\'][\"\'])$") -- match trailing double or single quote but not doubled single quotes (italic markup) if utilities.is_set (cap) then label = utilities.substitute (cfg.presentation['kern-right'], cap); end if 2 == wl_type then str = utilities.make_wikilink (link, label); -- reassemble the wikilink else str = label; end end return str; end --[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >---------------------------------------- |script-title= holds title parameters that are not written in Latin-based scripts: Chinese, Japanese, Arabic, Hebrew, etc. These scripts should not be italicized and may be written right-to-left. The value supplied by |script-title= is concatenated onto Title after Title has been wrapped in italic markup. Regardless of language, all values provided by |script-title= are wrapped in <bdi>...</bdi> tags to isolate RTL languages from the English left to right. |script-title= provides a unique feature. The value in |script-title= may be prefixed with a two-character ISO 639-1 language code and a colon: |script-title=ja:*** *** (where * represents a Japanese character) Spaces between the two-character code and the colon and the colon and the first script character are allowed: |script-title=ja : *** *** |script-title=ja: *** *** |script-title=ja :*** *** Spaces preceding the prefix are allowed: |script-title = ja:*** *** The prefix is checked for validity. If it is a valid ISO 639-1 language code, the lang attribute (lang="ja") is added to the <bdi> tag so that browsers can know the language the tag contains. This may help the browser render the script more correctly. If the prefix is invalid, the lang attribute is not added. At this time there is no error message for this condition. Supports |script-title=, |script-chapter=, |script-<periodical>= ]] local function format_script_value (script_value, script_param) local lang=''; -- initialize to empty string local name; if script_value:match('^%l%l%l?%s*:') then -- if first 3 or 4 non-space characters are script language prefix lang = script_value:match('^(%l%l%l?)%s*:%s*%S.*'); -- get the language prefix or nil if there is no script if not utilities.is_set (lang) then utilities.set_message ('err_script_parameter', {script_param, cfg.err_msg_supl['missing title part']}); -- prefix without 'title'; add error message return ''; -- script_value was just the prefix so return empty string end -- if we get this far we have prefix and script --#modified for faster name = cfg.lang_tag_remap[lang] or cfg.mw_languages_by_tag_t[lang]; -- get language name so that we can use it to categorize if utilities.is_set (name) then -- is prefix a proper ISO 639-1 language code? script_value = script_value:gsub ('^%l+%s*:%s*', ''); -- strip prefix from script -- is prefix one of these language codes? if utilities.in_array (lang, cfg.script_lang_codes) then utilities.add_prop_cat ('script', {name, lang}) else utilities.set_message ('err_script_parameter', {script_param, cfg.err_msg_supl['unknown language code']}); -- unknown script-language; add error message end lang = ' lang="' .. lang .. '" '; -- convert prefix into a lang attribute else utilities.set_message ('err_script_parameter', {script_param, cfg.err_msg_supl['invalid language code']}); -- invalid language code; add error message lang = ''; -- invalid so set lang to empty string end else utilities.set_message ('err_script_parameter', {script_param, cfg.err_msg_supl['missing prefix']}); -- no language code prefix; add error message end script_value = utilities.substitute (cfg.presentation['bdi'], {lang, script_value}); -- isolate in case script is RTL return script_value; end --[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------ Initially for |title= and |script-title=, this function concatenates those two parameter values after the script value has been wrapped in <bdi> tags. ]] local function script_concatenate (title, script, script_param) if utilities.is_set (script) then script = format_script_value (script, script_param); -- <bdi> tags, lang attribute, categorization, etc.; returns empty string on error if utilities.is_set (script) then title = title .. ' ' .. script; -- concatenate title and script title end end return title; end --[[--------------------------< W R A P _ M S G >-------------------------------------------------------------- Applies additional message text to various parameter values. Supplied string is wrapped using a message_list configuration taking one argument. Supports lower case text for {{citation}} templates. Additional text taken from citation_config.messages - the reason this function is similar to but separate from wrap_style(). ]] local function wrap_msg (key, str, lower) if not utilities.is_set ( str ) then return ""; end if true == lower then local msg; msg = cfg.messages[key]:lower(); -- set the message to lower case before return utilities.substitute ( msg, str ); -- including template text else return utilities.substitute ( cfg.messages[key], str ); end end --[[----------------< W I K I S O U R C E _ U R L _ M A K E >------------------- Makes a Wikisource URL from Wikisource interwiki-link. Returns the URL and appropriate label; nil else. str is the value assigned to |chapter= (or aliases) or |title= or |title-link= ]] local function wikisource_url_make (str) local wl_type, D, L; local ws_url, ws_label; local wikisource_prefix = table.concat ({'https://', cfg.this_wiki_code, '.wikisource.org/wiki/'}); wl_type, D, L = utilities.is_wikilink (str); -- wl_type is 0 (not a wikilink), 1 (simple wikilink), 2 (complex wikilink) if 0 == wl_type then -- not a wikilink; might be from |title-link= str = D:match ('^[Ww]ikisource:(.+)') or D:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace if utilities.is_set (str) then ws_url = table.concat ({ -- build a Wikisource URL wikisource_prefix, -- prefix str, -- article title }); ws_label = str; -- label for the URL end elseif 1 == wl_type then -- simple wikilink: [[Wikisource:ws article]] str = D:match ('^[Ww]ikisource:(.+)') or D:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace if utilities.is_set (str) then ws_url = table.concat ({ -- build a Wikisource URL wikisource_prefix, -- prefix str, -- article title }); ws_label = str; -- label for the URL end elseif 2 == wl_type then -- non-so-simple wikilink: [[Wikisource:ws article|displayed text]] ([[L|D]]) str = L:match ('^[Ww]ikisource:(.+)') or L:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace if utilities.is_set (str) then ws_label = D; -- get ws article name from display portion of interwiki link ws_url = table.concat ({ -- build a Wikisource URL wikisource_prefix, -- prefix str, -- article title without namespace from link portion of wikilink }); end end if ws_url then ws_url = mw.uri.encode (ws_url, 'WIKI'); -- make a usable URL ws_url = ws_url:gsub ('%%23', '#'); -- undo percent-encoding of fragment marker end return ws_url, ws_label, L or D; -- return proper URL or nil and a label or nil end --[[----------------< F O R M A T _ P E R I O D I C A L >----------------------- Format the three periodical parameters: |script-<periodical>=, |<periodical>=, and |trans-<periodical>= into a single Periodical meta-parameter. ]] local function format_periodical (script_periodical, script_periodical_source, periodical, trans_periodical) if not utilities.is_set (periodical) then periodical = ''; -- to be safe for concatenation else periodical = utilities.wrap_style ('italic-title', periodical); -- style end periodical = script_concatenate (periodical, script_periodical, script_periodical_source); -- <bdi> tags, lang attribute, categorization, etc.; must be done after title is wrapped if utilities.is_set (trans_periodical) then trans_periodical = utilities.wrap_style ('trans-italic-title', trans_periodical); if utilities.is_set (periodical) then periodical = periodical .. ' ' .. trans_periodical; else -- here when trans-periodical without periodical or script-periodical periodical = trans_periodical; utilities.set_message ('err_trans_missing_title', {'periodical'}); end end return periodical; end --[[------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------- Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single chapter meta- parameter (chapter_url_source used for error messages). ]] local function format_chapter_title (script_chapter, script_chapter_source, chapter, chapter_source, trans_chapter, trans_chapter_source, chapter_url, chapter_url_source, no_quotes, access) local ws_url, ws_label, L = wikisource_url_make (chapter); -- make a wikisource URL and label from a wikisource interwiki link if ws_url then ws_label = ws_label:gsub ('_', ' '); -- replace underscore separators with space characters chapter = ws_label; end if not utilities.is_set (chapter) then chapter = ''; -- to be safe for concatenation else if false == no_quotes then chapter = kern_quotes (chapter); -- if necessary, separate chapter title's leading and trailing quote marks from module provided quote marks chapter = utilities.wrap_style ('quoted-title', chapter); end end chapter = script_concatenate (chapter, script_chapter, script_chapter_source); -- <bdi> tags, lang attribute, categorization, etc.; must be done after title is wrapped if utilities.is_set (chapter_url) then chapter = external_link (chapter_url, chapter, chapter_url_source, access); -- adds bare_url_missing_title error if appropriate elseif ws_url then chapter = external_link (ws_url, chapter .. '&nbsp;', 'ws link in chapter'); -- adds bare_url_missing_title error if appropriate; space char to move icon away from chap text; TODO: better way to do this? chapter = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, chapter}); end if utilities.is_set (trans_chapter) then trans_chapter = utilities.wrap_style ('trans-quoted-title', trans_chapter); if utilities.is_set (chapter) then chapter = chapter .. ' ' .. trans_chapter; else -- here when trans_chapter without chapter or script-chapter chapter = trans_chapter; chapter_source = trans_chapter_source:match ('trans%-?(.+)'); -- when no chapter, get matching name from trans-<param> utilities.set_message ('err_trans_missing_title', {chapter_source}); end end return chapter; end --[[----------------< H A S _ I N V I S I B L E _ C H A R S >------------------- This function searches a parameter's value for non-printable or invisible characters. The search stops at the first match. This function will detect the visible replacement character when it is part of the Wikisource. Detects but ignores nowiki and math stripmarkers. Also detects other named stripmarkers (gallery, math, pre, ref) and identifies them with a slightly different error message. See also coins_cleanup(). Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the parameter value. ]] local function has_invisible_chars (param, v) local position = ''; -- position of invisible char or starting position of stripmarker local capture; -- used by stripmarker detection to hold name of the stripmarker local stripmarker; -- boolean set true when a stripmarker is found capture = string.match (v, '[%w%p ]*'); -- test for values that are simple ASCII text and bypass other tests if true if capture == v then -- if same there are no Unicode characters return; end for _, invisible_char in ipairs (cfg.invisible_chars) do local char_name = invisible_char[1]; -- the character or group name local pattern = invisible_char[2]; -- the pattern used to find it position, _, capture = mw.ustring.find (v, pattern); -- see if the parameter value contains characters that match the pattern if position and (cfg.invisible_defs.zwj == capture) then -- if we found a zero-width joiner character if mw.ustring.find (v, cfg.indic_script) then -- it's ok if one of the Indic scripts position = nil; -- unset position elseif cfg.emoji_t[mw.ustring.codepoint (v, position+1)] then -- is zwj followed by a character listed in emoji{}? position = nil; -- unset position end end if position then if 'nowiki' == capture or 'math' == capture or -- nowiki and math stripmarkers (not an error condition) ('templatestyles' == capture and utilities.in_array (param, {'id', 'quote','المعرف', 'اقتباس'})) then -- templatestyles stripmarker allowed in these parameters stripmarker = true; -- set a flag elseif true == stripmarker and cfg.invisible_defs.del == capture then -- because stripmakers begin and end with the delete char, assume that we've found one end of a stripmarker position = nil; -- unset else local err_msg; if capture and not (cfg.invisible_defs.del == capture or cfg.invisible_defs.zwj == capture) then err_msg = utilities.substitute (cfg.messages.invisible_1, {capture, char_name}); -- <capture> here is stripmarker name else err_msg = utilities.substitute (cfg.messages.invisible_2, char_name); end utilities.set_message ('err_invisible_char', {err_msg, utilities.wrap_style ('parameter', param), position}); -- add error message return; -- and done with this parameter end end end end --[[-------------------< A R G U M E N T _ W R A P P E R >---------------------- Argument wrapper. This function provides support for argument mapping defined in the configuration file so that multiple names can be transparently aliased to single internal variable. ]] local function argument_wrapper ( args ) local origin = {}; return setmetatable({ ORIGIN = function ( self, k ) local dummy = self[k]; -- force the variable to be loaded. return origin[k]; end }, { __index = function ( tbl, k ) if origin[k] ~= nil then return nil; end local args, list, v = args, cfg.aliases[k]; if type( list ) == 'table' then v, origin[k] = utilities.select_one ( args, list, 'err_redundant_parameters' ); if origin[k] == nil then origin[k] = ''; -- Empty string, not nil end elseif list ~= nil then v, origin[k] = args[list], list; else -- maybe let through instead of raising an error? -- v, origin[k] = args[k], k; error( cfg.messages['unknown_argument_map'] .. ': ' .. k); end -- Empty strings, not nil; if v == nil then v = ''; origin[k] = ''; end tbl = rawset( tbl, k, v ); return v; end, }); end --[[--------------------------< N O W R A P _ D A T E >------------------------- When date is YYYY-MM-DD format wrap in nowrap span: <span ...>YYYY-MM-DD</span>. When date is DD MMMM YYYY or is MMMM DD, YYYY then wrap in nowrap span: <span ...>DD MMMM</span> YYYY or <span ...>MMMM DD,</span> YYYY DOES NOT yet support MMMM YYYY or any of the date ranges. ]] local function nowrap_date (date) local cap = ''; local cap2 = ''; if date:match("^%d%d%d%d%-%d%d%-%d%d$") then date = utilities.substitute (cfg.presentation['nowrap1'], date); elseif date:match("^%a+%s*%d%d?,%s+%d%d%d%d$") or date:match ("^%d%d?%s*%a+%s+%d%d%d%d$") then cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$"); date = utilities.substitute (cfg.presentation['nowrap2'], {cap, cap2}); end return date; end --[[--------------------------< S E T _ T I T L E T Y P E >--------------------- This function sets default title types (equivalent to the citation including |type=<default value>) for those templates that have defaults. Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none). ]] local function set_titletype (cite_class, title_type) if utilities.is_set (title_type) then if 'none' == cfg.keywords_xlate[title_type] then title_type = ''; -- if |type=none then type parameter not displayed end return title_type; -- if |type= has been set to any other value use that value end return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation end --[[--------------------------< S A F E _ J O I N >----------------------------- Joins a sequence of strings together while checking for duplicate separation characters. ]] local function safe_join( tbl, duplicate_char ) local f = {}; -- create a function table appropriate to type of 'duplicate character' if 1 == #duplicate_char then -- for single byte ASCII characters use the string library functions f.gsub = string.gsub f.match = string.match f.sub = string.sub else -- for multi-byte characters use the ustring library functions f.gsub = mw.ustring.gsub f.match = mw.ustring.match f.sub = mw.ustring.sub end local str = ''; -- the output string local comp = ''; -- what does 'comp' mean? local end_chr = ''; local trim; for _, value in ipairs( tbl ) do if value == nil then value = ''; end if str == '' then -- if output string is empty str = value; -- assign value to it (first time through the loop) elseif value ~= '' then if value:sub(1, 1) == '<' then -- special case of values enclosed in spans and other markup. comp = value:gsub( "%b<>", "" ); -- remove HTML markup (<span>string</span> -> string) else comp = value; end -- typically duplicate_char is sepc if f.sub(comp, 1, 1) == duplicate_char then -- is first character same as duplicate_char? why test first character? -- Because individual string segments often (always?) begin with terminal punct for the -- preceding segment: 'First element' .. 'sepc next element' .. etc.? trim = false; end_chr = f.sub(str, -1, -1); -- get the last character of the output string -- str = str .. "<HERE(enchr=" .. end_chr .. ")" -- debug stuff? if end_chr == duplicate_char then -- if same as separator str = f.sub(str, 1, -2); -- remove it elseif end_chr == "'" then -- if it might be wiki-markup if f.sub(str, -3, -1) == duplicate_char .. "''" then -- if last three chars of str are sepc'' str = f.sub(str, 1, -4) .. "''"; -- remove them and add back '' elseif f.sub(str, -5, -1) == duplicate_char .. "]]''" then -- if last five chars of str are sepc]]'' trim = true; -- why? why do this and next differently from previous? elseif f.sub(str, -4, -1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]'' trim = true; -- same question end elseif end_chr == "]" then -- if it might be wiki-markup if f.sub(str, -3, -1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink trim = true; elseif f.sub(str, -3, -1) == duplicate_char .. '"]' then -- if last three chars of str are sepc"] quoted external link trim = true; elseif f.sub(str, -2, -1) == duplicate_char .. "]" then -- if last two chars of str are sepc] external link trim = true; elseif f.sub(str, -4, -1) == duplicate_char .. "'']" then -- normal case when |url=something & |title=Title. trim = true; end elseif end_chr == " " then -- if last char of output string is a space if f.sub(str, -2, -1) == duplicate_char .. " " then -- if last two chars of str are <sepc><space> str = f.sub(str, 1, -3); -- remove them both end end if trim then if value ~= comp then -- value does not equal comp when value contains HTML markup local dup2 = duplicate_char; if f.match(dup2, "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it value = f.gsub(value, "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows HTML markup else value = f.sub(value, 2, -1 ); -- remove duplicate_char when it is first character end end end str = str .. value; -- add it to the output string end end return str; end --[[--------------------------< I S _ S U F F I X >----------------------------- returns true if suffix is properly formed Jr, Sr, or ordinal in the range 1–9. Puncutation not allowed. ]] local function is_suffix (suffix) if utilities.in_array (suffix, {'Jr', 'Sr', 'Jnr', 'Snr', '1st', '2nd', '3rd'}) or suffix:match ('^%dth$') then return true; end return false; end --[[--------------------< I S _ G O O D _ V A N C _ N A M E >------------------- For Vancouver style, author/editor names are supposed to be rendered in Latin (read ASCII) characters. When a name uses characters that contain diacritical marks, those characters are to be converted to the corresponding Latin character. When a name is written using a non-Latin alphabet or logogram, that name is to be transliterated into Latin characters. The module doesn't do this so editors may/must. This test allows |first= and |last= names to contain any of the letters defined in the four Unicode Latin character sets [http://www.unicode.org/charts/PDF/U0000.pdf C0 Controls and Basic Latin] 0041–005A, 0061–007A [http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] 00C0–00D6, 00D8–00F6, 00F8–00FF [http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] 0100–017F [http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] 0180–01BF, 01C4–024F |lastn= also allowed to contain hyphens, spaces, and apostrophes. (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/) |firstn= also allowed to contain hyphens, spaces, apostrophes, and periods This original test: if nil == mw.ustring.find (last, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ-ɏ%-%s%']*$") or nil == mw.ustring.find (first, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ-ɏ%-%s%'%.]+[2-6%a]*$") then was written outside of the code editor and pasted here because the code editor gets confused between character insertion point and cursor position. The test has been rewritten to use decimal character escape sequence for the individual bytes of the Unicode characters so that it is not necessary to use an external editor to maintain this code. \195\128-\195\150 – À-Ö (U+00C0–U+00D6 – C0 controls) \195\152-\195\182 – Ø-ö (U+00D8-U+00F6 – C0 controls) \195\184-\198\191 – ø-ƿ (U+00F8-U+01BF – C0 controls, Latin extended A & B) \199\132-\201\143 – DŽ-ɏ (U+01C4-U+024F – Latin extended B) ]] local function is_good_vanc_name (last, first, suffix, position) if not suffix then if first:find ('[,%s]') then -- when there is a space or comma, might be first name/initials + generational suffix first = first:match ('(.-)[,%s]+'); -- get name/initials suffix = first:match ('[,%s]+(.+)$'); -- get generational suffix end end if utilities.is_set (suffix) then if not is_suffix (suffix) then add_vanc_error (cfg.err_msg_supl.suffix, position); return false; -- not a name with an appropriate suffix end end if nil == mw.ustring.find (last, "^[A-Za-z\195\128-\195\150\195\152-\195\182\195\184-\198\191\199\132-\201\143\225\184\128-\225\187\191%-%s%']*$") or nil == mw.ustring.find (first, "^[A-Za-z\195\128-\195\150\195\152-\195\182\195\184-\198\191\199\132-\201\143\225\184\128-\225\187\191%-%s%'%.]*$") then add_vanc_error (cfg.err_msg_supl['non-Latin char'], position); return false; -- not a string of Latin characters; Vancouver requires Romanization end; return true; end --[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------ Attempts to convert names to initials in support of |name-list-style=vanc. Names in |firstn= may be separated by spaces or hyphens, or for initials, a period. See http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35062/. Vancouver style requires family rank designations (Jr, II, III, etc.) to be rendered as Jr, 2nd, 3rd, etc. See http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35085/. This code only accepts and understands generational suffix in the Vancouver format because Roman numerals look like, and can be mistaken for, initials. This function uses ustring functions because firstname initials may be any of the Unicode Latin characters accepted by is_good_vanc_name (). ]] local function reduce_to_initials (first, position) if first:find (',', 1, true) then return first; -- commas not allowed; abandon end local name, suffix = mw.ustring.match (first, "^(%u+) ([%dJS][%drndth]+)$"); if not name then -- if not initials and a suffix name = mw.ustring.match (first, "^(%u+)$"); -- is it just initials? end if name then -- if first is initials with or without suffix if 3 > mw.ustring.len (name) then -- if one or two initials if suffix then -- if there is a suffix if is_suffix (suffix) then -- is it legitimate? return first; -- one or two initials and a valid suffix so nothing to do else add_vanc_error (cfg.err_msg_supl.suffix, position); -- one or two initials with invalid suffix so error message return first; -- and return first unmolested end else return first; -- one or two initials without suffix; nothing to do end end end -- if here then name has 3 or more uppercase letters so treat them as a word local initials_t, names_t = {}, {}; -- tables to hold name parts and initials local i = 1; -- counter for number of initials names_t = mw.text.split (first, '[%s%-]+'); -- split into a sequence of names and possible suffix while names_t[i] do -- loop through the sequence if 1 < i and names_t[i]:match ('[%dJS][%drndth]+%.?$') then -- if not the first name, and looks like a suffix (may have trailing dot) names_t[i] = names_t[i]:gsub ('%.', ''); -- remove terminal dot if present if is_suffix (names_t[i]) then -- if a legitimate suffix table.insert (initials_t, ' ' .. names_t[i]); -- add a separator space, insert at end of initials sequence break; -- and done because suffix must fall at the end of a name end -- no error message if not a suffix; possibly because of Romanization end if 3 > i then table.insert (initials_t, mw.ustring.sub (names_t[i], 1, 1)); -- insert the initial at end of initials sequence end i = i + 1; -- bump the counter end return table.concat (initials_t); -- Vancouver format does not include spaces. end --[[--------------------------< I N T E R W I K I _ P R E F I X E N _ G E T >---------------------------------- extract interwiki prefixen from <value>. Returns two one or two values: false – no prefixen nil – prefix exists but not recognized project prefix, language prefix – when value has either of: :<project>:<language>:<article> :<language>:<project>:<article> project prefix, nil – when <value> has only a known single-letter prefix nil, language prefix – when <value> has only a known language prefix accepts single-letter project prefixen: 'd' (wikidata), 's' (wikisource), and 'w' (wikipedia) prefixes; at this writing, the other single-letter prefixen (b (wikibook), c (commons), m (meta), n (wikinews), q (wikiquote), and v (wikiversity)) are not supported. ]] local function interwiki_prefixen_get (value, is_link) if not value:find (':%l+:') then -- if no prefix return false; -- abandon; boolean here to distinguish from nil fail returns later end local prefix_patterns_linked_t = { -- sequence of valid interwiki and inter project prefixen '^%[%[:([dsw]):(%l%l+):', -- wikilinked; project and language prefixes '^%[%[:(%l%l+):([dsw]):', -- wikilinked; language and project prefixes '^%[%[:([dsw]):', -- wikilinked; project prefix '^%[%[:(%l%l+):', -- wikilinked; language prefix } local prefix_patterns_unlinked_t = { -- sequence of valid interwiki and inter project prefixen '^:([dsw]):(%l%l+):', -- project and language prefixes '^:(%l%l+):([dsw]):', -- language and project prefixes '^:([dsw]):', -- project prefix '^:(%l%l+):', -- language prefix } local cap1, cap2; for _, pattern in ipairs ((is_link and prefix_patterns_linked_t) or prefix_patterns_unlinked_t) do cap1, cap2 = value:match (pattern); if cap1 then break; -- found a match so stop looking end end if cap1 and cap2 then -- when both then :project:language: or :language:project: (both forms allowed) if 1 == #cap1 then -- length == 1 then :project:language: if cfg.inter_wiki_map[cap2] then -- is language prefix in the interwiki map? return cap1, cap2; -- return interwiki project and interwiki language end else -- here when :language:project: if cfg.inter_wiki_map[cap1] then -- is language prefix in the interwiki map? return cap2, cap1; -- return interwiki project and interwiki language end end return nil; -- unknown interwiki language elseif not (cap1 or cap2) then -- both are nil? return nil; -- we got something that looks like a project prefix but isn't; return fail elseif 1 == #cap1 then -- here when one capture return cap1, nil; -- length is 1 so return project, nil language else -- here when one capture and its length it more than 1 if cfg.inter_wiki_map[cap1] then -- is language prefix in the interwiki map? return nil, cap1; -- return nil project, language end end end --[[--------------------------< L I S T _ P E O P L E >-------------------------- Formats a list of people (authors, contributors, editors, interviewers, translators) names in the list will be linked when |<name>-link= has a value |<name>-mask- does NOT have a value; masked names are presumed to have been rendered previously so should have been linked there when |<name>-mask=0, the associated name is not rendered ]] local function list_people (control, people, etal) local sep; local namesep; local format = control.format; local maximum = control.maximum; local name_list = {}; if 'vanc' == format then -- Vancouver-like name styling? sep = cfg.presentation['sep_nl_vanc']; -- name-list separator between names is a comma namesep = cfg.presentation['sep_name_vanc']; -- last/first separator is a space else sep = cfg.presentation['sep_nl']; -- name-list separator between names is a semicolon namesep = cfg.presentation['sep_name']; -- last/first separator is <comma><space> end if sep:sub (-1, -1) ~= " " then sep = sep .. " " end if utilities.is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for other names for i, person in ipairs (people) do if utilities.is_set (person.last) then local mask = person.mask; local one; local sep_one = sep; if utilities.is_set (maximum) and i > maximum then etal = true; break; end if mask then local n = tonumber (mask); -- convert to a number if it can be converted; nil else if n then one = 0 ~= n and string.rep("&mdash;", n) or nil; -- make a string of (n > 0) mdashes, nil else, to replace name person.link = nil; -- don't create link to name if name is replaces with mdash string or has been set nil else one = mask; -- replace name with mask text (must include name-list separator) sep_one = " "; -- modify name-list separator end else one = person.last; -- get surname local first = person.first -- get given name if utilities.is_set (first) then if ("vanc" == format) then -- if Vancouver format one = one:gsub ('%.', ''); -- remove periods from surnames (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/) if not person.corporate and is_good_vanc_name (one, first, nil, i) then -- and name is all Latin characters; corporate authors not tested first = reduce_to_initials (first, i); -- attempt to convert first name(s) to initials end end one = one .. namesep .. first; end end if utilities.is_set (person.link) then one = utilities.make_wikilink (person.link, one); -- link author/editor end if one then -- if <one> has a value (name, mdash replacement, or mask text replacement) local proj, tag = interwiki_prefixen_get (one, true); -- get the interwiki prefixen if present if 'w' == proj and ('Wikipedia' == mw.site.namespaces.Project['name']) then proj = nil; -- for stuff like :w:de:<article>, :w is unnecessary TODO: maint cat? end if proj then local proj_name = ({['d'] = cfg.messages.wikidata, ['s'] = cfg.messages.wikisource, ['w'] = cfg.messages.wikipedia})[proj]; -- :w (wikipedia) for linking from a non-wikipedia project if proj_name then one = one .. utilities.wrap_style ('interproj', proj_name); -- add resized leading space, brackets, static text, language name utilities.add_prop_cat ('interproj-linked-name', proj); -- categorize it; <proj> is sort key tag = nil; -- unset; don't do both project and language end end if tag == cfg.this_wiki_code then tag = nil; -- stuff like :en:<article> at en.wiki is pointless TODO: maint cat? end if tag then local lang = cfg.lang_tag_remap[tag] or cfg.mw_languages_by_tag_t[tag]; if lang then -- error messaging done in extract_names() where we know parameter names one = one .. utilities.wrap_style ('interwiki', lang); -- add resized leading space, brackets, static text, language name utilities.add_prop_cat ('interwiki-linked-name', tag); -- categorize it; <tag> is sort key end end table.insert (name_list, one); -- add it to the list of names table.insert (name_list, sep_one); -- add the proper name-list separator end end end local count = #name_list / 2; -- (number of names + number of separators) divided by 2 if 0 < count then if 1 < count and not etal then if 'amp' == format then name_list[#name_list-2] = " & "; -- replace last separator with ampersand text elseif 'and' == format then if 2 == count then name_list[#name_list-2] = cfg.presentation.sep_nl_and; -- replace last separator with 'and' text else name_list[#name_list-2] = cfg.presentation.sep_nl_end; -- replace last separator with '(sep) and' text end end end name_list[#name_list] = nil; -- erase the last separator end local result = table.concat (name_list); -- construct list if etal and utilities.is_set (result) then -- etal may be set by |display-authors=etal but we might not have a last-first list result = result .. sep .. cfg.messages['et al']; -- we've got a last-first list and etal so add et al. end return result, count; -- return name-list string and count of number of names (count used for editor names only) end --[[--------------------< M A K E _ C I T E R E F _ I D >----------------------- Generates a CITEREF anchor ID if we have at least one name or a date. Otherwise returns an empty string. namelist is one of the contributor-, author-, or editor-name lists chosen in that order. year is Year or anchor_year. ]] local function make_citeref_id (namelist, year) local names={}; -- a table for the one to four names and year for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names names[i] = v.last if i == 4 then break end -- if four then done end table.insert (names, year); -- add the year at the end local id = table.concat(names); -- concatenate names and year for CITEREF id if utilities.is_set (id) then -- if concatenation is not an empty string return "CITEREF" .. id; -- add the CITEREF portion else return ''; -- return an empty string; no reason to include CITEREF id in this citation end end --[[--------------------------< C I T E _ C L A S S _A T T R I B U T E _M A K E >------------------------------ construct <cite> tag class attribute for this citation. <cite_class> – config.CitationClass from calling template <mode> – value from |mode= parameter ]] local function cite_class_attribute_make (cite_class, mode) local class_t = {}; table.insert (class_t, 'citation'); -- required for blue highlight if 'citation' ~= cite_class then table.insert (class_t, cite_class); -- identify this template for user css table.insert (class_t, utilities.is_set (mode) and mode or 'cs1'); -- identify the citation style for user css or javascript else table.insert (class_t, utilities.is_set (mode) and mode or 'cs2'); -- identify the citation style for user css or javascript end for _, prop_key in ipairs (z.prop_keys_t) do table.insert (class_t, prop_key); -- identify various properties for user css or javascript end return table.concat (class_t, ' '); -- make a big string and done end --[[---------------------< N A M E _ H A S _ E T A L >-------------------------- Evaluates the content of name parameters (author, editor, etc.) for variations on the theme of et al. If found, the et al. is removed, a flag is set to true and the function returns the modified name and the flag. This function never sets the flag to false but returns its previous state because it may have been set by previous passes through this function or by the associated |display-<names>=etal parameter ]] local function name_has_etal (name, etal, nocat, param) if utilities.is_set (name) then -- name can be nil in which case just return local patterns = cfg.et_al_patterns; -- get patterns from configuration for _, pattern in ipairs (patterns) do -- loop through all of the patterns if name:match (pattern) then -- if this 'et al' pattern is found in name name = name:gsub (pattern, ''); -- remove the offending text etal = true; -- set flag (may have been set previously here or by |display-<names>=etal) if not nocat then -- no categorization for |vauthors= utilities.set_message ('err_etal', {param}); -- and set an error if not added end end end end return name, etal; end --[[---------------------< N A M E _ I S _ N U M E R I C >---------------------- Add an error message and category when <name> parameter value does not contain letters. Add a maintenance category when <name> parameter value has numeric characters mixed with characters that are not numeric characters; could be letters and/or punctuation characters. This function will only emit one error and one maint message for the current template. Does not emit both error and maint messages/categories for the same parameter value. returns nothing ]] local function name_is_numeric (name, name_alias, list_name) local patterns = { '^%D+%d', -- <name> must have digits preceded by other characters '^%D*%d+%D+', -- <name> must have digits followed by other characters } if not added_numeric_name_errs and mw.ustring.match (name, '^[%A]+$') then -- if we have not already set an error message and <name> does not have any alpha characters utilities.set_message ('err_numeric_names', name_alias); -- add an error message added_numeric_name_errs = true; -- set the flag so we emit only one error message return; -- when here no point in further testing; abandon end if not added_numeric_name_maint then -- if we have already set a maint message for _, pattern in ipairs (patterns) do -- spin through list of patterns if mw.ustring.match (name, pattern) then -- digits preceded or followed by anything but digits; %D+ includes punctuation utilities.set_message ('maint_numeric_names', cfg.special_case_translation [list_name]); -- add a maint cat for this template added_numeric_name_maint = true; -- set the flag so we emit only one maint message return; -- when here no point in further testing; abandon end end end end --[[-----------------< N A M E _ H A S _ M U L T _ N A M E S >------------------ Evaluates the content of last/surname (authors etc.) parameters for multiple names. Multiple names are indicated if there is more than one comma or any "unescaped" semicolons. Escaped semicolons are ones used as part of selected HTML entities. If the condition is met, the function adds the multiple name maintenance category. Same test for first except that commas should not appear in given names (MOS:JR says that the generational suffix does not take a separator character). Titles, degrees, postnominals, affiliations, all normally comma separated don't belong in a citation. <name> – name parameter value <list_name> – AuthorList, EditorList, etc <limit> – number of allowed commas; 1 (default) for surnames; 0 for given names returns nothing ]] local function name_has_mult_names (name, list_name, limit) local _, commas, semicolons, nbsps; limit = limit and limit or 1; if utilities.is_set (name) then _, commas = name:gsub (',', ''); -- count the number of commas _, semicolons = name:gsub (';', ''); -- count the number of semicolons -- nbsps probably should be its own separate count rather than merged in -- some way with semicolons because Lua patterns do not support the -- grouping operator that regex does, which means there is no way to add -- more entities to escape except by adding more counts with the new -- entities _, nbsps = name:gsub ('&nbsp;',''); -- count nbsps -- There is exactly 1 semicolon per &nbsp; entity, so subtract nbsps -- from semicolons to 'escape' them. If additional entities are added, -- they also can be subtracted. if limit < commas or 0 < (semicolons - nbsps) then utilities.set_message ('maint_mult_names', cfg.special_case_translation [list_name]); -- add a maint message end end end --[=[-------------------------< I S _ G E N E R I C >---------------------------------------------------------- Compares values assigned to various parameters according to the string provided as <item> in the function call. <item> can have on of two values: 'generic_names' – for name-holding parameters: |last=, |first=, |editor-last=, etc 'generic_titles' – for |title= There are two types of generic tests. The 'accept' tests look for a pattern that should not be rejected by the 'reject' test. For example, |author=[[John Smith (author)|Smith, John]] would be rejected by the 'author' reject test. But piped wikilinks with 'author' disambiguation should not be rejected so the 'accept' test prevents that from happening. Accept tests are always performed before reject tests. Each of the 'accept' and 'reject' sequence tables hold tables for en.wiki (['en']) and local.wiki (['local']) that each can hold a test sequence table The sequence table holds, at index [1], a test pattern, and, at index [2], a boolean control value. The control value tells string.find() or mw.ustring.find() to do plain-text search (true) or a pattern search (false). The intent of all this complexity is to make these searches as fast as possible so that we don't run out of processing time on very large articles. Returns true when a reject test finds the pattern or string false when an accept test finds the pattern or string nil else ]=] local function is_generic (item, value, wiki) local test_val; local str_lower = { -- use string.lower() for en.wiki (['en']) and use mw.ustring.lower() or local.wiki (['local']) ['en'] = string.lower, ['local'] = mw.ustring.lower, } local str_find = { -- use string.find() for en.wiki (['en']) and use mw.ustring.find() or local.wiki (['local']) ['en'] = string.find, ['local'] = mw.ustring.find, } local function test (val, test_t, wiki) -- local function to do the testing; <wiki> selects lower() and find() functions val = test_t[2] and str_lower[wiki](value) or val; -- when <test_t[2]> set to 'true', plaintext search using lowercase value return str_find[wiki] (val, test_t[1], 1, test_t[2]); -- return nil when not found or matched end local test_types_t = {'accept', 'reject'}; -- test accept patterns first, then reject patterns local wikis_t = {'en', 'local'}; -- do tests for each of these keys; en.wiki first, local.wiki second for _, test_type in ipairs (test_types_t) do -- for each test type for _, generic_value in pairs (cfg.special_case_translation[item][test_type]) do -- spin through the list of generic value fragments to accept or reject for _, wiki in ipairs (wikis_t) do if generic_value[wiki] then if test (value, generic_value[wiki], wiki) then -- go do the test return ('reject' == test_type); -- param value rejected, return true; false else end end end end end end --[[--------------------------< N A M E _ I S _ G E N E R I C >------------------------------------------------ calls is_generic() to determine if <name> is a 'generic name' listed in cfg.generic_names; <name_alias> is the parameter name used in error messaging ]] local function name_is_generic (name, name_alias) if not added_generic_name_errs and is_generic ('generic_names', name) then utilities.set_message ('err_generic_name', name_alias); -- set an error message added_generic_name_errs = true; end end --[[--------------------------< N A M E _ C H E C K S >-------------------------------------------------------- This function calls various name checking functions used to validate the content of the various name-holding parameters. ]] local function name_checks (last, first, list_name, last_alias, first_alias) local accept_name; if utilities.is_set (last) then last, accept_name = utilities.has_accept_as_written (last); -- remove accept-this-as-written markup when it wraps all of <last> if not accept_name then -- <last> not wrapped in accept-as-written markup name_has_mult_names (last, list_name); -- check for multiple names in the parameter name_is_numeric (last, last_alias, list_name); -- check for names that have no letters or are a mix of digits and other characters name_is_generic (last, last_alias); -- check for names found in the generic names list end end if utilities.is_set (first) then first, accept_name = utilities.has_accept_as_written (first); -- remove accept-this-as-written markup when it wraps all of <first> if not accept_name then -- <first> not wrapped in accept-as-written markup name_has_mult_names (first, list_name, 0); -- check for multiple names in the parameter; 0 is number of allowed commas in a given name name_is_numeric (first, first_alias, list_name); -- check for names that have no letters or are a mix of digits and other characters name_is_generic (first, first_alias); -- check for names found in the generic names list end local wl_type, D = utilities.is_wikilink (first); if 0 ~= wl_type then first = D; utilities.set_message ('err_bad_paramlink', first_alias); end end return last, first; -- done end --[[----------------------< E X T R A C T _ N A M E S >------------------------- Gets name list from the input arguments Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters. Stops searching when both |lastn= and |firstn= are not found in args after two sequential attempts: found |last1=, |last2=, and |last3= but doesn't find |last4= and |last5= then the search is done. This function emits an error message when there is a |firstn= without a matching |lastn=. When there are 'holes' in the list of last names, |last1= and |last3= are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=. When an author or editor parameter contains some form of 'et al.', the 'et al.' is stripped from the parameter and a flag (etal) returned that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration. This keeps 'et al.' out of the template's metadata. When this occurs, an error is emitted. ]] local function extract_names(args, list_name) local names = {}; -- table of names local last; -- individual name components local first; local link; local mask; local i = 1; -- loop counter/indexer local n = 1; -- output table indexer local count = 0; -- used to count the number of times we haven't found a |last= (or alias for authors, |editor-last or alias for editors) local etal = false; -- return value set to true when we find some form of et al. in an author parameter local last_alias, first_alias, link_alias; -- selected parameter aliases used in error messaging while true do last, last_alias = utilities.select_one ( args, cfg.aliases[list_name .. '-Last'], 'err_redundant_parameters', i ); -- search through args for name components beginning at 1 first, first_alias = utilities.select_one ( args, cfg.aliases[list_name .. '-First'], 'err_redundant_parameters', i ); link, link_alias = utilities.select_one ( args, cfg.aliases[list_name .. '-Link'], 'err_redundant_parameters', i ); mask = utilities.select_one ( args, cfg.aliases[list_name .. '-Mask'], 'err_redundant_parameters', i ); if last then -- error check |lastn= alias for unknown interwiki link prefix; done here because this is where we have the parameter name local project, language = interwiki_prefixen_get (last, true); -- true because we expect interwiki links in |lastn= to be wikilinked if nil == project and nil == language then -- when both are nil utilities.set_message ('err_bad_paramlink', last_alias); -- not known, emit an error message -- TODO: err_bad_interwiki? last = utilities.remove_wiki_link (last); -- remove wikilink markup; show display value only end end if link then -- error check |linkn= alias for unknown interwiki link prefix local project, language = interwiki_prefixen_get (link, false); -- false because wiki links in |author-linkn= is an error if nil == project and nil == language then -- when both are nil utilities.set_message ('err_bad_paramlink', link_alias); -- not known, emit an error message -- TODO: err_bad_interwiki? link = nil; -- unset so we don't link link_alias = nil; end end last, etal = name_has_etal (last, etal, false, last_alias); -- find and remove variations on et al. first, etal = name_has_etal (first, etal, false, first_alias); -- find and remove variations on et al. last, first = name_checks (last, first, list_name, last_alias, first_alias); -- multiple names, extraneous annotation, etc. checks --#Modified Localized if first and not last then -- if there is a firstn without a matching lastn local alias = first_alias:find ('اول', 1, true) and 'اول' or 'أول'; -- get first or given form of the alias utilities.set_message ('err_first_missing_last', { first_alias, -- param name of alias missing its mate first_alias:gsub (alias, {['أول'] ='أخير', ['اول'] = 'اخير'}), -- make param name appropriate to the alias form }); -- add this error message elseif not first and not last then -- if both firstn and lastn aren't found, are we done? count = count + 1; -- number of times we haven't found last and first if 2 <= count then -- two missing names and we give up break; -- normal exit or there is a two-name hole in the list; can't tell which end else -- we have last with or without a first local result; link = link_title_ok (link, link_alias, last, last_alias); -- check for improper wiki-markup if first then link = link_title_ok (link, link_alias, first, first_alias); -- check for improper wiki-markup end names[n] = {last = last, first = first, link = link, mask = mask, corporate = false}; -- add this name to our names list (corporate for |vauthors= only) n = n + 1; -- point to next location in the names table if 1 == count then -- if the previous name was missing utilities.set_message ('err_missing_name', {list_name:match ("(%w+)List"):lower(), i - 1}); -- add this error message end count = 0; -- reset the counter, we're looking for two consecutive missing names end i = i + 1; -- point to next args location end return names, etal; -- all done, return our list of names and the etal flag end --[[--------------------------< N A M E _ T A G _ G E T >------------------------------------------------------ attempt to decode |language=<lang_param> and return language name and matching tag; nil else. This function looks for: <lang_param> as a tag in cfg.lang_tag_remap{} <lang_param> as a name in cfg.lang_name_remap{} <lang_param> as a name in cfg.mw_languages_by_name_t <lang_param> as a tag in cfg.mw_languages_by_tag_t when those fail, presume that <lang_param> is an IETF-like tag that MediaWiki does not recognize. Strip all script, region, variant, whatever subtags from <lang_param> to leave just a two or three character language tag and look for the new <lang_param> in cfg.mw_languages_by_tag_t{} on success, returns name (in properly capitalized form) and matching tag (in lowercase); on failure returns nil ]] local function name_tag_get (lang_param) local lang_param_lc = mw.ustring.lower (lang_param); -- use lowercase as an index into the various tables local name; local tag; name = cfg.lang_tag_remap[lang_param_lc]; -- assume <lang_param_lc> is a tag; attempt to get remapped language name if name then -- when <name>, <lang_param> is a tag for a remapped language name if cfg.lang_name_remap[name:lower()][2] ~= lang_param_lc then utilities.set_message ('maint_unknown_lang'); -- add maint category if not already added return name, cfg.lang_name_remap[name:lower()][2]; -- so return name and tag from lang_name_remap[name]; special case to xlate sr-ec and sr-el to sr-cyrl and sr-latn end return name, lang_param_lc; -- so return <name> from remap and <lang_param_lc> end --#Modifed pattern tag = lang_param_lc:match ('^(%a%a%a?)[%-_].*'); -- still assuming that <lang_param_lc> is a tag; strip script, region, variant subtags name = cfg.lang_tag_remap[tag]; -- attempt to get remapped language name with language subtag only if name then -- when <name>, <tag> is a tag for a remapped language name return name, tag; -- so return <name> from remap and <tag> end if cfg.lang_name_remap[lang_param_lc] then -- not a remapped tag, assume <lang_param_lc> is a name; attempt to get remapped language tag return cfg.lang_name_remap[lang_param_lc][1], cfg.lang_name_remap[lang_param_lc][2]; -- for this <lang_param_lc>, return a (possibly) new name and appropriate tag end name = cfg.mw_languages_by_tag_t[lang_param_lc]; -- assume that <lang_param_lc> is a tag; attempt to get its matching language name if name then return name, lang_param_lc; -- <lang_param_lc> is a tag so return it and <name> end tag = cfg.mw_languages_by_name_t[lang_param_lc]; -- assume that <lang_param_lc> is a language name; attempt to get its matching tag if tag then return cfg.mw_languages_by_tag_t[tag], tag; -- <lang_param_lc> is a name so return the name from the table and <tag> end --#Modifed pattern tag = lang_param_lc:match ('^(%a%a%a?)[%-_].*'); -- is <lang_param_lc> an IETF-like tag that MediaWiki doesn't recognize? <tag> gets the language subtag; nil else if tag then name = cfg.mw_languages_by_tag_t[tag]; -- attempt to get a language name using the shortened <tag> if name then return name, tag; -- <lang_param_lc> is an unrecognized IETF-like tag so return <name> and language subtag end end end --[[-------------------< L A N G U A G E _ P A R A M E T E R >------------------ Gets language name from a provided two- or three-character ISO 639 code. If a code is recognized by MediaWiki, use the returned name; if not, then use the value that was provided with the language parameter. When |language= contains a recognized language (either code or name), the page is assigned to the category for that code: Category:Norwegian-language sources (no). For valid three-character code languages, the page is assigned to the single category for '639-2' codes: Category:CS1 ISO 639-2 language sources. Languages that are the same as the local wiki are not categorized. MediaWiki does not recognize three-character equivalents of two-character codes: code 'ar' is recognized but code 'ara' is not. This function supports multiple languages in the form |language=nb, French, th where the language names or codes are separated from each other by commas with optional space characters. ]] local function language_parameter (lang) local tag; -- some form of IETF-like language tag; language subtag with optional region, sript, vatiant, etc subtags local lang_subtag; -- ve populates |language= with mostly unecessary region subtags the MediaWiki does not recognize; this is the base language subtag local name; -- the language name local language_list = {}; -- table of language names to be rendered local names_t = {}; -- table made from the value assigned to |language= local this_wiki_name = mw.language.fetchLanguageName (cfg.this_wiki_code, cfg.this_wiki_code); -- get this wiki's language name names_t = mw.text.split (lang, '%s*,%s*'); -- names should be a comma separated list for _, lang in ipairs (names_t) do -- reuse lang here because we don't yet know if lang is a language name or a language tag name, tag = name_tag_get (lang); -- attempt to get name/tag pair for <lang>; <name> has proper capitalization; <tag> is lowercase if utilities.is_set (tag) then --#Modifed pattern lang_subtag = tag:gsub ('^(%a%a%a?)[%-_].*', '%1'); -- for categorization, strip any IETF-like tags from language tag if cfg.this_wiki_code ~= lang_subtag then -- when the language is not the same as this wiki's language if 2 == lang_subtag:len() then -- and is a two-character tag utilities.add_prop_cat ('foreign-lang-source', {name, tag}, lang_subtag); -- categorize it; tag appended to allow for multiple language categorization else -- or is a recognized language (but has a three-character tag) utilities.add_prop_cat ('foreign-lang-source-2', {lang_subtag}, lang_subtag); -- categorize it differently TODO: support multiple three-character tag categories per cs1|2 template? end elseif cfg.local_lang_cat_enable then -- when the language and this wiki's language are the same and categorization is enabled utilities.add_prop_cat ('local-lang-source', {name, lang_subtag}); -- categorize it end else name = lang; -- return whatever <lang> has so that we show something utilities.set_message ('maint_unknown_lang'); -- add maint category if not already added end table.insert (language_list, name); name = ''; -- so we can reuse it end --#Modified اللغات تظهر بالعربية دائمًا name = utilities.make_sep_list (#language_list, language_list, " و", " و", " و"); if (1 == #language_list) and (lang_subtag == cfg.this_wiki_code) then -- when only one language, find lang name in this wiki lang name; for |language=en-us, 'English' in 'American English' return ''; -- if one language and that language is this wiki's return an empty string (no annotation) end return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)' --[[ TODO: should only return blank or name rather than full list so we can clean up the bunched parenthetical elements Language, Type, Format ]] end --[[-----------------------< S E T _ C S _ S T Y L E >-------------------------- Gets the default CS style configuration for the given mode. Returns default separator and either postscript as passed in or the default. In CS1, the default postscript and separator are '.'. In CS2, the default postscript is the empty string and the default separator is ','. ]] local function set_cs_style (postscript, mode) if utilities.is_set(postscript) then -- emit a maintenance message if user postscript is the default cs1 postscript -- we catch the opposite case for cs2 in set_style if mode == 'cs1' and postscript == cfg.presentation['ps_' .. mode] then utilities.set_message ('maint_postscript'); end else postscript = cfg.presentation['ps_' .. mode]; end return cfg.presentation['sep_' .. mode], postscript; end --[[--------------------------< S E T _ S T Y L E >----------------------------- Sets the separator and postscript styles. Checks the |mode= first and the #invoke CitationClass second. Removes the postscript if postscript == none. ]] local function set_style (mode, postscript, cite_class) local sep; if 'cs2' == mode then sep, postscript = set_cs_style (postscript, 'cs2'); elseif 'cs1' == mode then sep, postscript = set_cs_style (postscript, 'cs1'); elseif 'citation' == cite_class then sep, postscript = set_cs_style (postscript, 'cs2'); else sep, postscript = set_cs_style (postscript, 'cs1'); end if cfg.keywords_xlate[postscript:lower()] == 'none' then -- emit a maintenance message if user postscript is the default cs2 postscript -- we catch the opposite case for cs1 in set_cs_style if 'cs2' == mode or ('cs1' ~= mode and 'citation' == cite_class) then -- {{citation |title=Title |mode=cs1 |postscript=none}} should not emit maint message utilities.set_message ('maint_postscript'); end postscript = ''; end return sep, postscript end --[=[-------------------------< I S _ P D F >----------------------------------- Determines if a URL has the file extension that is one of the PDF file extensions used by [[MediaWiki:Common.css]] when applying the PDF icon to external links. returns true if file extension is one of the recognized extensions, else false ]=] local function is_pdf (url) return url:match ('%.pdf$') or url:match ('%.PDF$') or url:match ('%.pdf[%?#]') or url:match ('%.PDF[%?#]') or url:match ('%.PDF&#035') or url:match ('%.pdf&#035'); end --[[--------------------------< S T Y L E _ F O R M A T >----------------------- Applies CSS style to |format=, |chapter-format=, etc. Also emits an error message if the format parameter does not have a matching URL parameter. If the format parameter is not set and the URL contains a file extension that is recognized as a PDF document by MediaWiki's commons.css, this code will set the format parameter to (PDF) with the appropriate styling. ]] local function style_format (format, url, fmt_param, url_param) if utilities.is_set (format) then format = utilities.wrap_style ('format', format); -- add leading space, parentheses, resize if not utilities.is_set (url) then utilities.set_message ('err_format_missing_url', {fmt_param, url_param}); -- add an error message end elseif is_pdf (url) then -- format is not set so if URL is a PDF file then format = utilities.wrap_style ('format', 'PDF'); -- set format to PDF else format = ''; -- empty string for concatenation end return format; end --[[---------------------< G E T _ D I S P L A Y _ N A M E S >------------------ Returns a number that defines the number of names displayed for author and editor name lists and a Boolean flag to indicate when et al. should be appended to the name list. When the value assigned to |display-xxxxors= is a number greater than or equal to zero, return the number and the previous state of the 'etal' flag (false by default but may have been set to true if the name list contains some variant of the text 'et al.'). When the value assigned to |display-xxxxors= is the keyword 'etal', return a number that is one greater than the number of authors in the list and set the 'etal' flag true. This will cause the list_people() to display all of the names in the name list followed by 'et al.' In all other cases, returns nil and the previous state of the 'etal' flag. inputs: max: A['DisplayAuthors'] or A['DisplayEditors'], etc; a number or some flavor of etal count: #a or #e list_name: 'authors' or 'editors' etal: author_etal or editor_etal This function sets an error message when |display-xxxxors= value greater than or equal to number of names but not when <max> comes from {{cs1 config}} global settings. When using global settings, <param> is set to the keyword 'cs1 config' which is used to supress the normal error. Error is suppressed because it is to be expected that some citations in an article will have the same or fewer names that the limit specified in {{cs1 config}}. ]] local function get_display_names (max, count, list_name, etal, param) if utilities.is_set (max) then if 'etal' == max:lower():gsub("[ '%.]", '') then -- the :gsub() portion makes 'etal' from a variety of 'et al.' spellings and stylings max = count + 1; -- number of authors + 1 so display all author name plus et al. etal = true; -- overrides value set by extract_names() elseif max:match ('^%d+$') then -- if is a string of numbers max = tonumber (max); -- make it a number if (max >= count) and ('cs1 config' ~= param) then -- error when local |display-xxxxors= value greater than or equal to number of names; not an error when using global setting utilities.set_message ('err_disp_name', {param, max}); -- add error message max = nil; end else -- not a valid keyword or number utilities.set_message ('err_disp_name', {param, max}); -- add error message max = nil; -- unset; as if |display-xxxxors= had not been set end end return max, etal; end --[[----------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >--------------- Adds error if |page=, |pages=, |quote-page=, |quote-pages= has what appears to be some form of p. or pp. abbreviation in the first characters of the parameter content. check page for extraneous p, p., pp, pp., pg, pg. at start of parameter value: good pattern: '^P[^%.P%l]' matches when page begins PX or P# but not Px where x and X are letters and # is a digit bad pattern: '^[Pp][PpGg]' matches when page begins pp, pP, Pp, PP, pg, pG, Pg, PG ]] local function extra_text_in_page_check (val, name) if not val:match (cfg.vol_iss_pg_patterns.good_ppattern) then for _, pattern in ipairs (cfg.vol_iss_pg_patterns.bad_ppatterns) do -- spin through the selected sequence table of patterns if val:match (pattern) then -- when a match, error so utilities.set_message ('err_extra_text_pages', name); -- add error message return; -- and done end end end end --[[--------------------------< E X T R A _ T E X T _ I N _ V O L _ I S S _ C H E C K >------------------------ Adds error if |volume= or |issue= has what appears to be some form of redundant 'type' indicator. Applies to both; this function looks for issue text in both |issue= and |volume= and looks for volume-like text in |voluem= and |issue=. For |volume=: 'V.', or 'Vol.' (with or without the dot) abbreviations or 'Volume' in the first characters of the parameter content (all case insensitive). 'V' and 'v' (without the dot) are presumed to be roman numerals so are allowed. For |issue=: 'No.', 'I.', 'Iss.' (with or without the dot) abbreviations, or 'Issue' in the first characters of the parameter content (all case insensitive); numero styling: 'n°' with degree sign U+00B0, and № precomposed numero sign U+2116. Single character values ('v', 'i', 'n') allowed when not followed by separator character ('.', ':', '=', or whitespace character) – param values are trimmed of whitespace by MediaWiki before delivered to the module. <val> is |volume= or |issue= parameter value <name> is |volume= or |issue= parameter name for error message <selector> is 'v' for |volume=, 'i' for |issue= sets error message on failure; returns nothing ]] local function extra_text_in_vol_iss_check (val, name, selector) if not utilities.is_set (val) then return; end local handler = 'v' == selector and 'err_extra_text_volume' or 'err_extra_text_issue'; local accept_val; val, accept_val = utilities.has_accept_as_written (val); --if accept_val then -- if uncomment this, |volume=((vol. 1)) suppresses error -- return; --end --val = mw.text.unstripNoWiki (val); -- if uncomment this, |volume=<nowiki>vol. 1</nowiki> shows error val = val:lower(); -- force parameter value to lower case for _, pattern in ipairs (cfg.vol_iss_pg_patterns.vi_patterns_t) do -- spin through the sequence table of patterns if val:match (pattern) then -- when a match, error so utilities.set_message (handler, name); -- add error message return; -- and done end end end --[=[-------------------------< G E T _ V _ N A M E _ T A B L E >---------------------------------------------- split apart a |vauthors= or |veditors= parameter. This function allows for corporate names, wrapped in doubled parentheses to also have commas; in the old version of the code, the doubled parentheses were included in the rendered citation and in the metadata. Individual author names may be wikilinked |vauthors=Jones AB, [[E. B. White|White EB]], ((Black, Brown, and Co.)) ]=] local function get_v_name_table (vparam, output_table, output_link_table) local _, accept = utilities.has_accept_as_written (vparam); if accept then utilities.add_prop_cat ('vanc-accept'); -- add properties category end --#Modified: next line local name_table = mw.text.split(vparam or "", "%s*,%s*") local wl_type, label, link; -- wl_type not used here; just a placeholder local i = 1; while name_table[i] do if name_table[i]:match ('^%(%(.*[^%)][^%)]$') then -- first segment of corporate with one or more commas; this segment has the opening doubled parentheses local name = name_table[i]; i = i + 1; -- bump indexer to next segment while name_table[i] do name = name .. ', ' .. name_table[i]; -- concatenate with previous segments if name_table[i]:match ('^.*%)%)$') then -- if this table member has the closing doubled parentheses break; -- and done reassembling so end i = i + 1; -- bump indexer end table.insert (output_table, name); -- and add corporate name to the output table table.insert (output_link_table, ''); -- no wikilink else wl_type, label, link = utilities.is_wikilink (name_table[i]); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]] table.insert (output_table, label); -- add this name if 1 == wl_type then table.insert (output_link_table, label); -- simple wikilink [[D]] else table.insert (output_link_table, link); -- no wikilink or [[L|D]]; add this link if there is one, else empty string end end i = i + 1; end return output_table; end --[[--------------------------< P A R S E _ V A U T H O R S _ V E D I T O R S >-------------------------------- This function extracts author / editor names from |vauthors= or |veditors= and finds matching |xxxxor-maskn= and |xxxxor-linkn= in args. It then returns a table of assembled names just as extract_names() does. Author / editor names in |vauthors= or |veditors= must be in Vancouver system style. Corporate or institutional names may sometimes be required and because such names will often fail the is_good_vanc_name() and other format compliance tests, are wrapped in doubled parentheses ((corporate name)) to suppress the format tests. Supports generational suffixes Jr, 2nd, 3rd, 4th–6th. This function sets the Vancouver error when a required comma is missing and when there is a space between an author's initials. ]] local function parse_vauthors_veditors (args, vparam, list_name) local names = {}; -- table of names assembled from |vauthors=, |author-maskn=, |author-linkn= local v_name_table = {}; local v_link_table = {}; -- when name is wikilinked, targets go in this table local etal = false; -- return value set to true when we find some form of et al. vauthors parameter local last, first, link, mask, suffix; local corporate = false; vparam, etal = name_has_etal (vparam, etal, true); -- find and remove variations on et al. do not categorize (do it here because et al. might have a period) v_name_table = get_v_name_table (vparam, v_name_table, v_link_table); -- names are separated by commas for i, v_name in ipairs(v_name_table) do first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor local accept_name; v_name, accept_name = utilities.has_accept_as_written (v_name); -- remove accept-this-as-written markup when it wraps all of <v_name> if accept_name then last = v_name; corporate = true; -- flag used in list_people() elseif string.find(v_name, "%s") then if v_name:find('[;%.]') then -- look for commonly occurring punctuation characters; add_vanc_error (cfg.err_msg_supl.punctuation, i); end local lastfirstTable = {} lastfirstTable = mw.text.split(v_name, "%s+") first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be initials or generational suffix if not mw.ustring.match (first, '^%u+$') then -- mw.ustring here so that later we will catch non-Latin characters suffix = first; -- not initials so assume that whatever we got is a generational suffix first = table.remove(lastfirstTable); -- get what should be the initials from the table end last = table.concat(lastfirstTable, ' ') -- returns a string that is the concatenation of all other names that are not initials and generational suffix if not utilities.is_set (last) then first = ''; -- unset last = v_name; -- last empty because something wrong with first add_vanc_error (cfg.err_msg_supl.name, i); end if mw.ustring.match (last, '%a+%s+%u+%s+%a+') then add_vanc_error (cfg.err_msg_supl['missing comma'], i); -- matches last II last; the case when a comma is missing end if mw.ustring.match (v_name, ' %u %u$') then -- this test is in the wrong place TODO: move or replace with a more appropriate test add_vanc_error (cfg.err_msg_supl.initials, i); -- matches a space between two initials end else last = v_name; -- last name or single corporate name? Doesn't support multiword corporate names? do we need this? end if utilities.is_set (first) then if not mw.ustring.match (first, "^%u?%u$") then -- first shall contain one or two upper-case letters, nothing else add_vanc_error (cfg.err_msg_supl.initials, i); -- too many initials; mixed case initials (which may be ok Romanization); hyphenated initials end is_good_vanc_name (last, first, suffix, i); -- check first and last before restoring the suffix which may have a non-Latin digit if utilities.is_set (suffix) then first = first .. ' ' .. suffix; -- if there was a suffix concatenate with the initials suffix = ''; -- unset so we don't add this suffix to all subsequent names end else if not corporate then is_good_vanc_name (last, '', nil, i); end end link = utilities.select_one ( args, cfg.aliases[list_name .. '-Link'], 'err_redundant_parameters', i ) or v_link_table[i]; mask = utilities.select_one ( args, cfg.aliases[list_name .. '-Mask'], 'err_redundant_parameters', i ); names[i] = {last = last, first = first, link = link, mask = mask, corporate = corporate}; -- add this assembled name to our names list end return names, etal; -- all done, return our list of names end --[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------ Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or select one of |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list. Only one of these appropriate three will be used. The hierarchy is: |authorn= (and aliases) highest and |authors= lowest; |editorn= (and aliases) highest and |veditors= lowest (support for |editors= withdrawn) When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second test which mimicks the test used in extract_names() when looking for a hole in the author name list. There may be a better way to do this, I just haven't discovered what that way is. Emits an error message when more than one xxxxor name source is provided. In this function, vxxxxors = vauthors or veditors; xxxxors = authors as appropriate. ]] local function select_author_editor_source (vxxxxors, xxxxors, args, list_name) local lastfirst = false; if utilities.select_one ( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice in case we have a |first1= without a |last1=; this ... utilities.select_one ( args, cfg.aliases[list_name .. '-First'], 'none', 1 ) or -- ... also catches the case where |first= is used with |vauthors= utilities.select_one ( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) or utilities.select_one ( args, cfg.aliases[list_name .. '-First'], 'none', 2 ) then lastfirst = true; end if (utilities.is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions (utilities.is_set (vxxxxors) and utilities.is_set (xxxxors)) or (true == lastfirst and utilities.is_set (xxxxors)) then utilities.set_message ('err_redundant_parameters', utilities.substitute (cfg.special_case_translation.name_list_params, ('AuthorList' == list_name) and cfg.special_case_translation.author or cfg.special_case_translation.editor)); -- add error message end if true == lastfirst then return 1 end; -- return a number indicating which author name source to use if utilities.is_set (vxxxxors) then return 2 end; if utilities.is_set (xxxxors) then return 3 end; return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last end --[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------ This function is used to validate a parameter's assigned value for those parameters that have only a limited number of allowable values (yes, y, true, live, dead, etc.). When the parameter value has not been assigned a value (missing or empty in the source template) the function returns the value specified by ret_val. If the parameter value is one of the list of allowed values returns the translated value; else, emits an error message and returns the value specified by ret_val. TODO: explain <invert> ]] local function is_valid_parameter_value (value, name, possible, ret_val, invert) if not utilities.is_set (value) then return ret_val; -- an empty parameter is ok end if (not invert and utilities.in_array (value, possible)) then -- normal; <value> is in <possible> table return cfg.keywords_xlate[value]; -- return translation of parameter keyword elseif invert and not utilities.in_array (value, possible) then -- invert; <value> is not in <possible> table return value; -- return <value> as it is else utilities.set_message ('err_invalid_param_val', {name, value}); -- not an allowed value so add error message return ret_val; end end --[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >---------------------------------------- This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space when the last character is not a sepc character or when the last three characters are not sepc followed by two closing square brackets (close of a wikilink). When either of these is true, the name_list is terminated with a single space character. ]] local function terminate_name_list (name_list, sepc) if (string.sub (name_list, -3, -1) == sepc .. '. ') then -- if already properly terminated return name_list; -- just return the name list elseif (string.sub (name_list, -1, -1) == sepc) or (string.sub (name_list, -3, -1) == sepc .. ']]') then -- if last name in list ends with sepc char return name_list .. " "; -- don't add another else return name_list .. sepc .. ' '; -- otherwise terminate the name list end end --[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------- returns the concatenation of the formatted volume and issue (or journal article number) parameters as a single string; or formatted volume or formatted issue, or an empty string if neither are set. ]] local function format_volume_issue (volume, issue, article, cite_class, origin, sepc, lower) if not utilities.is_set (volume) and not utilities.is_set (issue) and not utilities.is_set (article) then return ''; end -- same condition as in format_pages_sheets() local is_journal = 'journal' == cite_class or (utilities.in_array (cite_class, {'citation', 'map', 'interview'}) and 'journal' == origin); local is_numeric_vol = volume and (volume:match ('^[MDCLXVI]+$') or volume:match ('^%d+$')); -- is only uppercase roman numerals or only digits? local is_long_vol = volume and (4 < mw.ustring.len(volume)); -- is |volume= value longer than 4 characters? if volume and (not is_numeric_vol and is_long_vol) then -- when not all digits or Roman numerals, is |volume= longer than 4 characters? utilities.add_prop_cat ('long-vol'); -- yes, add properties cat end if is_journal then -- journal-style formatting local vol = ''; if utilities.is_set (volume) then if is_numeric_vol then -- |volume= value all digits or all uppercase Roman numerals? vol = utilities.substitute (cfg.presentation['vol-bold'], {sepc, volume}); -- render in bold face elseif is_long_vol then -- not all digits or Roman numerals; longer than 4 characters? vol = utilities.substitute (cfg.messages['j-vol'], {sepc, utilities.hyphen_to_dash (volume)}); -- not bold else -- four or fewer characters vol = utilities.substitute (cfg.presentation['vol-bold'], {sepc, utilities.hyphen_to_dash (volume)}); -- bold end end vol = vol .. (utilities.is_set (issue) and utilities.substitute (cfg.messages['j-issue'], issue) or '') vol = vol .. (utilities.is_set (article) and utilities.substitute (cfg.messages['j-article-num'], article) or '') return vol; end if 'podcast' == cite_class and utilities.is_set (issue) then return wrap_msg ('issue', {sepc, issue}, lower); end if 'conference' == cite_class and utilities.is_set (article) then -- |article-number= supported only in journal and conference cites if utilities.is_set (volume) and utilities.is_set (article) then -- both volume and article number return wrap_msg ('vol-art', {sepc, utilities.hyphen_to_dash (volume), article}, lower); elseif utilities.is_set (article) then -- article number alone; when volume alone, handled below return wrap_msg ('art', {sepc, article}, lower); end end -- all other types of citation if utilities.is_set (volume) and utilities.is_set (issue) then return wrap_msg ('vol-no', {sepc, utilities.hyphen_to_dash (volume), issue}, lower); elseif utilities.is_set (volume) then return wrap_msg ('vol', {sepc, utilities.hyphen_to_dash (volume)}, lower); else return wrap_msg ('issue', {sepc, issue}, lower); end end --[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >----------------------------------------- adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings. The return order is: page, pages, sheet, sheets Singular has priority over plural when both are provided. ]] local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower) if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators if utilities.is_set (sheet) then if 'journal' == origin then return '', '', wrap_msg ('j-sheet', sheet, lower), ''; else return '', '', wrap_msg ('sheet', {sepc, sheet}, lower), ''; end elseif utilities.is_set (sheets) then if 'journal' == origin then return '', '', '', wrap_msg ('j-sheets', sheets, lower); else return '', '', '', wrap_msg ('sheets', {sepc, sheets}, lower); end end end local is_journal = 'journal' == cite_class or (utilities.in_array (cite_class, {'citation', 'map', 'interview'}) and 'journal' == origin); if utilities.is_set (page) then if is_journal then return utilities.substitute (cfg.messages['j-page(s)'], page), '', '', ''; elseif not nopp then return utilities.substitute (cfg.messages['p-prefix'], {sepc, page}), '', '', ''; else return utilities.substitute (cfg.messages['nopp'], {sepc, page}), '', '', ''; end elseif utilities.is_set (pages) then if is_journal then return utilities.substitute (cfg.messages['j-page(s)'], pages), '', '', ''; elseif tonumber(pages) ~= nil and not nopp then -- if pages is only digits, assume a single page number return '', utilities.substitute (cfg.messages['p-prefix'], {sepc, pages}), '', ''; elseif not nopp then return '', utilities.substitute (cfg.messages['pp-prefix'], {sepc, pages}), '', ''; else return '', utilities.substitute (cfg.messages['nopp'], {sepc, pages}), '', ''; end end return '', '', '', ''; -- return empty strings end --[[--------------------------< I N S O U R C E _ L O C _ G E T >---------------------------------------------- returns one of the in-source locators: page, pages, or at. If any of these are interwiki links to Wikisource, returns the label portion of the interwiki-link as plain text for use in COinS. This COinS thing is done because here we convert an interwiki-link to an external link and add an icon span around that; get_coins_pages() doesn't know about the span. TODO: should it? TODO: add support for sheet and sheets?; streamline; TODO: make it so that this function returns only one of the three as the single in-source (the return value assigned to a new name)? ]] local function insource_loc_get (page, page_orig, pages, pages_orig, at) local ws_url, ws_label, coins_pages, L; -- for Wikisource interwiki-links; TODO: this corrupts page metadata (span remains in place after cleanup; fix there?) if utilities.is_set (page) then if utilities.is_set (pages) or utilities.is_set (at) then pages = ''; -- unset the others at = ''; end extra_text_in_page_check (page, page_orig); -- emit error message when |page= value begins with what looks like p., pp., etc. ws_url, ws_label, L = wikisource_url_make (page); -- make ws URL from |page= interwiki link; link portion L becomes tooltip label if ws_url then page = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in page'); -- space char after label to move icon away from in-source text; TODO: a better way to do this? page = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, page}); coins_pages = ws_label; end elseif utilities.is_set (pages) then if utilities.is_set (at) then at = ''; -- unset end extra_text_in_page_check (pages, pages_orig); -- emit error message when |page= value begins with what looks like p., pp., etc. ws_url, ws_label, L = wikisource_url_make (pages); -- make ws URL from |pages= interwiki link; link portion L becomes tooltip label if ws_url then pages = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in pages'); -- space char after label to move icon away from in-source text; TODO: a better way to do this? pages = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, pages}); coins_pages = ws_label; end elseif utilities.is_set (at) then ws_url, ws_label, L = wikisource_url_make (at); -- make ws URL from |at= interwiki link; link portion L becomes tooltip label if ws_url then at = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in at'); -- space char after label to move icon away from in-source text; TODO: a better way to do this? at = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, at}); coins_pages = ws_label; end end return page, pages, at, coins_pages; end --[[--------------------------< I S _ U N I Q U E _ A R C H I V E _ U R L >------------------------------------ add error message when |archive-url= value is same as |url= or chapter-url= (or alias...) value ]] local function is_unique_archive_url (archive, url, c_url, source, date) if utilities.is_set (archive) then if archive == url or archive == c_url then utilities.set_message ('err_bad_url', {utilities.wrap_style ('parameter', source)}); -- add error message return '', ''; -- unset |archive-url= and |archive-date= because same as |url= or |chapter-url= end end return archive, date; end --[=[-------------------------< A R C H I V E _ U R L _ C H E C K >-------------------------------------------- Check archive.org URLs to make sure they at least look like they are pointing at valid archives and not to the save snapshot URL or to calendar pages. When the archive URL is 'https://web.archive.org/save/' (or http://...) archive.org saves a snapshot of the target page in the URL. That is something that Wikipedia should not allow unwitting readers to do. When the archive.org URL does not have a complete timestamp, archive.org chooses a snapshot according to its own algorithm or provides a calendar 'search' result. [[WP:ELNO]] discourages links to search results. This function looks at the value assigned to |archive-url= and returns empty strings for |archive-url= and |archive-date= and an error message when: |archive-url= holds an archive.org save command URL |archive-url= is an archive.org URL that does not have a complete timestamp (YYYYMMDDhhmmss 14 digits) in the correct place otherwise returns |archive-url= and |archive-date= There are two mostly compatible archive.org URLs: //web.archive.org/<timestamp>... -- the old form //web.archive.org/web/<timestamp>... -- the new form The old form does not support or map to the new form when it contains a display flag. There are four identified flags ('id_', 'js_', 'cs_', 'im_') but since archive.org ignores others following the same form (two letters and an underscore) we don't check for these specific flags but we do check the form. This function supports a preview mode. When the article is rendered in preview mode, this function may return a modified archive URL: for save command errors, return undated wildcard (/*/) for timestamp errors when the timestamp has a wildcard, return the URL unmodified for timestamp errors when the timestamp does not have a wildcard, return with timestamp limited to six digits plus wildcard (/yyyymm*/) A secondary function is to return an archive-url timestamp from those archive urls that have them. The timestamp is used by validation.archive_date_check() to see if the value in |archive-date= matches the timestamp in the archive url. ]=] local function archive_url_check (url, date) local err_msg = ''; -- start with the error message empty local path, timestamp, flag; -- portions of the archive.org URL local function is_ts_date_valid (timestamp, origin_date) -- local function to validate archive url taimestamp date; time ignored local y, m, d = timestamp:match ('(%d%d%d%d)(%d%d)(%d%d)'); -- split into parts if y and validation.is_valid_date (y, m, d) and (tonumber (y) >= 1996) then return true, tonumber (y) >= origin_date; -- <origin_date> is 1996 for archive.org end end if (not url:match('//web%.archive%.org/')) and (not url:match('//liveweb%.archive%.org/')) then -- also deprecated liveweb Wayback machine URL return url, date; -- not an archive.org archive, return ArchiveURL and ArchiveDate end if url:match('//web%.archive%.org/save/') then -- if a save command URL, we don't want to allow saving of the target page err_msg = cfg.err_msg_supl.save; url = url:gsub ('(//web%.archive%.org)/save/', '%1/*/', 1); -- for preview mode: modify ArchiveURL elseif url:match('//liveweb%.archive%.org/') then err_msg = cfg.err_msg_supl.liveweb; else path, timestamp, flag = url:match('//web%.archive%.org/([^%d]*)(%d+)([^/]*)/'); -- split out some of the URL parts for evaluation if not path then -- malformed in some way; pattern did not match err_msg = cfg.err_msg_supl.timestamp; elseif 14 ~= timestamp:len() then -- path and flag optional, must have 14-digit timestamp here err_msg = cfg.err_msg_supl.timestamp; if '*' ~= flag then local replacement = timestamp:match ('^%d%d%d%d%d%d') or timestamp:match ('^%d%d%d%d'); -- get the first 6 (YYYYMM) or first 4 digits (YYYY) if replacement then -- nil if there aren't at least 4 digits (year) replacement = replacement .. string.rep ('0', 14 - replacement:len()); -- year or yearmo (4 or 6 digits) zero-fill to make 14-digit timestamp url=url:gsub ('(//web%.archive%.org/[^%d]*)%d[^/]*', '%1' .. replacement .. '*', 1) -- for preview, modify ts to 14 digits plus splat for calendar display end end elseif not is_ts_date_valid (timestamp, 1996) then -- is ymd portion of timestamp a valid date? err_msg = cfg.err_msg_supl.timestamp; timestamp = nil; -- unset because invalid elseif utilities.is_set (path) and 'web/' ~= path then -- older archive URLs do not have the extra 'web/' path element err_msg = cfg.err_msg_supl.path; elseif utilities.is_set (flag) and not utilities.is_set (path) then -- flag not allowed with the old form URL (without the 'web/' path element) err_msg = cfg.err_msg_supl.flag; elseif utilities.is_set (flag) and not flag:match ('%a%a_') then -- flag if present must be two alpha characters and underscore (requires 'web/' path element) err_msg = cfg.err_msg_supl.flag; else return url, date, timestamp; -- return ArchiveURL, ArchiveDate, and timestamp from |archive-url= end end -- if here, something not right so utilities.set_message ('err_archive_url', {err_msg}); -- add error message and if is_preview_mode then return url, date, timestamp; -- preview mode so return ArchiveURL, ArchiveDate, and timestamp from |archive-url= else return '', ''; -- return empty strings for ArchiveURL and ArchiveDate end end --[[--------------------------< P L A C E _ C H E C K >-------------------------------------------------------- check |place=, |publication-place=, |location= to see if these params include digits. This function added because many editors misuse location to specify the in-source location (|page(s)= and |at= are supposed to do that) returns the original parameter value without modification; added maint cat when parameter value contains digits ]] local function place_check (param_val) if not utilities.is_set (param_val) then -- parameter empty or omitted return param_val; -- return that empty state end if mw.ustring.find (param_val, '%d') then -- not empty, are there digits in the parameter value utilities.set_message ('maint_location'); -- yep, add maint cat end return param_val; -- and done end --[[--------------------------< I S _ A R C H I V E D _ C O P Y >---------------------------------------------- compares |title= to 'Archived copy' (placeholder added by bots that can't find proper title); if matches, return true; nil else ]] local function is_archived_copy (title) title = mw.ustring.lower(title); -- switch title to lower case if title:find (cfg.special_case_translation.archived_copy.en) then -- if title is 'Archived copy' return true; elseif cfg.special_case_translation.archived_copy['local'] then if mw.ustring.find (title, cfg.special_case_translation.archived_copy['local']) then -- mw.ustring() because might not be Latin script return true; end end end --[[--------------------------< D I S P L A Y _ N A M E S _ S E L E C T >-------------------------------------- for any of the |display-authors=, |display-editors=, etc parameters, select either the local or global setting. When both are present, look at <local_display_names> value. When the value is some sort of 'et al.'string, special handling is required. When {{cs1 config}} has |display-<namelist>= AND this template has |display-<namelist>=etal AND: the number of names specified by <number_of_names> is: greater than the number specified in the global |display-<namelist>= parameter (<global_display_names>) use global |display-<namelist>= parameter value set overridden maint category less than or equal to the number specified in the global |display-<namelist>= parameter use local |display-<namelist>= parameter value The purpose of this function is to prevent categorizing a template that has fewer names than the global setting to keep the etal annotation specified by <local_display_names>. ]] local function display_names_select (global_display_names, local_display_names, param_name, number_of_names, test) if global_display_names and utilities.is_set (local_display_names) then -- when both if 'etal' == local_display_names:lower():gsub("[ '%.]", '') then -- the :gsub() portion makes 'etal' from a variety of 'et al.' spellings and stylings number_of_names = tonumber (number_of_names); -- convert these to numbers for comparison local global_display_names_num = tonumber (global_display_names); -- <global_display_names> not set when parameter value is not digits if number_of_names > global_display_names_num then -- template has more names than global config allows to be displayed? utilities.set_message ('maint_overridden_setting'); -- set a maint message because global is overriding local |display-<namelist>=etal return global_display_names, 'cs1 config'; -- return global with spoof parameter name (for get_display_names()) else return local_display_names, param_name; -- return local because fewer names so let <local_display_names> control end end -- here when <global_display_names> and <local_display_names> both numbers; <global_display_names> controls utilities.set_message ('maint_overridden_setting'); -- set a maint message return global_display_names, 'cs1 config'; -- return global with spoof parameter name (for get_display_names()) end -- here when only one of <global_display_names> or <local_display_names> set if global_display_names then return global_display_names, 'cs1 config'; -- return global with spoof parameter name (for get_display_names()) else return local_display_names, param_name; -- return local end end --[[--------------------------< M O D E _ S E T >-------------------------------------------------------------- fetch global mode setting from {{cs1 config}} (if present) or from |mode= (if present); global setting overrides local |mode= parameter value. When both are present, emit maintenance message ]] local function mode_set (Mode, Mode_origin) local mode; if cfg.global_cs1_config_t['Mode'] then -- global setting in {{cs1 config}}; nil when empty or assigned value invalid mode = is_valid_parameter_value (cfg.global_cs1_config_t['Mode'], 'cs1 config: mode', cfg.keywords_lists['mode'], ''); -- error messaging 'param' here is a hoax else mode = is_valid_parameter_value (Mode, Mode_origin, cfg.keywords_lists['mode'], ''); end if cfg.global_cs1_config_t['Mode'] and utilities.is_set (Mode) then -- when template has |mode=<something> which global setting has overridden utilities.set_message ('maint_overridden_setting'); -- set a maint message end return mode; end --[[--------------------------< Q U O T E _ M A K E >---------------------------------------------------------- create quotation from |quote=, |trans-quote=, and/or script-quote= with or without |quote-page= or |quote-pages= when any of those three quote parameters are set, this function unsets <PostScript>. When none of those parameters are set, |quote-page= and |quote-pages= are unset to nil so that they are not included in the template's metadata ]] local function quote_make (quote, trans_quote, script_quote, quote_page, quote_pages, nopp, sepc, postscript) if utilities.is_set (quote) or utilities.is_set (trans_quote) or utilities.is_set (script_quote) then if utilities.is_set (quote) then if quote:sub(1, 1) == '"' and quote:sub(-1, -1) == '"' then -- if first and last characters of quote are quote marks quote = quote:sub(2, -2); -- strip them off end end quote = kern_quotes (quote); -- kern if needed quote = utilities.wrap_style ('quoted-text', quote ); -- wrap in <q>...</q> tags if utilities.is_set (script_quote) then quote = script_concatenate (quote, script_quote, 'script-quote'); -- <bdi> tags, lang attribute, categorization, etc.; must be done after quote is wrapped end if utilities.is_set (trans_quote) then if trans_quote:sub(1, 1) == '"' and trans_quote:sub(-1, -1) == '"' then -- if first and last characters of |trans-quote are quote marks trans_quote = trans_quote:sub(2, -2); -- strip them off end quote = quote .. " " .. utilities.wrap_style ('trans-quoted-title', trans_quote ); end if utilities.is_set (quote_page) or utilities.is_set (quote_pages) then -- add page prefix local quote_prefix = ''; if utilities.is_set (quote_page) then extra_text_in_page_check (quote_page, 'quote-page'); -- add to maint cat if |quote-page= value begins with what looks like p., pp., etc. if not nopp then quote_prefix = utilities.substitute (cfg.messages['p-prefix'], {sepc, quote_page}), '', '', ''; else quote_prefix = utilities.substitute (cfg.messages['nopp'], {sepc, quote_page}), '', '', ''; end elseif utilities.is_set (quote_pages) then extra_text_in_page_check (quote_pages, 'quote-pages'); -- add to maint cat if |quote-pages= value begins with what looks like p., pp., etc. if tonumber(quote_pages) ~= nil and not nopp then -- if only digits, assume single page quote_prefix = utilities.substitute (cfg.messages['p-prefix'], {sepc, quote_pages}), '', ''; elseif not nopp then quote_prefix = utilities.substitute (cfg.messages['pp-prefix'], {sepc, quote_pages}), '', ''; else quote_prefix = utilities.substitute (cfg.messages['nopp'], {sepc, quote_pages}), '', ''; end end quote = quote_prefix .. ": " .. quote; else quote = sepc .. " " .. quote; end postscript = ""; -- cs1|2 does not supply terminal punctuation when |quote= is set elseif utilities.is_set (quote_page) or utilities.is_set (quote_pages) then quote_page = nil; -- unset; these require |quote=; TODO: error message? quote_pages = nil; end return quote, quote_page, quote_pages, postscript; end --[[--------------------------< C H E C K _ P U B L I S H E R _ N A M E >-------------------------------------- look for variations of '<text>: <text>' that might be '<location>: <publisher>' in |publisher= parameter value. when found, emit a maintenance message; return nil else <publisher> is the value assigned to |publisher= or |institution= ]] local function check_publisher_name (publisher) local patterns_t = { '^[%w%s]+%s*:%s*[%w%s]+$', -- plain text <location>: <publisher> '^%[+[%w%s:|]+%]+%s*:%s*[%w%s]+$', -- partially wikilinked [[<location>]]: <publisher> '^[%w%s]+%s*:%s*%[+[%w%s:|]+%]+$', -- partially wikilinked <location>: [[<publisher>]] '^%[+[%w%s:|]+%]+%s*:%s*%[+[%w%s:|]+%]+$', -- wikilinked [[<location>]]: [[<publisher>]] } for _, pattern in ipairs (patterns_t) do -- spin through the patterns_t sequence if mw.ustring.match (publisher, pattern) then -- does this pattern match? utilities.set_message ('maint_publisher_location'); -- set a maint message return; -- and done end end end --[[--------------------------< I S _ P A G E _ A R T _ N U M >------------------------------------------------ compare the trailing (rightmost) characters of the |doi= value against the whole value assigned to |page(s)=. return boolean true when: |page(s)= has exactly 8 digits and a dot between the fourth and fifth digits matches the trailing 9 characters of the |doi= value: |page=12345678 → |page=1234.5678 matches |doi=10.xxxx/yyyy1234.5678 |page(s)= is 5 or more characters and matches |doi= values's trailing characters |page(s)= begins with a lowercase 'e' and |page(s)= without the 'e' matches |doi= values's trailing characters: |page=e12345 → |page=12345 matches |doi=10.xxxx/yyyy12345 |page(s)= begins with a uppercase 'CD' followed by (typically) six digits matches |doi= values that ends with 'CDxxxxxx.pubx' (where 'x' is any single digit) return nil when |page(s)= values: are ranges separated by underscore, hyphen, emdash, endash, figure dash, or minus character are comma- or semicolon-separated lists of pages have external urls (has text 'http') are digit-only values less than 10000 do not match |doi= values's trailing characters ]] local function is_page_art_num (page, doi) if not (utilities.is_set (page) and utilities.is_set (doi)) then -- both required return; -- abandon; nothing to do end if page:match ('[,;_−–—‒%-]') then -- when |page(s)= might be a page range or a separated list of pages return; -- abandon end page = page:lower(); -- because doi names are case insensitive doi = doi:lower(); -- force these to lowercase for testing if page:match ('http') then -- when |page(s)= appears to hold a url return; -- abandon end if tonumber (page) then -- is |page(s)= digits only if 10000 > tonumber (page) then -- when |page(s)= less than 10000 return; -- abandon end if doi:match (page .. '$') then -- digits only page number match the last digits in |doi=? return true; end if 8 == page:len() then -- special case when |page(s)= is exactly 8 digits local dot_page = page:gsub ('(%d%d%d%d)(%d%d%d%d)', '%1.%2'); -- make a |page=xxxx.yyyy version commonly used in |doi= if doi:match (dot_page .. '$') then -- 8-digit dotted page number match the last characters in |doi=? return true; end end else -- here when |page(s)= is alpha-numeric if 4 < page:len() then -- when |page(s)= is five or more characters if doi:match (page .. '$') then -- alpha-numeric page match the last characters in |doi=? return true; end local epage = page:match ('^e([%w]+)$'); -- if first character of |page= is 'e', remove it if epage and doi:match (epage .. '$') then -- page number match the last characters in |doi=? return true; end local cdpage = page:match ('^cd%d+$'); -- if first characters of |page= are 'CD' and last characters are digits (typically 6 digits) if cdpage and doi:match (cdpage .. '%.pub%d$') then -- page number matches doi 'CDxxxxxx.pubx' where 'x' is a digit return true; end end end end --[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------ This is the main function doing the majority of the citation formatting. ]] local function citation0( config, args ) --[[ Load Input Parameters The argument_wrapper facilitates the mapping of multiple aliases to single internal variable. ]] local A = argument_wrapper ( args ); local i -- Pick out the relevant fields from the arguments. Different citation templates -- define different field names for the same underlying things. local author_etal; local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors= local Authors; local NameListStyle; if cfg.global_cs1_config_t['NameListStyle'] then -- global setting in {{cs1 config}} overrides local |name-list-style= parameter value; nil when empty or assigned value invalid NameListStyle = is_valid_parameter_value (cfg.global_cs1_config_t['NameListStyle'], 'cs1 config: name-list-style', cfg.keywords_lists['name-list-style'], ''); -- error messaging 'param' here is a hoax else NameListStyle = is_valid_parameter_value (A['NameListStyle'], A:ORIGIN('NameListStyle'), cfg.keywords_lists['name-list-style'], ''); end if cfg.global_cs1_config_t['NameListStyle'] and utilities.is_set (A['NameListStyle']) then -- when template has |name-list-style=<something> which global setting has overridden utilities.set_message ('maint_overridden_setting'); -- set a maint message end local Collaboration = A['Collaboration']; do -- to limit scope of selected local selected = select_author_editor_source (A['Vauthors'], A['Authors'], args, 'AuthorList'); if 1 == selected then a, author_etal = extract_names (args, 'AuthorList'); -- fetch author list from |authorn= / |lastn= / |firstn=, |author-linkn=, and |author-maskn= elseif 2 == selected then NameListStyle = 'vanc'; -- override whatever |name-list-style= might be a, author_etal = parse_vauthors_veditors (args, A['Vauthors'], 'AuthorList'); -- fetch author list from |vauthors=, |author-linkn=, and |author-maskn= elseif 3 == selected then Authors = A['Authors']; -- use content of |people= or |credits=; |authors= is deprecated; TODO: constrain |people= and |credits= to cite av media, episode, serial? end if utilities.is_set (Collaboration) then author_etal = true; -- so that |display-authors=etal not required end end local editor_etal; local e = {}; -- editors list from |editor-lastn= / |editor-firstn= pairs or |veditors= do -- to limit scope of selected local selected = select_author_editor_source (A['Veditors'], nil, args, 'EditorList'); -- support for |editors= withdrawn if 1 == selected then e, editor_etal = extract_names (args, 'EditorList'); -- fetch editor list from |editorn= / |editor-lastn= / |editor-firstn=, |editor-linkn=, and |editor-maskn= elseif 2 == selected then NameListStyle = 'vanc'; -- override whatever |name-list-style= might be e, editor_etal = parse_vauthors_veditors (args, args.veditors, 'EditorList'); -- fetch editor list from |veditors=, |editor-linkn=, and |editor-maskn= end end local Chapter = A['Chapter']; -- done here so that we have access to |contribution= from |chapter= and |script-chapter= aliases local Chapter_origin = A:ORIGIN ('Chapter'); local ScriptChapter = A['ScriptChapter']; local ScriptChapter_origin = A:ORIGIN ('ScriptChapter'); local Contribution; -- because contribution is required for contributor(s) if 'contribution' == Chapter_origin then -- is there a |contribution= parameter? Contribution = Chapter; -- get the name of the contribution elseif 'script-contribution' == ScriptChapter_origin then -- no, so is there a |script-contribution= parameter? Contribution = ScriptChapter; -- get the name of the contribution end local c = {}; -- contributors list from |contributor-lastn= / contributor-firstn= pairs if utilities.in_array (config.CitationClass, {"book", "citation"}) and not utilities.is_set (A['Periodical']) then -- |contributor= and |contribution= only supported in book cites c = extract_names (args, 'ContributorList'); -- fetch contributor list from |contributorn= / |contributor-lastn=, -firstn=, -linkn=, -maskn= if 0 < #c then if not utilities.is_set (Contribution) then -- |contributor= requires |contribution= utilities.set_message ('err_contributor_missing_required_param', 'contribution'); -- add missing contribution error message c = {}; -- blank the contributors' table; it is used as a flag later end if 0 == #a then -- |contributor= requires |author= utilities.set_message ('err_contributor_missing_required_param', 'author'); -- add missing author error message c = {}; -- blank the contributors' table; it is used as a flag later end end else -- if not a book cite if utilities.select_one (args, cfg.aliases['ContributorList-Last'], 'err_redundant_parameters', 1 ) then -- are there contributor name list parameters? utilities.set_message ('err_contributor_ignored'); -- add contributor ignored error message end Contribution = nil; -- unset end local Title = A['Title']; local TitleLink = A['TitleLink']; local auto_select = ''; -- default is auto local accept_link; TitleLink, accept_link = utilities.has_accept_as_written (TitleLink, true); -- test for accept-this-as-written markup if (not accept_link) and utilities.in_array (TitleLink, {'none', 'pmc', 'doi'}) then -- check for special keywords auto_select = TitleLink; -- remember selection for later TitleLink = ''; -- treat as if |title-link= would have been empty end TitleLink = link_title_ok (TitleLink, A:ORIGIN ('TitleLink'), Title, 'title'); -- check for wiki-markup in |title-link= or wiki-markup in |title= when |title-link= is set local Section = ''; -- {{cite map}} only; preset to empty string for concatenation if not used if 'map' == config.CitationClass and 'section' == Chapter_origin then Section = A['Chapter']; -- get |section= from |chapter= alias list; |chapter= and the other aliases not supported in {{cite map}} Chapter = ''; -- unset for now; will be reset later from |map= if present end local Periodical = A['Periodical']; local Periodical_origin = A:ORIGIN('Periodical'); local ScriptPeriodical = A['ScriptPeriodical']; local ScriptPeriodical_origin = A:ORIGIN('ScriptPeriodical'); local TransPeriodical = A['TransPeriodical']; local TransPeriodical_origin = A:ORIGIN ('TransPeriodical'); if (utilities.in_array (config.CitationClass, {'book', 'encyclopaedia'}) and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical) or utilities.is_set (TransPeriodical))) then local param; if utilities.is_set (Periodical) then -- get a parameter name from one of these periodical related meta-parameters Periodical = ''; -- unset because not valid {{cite book}} or {{cite encyclopedia}} parameters param = Periodical_origin -- get parameter name for error messaging elseif utilities.is_set (TransPeriodical) then TransPeriodical = ''; -- unset because not valid {{cite book}} or {{cite encyclopedia}} parameters param = TransPeriodical_origin; -- get parameter name for error messaging elseif utilities.is_set (ScriptPeriodical) then ScriptPeriodical = ''; -- unset because not valid {{cite book}} or {{cite encyclopedia}} parameters param = ScriptPeriodical_origin; -- get parameter name for error messaging end if utilities.is_set (param) then -- if we found one utilities.set_message ('err_periodical_ignored', {param}); -- emit an error message end end if utilities.is_set (Periodical) then local i; Periodical, i = utilities.strip_apostrophe_markup (Periodical); -- strip apostrophe markup so that metadata isn't contaminated if i then -- non-zero when markup was stripped so emit an error message utilities.set_message ('err_apostrophe_markup', {Periodical_origin}); end end if 'mailinglist' == config.CitationClass then -- special case for {{cite mailing list}} if utilities.is_set (Periodical) and utilities.is_set (A ['MailingList']) then -- both set emit an error TODO: make a function for this and similar? utilities.set_message ('err_redundant_parameters', {utilities.wrap_style ('parameter', Periodical_origin) .. cfg.presentation['sep_list_pair'] .. utilities.wrap_style ('parameter', 'mailinglist')}); end Periodical = A ['MailingList']; -- error or no, set Periodical to |mailinglist= value because this template is {{cite mailing list}} Periodical_origin = A:ORIGIN('MailingList'); end -- web and news not tested for now because of -- Wikipedia:Administrators%27_noticeboard#Is_there_a_semi-automated_tool_that_could_fix_these_annoying_"Cite_Web"_errors? if not (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)) then -- 'periodical' templates require periodical parameter -- local p = {['journal'] = 'journal', ['magazine'] = 'magazine', ['news'] = 'newspaper', ['web'] = 'website'}; -- for error message local p = {['journal'] = 'journal', ['magazine'] = 'magazine'}; -- for error message if p[config.CitationClass] then --#Modifed utilities.set_message ('err_missing_periodical', {cfg.err_msg_supl[config.CitationClass], cfg.err_msg_supl[p[config.CitationClass]]}); end end local Volume; if 'citation' == config.CitationClass then if utilities.is_set (Periodical) then if not utilities.in_array (Periodical_origin, cfg.citation_no_volume_t) then -- {{citation}} does not render |volume= when these parameters are used Volume = A['Volume']; -- but does for all other 'periodicals' end elseif utilities.is_set (ScriptPeriodical) then if 'script-website' ~= ScriptPeriodical_origin then -- {{citation}} does not render volume for |script-website= Volume = A['Volume']; -- but does for all other 'periodicals' end else Volume = A['Volume']; -- and does for non-'periodical' cites end elseif utilities.in_array (config.CitationClass, cfg.templates_using_volume) then -- render |volume= for cs1 according to the configuration settings Volume = A['Volume']; end extra_text_in_vol_iss_check (Volume, A:ORIGIN ('Volume'), 'v'); local Issue; if 'citation' == config.CitationClass then if utilities.is_set (Periodical) and utilities.in_array (Periodical_origin, cfg.citation_issue_t) then -- {{citation}} may render |issue= when these parameters are used Issue = A['Issue']; end elseif utilities.in_array (config.CitationClass, cfg.templates_using_issue) then -- conference & map books do not support issue; {{citation}} listed here because included in settings table if not (utilities.in_array (config.CitationClass, {'conference', 'map', 'citation'}) and not (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical))) then Issue = A['Issue']; end end local ArticleNumber; if utilities.in_array (config.CitationClass, {'journal', 'conference'}) or ('citation' == config.CitationClass and utilities.is_set (Periodical) and 'journal' == Periodical_origin) then ArticleNumber = A['ArticleNumber']; end extra_text_in_vol_iss_check (Issue, A:ORIGIN ('Issue'), 'i'); Issue = utilities.hyphen_to_dash (Issue); local Page; local Pages; local At; local QuotePage; local QuotePages; if not utilities.in_array (config.CitationClass, cfg.templates_not_using_page) then -- TODO: rewrite to emit ignored parameter error message? Page = A['Page']; Pages = utilities.hyphen_to_dash (A['Pages']); At = A['At']; QuotePage = A['QuotePage']; QuotePages = utilities.hyphen_to_dash (A['QuotePages']); end local NoPP = is_valid_parameter_value (A['NoPP'], A:ORIGIN('NoPP'), cfg.keywords_lists['yes_true_y'], nil); local Mode = mode_set (A['Mode'], A:ORIGIN('Mode')); -- separator character and postscript local sepc, PostScript = set_style (Mode:lower(), A['PostScript'], config.CitationClass); local Quote; Quote, QuotePage, QuotePages, PostScript = quote_make (A['Quote'], A['TransQuote'], A['ScriptQuote'], QuotePage, QuotePages, NoPP, sepc, PostScript); local Edition = A['Edition']; local PublicationPlace = place_check (A['PublicationPlace'], A:ORIGIN('PublicationPlace')); local Place = place_check (A['Place'], A:ORIGIN('Place')); local PublisherName = A['PublisherName']; local PublisherName_origin = A:ORIGIN('PublisherName'); if utilities.is_set (PublisherName) and (cfg.keywords_xlate['none'] ~= PublisherName) then local i = 0; PublisherName, i = utilities.strip_apostrophe_markup (PublisherName); -- strip apostrophe markup so that metadata isn't contaminated; publisher is never italicized if i and (0 < i) then -- non-zero when markup was stripped so emit an error message utilities.set_message ('err_apostrophe_markup', {PublisherName_origin}); end end if ('document' == config.CitationClass) and not utilities.is_set (PublisherName) then utilities.set_message ('err_missing_publisher', {config.CitationClass, 'publisher'}); end local Newsgroup = A['Newsgroup']; -- TODO: strip apostrophe markup? local Newsgroup_origin = A:ORIGIN('Newsgroup'); if 'newsgroup' == config.CitationClass then if utilities.is_set (PublisherName) and (cfg.keywords_xlate['none'] ~= PublisherName) then -- general use parameter |publisher= not allowed in cite newsgroup utilities.set_message ('err_parameter_ignored', {PublisherName_origin}); end PublisherName = nil; -- ensure that this parameter is unset for the time being; will be used again after COinS end if 'book' == config.CitationClass or 'encyclopaedia' == config.CitationClass or ('citation' == config.CitationClass and not utilities.is_set (Periodical)) then local accept; PublisherName, accept = utilities.has_accept_as_written (PublisherName); -- check for and remove accept-as-written markup from |publisher= wrapped if not accept then -- when no accept-as-written markup check_publisher_name (PublisherName); -- emit maint message when |publisher= might be prefixed with publisher's location end end local URL = A['URL']; -- TODO: better way to do this for URL, ChapterURL, and MapURL? local UrlAccess = is_valid_parameter_value (A['UrlAccess'], A:ORIGIN('UrlAccess'), cfg.keywords_lists['url-access'], nil); if not utilities.is_set (URL) and utilities.is_set (UrlAccess) then UrlAccess = nil; utilities.set_message ('err_param_access_requires_param', 'url'); end local ChapterURL = A['ChapterURL']; local ChapterUrlAccess = is_valid_parameter_value (A['ChapterUrlAccess'], A:ORIGIN('ChapterUrlAccess'), cfg.keywords_lists['url-access'], nil); if not utilities.is_set (ChapterURL) and utilities.is_set (ChapterUrlAccess) then ChapterUrlAccess = nil; utilities.set_message ('err_param_access_requires_param', {A:ORIGIN('ChapterUrlAccess'):gsub ('%-access', '')}); end local MapUrlAccess = is_valid_parameter_value (A['MapUrlAccess'], A:ORIGIN('MapUrlAccess'), cfg.keywords_lists['url-access'], nil); if not utilities.is_set (A['MapURL']) and utilities.is_set (MapUrlAccess) then MapUrlAccess = nil; utilities.set_message ('err_param_access_requires_param', {'map-url'}); end local this_page = mw.title.getCurrentTitle(); -- also used for COinS and for language local no_tracking_cats = is_valid_parameter_value (A['NoTracking'], A:ORIGIN('NoTracking'), cfg.keywords_lists['yes_true_y'], nil); -- check this page to see if it is in one of the namespaces that cs1 is not supposed to add to the error categories if not utilities.is_set (no_tracking_cats) then -- ignore if we are already not going to categorize this page if cfg.uncategorized_namespaces[this_page.namespace] then -- is this page's namespace id one of the uncategorized namespace ids? no_tracking_cats = "true"; -- set no_tracking_cats end for _, v in ipairs (cfg.uncategorized_subpages) do -- cycle through page name patterns if this_page.text:match (v) then -- test page name against each pattern no_tracking_cats = "true"; -- set no_tracking_cats break; -- bail out if one is found end end end -- check for extra |page=, |pages= or |at= parameters. (also sheet and sheets while we're at it) utilities.select_one (args, {'page', 'p', 'pp', 'pages', 'at', 'sheet', 'sheets'}, 'err_redundant_parameters'); -- this is a dummy call simply to get the error message and category local coins_pages; Page, Pages, At, coins_pages = insource_loc_get (Page, A:ORIGIN('Page'), Pages, A:ORIGIN('Pages'), At); if utilities.is_set (PublicationPlace) and utilities.is_set (Place) then -- both |publication-place= and |place= (|location=) allowed if different utilities.add_prop_cat ('location-test'); -- add property cat to evaluate how often PublicationPlace and Place are used together if PublicationPlace == Place then Place = ''; -- unset; don't need both if they are the same end elseif not utilities.is_set (PublicationPlace) and utilities.is_set (Place) then -- when only |place= (|location=) is set ... PublicationPlace = Place; -- promote |place= (|location=) to |publication-place end if PublicationPlace == Place then Place = ''; end -- don't need both if they are the same local URL_origin = A:ORIGIN('URL'); -- get name of parameter that holds URL local ChapterURL_origin = A:ORIGIN('ChapterURL'); -- get name of parameter that holds ChapterURL -- local ScriptChapter = A['ScriptChapter']; -- local ScriptChapter_origin = A:ORIGIN ('ScriptChapter'); local Format = A['Format']; local ChapterFormat = A['ChapterFormat']; local TransChapter = A['TransChapter']; local TransChapter_origin = A:ORIGIN ('TransChapter'); local TransTitle = A['TransTitle']; local ScriptTitle = A['ScriptTitle']; --[[ Parameter remapping for cite encyclopedia: When the citation has these parameters: |encyclopedia= and |title= then map |title= to |article= and |encyclopedia= to |title= for rendering |encyclopedia= and |article= then map |encyclopedia= to |title= for rendering |trans-title= maps to |trans-chapter= when |title= is re-mapped |url= maps to |chapter-url= when |title= is remapped All other combinations of |encyclopedia=, |title=, and |article= are not modified ]] local Encyclopedia = A['Encyclopedia']; -- used as a flag by this module and by ~/COinS local ScriptEncyclopedia = A['ScriptEncyclopedia']; local TransEncyclopedia = A['TransEncyclopedia']; if utilities.is_set (Encyclopedia) or utilities.is_set (ScriptEncyclopedia) then -- emit error message when Encyclopedia set but template is other than {{cite encyclopedia}} or {{citation}} if 'encyclopaedia' ~= config.CitationClass and 'citation' ~= config.CitationClass then if utilities.is_set (Encyclopedia) then utilities.set_message ('err_parameter_ignored', {A:ORIGIN ('Encyclopedia')}); else utilities.set_message ('err_parameter_ignored', {A:ORIGIN ('ScriptEncyclopedia')}); end Encyclopedia = nil; -- unset these because not supported by this template ScriptEncyclopedia = nil; TransEncyclopedia = nil; end elseif utilities.is_set (TransEncyclopedia) then utilities.set_message ('err_trans_missing_title', {'encyclopedia'}); end if ('encyclopaedia' == config.CitationClass) or ('citation' == config.CitationClass and utilities.is_set (Encyclopedia)) then if utilities.is_set (Periodical) and utilities.is_set (Encyclopedia) then -- when both parameters set emit an error message; {{citation}} only; Periodical not allowed in {{cite encyclopedia}} utilities.set_message ('err_periodical_ignored', {Periodical_origin}); end if utilities.is_set (Encyclopedia) or utilities.is_set (ScriptEncyclopedia) then Periodical = Encyclopedia; -- error or no, set Periodical to Encyclopedia for rendering; {{citation}} could (not legitimately) have both; use Encyclopedia Periodical_origin = A:ORIGIN ('Encyclopedia'); ScriptPeriodical = ScriptEncyclopedia; ScriptPeriodical_origin = A:ORIGIN ('ScriptEncyclopedia'); if utilities.is_set (Title) or utilities.is_set (ScriptTitle) then if not utilities.is_set (Chapter) then Chapter = Title; -- |encyclopedia= and |title= are set so map |title= params to |article= params for rendering ScriptChapter = ScriptTitle; ScriptChapter_origin = A:ORIGIN('ScriptTitle') TransChapter = TransTitle; ChapterURL = URL; ChapterURL_origin = URL_origin; ChapterUrlAccess = UrlAccess; ChapterFormat = Format; if not utilities.is_set (ChapterURL) and utilities.is_set (TitleLink) then Chapter = utilities.make_wikilink (TitleLink, Chapter); end Title = Periodical; -- now map |encyclopedia= params to |title= params for rendering ScriptTitle = ScriptPeriodical or ''; TransTitle = TransEncyclopedia or ''; Periodical = ''; -- redundant so unset ScriptPeriodical = ''; URL = ''; Format = ''; TitleLink = ''; end elseif utilities.is_set (Chapter) or utilities.is_set (ScriptChapter) then -- |title= not set Title = Periodical; -- |encyclopedia= set and |article= set so map |encyclopedia= to |title= for rendering ScriptTitle = ScriptPeriodical or ''; TransTitle = TransEncyclopedia or ''; Periodical = ''; -- redundant so unset ScriptPeriodical = ''; end end end -- special case for cite techreport. local ID = A['ID']; if (config.CitationClass == "techreport") then -- special case for cite techreport if utilities.is_set (A['Number']) then -- cite techreport uses 'number', which other citations alias to 'issue' if not utilities.is_set (ID) then -- can we use ID for the "number"? ID = A['Number']; -- yes, use it else -- ID has a value so emit error message utilities.set_message ('err_redundant_parameters', {utilities.wrap_style ('parameter', 'id') .. cfg.presentation['sep_list_pair'] .. utilities.wrap_style ('parameter', 'number')}); end end end -- Account for the oddity that is {{cite conference}}, before generation of COinS data. local ChapterLink -- = A['ChapterLink']; -- deprecated as a parameter but still used internally by cite episode local Conference = A['Conference']; local BookTitle = A['BookTitle']; local TransTitle_origin = A:ORIGIN ('TransTitle'); if 'conference' == config.CitationClass then if utilities.is_set (BookTitle) then Chapter = Title; Chapter_origin = 'title'; -- ChapterLink = TitleLink; -- |chapter-link= is deprecated ChapterURL = URL; ChapterUrlAccess = UrlAccess; ChapterURL_origin = URL_origin; URL_origin = ''; ChapterFormat = Format; TransChapter = TransTitle; TransChapter_origin = TransTitle_origin; Title = BookTitle; Format = ''; -- TitleLink = ''; TransTitle = ''; URL = ''; end elseif 'speech' ~= config.CitationClass then Conference = ''; -- not cite conference or cite speech so make sure this is empty string end local cs2_lower = function (text) -- if mode is cs2, use lower case if (sepc ~= '.') then return text:lower(); end return text; end local use_lowercase = ( sepc == ',' ); -- controls capitalization of certain static text -- cite map oddities local Cartography = ""; local Scale = ""; local Sheet = A['Sheet'] or ''; local Sheets = A['Sheets'] or ''; if config.CitationClass == "map" then if utilities.is_set (Chapter) then --TODO: make a function for this and similar? utilities.set_message ('err_redundant_parameters', {utilities.wrap_style ('parameter', 'map') .. cfg.presentation['sep_list_pair'] .. utilities.wrap_style ('parameter', Chapter_origin)}); -- add error message end Chapter = A['Map']; Chapter_origin = A:ORIGIN('Map'); ChapterURL = A['MapURL']; ChapterURL_origin = A:ORIGIN('MapURL'); TransChapter = A['TransMap']; ScriptChapter = A['ScriptMap'] ScriptChapter_origin = A:ORIGIN('ScriptMap') ChapterUrlAccess = MapUrlAccess; ChapterFormat = A['MapFormat']; Cartography = A['Cartography']; if utilities.is_set ( Cartography ) then Cartography = sepc .. " " .. wrap_msg ('cartography', Cartography, use_lowercase); end Scale = A['Scale']; if utilities.is_set ( Scale ) then Scale = sepc .. " " .. Scale; end end -- Account for the oddities that are {{cite episode}} and {{cite serial}}, before generation of COinS data. local Series = A['Series']; if 'episode' == config.CitationClass or 'serial' == config.CitationClass then local SeriesLink = A['SeriesLink']; SeriesLink = link_title_ok (SeriesLink, A:ORIGIN ('SeriesLink'), Series, 'series'); -- check for wiki-markup in |series-link= or wiki-markup in |series= when |series-link= is set local Network = A['Network']; local Station = A['Station']; local s, n = {}, {}; -- do common parameters first if utilities.is_set (Network) then table.insert(n, Network); end if utilities.is_set (Station) then table.insert(n, Station); end ID = table.concat(n, sepc .. ' '); if 'episode' == config.CitationClass then -- handle the oddities that are strictly {{cite episode}} local Season = A['Season']; local SeriesNumber = A['SeriesNumber']; if utilities.is_set (Season) and utilities.is_set (SeriesNumber) then -- these are mutually exclusive so if both are set TODO: make a function for this and similar? utilities.set_message ('err_redundant_parameters', {utilities.wrap_style ('parameter', 'season') .. cfg.presentation['sep_list_pair'] .. utilities.wrap_style ('parameter', 'seriesno')}); -- add error message SeriesNumber = ''; -- unset; prefer |season= over |seriesno= end -- assemble a table of parts concatenated later into Series if utilities.is_set (Season) then table.insert(s, wrap_msg ('season', Season, use_lowercase)); end if utilities.is_set (SeriesNumber) then table.insert(s, wrap_msg ('seriesnum', SeriesNumber, use_lowercase)); end if utilities.is_set (Issue) then table.insert(s, wrap_msg ('episode', Issue, use_lowercase)); end Issue = ''; -- unset because this is not a unique parameter Chapter = Title; -- promote title parameters to chapter ScriptChapter = ScriptTitle; ScriptChapter_origin = A:ORIGIN('ScriptTitle'); ChapterLink = TitleLink; -- alias |episode-link= TransChapter = TransTitle; ChapterURL = URL; ChapterUrlAccess = UrlAccess; ChapterURL_origin = URL_origin; ChapterFormat = Format; Title = Series; -- promote series to title TitleLink = SeriesLink; Series = table.concat(s, sepc .. ' '); -- this is concatenation of season, seriesno, episode number if utilities.is_set (ChapterLink) and not utilities.is_set (ChapterURL) then -- link but not URL Chapter = utilities.make_wikilink (ChapterLink, Chapter); elseif utilities.is_set (ChapterLink) and utilities.is_set (ChapterURL) then -- if both are set, URL links episode; Series = utilities.make_wikilink (ChapterLink, Series); end URL = ''; -- unset TransTitle = ''; ScriptTitle = ''; Format = ''; else -- now oddities that are cite serial Issue = ''; -- unset because this parameter no longer supported by the citation/core version of cite serial Chapter = A['Episode']; -- TODO: make |episode= available to cite episode someday? if utilities.is_set (Series) and utilities.is_set (SeriesLink) then Series = utilities.make_wikilink (SeriesLink, Series); end Series = utilities.wrap_style ('italic-title', Series); -- series is italicized end end -- end of {{cite episode}} stuff -- handle type parameter for those CS1 citations that have default values local TitleType = A['TitleType']; local Degree = A['Degree']; if utilities.in_array (config.CitationClass, {'AV-media-notes', 'document', 'interview', 'mailinglist', 'map', 'podcast', 'pressrelease', 'report', 'speech', 'techreport', 'thesis'}) then TitleType = set_titletype (config.CitationClass, TitleType); if utilities.is_set (Degree) and ("Thesis" == TitleType or cfg.title_types ['thesis'] == TitleType) then -- special case for cite thesis TitleType = utilities.substitute (cfg.presentation['thesis-degree'], {Degree, cfg.title_types ['thesis']:lower()}); end end if utilities.is_set (TitleType) then -- if type parameter is specified TitleType = utilities.substitute ( cfg.messages['type'], TitleType); -- display it in parentheses -- TODO: Hack on TitleType to fix bunched parentheses problem end -- legacy: promote PublicationDate to Date if neither Date nor Year are set. local Date = A['Date']; local Date_origin; -- to hold the name of parameter promoted to Date; required for date error messaging local PublicationDate = A['PublicationDate']; local Year = A['Year']; -- #Added local HijriDate = A['HijriDate']; if utilities.is_set (Year) then validation.year_check (Year); -- returns nothing; emits maint message when |year= doesn't hold a 'year' value end if not utilities.is_set (Date) then Date = Year; -- promote Year to Date Year = nil; -- make nil so Year as empty string isn't used for CITEREF if not utilities.is_set (Date) and utilities.is_set (PublicationDate) then -- use PublicationDate when |date= and |year= are not set Date = PublicationDate; -- promote PublicationDate to Date PublicationDate = ''; -- unset, no longer needed Date_origin = A:ORIGIN('PublicationDate'); -- save the name of the promoted parameter else Date_origin = A:ORIGIN('Year'); -- save the name of the promoted parameter end else Date_origin = A:ORIGIN('Date'); -- not a promotion; name required for error messaging end if PublicationDate == Date then PublicationDate = ''; end -- if PublicationDate is same as Date, don't display in rendered citation --[[ Go test all of the date-holding parameters for valid MOS:DATE format and make sure that dates are real dates. This must be done before we do COinS because here is where we get the date used in the metadata. Date validation supporting code is in Module:Citation/CS1/Date_validation ]] local DF = is_valid_parameter_value (A['DF'], A:ORIGIN('DF'), cfg.keywords_lists['df'], ''); if not utilities.is_set (DF) then DF = cfg.global_df; -- local |df= if present overrides global df set by {{use xxx date}} template end local ArchiveURL; local ArchiveDate; local ArchiveFormat = A['ArchiveFormat']; local archive_url_timestamp; -- timestamp from wayback machine url --#Added Block to add category for unarchived URLs if utilities.is_set(URL) and not utilities.is_set(A['ArchiveURL']) and not wikidatacite and not URL:match('//archive%.org/') -- no need for ArchiveURL if the URL is from archive cite! by User:Mr. Ibrahem then utilities.add_prop_cat ('no_archive'); end ArchiveURL, ArchiveDate, archive_url_timestamp = archive_url_check (A['ArchiveURL'], A['ArchiveDate']) ArchiveFormat = style_format (ArchiveFormat, ArchiveURL, 'archive-format', 'archive-url'); ArchiveURL, ArchiveDate = is_unique_archive_url (ArchiveURL, URL, ChapterURL, A:ORIGIN('ArchiveURL'), ArchiveDate); -- add error message when URL or ChapterURL == ArchiveURL local AccessDate = A['AccessDate']; local COinS_date = {}; -- holds date info extracted from |date= for the COinS metadata by Module:Date verification local DoiBroken = A['DoiBroken']; local Embargo = A['Embargo']; local anchor_year; -- used in the CITEREF identifier do -- create defined block to contain local variables error_message, date_parameters_list, mismatch local error_message = ''; -- AirDate has been promoted to Date so not necessary to check it local date_parameters_list = { ['access-date'] = {val = AccessDate, name = A:ORIGIN ('AccessDate')}, ['archive-date'] = {val = ArchiveDate, name = A:ORIGIN ('ArchiveDate')}, ['hijri-date'] = {val = HijriDate, name = A:ORIGIN ('HijriDate')}, ['date'] = {val = Date, name = Date_origin}, ['doi-broken-date'] = {val = DoiBroken, name = A:ORIGIN ('DoiBroken')}, ['pmc-embargo-date'] = {val = Embargo, name = A:ORIGIN ('Embargo')}, ['publication-date'] = {val = PublicationDate, name = A:ORIGIN ('PublicationDate')}, ['year'] = {val = Year, name = A:ORIGIN ('Year')}, }; local error_list = {}; anchor_year, Embargo = validation.dates(date_parameters_list, COinS_date, error_list); if utilities.is_set (Year) and utilities.is_set (Date) then -- both |date= and |year= not normally needed; validation.year_date_check (Year, A:ORIGIN ('Year'), Date, A:ORIGIN ('Date'), error_list); end --#Added Block if date_parameters_list['hijri-date'].tt and date_parameters_list['date'].tt then validation.hijri_and_gregorian_corresponded (date_parameters_list['hijri-date'].tt, date_parameters_list['date'].tt); end if 0 == #error_list then -- error free dates only; 0 when error_list is empty local modified = false; -- flag if utilities.is_set (DF) then -- if we need to reformat dates modified = validation.reformat_dates (date_parameters_list, DF); -- reformat to DF format, use long month names if appropriate end if true == validation.date_hyphen_to_dash (date_parameters_list) then -- convert hyphens to dashes where appropriate modified = true; utilities.set_message ('maint_date_format'); -- hyphens were converted so add maint category end -- for those wikis that can and want to have English date names translated to the local language; not supported at en.wiki if cfg.date_name_auto_xlate_enable and validation.date_name_xlate (date_parameters_list, cfg.date_digit_auto_xlate_enable ) then utilities.set_message ('maint_date_auto_xlated'); -- add maint cat modified = true; end if modified then -- if the date_parameters_list values were modified AccessDate = date_parameters_list['access-date'].val; -- overwrite date holding parameters with modified values ArchiveDate = date_parameters_list['archive-date'].val; Date = date_parameters_list['date'].val; DoiBroken = date_parameters_list['doi-broken-date'].val; PublicationDate = date_parameters_list['publication-date'].val; HijriDate = date_parameters_list['hijri-date'].val; end if archive_url_timestamp and utilities.is_set (ArchiveDate) then validation.archive_date_check (ArchiveDate, archive_url_timestamp, DF); -- does YYYYMMDD in archive_url_timestamp match date in ArchiveDate end else utilities.set_message ('err_bad_date', {utilities.make_sep_list (#error_list, error_list)}); -- add this error message end --#Added block if date_parameters_list['date'].val == "nd" then Date = cfg.messages["unknown_date"]; end end -- end of do if utilities.in_array (config.CitationClass, {'book', 'encyclopaedia'}) or -- {{cite book}}, {{cite encyclopedia}}; TODO: {{cite conference}} and others? ('citation' == config.CitationClass and utilities.is_set (Encyclopedia)) or -- {{citation}} as an encylopedia citation ('citation' == config.CitationClass and not utilities.is_set (Periodical)) then -- {{citation}} as a book citation if utilities.is_set (PublicationPlace) then if not utilities.is_set (PublisherName) then local date = COinS_date.rftdate and tonumber (COinS_date.rftdate:match ('%d%d%d%d')); -- get year portion of COinS date (because in Arabic numerals); convert string to number if date and (1850 <= date) then -- location has no publisher; if date is 1850 or later utilities.set_message ('maint_location_no_publisher'); -- add maint cat end else -- PublisherName has a value if cfg.keywords_xlate['none'] == PublisherName then -- if that value is 'none' (only for book and encyclopedia citations) PublisherName = ''; -- unset end end end end local ID_list = {}; -- sequence table of rendered identifiers local ID_list_coins = {}; -- table of identifiers and their values from args; key is same as cfg.id_handlers's key local Class = A['Class']; -- arxiv class identifier local ID_support = { {A['ASINTLD'], 'ASIN', 'err_asintld_missing_asin', A:ORIGIN ('ASINTLD')}, {DoiBroken, 'DOI', 'err_doibroken_missing_doi', A:ORIGIN ('DoiBroken')}, {Embargo, 'PMC', 'err_embargo_missing_pmc', A:ORIGIN ('Embargo')}, } ID_list, ID_list_coins = identifiers.identifier_lists_get (args, { DoiBroken = DoiBroken, -- for |doi= ASINTLD = A['ASINTLD'], -- for |asin= Embargo = Embargo, -- for |pmc= Class = Class, -- for |arxiv= CitationClass = config.CitationClass, -- for |arxiv= Year=anchor_year, -- for |isbn= }, ID_support); if 'citation' == config.CitationClass then -- catch ve/citoid malformed book cites in {{citation}} templates for _, id in ipairs (ID_list) do -- search through the ID_list sequence looking for an ISBN identifier string if id:find ('ISBN', 3, true) then -- plain find, start at index 3 ([[ISBN) if utilities.is_set (Periodical) then -- when a |work= alias is set utilities.set_message ('maint_work_isbn'); -- add maint cat end end end end -- Account for the oddities that are {{cite arxiv}}, {{cite biorxiv}}, {{cite citeseerx}}, {{cite medrxiv}}, {{cite ssrn}}, before generation of COinS data. if utilities.in_array (config.CitationClass, whitelist.preprint_template_list_t) then -- |arxiv= or |eprint= required for cite arxiv; |biorxiv=, |citeseerx=, |medrxiv=, |ssrn= required for their templates if not (args[cfg.id_handlers[config.CitationClass:upper()].parameters[1]] or -- can't use ID_list_coins k/v table here because invalid parameters omitted args[cfg.id_handlers[config.CitationClass:upper()].parameters[2]]) then -- which causes unexpected parameter missing error message utilities.set_message ('err_' .. config.CitationClass .. '_missing'); -- add error message end Periodical = ({['arxiv'] = 'arXiv', ['biorxiv'] = 'bioRxiv', ['citeseerx'] = 'CiteSeerX', ['medrxiv'] = 'medRxiv', ['ssrn'] = 'Social Science Research Network'})[config.CitationClass]; end -- Link the title of the work if no |url= was provided, but we have a |pmc= or a |doi= with |doi-access=free if config.CitationClass == "journal" and not utilities.is_set (URL) and not utilities.is_set (TitleLink) and not utilities.in_array (cfg.keywords_xlate[Title], {'off', 'none'}) then -- TODO: remove 'none' once existing citations have been switched to 'off', so 'none' can be used as token for "no title" instead if 'none' ~= cfg.keywords_xlate[auto_select] then -- if auto-linking not disabled if identifiers.auto_link_urls[auto_select] then -- manual selection URL = identifiers.auto_link_urls[auto_select]; -- set URL to be the same as identifier's external link URL_origin = cfg.id_handlers[auto_select:upper()].parameters[1]; -- set URL_origin to parameter name for use in error message if citation is missing a |title= elseif identifiers.auto_link_urls['pmc'] then -- auto-select PMC URL = identifiers.auto_link_urls['pmc']; -- set URL to be the same as the PMC external link if not embargoed URL_origin = cfg.id_handlers['PMC'].parameters[1]; -- set URL_origin to parameter name for use in error message if citation is missing a |title= elseif identifiers.auto_link_urls['doi'] then -- auto-select DOI URL = identifiers.auto_link_urls['doi']; URL_origin = cfg.id_handlers['DOI'].parameters[1]; end end if utilities.is_set (URL) then -- set when using an identifier-created URL if utilities.is_set (AccessDate) then -- |access-date= requires |url=; identifier-created URL is not |url= --#Modified utilities.set_message ('err_accessdate_missing_url', A:ORIGIN('AccessDate')); -- add an error message AccessDate = ''; -- unset end if utilities.is_set (ArchiveURL) then -- |archive-url= requires |url=; identifier-created URL is not |url= --#Modified utilities.set_message ('err_archive_missing_url', {A:ORIGIN('ArchiveURL')}); -- add an error message ArchiveURL = ''; -- unset end end end -- At this point fields may be nil if they weren't specified in the template use. We can use that fact. -- Test if citation has no title if not utilities.is_set (Title) and not utilities.is_set (TransTitle) and not utilities.is_set (ScriptTitle) then -- has special case for cite episode --#Modified Block utilities.set_message (wikidatacite and 'err_wikidata_citation_missing_title' or 'err_citation_missing_title', {'episode' == config.CitationClass and 'series' or 'title'}); end if utilities.in_array (cfg.keywords_xlate[Title], {'off', 'none'}) and utilities.in_array (config.CitationClass, {'journal', 'citation'}) and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)) and ('journal' == Periodical_origin or 'script-journal' == ScriptPeriodical_origin) then -- special case for journal cites Title = ''; -- set title to empty string utilities.set_message ('maint_untitled'); -- add maint cat end if 'journal' == config.CitationClass or ('citation' == config.CitationClass and utilities.is_set (Periodical) and 'journal' == Periodical_origin) then if is_page_art_num (((utilities.is_set (Page) and Page) or (utilities.is_set (Pages) and Pages)) or nil, ID_list_coins['DOI']) then -- does |page(s)= look like it holds an article number utilities.set_message ('maint_page_art_num'); -- add maint cat end end -- COinS metadata (see <http://ocoins.info/>) for automated parsing of citation information. -- handle the oddity that is cite encyclopedia and {{citation |encyclopedia=something}}. Here we presume that -- when Periodical, Title, and Chapter are all set, then Periodical is the book (encyclopedia) title, Title -- is the article title, and Chapter is a section within the article. So, we remap local coins_chapter = Chapter; -- default assuming that remapping not required local coins_title = Title; -- et tu if 'encyclopaedia' == config.CitationClass or ('citation' == config.CitationClass and utilities.is_set (Encyclopedia)) then if utilities.is_set (Chapter) and utilities.is_set (Title) and utilities.is_set (Periodical) then -- if all are used then coins_chapter = Title; -- remap coins_title = Periodical; end end local coins_author = a; -- default for coins rft.au if 0 < #c then -- but if contributor list coins_author = c; -- use that instead end -- this is the function call to COinS() local OCinSoutput = metadata.COinS({ ['Periodical'] = utilities.strip_apostrophe_markup (Periodical), -- no markup in the metadata ['Encyclopedia'] = Encyclopedia, -- just a flag; content ignored by ~/COinS ['Chapter'] = metadata.make_coins_title (coins_chapter, ScriptChapter), -- Chapter and ScriptChapter stripped of bold / italic / accept-as-written markup ['Degree'] = Degree; -- cite thesis only ['Title'] = metadata.make_coins_title (coins_title, ScriptTitle), -- Title and ScriptTitle stripped of bold / italic / accept-as-written markup ['PublicationPlace'] = PublicationPlace, ['Date'] = COinS_date.rftdate, -- COinS_date.* has correctly formatted date values if Date is valid; ['Season'] = COinS_date.rftssn, ['Quarter'] = COinS_date.rftquarter, ['Chron'] = COinS_date.rftchron, ['Series'] = Series, ['Volume'] = Volume, ['Issue'] = Issue, ['ArticleNumber'] = ArticleNumber, ['Pages'] = coins_pages or metadata.get_coins_pages (first_set ({Sheet, Sheets, Page, Pages, At, QuotePage, QuotePages}, 7)), -- pages stripped of external links ['Edition'] = Edition, ['PublisherName'] = PublisherName or Newsgroup, -- any apostrophe markup already removed from PublisherName ['URL'] = first_set ({ChapterURL, URL}, 2), ['Authors'] = coins_author, ['ID_list'] = ID_list_coins, ['RawPage'] = this_page.prefixedText, }, config.CitationClass); -- Account for the oddities that are {{cite arxiv}}, {{cite biorxiv}}, {{cite citeseerx}}, {{cite medrxiv}}, and {{cite ssrn}} AFTER generation of COinS data. if utilities.in_array (config.CitationClass, whitelist.preprint_template_list_t) then -- we have set rft.jtitle in COinS to arXiv, bioRxiv, CiteSeerX, medRxiv, or ssrn now unset so it isn't displayed Periodical = ''; -- periodical not allowed in these templates; if article has been published, use cite journal end -- special case for cite newsgroup. Do this after COinS because we are modifying Publishername to include some static text if 'newsgroup' == config.CitationClass and utilities.is_set (Newsgroup) then PublisherName = utilities.substitute (cfg.messages['newsgroup'], external_link( 'news:' .. Newsgroup, Newsgroup, Newsgroup_origin, nil )); end local Editors; local EditorCount; -- used only for choosing {ed.) or (eds.) annotation at end of editor name-list local Contributors; -- assembled contributors name list local contributor_etal; local Translators; -- assembled translators name list local translator_etal; local t = {}; -- translators list from |translator-lastn= / translator-firstn= pairs t = extract_names (args, 'TranslatorList'); -- fetch translator list from |translatorn= / |translator-lastn=, -firstn=, -linkn=, -maskn= local Interviewers; local interviewers_list = {}; interviewers_list = extract_names (args, 'InterviewerList'); -- process preferred interviewers parameters local interviewer_etal; -- Now perform various field substitutions. -- We also add leading spaces and surrounding markup and punctuation to the -- various parts of the citation, but only when they are non-nil. do local last_first_list; local control = { format = NameListStyle, -- empty string, '&', 'amp', 'and', or 'vanc' maximum = nil, -- as if display-authors or display-editors not set mode = Mode }; do -- do editor name list first because the now unsupported coauthors used to modify control table local display_names, param = display_names_select (cfg.global_cs1_config_t['DisplayEditors'], A['DisplayEditors'], A:ORIGIN ('DisplayEditors'), #e); control.maximum, editor_etal = get_display_names (display_names, #e, 'editors', editor_etal, param); Editors, EditorCount = list_people (control, e, editor_etal); if 1 == EditorCount and (true == editor_etal or 1 < #e) then -- only one editor displayed but includes etal then EditorCount = 2; -- spoof to display (eds.) annotation end end do -- now do interviewers local display_names, param = display_names_select (cfg.global_cs1_config_t['DisplayInterviewers'], A['DisplayInterviewers'], A:ORIGIN ('DisplayInterviewers'), #interviewers_list); control.maximum, interviewer_etal = get_display_names (display_names, #interviewers_list, 'interviewers', interviewer_etal, param); Interviewers = list_people (control, interviewers_list, interviewer_etal); end do -- now do translators local display_names, param = display_names_select (cfg.global_cs1_config_t['DisplayTranslators'], A['DisplayTranslators'], A:ORIGIN ('DisplayTranslators'), #t); control.maximum, translator_etal = get_display_names (display_names, #t, 'translators', translator_etal, param); Translators = list_people (control, t, translator_etal); end do -- now do contributors local display_names, param = display_names_select (cfg.global_cs1_config_t['DisplayContributors'], A['DisplayContributors'], A:ORIGIN ('DisplayContributors'), #c); control.maximum, contributor_etal = get_display_names (display_names, #c, 'contributors', contributor_etal, param); Contributors = list_people (control, c, contributor_etal); end do -- now do authors local display_names, param = display_names_select (cfg.global_cs1_config_t['DisplayAuthors'], A['DisplayAuthors'], A:ORIGIN ('DisplayAuthors'), #a, author_etal); control.maximum, author_etal = get_display_names (display_names, #a, 'authors', author_etal, param); last_first_list = list_people (control, a, author_etal); if utilities.is_set (Authors) then Authors, author_etal = name_has_etal (Authors, author_etal, false, 'authors'); -- find and remove variations on et al. if author_etal then Authors = Authors .. ' ' .. cfg.messages['et al']; -- add et al. to authors parameter end else Authors = last_first_list; -- either an author name list or an empty string end end -- end of do if utilities.is_set (Authors) and utilities.is_set (Collaboration) then Authors = Authors .. ' (' .. Collaboration .. ')'; -- add collaboration after et al. end end local ConferenceFormat = A['ConferenceFormat']; local ConferenceURL = A['ConferenceURL']; ConferenceFormat = style_format (ConferenceFormat, ConferenceURL, 'conference-format', 'conference-url'); Format = style_format (Format, URL, 'format', 'url'); -- special case for chapter format so no error message or cat when chapter not supported if not (utilities.in_array (config.CitationClass, {'web', 'news', 'journal', 'magazine', 'pressrelease', 'podcast', 'newsgroup', 'arxiv', 'biorxiv', 'citeseerx', 'medrxiv', 'ssrn'}) or ('citation' == config.CitationClass and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)) and not utilities.is_set (Encyclopedia))) then ChapterFormat = style_format (ChapterFormat, ChapterURL, 'chapter-format', 'chapter-url'); end if not utilities.is_set (URL) then if utilities.in_array (config.CitationClass, {"web", "podcast", "mailinglist"}) or -- |url= required for {{cite web}}, {{cite podcast}}, and {{cite mailing list}} ('citation' == config.CitationClass and ('website' == Periodical_origin or 'script-website' == ScriptPeriodical_origin)) then -- and required for {{citation}} with |website= or |script-website= if not url_was_archive_today then -- is <URL> empty/missing because it was a deprecated archive.today url utilities.set_message ('err_cite_web_url'); -- no; emit an error message for an empty/ missing |url= end end -- do we have |accessdate= without either |url= or |chapter-url=? if utilities.is_set (AccessDate) and not utilities.is_set (ChapterURL) then -- ChapterURL may be set when URL is not set; --#Modified utilities.set_message ('err_accessdate_missing_url', A:ORIGIN('AccessDate')); AccessDate = ''; end end local UrlStatus = is_valid_parameter_value (A['UrlStatus'], A:ORIGIN('UrlStatus'), cfg.keywords_lists['url-status'], ''); local OriginalURL local OriginalURL_origin local OriginalFormat local OriginalAccess; UrlStatus = UrlStatus:lower(); -- used later when assembling archived text if utilities.is_set ( ArchiveURL ) then if utilities.is_set (ChapterURL) then -- if chapter-url= is set apply archive url to it OriginalURL = ChapterURL; -- save copy of source chapter's url for archive text OriginalURL_origin = ChapterURL_origin; -- name of |chapter-url= parameter for error messages OriginalFormat = ChapterFormat; -- and original |chapter-format= if 'live' ~= UrlStatus then ChapterURL = ArchiveURL -- swap-in the archive's URL ChapterURL_origin = A:ORIGIN('ArchiveURL') -- name of |archive-url= parameter for error messages ChapterFormat = ArchiveFormat or ''; -- swap in archive's format ChapterUrlAccess = nil; -- restricted access levels do not make sense for archived URLs end elseif utilities.is_set (URL) then OriginalURL = URL; -- save copy of original source URL OriginalURL_origin = URL_origin; -- name of URL parameter for error messages OriginalFormat = Format; -- and original |format= OriginalAccess = UrlAccess; if 'live' ~= UrlStatus then -- if URL set then |archive-url= applies to it URL = ArchiveURL -- swap-in the archive's URL URL_origin = A:ORIGIN('ArchiveURL') -- name of archive URL parameter for error messages Format = ArchiveFormat or ''; -- swap in archive's format UrlAccess = nil; -- restricted access levels do not make sense for archived URLs end end elseif utilities.is_set (UrlStatus) then -- if |url-status= is set when |archive-url= is not set utilities.set_message ('maint_url_status'); -- add maint cat end --******* if utilities.in_array (config.CitationClass, {'web', 'news', 'journal', 'magazine', 'pressrelease', 'podcast', 'newsgroup', 'arxiv', 'biorxiv', 'citeseerx', 'medrxiv', 'ssrn'}) or -- if any of the 'periodical' cites except encyclopedia ('citation' == config.CitationClass and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)) and not utilities.is_set (Encyclopedia)) then local chap_param; if utilities.is_set (Chapter) then -- get a parameter name from one of these chapter related meta-parameters chap_param = A:ORIGIN ('Chapter') elseif utilities.is_set (TransChapter) then chap_param = A:ORIGIN ('TransChapter') elseif utilities.is_set (ChapterURL) then chap_param = A:ORIGIN ('ChapterURL') elseif utilities.is_set (ScriptChapter) then chap_param = ScriptChapter_origin; else utilities.is_set (ChapterFormat) chap_param = A:ORIGIN ('ChapterFormat') end if utilities.is_set (chap_param) then -- if we found one utilities.set_message ('err_chapter_ignored', {chap_param}); -- add error message Chapter = ''; -- and set them to empty string to be safe with concatenation TransChapter = ''; ChapterURL = ''; ScriptChapter = ''; ChapterFormat = ''; end else -- otherwise, format chapter / article title local no_quotes = false; -- default assume that we will be quoting the chapter parameter value if utilities.is_set (Contribution) and 0 < #c then -- if this is a contribution with contributor(s) if utilities.in_array (Contribution:lower(), cfg.keywords_lists.contribution) then -- and a generic contribution title no_quotes = true; -- then render it unquoted end end Chapter = format_chapter_title (ScriptChapter, ScriptChapter_origin, Chapter, Chapter_origin, TransChapter, TransChapter_origin, ChapterURL, ChapterURL_origin, no_quotes, ChapterUrlAccess); -- Contribution is also in Chapter if utilities.is_set (Chapter) then Chapter = Chapter .. ChapterFormat ; if 'map' == config.CitationClass and utilities.is_set (TitleType) then Chapter = Chapter .. ' ' .. TitleType; -- map annotation here; not after title end Chapter = Chapter .. sepc .. ' '; elseif utilities.is_set (ChapterFormat) then -- |chapter= not set but |chapter-format= is so ... Chapter = ChapterFormat .. sepc .. ' '; -- ... ChapterFormat has error message, we want to see it end end -- Format main title local plain_title = false; local accept_title; Title, accept_title = utilities.has_accept_as_written (Title, true); -- remove accept-this-as-written markup when it wraps all of <Title> if accept_title and ('' == Title) then -- only support forced empty for now "(())" Title = cfg.messages['notitle']; -- replace by predefined "No title" message -- TODO: utilities.set_message ( 'err_redundant_parameters', ...); -- issue proper error message instead of muting ScriptTitle = ''; -- just mute for now TransTitle = ''; -- just mute for now plain_title = true; -- suppress text decoration for descriptive title utilities.set_message ('maint_untitled'); -- add maint cat end if not accept_title then -- <Title> not wrapped in accept-as-written markup if '...' == Title:sub (-3) then -- if ellipsis is the last three characters of |title= Title = Title:gsub ('(%.%.%.)%.+$', '%1'); -- limit the number of dots to three elseif not mw.ustring.find (Title, '%.%s*%a%.$') and -- end of title is not a 'dot-(optional space-)letter-dot' initialism ... not mw.ustring.find (Title, '%s+%a%.$') then -- ...and not a 'space-letter-dot' initial (''Allium canadense'' L.) Title = mw.ustring.gsub(Title, '%' .. sepc .. '$', ''); -- remove any trailing separator character; sepc and ms.ustring() here for languages that use multibyte separator characters end if utilities.is_set (ArchiveURL) and is_archived_copy (Title) then utilities.set_message ('maint_archived_copy'); -- add maintenance category before we modify the content of Title end if is_generic ('generic_titles', Title) then utilities.set_message ('err_generic_title'); -- set an error message end end if (not plain_title) and (utilities.in_array (config.CitationClass, {'web', 'news', 'journal', 'magazine', 'document', 'pressrelease', 'podcast', 'newsgroup', 'mailinglist', 'interview', 'arxiv', 'biorxiv', 'citeseerx', 'medrxiv', 'ssrn'}) or ('citation' == config.CitationClass and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)) and not utilities.is_set (Encyclopedia)) or ('map' == config.CitationClass and (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical)))) then -- special case for cite map when the map is in a periodical treat as an article Title = kern_quotes (Title); -- if necessary, separate title's leading and trailing quote marks from module provided quote marks Title = utilities.wrap_style ('quoted-title', Title); Title = script_concatenate (Title, ScriptTitle, 'script-title'); -- <bdi> tags, lang attribute, categorization, etc.; must be done after title is wrapped TransTitle = utilities.wrap_style ('trans-quoted-title', TransTitle ); elseif plain_title or ('report' == config.CitationClass) then -- no styling for cite report and descriptive titles (otherwise same as above) Title = script_concatenate (Title, ScriptTitle, 'script-title'); -- <bdi> tags, lang attribute, categorization, etc.; must be done after title is wrapped TransTitle = utilities.wrap_style ('trans-quoted-title', TransTitle ); -- for cite report, use this form for trans-title else Title = utilities.wrap_style ('italic-title', Title); Title = script_concatenate (Title, ScriptTitle, 'script-title'); -- <bdi> tags, lang attribute, categorization, etc.; must be done after title is wrapped TransTitle = utilities.wrap_style ('trans-italic-title', TransTitle); end if utilities.is_set (TransTitle) then if utilities.is_set (Title) then TransTitle = " " .. TransTitle; else utilities.set_message ('err_trans_missing_title', {'title'}); end end if utilities.is_set (Title) then -- TODO: is this the right place to be making Wikisource URLs? if utilities.is_set (TitleLink) and utilities.is_set (URL) then utilities.set_message ('err_wikilink_in_url'); -- set an error message because we can't have both TitleLink = ''; -- unset end if not utilities.is_set (TitleLink) and utilities.is_set (URL) then Title = external_link (URL, Title, URL_origin, UrlAccess) .. TransTitle .. Format; URL = ''; -- unset these because no longer needed Format = ""; elseif utilities.is_set (TitleLink) and not utilities.is_set (URL) then local ws_url; ws_url = wikisource_url_make (TitleLink); -- ignore ws_label return; not used here if ws_url then Title = external_link (ws_url, Title .. '&nbsp;', 'ws link in title-link'); -- space char after Title to move icon away from italic text; TODO: a better way to do this? Title = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], TitleLink, Title}); Title = Title .. TransTitle; else Title = utilities.make_wikilink (TitleLink, Title) .. TransTitle; end else local ws_url, ws_label, L; -- Title has italic or quote markup by the time we get here which causes is_wikilink() to return 0 (not a wikilink) ws_url, ws_label, L = wikisource_url_make (Title:gsub('^[\'"]*(.-)[\'"]*$', '%1')); -- make ws URL from |title= interwiki link (strip italic or quote markup); link portion L becomes tooltip label if ws_url then Title = Title:gsub ('%b[]', ws_label); -- replace interwiki link with ws_label to retain markup Title = external_link (ws_url, Title .. '&nbsp;', 'ws link in title'); -- space char after Title to move icon away from italic text; TODO: a better way to do this? Title = utilities.substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, Title}); Title = Title .. TransTitle; else Title = Title .. TransTitle; end end else Title = TransTitle; end if utilities.is_set (Place) then Place = " " .. wrap_msg ('written', Place, use_lowercase) .. sepc .. " "; end local ConferenceURL_origin = A:ORIGIN('ConferenceURL'); -- get name of parameter that holds ConferenceURL if utilities.is_set (Conference) then if utilities.is_set (ConferenceURL) then Conference = external_link( ConferenceURL, Conference, ConferenceURL_origin, nil ); end Conference = sepc .. " " .. Conference .. ConferenceFormat; elseif utilities.is_set (ConferenceURL) then Conference = sepc .. " " .. external_link( ConferenceURL, nil, ConferenceURL_origin, nil ); end local Position = ''; if not utilities.is_set (Position) then local Minutes = A['Minutes']; local Time = A['Time']; if utilities.is_set (Minutes) then if utilities.is_set (Time) then --TODO: make a function for this and similar? utilities.set_message ('err_redundant_parameters', {utilities.wrap_style ('parameter', 'minutes') .. cfg.presentation['sep_list_pair'] .. utilities.wrap_style ('parameter', 'time')}); end Position = utilities.substitute (cfg.messages.minutes, Minutes); else if utilities.is_set (Time) then local TimeCaption = A['TimeCaption'] if not utilities.is_set (TimeCaption) then TimeCaption = cs2_lower(cfg.messages['event']); end Position = " " .. TimeCaption .. " " .. Time; end end else Position = " " .. Position; At = ''; end Page, Pages, Sheet, Sheets = format_pages_sheets (Page, Pages, Sheet, Sheets, config.CitationClass, Periodical_origin, sepc, NoPP, use_lowercase); At = utilities.is_set (At) and (sepc .. " " .. At) or ""; Position = utilities.is_set (Position) and (sepc .. " " .. Position) or ""; if config.CitationClass == 'map' then local Sections = A['Sections']; -- Section (singular) is an alias of Chapter so set earlier local Inset = A['Inset']; if utilities.is_set ( Inset ) then Inset = sepc .. " " .. wrap_msg ('inset', Inset, use_lowercase); end if utilities.is_set ( Sections ) then Section = sepc .. " " .. wrap_msg ('sections', Sections, use_lowercase); elseif utilities.is_set ( Section ) then Section = sepc .. " " .. wrap_msg ('section', Section, use_lowercase); end At = At .. Inset .. Section; end local Others = A['Others']; if utilities.is_set (Others) and 0 == #a and 0 == #e then -- add maint cat when |others= has value and used without |author=, |editor= if config.CitationClass == "AV-media-notes" or config.CitationClass == "audio-visual" then -- special maint for AV/M which has a lot of 'false' positives right now utilities.set_message ('maint_others_avm') else utilities.set_message ('maint_others'); end end Others = utilities.is_set (Others) and (sepc .. " " .. Others) or ""; if utilities.is_set (Translators) then Others = safe_join ({sepc .. ' ', wrap_msg ('translated', Translators, use_lowercase), Others}, sepc); end if utilities.is_set (Interviewers) then Others = safe_join ({sepc .. ' ', wrap_msg ('interview', Interviewers, use_lowercase), Others}, sepc); end local TitleNote = A['TitleNote']; TitleNote = utilities.is_set (TitleNote) and (sepc .. " " .. TitleNote) or ""; if utilities.is_set (Edition) then if Edition:match ('%f[%a][Ee]d%n?%.?$') or Edition:match ('%f[%a][Ee]dition$') then -- Ed, ed, Ed., ed., Edn, edn, Edn., edn. utilities.set_message ('err_extra_text_edition'); -- add error message end Edition = " " .. wrap_msg ('edition', Edition); else Edition = ''; end Series = utilities.is_set (Series) and wrap_msg ('series', {sepc, Series}) or ""; -- not the same as SeriesNum local Agency = A['Agency'] or ''; -- |agency= is supported by {{cite magazine}}, {{cite news}}, {{cite press release}}, {{cite web}}, and certain {{citation}} templates if utilities.is_set (Agency) then -- this testing done here because {{citation}} supports 'news' citations if utilities.in_array (config.CitationClass, {'magazine', 'news', 'pressrelease', 'web'}) or ('citation' == config.CitationClass and utilities.in_array (Periodical_origin, {"magazine", "newspaper", "work"})) then Agency = wrap_msg ('agency', {sepc, Agency}); -- format for rendering else Agency = ''; -- unset; not supported utilities.set_message ('err_parameter_ignored', {'agency'}); -- add error message end end Volume = format_volume_issue (Volume, Issue, ArticleNumber, config.CitationClass, Periodical_origin, sepc, use_lowercase); if utilities.is_set (AccessDate) then local retrv_text = " " .. cfg.messages['retrieved']; AccessDate = nowrap_date (AccessDate); -- wrap in nowrap span if date in appropriate format retrv_text = cs2_lower(retrv_text); -- if mode is cs2, lower case AccessDate = utilities.substitute (retrv_text, AccessDate); -- add retrieved text AccessDate = utilities.substitute (cfg.presentation['accessdate'], {sepc, AccessDate}); -- allow editors to hide accessdates end if utilities.is_set (ID) then ID = sepc .. " " .. ID; end local Docket = A['Docket']; if "thesis" == config.CitationClass and utilities.is_set (Docket) then ID = sepc .. " Docket " .. Docket .. ID; end if "report" == config.CitationClass and utilities.is_set (Docket) then -- for cite report when |docket= is set ID = sepc .. ' ' .. Docket; -- overwrite ID even if |id= is set end if utilities.is_set (URL) then URL = " " .. external_link( URL, nil, URL_origin, UrlAccess ); end -- We check length of PostScript here because it will have been nuked by -- the quote parameters. We'd otherwise emit a message even if there wasn't -- a displayed postscript. -- TODO: Should the max size (1) be configurable? -- TODO: Should we check a specific pattern? if utilities.is_set(PostScript) and mw.ustring.len(PostScript) > 1 then utilities.set_message ('maint_postscript') end local Archived; if utilities.is_set (ArchiveURL) then if not utilities.is_set (ArchiveDate) then -- ArchiveURL set but ArchiveDate not set --#Modified utilities.set_message ('err_archive_missing_date', {A:ORIGIN('ArchiveURL')}); ArchiveURL = ''; -- empty string for concatenation ArchiveDate = ''; -- empty string for concatenation end else if utilities.is_set (ArchiveDate) then -- ArchiveURL not set but ArchiveDate is set --#Modified utilities.set_message ('err_archive_date_missing_url', {A:ORIGIN('ArchiveDate')}); -- emit an error message ArchiveURL = ''; -- empty string for concatenation ArchiveDate = ''; -- empty string for concatenation end end if utilities.is_set (ArchiveURL) then local arch_text; if "live" == UrlStatus then if utilities.is_set (ArchiveDate) then Archived = sepc .. ' ' .. utilities.substitute ( cfg.messages['archived-live'], { external_link( ArchiveURL, cs2_lower(cfg.messages['archived']), A:ORIGIN('ArchiveURL'), nil) .. ArchiveFormat, ArchiveDate } ); else Archived = ''; end if not utilities.is_set (OriginalURL) then --#Modified utilities.set_message ('err_archive_missing_url', {A:ORIGIN('ArchiveURL')}); Archived = ''; -- empty string for concatenation end elseif utilities.is_set (OriginalURL) then -- UrlStatus is empty, 'dead', 'unfit', 'usurped', 'bot: unknown' if utilities.in_array (UrlStatus, {'unfit', 'usurped', 'bot: unknown'}) then Archived = sepc .. ' ' .. utilities.substitute ( cs2_lower(cfg.messages['archived-unfit']), ArchiveDate); if 'bot: unknown' == UrlStatus then utilities.set_message ('maint_bot_unknown'); -- and add a category if not already added else utilities.add_prop_cat ('unfit'); -- and add a category if not already added end else -- UrlStatus is empty, 'dead' if utilities.is_set (ArchiveDate) then Archived = sepc .. ' ' .. utilities.substitute ( cs2_lower(cfg.messages['archived-dead']), { external_link( OriginalURL, cfg.messages['original'], OriginalURL_origin, OriginalAccess ) .. OriginalFormat, ArchiveDate } ); else Archived = ''; -- unset for concatenation end end else -- OriginalUrl not set --#Modified utilities.set_message ('err_archive_missing_url', {A:ORIGIN('ArchiveURL')}); Archived = ''; -- empty string for concatenation end elseif utilities.is_set (ArchiveFormat) then Archived = ArchiveFormat; -- if set and ArchiveURL not set ArchiveFormat has error message else Archived = ''; end local TranscriptURL = A['TranscriptURL'] local TranscriptFormat = A['TranscriptFormat']; TranscriptFormat = style_format (TranscriptFormat, TranscriptURL, 'transcript-format', 'transcripturl'); local Transcript = A['Transcript']; local TranscriptURL_origin = A:ORIGIN('TranscriptURL'); -- get name of parameter that holds TranscriptURL if utilities.is_set (Transcript) then if utilities.is_set (TranscriptURL) then Transcript = external_link( TranscriptURL, Transcript, TranscriptURL_origin, nil ); end Transcript = sepc .. ' ' .. Transcript .. TranscriptFormat; elseif utilities.is_set (TranscriptURL) then Transcript = external_link( TranscriptURL, nil, TranscriptURL_origin, nil ); end local Publisher; if utilities.is_set (PublicationDate) then PublicationDate = wrap_msg ('published', PublicationDate); end if utilities.is_set (PublisherName) then if utilities.is_set (PublicationPlace) then Publisher = sepc .. " " .. PublicationPlace .. ": " .. PublisherName .. PublicationDate; else Publisher = sepc .. " " .. PublisherName .. PublicationDate; end elseif utilities.is_set (PublicationPlace) then Publisher= sepc .. " " .. PublicationPlace .. PublicationDate; else Publisher = PublicationDate; end -- Several of the above rely upon detecting this as nil, so do it last. if (utilities.is_set (Periodical) or utilities.is_set (ScriptPeriodical) or utilities.is_set (TransPeriodical)) then if utilities.is_set (Title) or utilities.is_set (TitleNote) then Periodical = sepc .. " " .. format_periodical (ScriptPeriodical, ScriptPeriodical_origin, Periodical, TransPeriodical, TransPeriodical_origin); else Periodical = format_periodical (ScriptPeriodical, ScriptPeriodical_origin, Periodical, TransPeriodical, TransPeriodical_origin); end end local Language = A['Language']; if utilities.is_set (Language) then Language = language_parameter (Language); -- format, categories, name from ISO639-1, etc. else Language=''; -- language not specified so make sure this is an empty string; --[[ TODO: need to extract the wrap_msg from language_parameter so that we can solve parentheses bunching problem with Format/Language/TitleType ]] end --[[ Handle the oddity that is cite speech. This code overrides whatever may be the value assigned to TitleNote (through |department=) and forces it to be " (Speech)" so that the annotation directly follows the |title= parameter value in the citation rather than the |event= parameter value (if provided). ]] if "speech" == config.CitationClass then -- cite speech only TitleNote = TitleType; -- move TitleType to TitleNote so that it renders ahead of |event= TitleType = ''; -- and unset if utilities.is_set (Periodical) then -- if Periodical, perhaps because of an included |website= or |journal= parameter if utilities.is_set (Conference) then -- and if |event= is set Conference = Conference .. sepc .. " "; -- then add appropriate punctuation to the end of the Conference variable before rendering end end end -- Piece all bits together at last. Here, all should be non-nil. -- We build things this way because it is more efficient in LUA -- not to keep reassigning to the same string variable over and over. local tcommon; local tcommon2; -- used for book cite when |contributor= is set local pgtext = Position .. Sheet .. Sheets .. Page .. Pages .. At; -- Volume = safe_join ({Volume, pgtext}, sepc); if utilities.in_array (config.CitationClass, {"book", "citation"}) and not utilities.is_set (Periodical) then -- special cases for book cites if utilities.is_set (Contributors) then -- when we are citing foreword, preface, introduction, etc. tcommon = safe_join ({Title, TitleNote}, sepc); -- author and other stuff will come after this and before tcommon2 --#Mdified Arabic style make vol last tcommon2 = safe_join({TitleType, Series, Language, Others, Edition, Publisher, Volume}, sepc); else --#Mdified Arabic style make vol last tcommon = safe_join( {Title, TitleNote, TitleType, Series, Language, Others, Edition, Publisher, Volume}, sepc); end elseif 'map' == config.CitationClass then -- special cases for cite map if utilities.is_set (Chapter) then -- map in a book; TitleType is part of Chapter tcommon = safe_join ({Title, Edition, Scale, Series, Language, Cartography, Others, Publisher, Volume}, sepc); elseif utilities.is_set (Periodical) then -- map in a periodical tcommon = safe_join ({Title, TitleType, Periodical, Scale, Series, Language, Cartography, Others, Publisher, Volume}, sepc); else -- a sheet or stand-alone map tcommon = safe_join ({Title, TitleType, Edition, Scale, Series, Language, Cartography, Others, Publisher}, sepc); end elseif 'episode' == config.CitationClass then -- special case for cite episode tcommon = safe_join ({Title, TitleNote, TitleType, Series, Language, Edition, Publisher}, sepc); else -- all other CS1 templates --#Mdified Arabic style make vol last tcommon = safe_join ({Title, TitleNote, Conference, Periodical, TitleType, Series, Language, Others, Edition, Publisher, Agency, Volume}, sepc); end if #ID_list > 0 then ID_list = safe_join( { sepc .. " ", table.concat( ID_list, sepc .. " " ), ID }, sepc ); else ID_list = ID; end local Via = A['Via']; Via = utilities.is_set (Via) and wrap_msg ('via', Via) or ''; local idcommon; if 'audio-visual' == config.CitationClass or 'episode' == config.CitationClass then -- special case for cite AV media & cite episode position transcript idcommon = safe_join( { ID_list, URL, Archived, Transcript, AccessDate, Via, Quote }, sepc ); else idcommon = safe_join( { ID_list, URL, Archived, AccessDate, Via, Quote }, sepc ); end local text; -- local pgtext = Position .. Sheet .. Sheets .. Page .. Pages .. At; --#Added block if utilities.is_set (HijriDate) then if not mw.ustring.match(HijriDate,'%d *هـ?$') then HijriDate = HijriDate .. 'هـ' end if utilities.is_set (Date) then Date = HijriDate .. ' – ' .. Date .. 'م' else Date = HijriDate end end local OrigDate = A['OrigDate']; OrigDate = utilities.is_set (OrigDate) and wrap_msg ('origdate', OrigDate) or ''; if utilities.is_set (Date) then if utilities.is_set (Authors) or utilities.is_set (Editors) then -- date follows authors or editors when authors not set Date = " (" .. Date .. ")" .. OrigDate .. sepc .. " "; -- in parentheses else -- neither of authors and editors set if (string.sub(tcommon, -1, -1) == sepc) then -- if the last character of tcommon is sepc Date = " " .. Date .. OrigDate; -- Date does not begin with sepc else Date = sepc .. " " .. Date .. OrigDate; -- Date begins with sepc end end end if utilities.is_set (Authors) then if (not utilities.is_set (Date)) then -- when date is set it's in parentheses; no Authors termination Authors = terminate_name_list (Authors, sepc); -- when no date, terminate with 0 or 1 sepc and a space end if utilities.is_set (Editors) then local in_text = ''; local post_text = ''; if utilities.is_set (Chapter) and 0 == #c then in_text = cfg.messages['in'] .. ' '; in_text = cs2_lower(in_text); -- lowercase for cs2 end if EditorCount <= 1 then post_text = ' (' .. cfg.messages['editor'] .. ')'; -- be consistent with no-author, no-date case else post_text = ' (' .. cfg.messages['editors'] .. ')'; end Editors = terminate_name_list (in_text .. Editors .. post_text, sepc); -- terminate with 0 or 1 sepc and a space end if utilities.is_set (Contributors) then -- book cite and we're citing the intro, preface, etc. local by_text = sepc .. ' ' .. cfg.messages['by'] .. ' '; by_text = cs2_lower(by_text); -- lowercase for cs2 Authors = by_text .. Authors; -- author follows title so tweak it here if utilities.is_set (Editors) and utilities.is_set (Date) then -- when Editors make sure that Authors gets terminated Authors = terminate_name_list (Authors, sepc); -- terminate with 0 or 1 sepc and a space end if (not utilities.is_set (Date)) then -- when date is set it's in parentheses; no Contributors termination Contributors = terminate_name_list (Contributors, sepc); -- terminate with 0 or 1 sepc and a space end text = safe_join( {Contributors, Date, Chapter, tcommon, Authors, Place, Editors, tcommon2, pgtext, idcommon }, sepc ); -- text = safe_join( {Contributors, Date, Chapter, tcommon, Authors, Place, Editors, tcommon2, idcommon }, sepc ); else text = safe_join( {Authors, Date, Chapter, Place, Editors, tcommon, pgtext, idcommon }, sepc ); -- text = safe_join( {Authors, Date, Chapter, Place, Editors, tcommon, idcommon }, sepc ); end elseif utilities.is_set (Editors) then if utilities.is_set (Date) then if EditorCount <= 1 then Editors = Editors .. cfg.presentation['sep_name'] .. cfg.messages['editor']; else Editors = Editors .. cfg.presentation['sep_name'] .. cfg.messages['editors']; end else if EditorCount <= 1 then Editors = Editors .. " (" .. cfg.messages['editor'] .. ")" .. sepc .. " " else Editors = Editors .. " (" .. cfg.messages['editors'] .. ")" .. sepc .. " " end end text = safe_join( {Editors, Date, Chapter, Place, tcommon, pgtext, idcommon}, sepc ); -- text = safe_join( {Editors, Date, Chapter, Place, tcommon, idcommon}, sepc ); else if utilities.in_array (config.CitationClass, {"journal", "citation"}) and utilities.is_set (Periodical) then text = safe_join( {Chapter, Place, tcommon, pgtext, Date, idcommon}, sepc ); -- text = safe_join( {Chapter, Place, tcommon, Date, idcommon}, sepc ); else text = safe_join( {Chapter, Place, tcommon, Date, pgtext, idcommon}, sepc ); -- text = safe_join( {Chapter, Place, tcommon, Date, idcommon}, sepc ); end end if utilities.is_set (PostScript) and PostScript ~= sepc then text = safe_join( {text, sepc}, sepc ); -- Deals with italics, spaces, etc. if '.' == sepc then -- remove final seperator if present text = text:gsub ('%' .. sepc .. '$', ''); -- dot must be escaped here else text = mw.ustring.gsub (text, sepc .. '$', ''); -- using ustring for non-dot sepc (likely a non-Latin character) end end text = safe_join( {text, PostScript}, sepc ); -- Now enclose the whole thing in a <cite> element local options_t = {}; options_t.class = cite_class_attribute_make (config.CitationClass, Mode); local Ref = is_valid_parameter_value (A['Ref'], A:ORIGIN('Ref'), cfg.keywords_lists['ref'], nil, true); -- nil when |ref=harv; A['Ref'] else --#Added Block auto add CITEREF if Ref and mw.ustring.sub(Ref,1,7) ~= 'CITEREF' then Ref = 'CITEREF' .. Ref end if 'none' ~= cfg.keywords_xlate[(Ref and Ref:lower()) or ''] then local namelist_t = {}; -- holds selected contributor, author, editor name list local year = first_set ({Year, anchor_year}, 2); -- Year first for legacy citations and for YMD dates that require disambiguation if #c > 0 then -- if there is a contributor list namelist_t = c; -- select it elseif #a > 0 then -- or an author list namelist_t = a; elseif #e > 0 then -- or an editor list namelist_t = e; end local citeref_id; if #namelist_t > 0 then -- if there are names in namelist_t citeref_id = make_citeref_id (namelist_t, year); -- go make the CITEREF anchor if mw.uri.anchorEncode (citeref_id) == ((Ref and mw.uri.anchorEncode (Ref)) or '') then -- Ref may already be encoded (by {{sfnref}}) so citeref_id must be encoded before comparison utilities.set_message ('maint_ref_duplicates_default'); end else citeref_id = ''; -- unset end options_t.id = Ref or citeref_id; end if string.len (text:gsub('%b<>', '')) <= 2 then -- remove html and html-like tags; then get length of what remains; z.error_cats_t = {}; -- blank the categories list z.error_msgs_t = {}; -- blank the error messages list OCinSoutput = nil; -- blank the metadata string text = ''; -- blank the the citation utilities.set_message ('err_empty_citation'); -- set empty citation message and category end local render_t = {}; -- here we collect the final bits for concatenation into the rendered citation if is_preview_mode and (0 ~= #z.error_msgs_t or 0 ~= #z.maint_cats_t) then -- when in preview mode if not utilities.is_set (options_t.id) then -- and when this template has not created an anchor ID math.randomseed (tonumber (mw.getLanguage ('en'):formatDate ('U')) * 10000 + os.clock() * 10000); options_t.id = 'CITEREF_temp_preview_id_' .. math.random (100000, 999999); -- create temporary anchor ID for preview mode end end if utilities.is_set (options_t.id) then -- here we wrap the rendered citation in <cite ...>...</cite> tags options_t.id = mw.uri.anchorEncode(options_t.id); -- do this here; also used when adding preview warning messages table.insert (render_t, utilities.substitute (cfg.presentation['cite-id'], {options_t.id, mw.text.nowiki(options_t.class), text})); -- when |ref= is set or when there is a namelist else table.insert (render_t, utilities.substitute (cfg.presentation['cite'], {mw.text.nowiki(options_t.class), text})); -- when |ref=none or when namelist_t empty and |ref= is missing or is empty end if OCinSoutput then -- blanked when citation is 'empty' so don't bother to add boilerplate metadata span table.insert (render_t, utilities.substitute (cfg.presentation['ocins'], OCinSoutput)); -- format and append metadata to the citation end local template_name = cfg.citation_class_map_t[config.CitationClass]; local template_link = '[[قالب:' .. template_name .. '|' .. template_name .. ']]'; local msg_prefix = '<code class="cs1-code">{{' .. template_link .. '}}</code>: '; if 0 ~= #z.error_msgs_t then mw.addWarning (utilities.substitute (cfg.messages.warning_msg_e, {template_link, options_t.id})); table.insert (render_t, ' '); -- insert a space between citation and its error messages table.sort (z.error_msgs_t); -- sort the error messages list; sorting includes wrapping <span> and <code> tags; hidden-error sorts ahead of visible-error local hidden = true; -- presume that the only error messages emited by this template are hidden for _, v in ipairs (z.error_msgs_t) do -- spin through the list of error messages if v:find ('cs1-visible-error', 1, true) then -- look for the visible error class name hidden = false; -- found one; so don't hide the error message prefix break; -- and done because no need to look further end end z.error_msgs_t[1] = table.concat ({utilities.error_comment (msg_prefix, hidden), z.error_msgs_t[1]}); -- add error message prefix to first error message to prevent extraneous punctuation --#Modified table.insert (render_t, utilities.make_sep_error_list(#z.error_msgs_t, z.error_msgs_t)); -- make a big string of error messages and add it to the rendering end if 0 ~= #z.maint_cats_t then mw.addWarning (utilities.substitute (cfg.messages.warning_msg_m, {template_link, options_t.id})); table.sort (z.maint_cats_t); -- sort the maintenance messages list local maint_msgs_t = {}; -- here we collect all of the maint messages if 0 == #z.error_msgs_t then -- if no error messages table.insert (maint_msgs_t, msg_prefix); -- insert message prefix in maint message livery end for _, v in ipairs( z.maint_cats_t ) do -- append maintenance categories table.insert (maint_msgs_t, -- assemble new maint message and add it to the maint_msgs_t table table.concat ({v, ' (', utilities.substitute (cfg.messages[':cat wikilink'], v), ')'}) ); end table.insert (render_t, utilities.substitute (cfg.presentation['hidden-maint'], table.concat (maint_msgs_t, ' '))); -- wrap the group of maint messages with proper presentation and save end if not no_tracking_cats then local sort_key; local cat_wikilink = 'cat wikilink'; if cfg.enable_sort_keys then -- when namespace sort keys enabled local namespace_number = mw.title.getCurrentTitle().namespace; -- get namespace number for this wikitext sort_key = (0 ~= namespace_number and (cfg.name_space_sort_keys[namespace_number] or cfg.name_space_sort_keys.other)) or nil; -- get sort key character; nil for mainspace cat_wikilink = (not sort_key and 'cat wikilink') or 'cat wikilink sk'; -- make <cfg.messages> key end for _, v in ipairs (z.error_cats_t) do -- append error categories table.insert (render_t, utilities.substitute (cfg.messages[cat_wikilink], {v, sort_key})); end if cfg.id_limits_data_load_fail then -- boolean true when load failed utilities.set_message ('maint_id_limit_load_fail'); -- done here because this maint cat emits no message end for _, v in ipairs (z.maint_cats_t) do -- append maintenance categories table.insert (render_t, utilities.substitute (cfg.messages[cat_wikilink], {v, sort_key})); end for _, v in ipairs (z.prop_cats_t) do -- append properties categories table.insert (render_t, utilities.substitute (cfg.messages['cat wikilink'], v)); -- no sort keys end end return table.concat (render_t); -- make a big string and done end --[[--------------------------< V A L I D A T E >-------------------------------------------------------------- Looks for a parameter's name in one of several whitelists. Parameters in the whitelist can have three values: true - active, supported parameters false - deprecated, supported parameters nil - unsupported parameters ]] local function validate (name, cite_class, empty) local name = tostring (name); local enum_name; -- parameter name with enumerator (if any) replaced with '#' local state; local function state_test (state, name) -- local function to do testing of state values if true == state then return true; end -- valid actively supported parameter if false == state then if empty then return nil; end -- empty deprecated parameters are treated as unknowns deprecated_parameter (name); -- parameter is deprecated but still supported return true; end if 'tracked' == state then local base_name = name:gsub ('%d', ''); -- strip enumerators from parameter names that have them to get the base name utilities.add_prop_cat ('tracked-param', {base_name}, base_name); -- add a properties category; <base_name> modifies <key> return true; end return nil; end if name:find ('#') then -- # is a cs1|2 reserved character so parameters with # not permitted return nil; end -- replace enumerator digit(s) with # (|last25= becomes |last#=) (mw.ustring because non-Western 'local' digits) enum_name = mw.ustring.gsub (name, '%d+$', '#'); -- where enumerator is last charaters in parameter name (these to protect |s2cid=) enum_name = mw.ustring.gsub (enum_name, '%d+([%-l])', '#%1'); -- where enumerator is in the middle of the parameter name; |author#link= is the oddity if 'document' == cite_class then -- special case for {{cite document}} state = whitelist.document_parameters_t[enum_name]; -- this list holds enumerated and nonenumerated parameters if true == state_test (state, name) then return true; end return false; end if utilities.in_array (cite_class, whitelist.preprint_template_list_t) then -- limited parameter sets allowed for these templates state = whitelist.limited_parameters_t[enum_name]; -- this list holds enumerated and nonenumerated parameters if true == state_test (state, name) then return true; end state = whitelist.preprint_arguments_t[cite_class][name]; -- look in the parameter-list for the template identified by cite_class if true == state_test (state, name) then return true; end return false; -- not supported because not found or name is set to nil end -- end limited parameter-set templates if utilities.in_array (cite_class, whitelist.unique_param_template_list_t) then -- template-specific parameters for templates that accept parameters from the basic argument list state = whitelist.unique_arguments_t[cite_class][name]; -- look in the template-specific parameter-lists for the template identified by cite_class if true == state_test (state, name) then return true; end end -- if here, fall into general validation state = whitelist.common_parameters_t[enum_name]; -- all other templates; all normal parameters allowed; this list holds enumerated and nonenumerated parameters if true == state_test (state, name) then return true; end return false; -- not supported because not found or name is set to nil end --[=[-------------------------< I N T E R _ W I K I _ C H E C K >---------------------------------------------- check <value> for inter-language interwiki-link markup. <prefix> must be a MediaWiki-recognized language code. when these values have the form (without leading colon): [[<prefix>:link|label]] return label as plain-text [[<prefix>:link]] return <prefix>:link as plain-text return value as is else ]=] local function inter_wiki_check (parameter, value) local prefix = value:match ('%[%[(%a+):'); -- get an interwiki prefix if one exists local _; if prefix and cfg.inter_wiki_map[prefix:lower()] then -- if prefix is in the map, needs preceding colon so utilities.set_message ('err_bad_paramlink', parameter); -- emit an error message _, value, _ = utilities.is_wikilink (value); -- extract label portion from wikilink end return value; end --[[--------------------------< M I S S I N G _ P I P E _ C H E C K >------------------------------------------ Look at the contents of a parameter. If the content has a string of characters and digits followed by an equal sign, compare the alphanumeric string to the list of cs1|2 parameters. If found, then the string is possibly a parameter that is missing its pipe. There are two tests made: {{cite ... |title=Title access-date=2016-03-17}} -- the first parameter has a value and whitespace separates that value from the missing pipe parameter name {{cite ... |title=access-date=2016-03-17}} -- the first parameter has no value (whitespace after the first = is trimmed by MediaWiki) cs1|2 shares some parameter names with XML/HTML attributes: class=, title=, etc. To prevent false positives XML/HTML tags are removed before the search. If a missing pipe is detected, this function adds the missing pipe maintenance category. ]] local function missing_pipe_check (parameter, value) local capture; value = value:gsub ('%b<>', ''); -- remove XML/HTML tags because attributes: class=, title=, etc. capture = value:match ('%s+(%a[%w%-]+)%s*=') or value:match ('^(%a[%w%-]+)%s*='); -- find and categorize parameters with possible missing pipes if capture and validate (capture) then -- if the capture is a valid parameter name utilities.set_message ('err_missing_pipe', parameter); end end --[[--------------------------< H A S _ E X T R A N E O U S _ P U N C T >-------------------------------------- look for extraneous terminal punctuation in most parameter values; parameters listed in skip table are not checked ]] local function has_extraneous_punc (param, value) if 'number' == type (param) then return; end param = param:gsub ('%d+', '#'); -- enumerated name-list mask params allow terminal punct; normalize if cfg.punct_skip[param] then return; -- parameter name found in the skip table so done end if value:match ('[,;:]$') then utilities.set_message ('maint_extra_punct'); -- has extraneous punctuation; add maint cat end if value:match ('^=') then -- sometimes an extraneous '=' character appears ... utilities.set_message ('maint_extra_punct'); -- has extraneous punctuation; add maint cat end end --[[--------------------------< H A S _ T W L _ U R L >-------------------------------------------------------- look for The Wikipedia Library urls in url-holding parameters. TWL urls are accessible only for readers who are active extended confirmed Wikipedia editors. This function sets an error message when such urls are discovered and when appropriate, sets the |<param>-url-access=subscription. returns nothing. looks for: '.wikipedialibrary.idm.oclc.org' ]] local function has_twl_url (url_params_t, cite_args_t) local url_error_t = {}; -- sequence of url-holding parameters that have a TWL url for param, value in pairs (url_params_t) do if value:find ('%.wikipedialibrary%.idm%.oclc%.org') then -- has the TWL base url? table.insert (url_error_t, param); -- add parameter name to the error list end end if 0 ~= #url_error_t then -- non-zero when there are errors table.sort (url_error_t); -- sor for error messaging for i, param in ipairs (url_error_t) do if cfg.url_access_map_t[param] then -- if <param> has a matching -access parameter cite_args_t[cfg.url_access_map_t[param]] = cfg.keywords_xlate.subscription; -- set |<param>-url-access=subscription end url_error_t[i] = utilities.wrap_style ('parameter', param); -- make the parameter pretty for error message end utilities.set_message ('err_param_has_twl_url', {utilities.make_sep_list (#url_error_t, url_error_t)}); -- add this error message end end --[[--------------------------< H A S _ E X T R A N E O U S _ U R L >------------------------------------------ look for extraneous url parameter values; parameters listed in skip table are not checked ]] local function has_extraneous_url (non_url_param_t) local url_error_t = {}; check_for_url (non_url_param_t, url_error_t); -- extraneous url check if 0 ~= #url_error_t then -- non-zero when there are errors table.sort (url_error_t); utilities.set_message ('err_param_has_ext_link', {utilities.make_sep_list (#url_error_t, url_error_t)}); -- add this error message end end --[[--------------------------< _ C I T A T I O N >------------------------------------------------------------ Module entry point frame – from template call (citation()); may be nil when called from another module args_t – table of all cs1|2 parameters in the template (the parent frame) config_t – table of template-supplied parameter (the #invoke frame) ]] local function _citation (frame, args_t, config_t) -- save a copy in case we need to display an error message in preview mode if not frame then frame = mw.getCurrentFrame(); -- if called from another module, get a frame for frame-provided functions end -- i18n: set the name that your wiki uses to identify sandbox subpages from sandbox template invoke (or can be set here) --#Modified line local sandbox = ((config_t.SandboxPath and '' ~= config_t.SandboxPath) and config_t.SandboxPath) or '/ملعب'; -- sandbox path from {{#invoke:Citation/CS1/sandbox|citation|SandboxPath=/...}} is_sandbox = nil ~= string.find (frame:getTitle(), sandbox, 1, true); -- is this invoke the sandbox module? sandbox = is_sandbox and sandbox or ''; -- use i18n sandbox to load sandbox modules when this module is the sandox; live modules else --#Added for ltr local tmp_cfg; tmp_cfg = mw.loadData ('Module:Citation/CS1/Configuration' .. sandbox); -- load sandbox versions of support modules when {{#invoke:Citation/CS1/sandbox|...}}; live modules else whitelist = mw.loadData ('Module:Citation/CS1/Whitelist' .. sandbox); utilities = require ('Module:Citation/CS1/Utilities' .. sandbox); validation = require ('Module:Citation/CS1/Date_validation' .. sandbox); identifiers = require ('Module:Citation/CS1/Identifiers' .. sandbox); metadata = require ('Module:Citation/CS1/COinS' .. sandbox); --#Added block for ltr for k,v in pairs(tmp_cfg) do cfg[k] = v end utilities.set_selected_modules (cfg); -- so that functions in Utilities can see the selected cfg tables identifiers.set_selected_modules (cfg, utilities); -- so that functions in Identifiers can see the selected cfg tables and selected Utilities module validation.set_selected_modules (cfg, utilities); -- so that functions in Date validataion can see selected cfg tables and the selected Utilities module metadata.set_selected_modules (cfg, utilities); -- so that functions in COinS can see the selected cfg tables and selected Utilities module z = utilities.z; -- table of error and category tables in Module:Citation/CS1/Utilities local cite_args_t = {}; -- because args_t is the parent (template) frame args (which cannot be modified); params and their values will be placed here is_preview_mode = not utilities.is_set (frame:preprocess ('{{REVISIONID}}')); local suggestions = {}; -- table where we store suggestions if we need to loadData them local error_text; -- used as a flag --#Added block for filtering some parameters if args_t['wikidatacite'] == '1' then wikidatacite = true args_t['wikidatacite'] = nil end --------------------- --#Added block for ltr local ch_lang = args_t['اللغة'] or args_t['لغة'] or args_t['language'] or args_t['lang'] --# Added by Mr. Ibrahem to stop ltr formatting local no_ltr = args_t['noltr'] if ch_lang and ch_lang ~= '' and no_ltr == nil then local isltr = true local rtl_lang = { 'ar', 'العربية', 'arabic', 'fa', 'الفارسية', 'persian', 'ur', 'urdu', 'الأردية', 'ku', 'kurdish', 'الكردية', 'he', 'hebrew', 'arc', 'aramaic', 'dv', 'divehi', 'ha', 'hausa', 'khw', 'khowar', 'ks', 'kashmiri', 'ps', 'pashto', 'yi', 'yiddish', 'ota', -- تركية عثمانية } local ch_langs = mw.text.split(ch_lang, '[,، ]') for k, v in pairs(ch_langs) do for k2, v2 in pairs(rtl_lang) do if (mw.ustring.lower(v) == v2) then isltr = false break end end if (not isltr) then break end end if isltr then local cfg2 cfg2 = mw.loadData('Module:Citation/CS1/Configuration ltr' .. sandbox); for k, v in pairs(cfg2) do cfg[k] = v end end end ---------------------- local capture; -- the single supported capture when matching unknown parameters using patterns local empty_unknowns = {}; -- sequence table to hold empty unknown params for error message listing for k, v in pairs (args_t) do -- get parameters from the parent (template) frame v = mw.ustring.gsub (v, '^%s*(.-)%s*$', '%1'); -- trim leading/trailing whitespace; when v is only whitespace, becomes empty string if v ~= '' then if cfg.enum_needs_xlation and 'string' == type (k) then -- for wikis that set date_names['local_digits'] to non-western digits k = mw.ustring.gsub (k, '%d', cfg.date_names.local_digits); -- translate 'local' digits enumerators to Western 0-9 end if not validate( k, config_t.CitationClass ) then if type (k) ~= 'string' then -- exclude empty numbered parameters if v:match("%S+") ~= nil then error_text = utilities.set_message ('err_text_ignored', {v}); end elseif validate (k:lower(), config_t.CitationClass) then error_text = utilities.set_message ('err_parameter_ignored_suggest', {k, k:lower()}); -- suggest the lowercase version of the parameter else if nil == suggestions.suggestions then -- if this table is nil then we need to load it suggestions = mw.loadData ('Module:Citation/CS1/Suggestions' .. sandbox); --load sandbox version of suggestion module when {{#invoke:Citation/CS1/sandbox|...}}; live module else end for pattern, param in pairs (suggestions.patterns) do -- loop through the patterns to see if we can suggest a proper parameter capture = k:match (pattern); -- the whole match if no capture in pattern else the capture if a match if capture then -- if the pattern matches param = utilities.substitute (param, capture); -- add the capture to the suggested parameter (typically the enumerator) if validate (param, config_t.CitationClass) then -- validate the suggestion to make sure that the suggestion is supported by this template (necessary for limited parameter lists) error_text = utilities.set_message ('err_parameter_ignored_suggest', {k, param}); -- set the suggestion error message else error_text = utilities.set_message ('err_parameter_ignored', {k}); -- suggested param not supported by this template v = ''; -- unset end end end if not utilities.is_set (error_text) then -- couldn't match with a pattern, is there an explicit suggestion? if (suggestions.suggestions[ k:lower() ] ~= nil) and validate (suggestions.suggestions[ k:lower() ], config_t.CitationClass) then utilities.set_message ('err_parameter_ignored_suggest', {k, suggestions.suggestions[ k:lower() ]}); else utilities.set_message ('err_parameter_ignored', {k}); v = ''; -- unset value assigned to unrecognized parameters (this for the limited parameter lists) end end end end cite_args_t[k] = v; -- save this parameter and its value elseif not utilities.is_set (v) then -- for empty parameters if not validate (k, config_t.CitationClass, true) then -- is this empty parameter a valid parameter k = ('' == k) and cfg.err_msg_supl.empty_str or k; -- when k is empty string (or was space(s) trimmed to empty string), replace with descriptive text table.insert (empty_unknowns, utilities.wrap_style ('parameter', k)); -- format for error message and add to the list end end end if 0 ~= #empty_unknowns then -- create empty unknown error message utilities.set_message ('err_param_unknown_empty', { 1 == #empty_unknowns and '' or 's', utilities.make_sep_list (#empty_unknowns, empty_unknowns) }); end local non_url_param_t = {}; -- table of parameters and values that are not url-holding parameters local url_param_t = {}; -- table of url-holding paramters and their values for k, v in pairs (cite_args_t) do if 'string' == type (k) then -- don't evaluate positional parameters has_invisible_chars (k, v); -- look for invisible characters end has_extraneous_punc (k, v); -- look for extraneous terminal punctuation in parameter values missing_pipe_check (k, v); -- do we think that there is a parameter that is missing a pipe? cite_args_t[k] = inter_wiki_check (k, v); -- when language interwiki-linked parameter missing leading colon replace with wiki-link label if 'string' == type (k) then -- when parameter k is not positional if not cfg.url_skip[k] then -- and not in url skip table non_url_param_t[k] = v; -- make a parameter/value list for extraneous url check else -- and is in url skip table (a url-holding parameter) url_param_t[k] = v; -- make a parameter/value list to check for values that are The Wikipedia Library url end end end if cfg.suppress_archive_today_urls then has_archive_today_url (cite_args_t); -- look for and suppress archive.today urls and dependent parameters for all templates end has_extraneous_url (non_url_param_t); -- look for url in parameter values where a url does not belong has_twl_url (url_param_t, cite_args_t); -- look for url-holding parameters that hold a The Wikipedia Library url return table.concat ({ frame:extensionTag ('templatestyles', '', {src='Module:Citation/CS1' .. sandbox .. '/styles.css'}), citation0 (config_t, cite_args_t) }); end --[[--------------------------< C I T A T I O N >-------------------------------------------------------------- Template entry point ]] local function citation (frame) local config_t = {}; -- table to store parameters from the module {{#invoke:}} local args_t = frame:getParent().args; -- get template's preset parameters for k, v in pairs (frame.args) do -- get parameters from the {{#invoke}} frame config_t[k] = v; -- args_t[k] = v; -- crude debug support that allows us to render a citation from module {{#invoke:}}; skips parameter validation; TODO: keep? end return _citation (frame, args_t, config_t) end --[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------ ]] return { citation = citation, -- template entry point _citation = _citation, -- module entry point } 9p0251m5n4ty49ymeozozlcx9pzy0bm 2 231945 1158954 1096477 2026-09-10T10:43:16Z حبيشان 31560 1158954 wikitext text/x-wiki * {{استشهاد بويكي بيانات|Q115641715|مرجع=الشربيني1868|سنة النشر الأصلية=unset}} * {{استشهاد بويكي بيانات/ملعب|Q115782668|مرجع=النحاس1985|سنة النشر الأصلية=unset}} jyiuevqyrptrujzi2uskp5cypuai6h5 828 235800 1158956 1097957 2026-09-10T10:47:47Z حبيشان 31560 1158956 Scribunto text/plain local config_t = { linksproject = "", wikilinkprefix = "" } local aliases_t = { ['accessdate'] = {'تاريخ-الوصول','تاريخ الوصول', 'تاريخ وصول', 'access-date', 'accessdate'}, ['agency'] = {'وكالة', 'agency'}, ['archivedate'] = {'تاريخ-الأرشيف', 'تاريخ أرشيف', 'تاريخ الأرشيف', 'archive-date', 'archivedate'}, ['archive-format'] = {'صيغة-الأرشيف', 'صيغة الأرشيف', 'archive-format'}, ['archive-url'] = {'مسار-الأرشيف', 'مسار أرشيف', 'مسار الأرشيف', 'archive-url', 'archiveurl'}, ['article-number'] = {'رقم-المقالة', 'article-number'}, ['asin-tld'] = 'asin-tld', ['at'] = {'في', 'at'}, --##TOREVISE ['authors'] = {'المؤلفون', 'المؤلفين', 'authors', 'people', 'host', 'credits', 'أشخاص'}, ['author'] = {'author', 'author1', 'subject', 'subject1', 'host', 'host1', 'last', 'last1', 'surname', 'surname1', 'author-last', 'author-last1', 'author1-last', 'author-surname', 'author-surname1', 'author1-surname1'}, ['book-title'] = {'عنوان الكتاب', 'book-title', 'booktitle', 'عنوان المؤتمر'}, ['cartography'] = 'cartography', ['chapter'] = {'الفصل', 'chapter', 'contribution', 'entry', 'article', 'section'}, ['chapter-format'] = {'صيغة-الفصل', 'صيغة الفصل', 'تنسيق الفصل', 'chapter-format', 'contribution-format', 'entry-format', 'article-format', 'section-format'}, ['chapter-url'] = {'مسار-الفصل', 'مسار الفصل', 'chapter-url', 'contribution-url', 'entry-url', 'article-url', 'section-url'}, -- Used by InternetArchiveBot ['chapter-url-access'] = {'chapter-url-access', 'contribution-url-access', 'entry-url-access', 'article-url-access', 'section-url-access'}, -- Used by InternetArchiveBot ['class'] = {'صنف', 'class'}, -- cite arxiv and arxiv identifier ['collaboration'] = 'collaboration', ['conference'] = {'اسم-المؤتمر', 'اسم المؤتمر', 'conference'}, ['conference-format'] = 'conference-format', ['conference-url'] = {'مسار-المؤتمر', 'مسار المؤتمر', 'conference-url'}, -- Used by InternetArchiveBot ['date'] = {'تاريخ', 'التاريخ', 'date', 'air-date', 'airdate'}, ['degree'] = {'الدرجة', 'degree'}, ['df'] = 'df', ['display-authors'] = {'إظهار-المؤلفين', 'إظهار المؤلفين', 'display-authors', 'display-subjects'}, ['display-contributors'] = 'display-contributors', ['display-editors'] = {'إظهار-المحررين', 'إظهار المحررين', 'display-editors'}, ['display-interviewers'] = 'display-interviewers', ['display-translators'] = 'display-translators', ['docket'] = 'docket', ['doi-broken-date'] = 'doi-broken-date', ['edition'] = {'طبعة', 'الإصدار', 'الاصدار', 'إصدار', 'اصدار', 'الطبعة', 'رقم الطبعة', 'edition'}, ['pmc-embargo-date'] = 'pmc-embargo-date', ['encyclopedia'] = {'موسوعة', 'encyclopedia', 'encyclopaedia', 'اسم المعجم', 'dictionary', 'اسم الموسوعة'}, -- cite encyclopedia only ['episode'] = {'الحلقة', 'episode'}, -- cite serial only TODO: make available to cite episode? ['format'] = {'صيغة', 'التنسيق', 'format', 'تنسيق', 'الصيغة'}, ['hijri-date'] = {'التاريخ الهجري', 'تاريخ هجري', 'hijri-date'}, ['id'] = {'id', 'ID', 'المعرف'}, ['inset'] = {'inset', 'مجموعة'}, ['issue'] = {'العدد', 'issue', 'number'}, ['language'] = {'لغة', 'اللغة', 'language', 'lang'}, ['mailing-list'] = {'القائمة البريدية', 'mailing-list'}, -- cite mailing list only ['map'] = {'الخريطة', 'map'}, -- cite map only ['map-format'] = {'صيغة الخريطة', 'map-format'}, -- cite map only ['map-url'] = {'مسار الخريطة', 'map-url'}, -- cite map only -- Used by InternetArchiveBot ['map-url-access'] = 'map-url-access', -- cite map only -- Used by InternetArchiveBot ['minutes'] = {'دقيقة', 'minutes'}, ['mode'] = 'mode', ['name-list-style'] = 'name-list-style', ['network'] = {'الشبكة', 'network'}, ['newsgroup'] = {'اسم المجموعة', 'newsgroup'}, -- cite newsgroup only ['nopp'] = {'no-pp', 'nopp'}, ['no-tracking'] = {'no-tracking', 'template-doc-demo'}, ['number'] = 'number', -- this case only for cite techreport ['orig-date'] = {'orig-date', 'orig-year', 'origyear', 'سنة النشر الأصلية'}, ['others'] = {'آخرون','others', 'المحاور', 'المحاورون', 'المحقق'}, ['page'] = {'صفحة', 'p', 'الصفحة', 'page', 'الصفحة', 'رقم الصفحة', 'رقم صفحة','ص'}, -- Used by InternetArchiveBot ['pages'] = {'صفحات', 'الصفحات', 'pp', 'أرقام الصفحات', 'أرقم صفحات', 'pages'}, -- Used by InternetArchiveBot ['journal'] = { 'صحيفة', 'journal', 'دورية محكمة', 'مجلة', 'magazine', 'اسم المجلة', 'newspaper', 'جريدة', 'اسم الصحيفة', 'دورية', 'periodical', 'موقع', 'website', 'الموقع', 'عمل', 'work', 'العمل', 'من عمل', }, ['place'] = { 'مكان', 'place', 'المكان', 'مقر', 'المقر', 'location'}, ['postscript'] = 'postscript', ['publication-date'] = {'تاريخ-النشر', 'تاريخ النشر', 'publication-date'}, ['publication-place'] = {'مكان-النشر', 'مكان النشر', 'publication-place'}, ['publisher'] = {'ناشر', 'publisher', 'الناشر', 'institution', 'مصدر', 'الموزع', 'المصدر', 'المؤسسة'}, ['quote'] = {'اقتباس', 'quote', 'quotation'}, ['quote-page'] = 'quote-page', ['quote-pages'] = 'quote-pages', ['ref'] = {'مرجع', 'ref'}, ['scale'] = {'المقياس', 'scale'}, ['script-chapter'] = {'script-chapter', 'script-contribution', 'script-entry', 'script-article', 'script-section'}, ['script-encyclopedia'] = {'script-encyclopedia', 'script-encyclopaedia'}, -- cite encyclopedia only ['script-map'] = 'script-map', ['script-journal'] = {'script-journal', 'script-magazine', 'script-newspaper', 'script-periodical', 'script-website', 'script-work'}, ['script-quote'] = 'script-quote', ['script-title'] = 'script-title', -- Used by InternetArchiveBot ['season'] = {'الموسم', 'season'}, ['sections'] = {'الأقسام', 'sections'}, -- cite map only ['series'] = {'سلسلة', 'series', 'version'}, ['series-link'] = {'وصلة-سلسلة', 'وصلة سلسلة', 'series-link', 'مسار المسلسل'}, ['series-number'] = {'رقم-سلسلة', 'رقم سلسلة', 'series-number', 'series-no'}, ['sheet'] = {'ورقة', 'sheet'}, -- cite map only ['sheets'] = {'أوراق', 'sheets'}, -- cite map only ['station'] = {'محطة', 'station'}, ['time'] = {'الوقت', 'time'}, ['time-caption'] = 'time-caption', ['title'] = {'عنوان', 'title', 'العنوان'}, ['title-link'] = {'وصلة-العنوان', 'وصلة العنوان', 'title-link', 'episode-link', 'وصلة الحلقة'}, ['title-note'] = {'title-note', 'department'}, ['type'] = {'النوع', 'type', 'medium'}, ['trans-chapter'] = {'trans-article', 'trans-chapter', 'trans-contribution', 'trans-entry', 'trans-section'}, ['transcript'] = 'transcript', ['transcript-format'] = 'transcript-format', ['transcript-url'] = 'transcript-url', -- Used by InternetArchiveBot ['trans-encyclopedia'] = {'trans-encyclopedia', 'trans-encyclopaedia'}, -- cite encyclopedia only ['trans-map'] = 'trans-map', -- cite map only ['trans-journal'] = {'trans-journal', 'trans-magazine', 'trans-newspaper', 'trans-periodical', 'trans-website', 'trans-work'}, ['trans-quote'] = 'trans-quote', ['trans-title'] = {'عنوان مترجم', 'العنوان بالعربي', 'العنوان الأجنبي', 'عنوان أجنبي', 'trans-title', 'العنوان المترجم'}, -- Used by InternetArchiveBot ['url'] = {'مسار', 'المسار','وصلة', 'url', 'URL'}, -- Used by InternetArchiveBot ['url-access'] = {'url-access', 'التسجيل'}, -- Used by InternetArchiveBot ['url-status'] = {'url-status','حالة المسار'}, -- Used by InternetArchiveBot ['vauthors'] = {'vauthors', 'عدة مؤلفين'}, ['veditors'] = {'veditors', 'عدة محررين' }, ['via'] = {'عبر', 'via'}, ['volume'] = {'المجلد', 'volume', 'الجزء', 'ج'}, ['wikidatacite'] = 'wikidatacite', ['year'] = {'سنة', 'year', 'السنة'}, } local numbered_aliases_t = { ['author#'] = {'مؤلف#', 'الأخير#', 'مؤلف#-الأخير', 'مؤلف# الأخير', 'last#', 'author-last#', 'مؤلف-الأخير#', 'author#-last', 'surname#', 'الأخير#', 'author-surname#', 'author#-surname', 'author#', 'المؤلف#', 'أخير#', 'اخير#', 'مؤلف الأخير#', 'subject#', 'host#', 'الموضوع'}, ['author-link#'] = { 'وصلة مؤلف#', 'مؤلف#-وصلة', 'author#-link', 'subject-link#', 'subject#-link', 'authorlink#', 'author#link', 'مؤلف# وصلة', 'وصلة-مؤلف#', 'وصلة المؤلف#', 'وصلة-المؤلف#' }, ['editor#'] = { 'محرر#', 'محرر#-الأخير', 'محرر# الأخير', 'محرر-الأخير#', 'محرر الأخير#', 'editor-last#', 'editor#-last', 'editor-surname#', 'editor#-surname', 'المحرر#' }, ['editor-link#'] = { 'وصلة محرر#', 'محرر#-وصلة', 'محرر# وصلة', 'وصلة-محرر#', 'editor#-link', 'محرر-وصلة#', 'وصلة-المحرر#', 'وصلة المحرر#' }, ['translator#'] = { 'مترجم#', 'مترجم#-الأخير', 'المترجم#', 'translator-last#', 'المترجم-الأخير#', 'المترجم#-الأخير', 'مترجم الأخير#', 'مترجم# الأخير', 'مترجم-الأخير#', 'translator#-last', 'translator-surname#', 'translator#-surname' }, ['translator-link#'] = { 'وصلة مترجم#', 'مترجم#-وصلة', 'مترجم# وصلة', 'وصلة-مترجم#', 'وصلة المترجم#', 'وصلة-المترجم#', 'translator#-link', 'المترجم#-وصلة' }, ['translator-first#'] = { 'مترجم#-الأول', 'translator#-first', 'مترجم-الأول#', 'مترجم# الأول', 'مترجم-الأول#', 'مترجم الأول#', 'المترجم-الأول#', 'المترجم#-الأول', 'translator-given#', 'translator#-given' }, ['editor-first#'] = { 'محرر#-الأول', 'محرر-الأول#', 'محرر الأول#', 'editor-first#', 'editor#-first', 'editor-given#', 'محرر# الأول', 'محرر# الأول', 'editor#-given' }, ['author-first#'] = { 'الأول#', 'مؤلف#-الأول', 'مؤلف# الأول', 'أول#', 'first#', 'اول#', 'الاول#', 'author-first#', 'مؤلف الأول#', 'author#-first', 'given#', 'author-given#', 'مؤلف-الأول#', 'author#-given' }, } local i18n_t = { ["ar"] = { illustrator = "رسم", composer = "تلحين", animator = "رسم الرسوم المتحركة", director = "إخراج", screenwriter = "سيناريو", signatory = "توقيع", presenter = "تقديم", performer = "أداء", annotator = "تحقيق", reviewer = "مراجعة", unknown = "مجهول", supervisor = "إشراف", ["unknown-publication-date"] = 'nd', ["unknown-publisher"] = '<abbr class="abbr" title="دون ناشر">د.ن.</abbr>', wikidata = "ويكي بيانات" }, ["cite_q"] = "{{استشهاد بويكي بيانات|", ["erratum"] = "(erratum)", ["unknown-author"] = { ar = "غير معروف", en = "Anonymous", fr = "Anonyme", de = "Anonymus", es = "Anónimo" }, ["wdq"] = "[[ويكي بيانات|QID]]", -- go through "WDQ (identifier)" redirect to reduce clutter in "What links here" and improve reverse lookup. Keep in sync with {{QID}}. ["wikisource"] = "arwikisource", ["errors"] = { ["local-article-not-found"] = "المقال المحلي غير موجود.", ["no-label-or-title"] = "لا عنون أو تسمية: ", }, -- Next Line Edited ["unknown-author-trackingcat"] = "", -- [[تصنيف:Cite Q - author unknown]] ["months"] = { "January", "February", "March", "April", "May", "June", "July", "August", "September", "October", "November", "December" }, ["ordinal"] = { [1] = "st", [2] = "nd", [3] = "rd", ["default"] = "th" }, ["trackingcats"] = { ["erratum"] = "[[تصنيف:استشهاد بويكي بيانات:الاستشهاد يعمل مع خطأ]]", ["missing-wikidata"] = "[[تصنيف:مقالات مع بيانات ويكي بيانات نقاصة]]", -- TODO: point to a {{cite q}}-specific category; this one is specific to Module:WikidataIB ['no-label-or-title'] = '[[تصنيف:استشهاد بويكي بيانات:يفتقر إلى عنوان أو تسمية]]', ["retracted"] = "[[تصنيف:الاستشهاد بويكي بيانات: الاستشهاد بعمل مسحوب]]<!-- retracted -->", ["replaced"] = "[[تصنيف:الاستشهاد بويكي بيانات: الاستشهاد بعمل مستبدل]]<!-- replaced -->", ["unknown-author"] = "", -- [[Category:Cite Q - author unknown]] removed as misplaced; someone is welcome to restore it if it is placed outside of the template parameter value ["unknown-template"] = "[[تصنيف:الاستشهاد بويكي بيانات:قالب مجهول]]", }, } local citation_class_map_t = { arxiv = 'استشهاد بأرخايف', ['audio-visual'] = 'استشهاد بوسائط مرئية ومسموعة', ['AV-media-notes'] = 'استشهاد بتسجيلات', biorxiv = 'استشهاد ببيوركسيف', book = 'استشهاد بكتاب', citation = 'استشهاد', citeseerx = 'Cite CiteSeerX', conference = 'استشهاد بمنشورات مؤتمر', document = 'استشهاد بوثيقة', encyclopaedia = 'استشهاد بموسوعة', episode = 'استشهاد بحلقة', interview = 'استشهاد بمقابلة', journal = 'استشهاد بدورية محكمة', magazine = 'استشهاد بمجلة', mailinglist = 'استشهاد بقائمة بريدية', map = 'استشهاد بخريطة', medrxiv = 'cite medRxiv', news = 'استشهاد بخبر', newsgroup = 'استشهاد بمجموعة أخبار', podcast = 'استشهاد بتدوين صوتي', pressrelease = 'استشهاد ببيان صحفي', report = 'استشهاد بتقرير', serial = 'استشهاد بمسلسل', sign = 'استشهاد بإشارة توضيحية', speech = 'استشهاد بخطاب', ssrn = 'cite ssrn', techreport = 'استشهاد بتقرير فني', thesis = 'استشهاد بأطروحة', web = 'استشهاد بويب', } local rejected_t = { sources = { Q465 = true, Q15241312 = true, Q52 = true }, properties = { "P1800", "P1566", "P21"}, instances = { Q5 = true, -- Human is not a direct reference Q184188 = true, Q1250464 = true, -- inatance of refference DBs Q8513 = true, Q17152639 = true, Q7094076 = true, Q5535082 = true, Q856638= true, Q1982918 = true, Q36524 = true, Q36218176 = true, Q526334 = true, Q5227411 = true } } -- Table of simple properties that can be fetched in roughly the same way: -- id = PXXX -- maxvals = maximum number of multiple values (0 for all) -- linked = "no" suppresses linking -- populate_from_journal = true/false determines whether to look in a journal where the source is published -- rank = "best", "preferred", normal, etc. determines how Wikidata ranks are treated -- others = true - the value for the property goes to "others" section local simple_properties_t = { -- Next Line Edited publisher = {id = "P123", maxvals = 3}, oclc = {id = "P243", maxvals = 1}, ['publication-place'] = {id = "P291", maxvals = 0, linked = false}, -- publication place (don't put into |place=; is treated specially in {{citation}} if both are given) doi = {id = "P356", maxvals = 1}, -- take care of |doi-broken-date= (WD "reason for deprecation"/"stated as") and |doi-access= (WD "access status")? issue = {id = "P433", maxvals = 0, populate_from_journal = true}, -- distinguish from |number= ("P1545"?) if both are given (still blocked by {{citation}}, but will be supported in the future) pmid = {id = "P698", maxvals = 1}, -- gbooks = {id = "P675", maxvals = 1}, -- to be added to {{citation}} -- ia = {id = "P724", maxvals = 1}, -- to be added to {{citation}} arxiv = {id = "P818", maxvals = 1}, bibcode = {id = "P819", maxvals = 1}, -- take care of |bibcode-access=? jstor = {id = "P888", maxvals = 1}, -- take care of |jstor-access=? mr = {id = "P889", maxvals = 1}, rfc = {id = "P892", maxvals = 1}, zbl = {id = "P894", maxvals = 1}, ssrn = {id = "P893", maxvals = 1}, place = {id = "P1071", maxvals = 0, linked = false}, -- written-at place -- ['total-pages'] = {id = "P1104", maxvals = 0, linked = 'no'}, -- to be added to {{citation}} / COinS &rft.tpages= -- coden = {id = "P1159", maxvals = 1}, -- to be added to {{citation}} / COinS &rft.coden= s2cid = {id = "P8299", maxvals = 1}, -- take care of |s2cid-access=? pmc = {id = "P932", maxvals = 1}, -- take care of |pmc-embargo-date= (WD "reason for deprecation")? lccn = {id = "P1144", maxvals = 1}, hdl = {id = "P1184", maxvals = 1}, -- take care of |hdl-access=? ismn = {id = "P1208", maxvals = 1}, journal = {id = "P1433", maxvals = 1}, citeseerx = {id = "P3784", maxvals = 1}, osti = {id = "P3894", maxvals = 1}, -- take care of |osti-access=? biorxiv = {id = "P3951", maxvals = 1}, asin = {id = "P5749", maxvals = 1}, -- What about |asin-tld=? (WD examples resolve to .com at present, but may change) -- ['catalog-number'] = {id = "P528", maxvals = 0}, -- to be added to {{citation}} / COinS &rft.artnum= isbn = {id = "P212", maxvals = 1, populate_from_journal = true}, -- ISBN 13 issn = {id = "P236", maxvals = 1, populate_from_journal = true}, -- distinguish from |eissn= for electronic issues? -- jfm = {id = "P?", maxvals = 1}, -- Jahrbuch über die Fortschritte der Mathematik (not Zbl) -- sbn = {id = "P?", maxvals = 1}, -- Standard Book Number (predecessor of ISBN, not ICCU) -- message-id = {id = "P?", maxvals = 1}, -- Usenet message ID chapter = {id = "P792", maxvals = 1, no_vol_qual=true}, ['publication-date'] = {id = "P577", maxvals = 1, populate_from_journal = true}, -- publication date (don't use |date=; is treated specially in {{citation}} if both are given.) series = {id = "P179", maxvals = 1, populate_from_journal = true, qual = "P1545"}, version = {id = "P348", maxvals = 0}, edition = {id = "P393", maxvals = 0}, volume = {id = "P478", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, -- part = {id = "P1545"?, maxvals = 0}, -- to be added to {{citation}} / COinS &rft.part= title = {id = "P1476", rank="p n"}, -- url = {id = "P953", maxvals = 1}, -- deal with this along with archive-url pages = {id = "P304", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, at = {id = "P958", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, -- also incorporate lines (P7421) and columns (P3903) into this (cite map also supports |section=) -- sheets = {id = "P7416", maxvals = 0, populate_from_journal = true}, -- interviewer = {id = "P?", maxvals = 0}, -- does **not** go to "others" section! Multiple interviewers should be n-enumerated illustrator = {id = "P110", maxvals = 10, others = true}, -- goes to "others" section -- foreword and afterword, when contributions to another author's work, are contributions so belong in |contribution=; -- the writer's name goes in |contributor=; requires |title= and |author= -- However, this might need to add support for multiple contributors and their roles to {{citation}}, see Help_talk:Citation_Style_1#Others -- foreword = {id = "P2679", maxvals = 10, others = true}, -- goes to "others" section -- afterword = {id = "P2680", maxvals = 10, others = true}, -- goes to "others" section composer = {id = "P86", maxvals = 10, others = true}, -- goes to "others" section animator = {id = "P6942", maxvals = 10, others = true}, -- goes to "others" section director = {id = "P57", maxvals = 10, others = true}, -- goes to "others" section screenwriter = {id = "P58", maxvals = 10, others = true}, -- goes to "others" section signatory = {id = "P1891", maxvals = 10, others = true}, -- goes to "others" section presenter = {id = "P371", maxvals = 10, others = true}, -- goes to "others" section performer = {id = "P175", maxvals = 10, others = true}, -- goes to "others" section -- next 2 lines added for Arabic Wikipedia annotator = {id ="P11105", maxvals = 10, others = true}, -- goes to "others" section reviewer = {id ="P4032", maxvals = 10, others = true}, -- goes to "others" section supervisor = {id ="P7604", maxvals = 3, others = true}, -- goes to "others" section } local used_args_t = {accessdate=true, language=true, mode=true, ol=true, other=true, ['display-authors']=true, ['display-editors']=true, ['archive-url']=true, ['archive-date']=true, ['orig-date']=true, url=true, title=true, page=true, pages = true, volume=true, } for i = 1, 9 do used_args_t['author' .. i] = true used_args_t['author-link' .. i] = true used_args_t['author-first' .. i] = true used_args_t['editor' .. i] = true used_args_t['editor-link' .. i] = true used_args_t['editor-first' .. i] = true used_args_t['translator' .. i] = true used_args_t['translator-link' .. i] = true used_args_t['translator-first' .. i] = true end -- all_used_properties local used_properties_t = { P1680 = true, P1932 = true, P2699 = true, P407 = true, P2093 = true, -- author string P50 = true, -- author P5769 = true, --editor P655 = true, P856 = true, P953 = true, P98 = true, P854 = true, } for k,v in pairs(simple_properties_t) do used_properties_t[v.id] = { arg = k, maxvals = v.maxvals, linked = v.linked, others = v.others, populate_from_journal = v.populate_from_journal } end simple_properties_t['hijri-date'] = {id = "P577", maxvals = 1, hijri=true} return { aliases_t = aliases_t, numbered_aliases_t = numbered_aliases_t, i18n_t = i18n_t, rejected_t = rejected_t, simple_properties_t = simple_properties_t, used_args_t = used_args_t, used_properties_t = used_properties_t, config_t = config_t, citation_class_map_t=citation_class_map_t } 89a3ngigl1bbmkbtq1fgzscdr73s3y4 828 240415 1158958 1097961 2026-09-10T10:48:12Z حبيشان 31560 1158958 Scribunto text/plain local config_t = { linksproject = "", wikilinkprefix = "" } local aliases_t = { ['accessdate'] = {'تاريخ-الوصول','تاريخ الوصول', 'تاريخ وصول', 'access-date', 'accessdate'}, ['agency'] = {'وكالة', 'agency'}, ['archivedate'] = {'تاريخ-الأرشيف', 'تاريخ أرشيف', 'تاريخ الأرشيف', 'archive-date', 'archivedate'}, ['archive-format'] = {'صيغة-الأرشيف', 'صيغة الأرشيف', 'archive-format'}, ['archive-url'] = {'مسار-الأرشيف', 'مسار أرشيف', 'مسار الأرشيف', 'archive-url', 'archiveurl'}, ['article-number'] = {'رقم-المقالة', 'article-number'}, ['asin-tld'] = 'asin-tld', ['at'] = {'في', 'at'}, --##TOREVISE ['authors'] = {'المؤلفون', 'المؤلفين', 'authors', 'people', 'host', 'credits', 'أشخاص'}, ['author'] = {'author', 'author1', 'subject', 'subject1', 'host', 'host1', 'last', 'last1', 'surname', 'surname1', 'author-last', 'author-last1', 'author1-last', 'author-surname', 'author-surname1', 'author1-surname1'}, ['book-title'] = {'عنوان الكتاب', 'book-title', 'booktitle', 'عنوان المؤتمر'}, ['cartography'] = 'cartography', ['chapter'] = {'الفصل', 'chapter', 'contribution', 'entry', 'article', 'section'}, ['chapter-format'] = {'صيغة-الفصل', 'صيغة الفصل', 'تنسيق الفصل', 'chapter-format', 'contribution-format', 'entry-format', 'article-format', 'section-format'}, ['chapter-url'] = {'مسار-الفصل', 'مسار الفصل', 'chapter-url', 'contribution-url', 'entry-url', 'article-url', 'section-url'}, -- Used by InternetArchiveBot ['chapter-url-access'] = {'chapter-url-access', 'contribution-url-access', 'entry-url-access', 'article-url-access', 'section-url-access'}, -- Used by InternetArchiveBot ['class'] = {'صنف', 'class'}, -- cite arxiv and arxiv identifier ['collaboration'] = 'collaboration', ['conference'] = {'اسم-المؤتمر', 'اسم المؤتمر', 'conference'}, ['conference-format'] = 'conference-format', ['conference-url'] = {'مسار-المؤتمر', 'مسار المؤتمر', 'conference-url'}, -- Used by InternetArchiveBot ['date'] = {'تاريخ', 'التاريخ', 'date', 'air-date', 'airdate'}, ['degree'] = {'الدرجة', 'degree'}, ['df'] = 'df', ['display-authors'] = {'إظهار-المؤلفين', 'إظهار المؤلفين', 'display-authors', 'display-subjects'}, ['display-contributors'] = 'display-contributors', ['display-editors'] = {'إظهار-المحررين', 'إظهار المحررين', 'display-editors'}, ['display-interviewers'] = 'display-interviewers', ['display-translators'] = 'display-translators', ['docket'] = 'docket', ['doi-broken-date'] = 'doi-broken-date', ['edition'] = {'طبعة', 'الإصدار', 'الاصدار', 'إصدار', 'اصدار', 'الطبعة', 'رقم الطبعة', 'edition'}, ['pmc-embargo-date'] = 'pmc-embargo-date', ['encyclopedia'] = {'موسوعة', 'encyclopedia', 'encyclopaedia', 'اسم المعجم', 'dictionary', 'اسم الموسوعة'}, -- cite encyclopedia only ['episode'] = {'الحلقة', 'episode'}, -- cite serial only TODO: make available to cite episode? ['format'] = {'صيغة', 'التنسيق', 'format', 'تنسيق', 'الصيغة'}, ['hijri-date'] = {'التاريخ الهجري', 'تاريخ هجري', 'hijri-date'}, ['id'] = {'id', 'ID', 'المعرف'}, ['inset'] = {'inset', 'مجموعة'}, ['issue'] = {'العدد', 'issue', 'number'}, ['language'] = {'لغة', 'اللغة', 'language', 'lang'}, ['mailing-list'] = {'القائمة البريدية', 'mailing-list'}, -- cite mailing list only ['map'] = {'الخريطة', 'map'}, -- cite map only ['map-format'] = {'صيغة الخريطة', 'map-format'}, -- cite map only ['map-url'] = {'مسار الخريطة', 'map-url'}, -- cite map only -- Used by InternetArchiveBot ['map-url-access'] = 'map-url-access', -- cite map only -- Used by InternetArchiveBot ['minutes'] = {'دقيقة', 'minutes'}, ['mode'] = 'mode', ['name-list-style'] = 'name-list-style', ['network'] = {'الشبكة', 'network'}, ['newsgroup'] = {'اسم المجموعة', 'newsgroup'}, -- cite newsgroup only ['nopp'] = {'no-pp', 'nopp'}, ['no-tracking'] = {'no-tracking', 'template-doc-demo'}, ['number'] = 'number', -- this case only for cite techreport ['orig-date'] = {'orig-date', 'orig-year', 'origyear', 'سنة النشر الأصلية'}, ['others'] = {'آخرون','others', 'المحاور', 'المحاورون', 'المحقق'}, ['page'] = {'صفحة', 'p', 'الصفحة', 'page', 'الصفحة', 'رقم الصفحة', 'رقم صفحة','ص'}, -- Used by InternetArchiveBot ['pages'] = {'صفحات', 'الصفحات', 'pp', 'أرقام الصفحات', 'أرقم صفحات', 'pages'}, -- Used by InternetArchiveBot ['journal'] = { 'صحيفة', 'journal', 'دورية محكمة', 'مجلة', 'magazine', 'اسم المجلة', 'newspaper', 'جريدة', 'اسم الصحيفة', 'دورية', 'periodical', 'موقع', 'website', 'الموقع', 'عمل', 'work', 'العمل', 'من عمل', }, ['place'] = { 'مكان', 'place', 'المكان', 'مقر', 'المقر', 'location'}, ['postscript'] = 'postscript', ['publication-date'] = {'تاريخ-النشر', 'تاريخ النشر', 'publication-date'}, ['publication-place'] = {'مكان-النشر', 'مكان النشر', 'publication-place'}, ['publisher'] = {'ناشر', 'publisher', 'الناشر', 'institution', 'مصدر', 'الموزع', 'المصدر', 'المؤسسة'}, ['quote'] = {'اقتباس', 'quote', 'quotation'}, ['quote-page'] = 'quote-page', ['quote-pages'] = 'quote-pages', ['ref'] = {'مرجع', 'ref'}, ['scale'] = {'المقياس', 'scale'}, ['script-chapter'] = {'script-chapter', 'script-contribution', 'script-entry', 'script-article', 'script-section'}, ['script-encyclopedia'] = {'script-encyclopedia', 'script-encyclopaedia'}, -- cite encyclopedia only ['script-map'] = 'script-map', ['script-journal'] = {'script-journal', 'script-magazine', 'script-newspaper', 'script-periodical', 'script-website', 'script-work'}, ['script-quote'] = 'script-quote', ['script-title'] = 'script-title', -- Used by InternetArchiveBot ['season'] = {'الموسم', 'season'}, ['sections'] = {'الأقسام', 'sections'}, -- cite map only ['series'] = {'سلسلة', 'series', 'version'}, ['series-link'] = {'وصلة-سلسلة', 'وصلة سلسلة', 'series-link', 'مسار المسلسل'}, ['series-number'] = {'رقم-سلسلة', 'رقم سلسلة', 'series-number', 'series-no'}, ['sheet'] = {'ورقة', 'sheet'}, -- cite map only ['sheets'] = {'أوراق', 'sheets'}, -- cite map only ['station'] = {'محطة', 'station'}, ['time'] = {'الوقت', 'time'}, ['time-caption'] = 'time-caption', ['title'] = {'عنوان', 'title', 'العنوان'}, ['title-link'] = {'وصلة-العنوان', 'وصلة العنوان', 'title-link', 'episode-link', 'وصلة الحلقة'}, ['title-note'] = {'title-note', 'department'}, ['type'] = {'النوع', 'type', 'medium'}, ['trans-chapter'] = {'trans-article', 'trans-chapter', 'trans-contribution', 'trans-entry', 'trans-section'}, ['transcript'] = 'transcript', ['transcript-format'] = 'transcript-format', ['transcript-url'] = 'transcript-url', -- Used by InternetArchiveBot ['trans-encyclopedia'] = {'trans-encyclopedia', 'trans-encyclopaedia'}, -- cite encyclopedia only ['trans-map'] = 'trans-map', -- cite map only ['trans-journal'] = {'trans-journal', 'trans-magazine', 'trans-newspaper', 'trans-periodical', 'trans-website', 'trans-work'}, ['trans-quote'] = 'trans-quote', ['trans-title'] = {'عنوان مترجم', 'العنوان بالعربي', 'العنوان الأجنبي', 'عنوان أجنبي', 'trans-title', 'العنوان المترجم'}, -- Used by InternetArchiveBot ['url'] = {'مسار', 'المسار','وصلة', 'url', 'URL'}, -- Used by InternetArchiveBot ['url-access'] = {'url-access', 'التسجيل'}, -- Used by InternetArchiveBot ['url-status'] = {'url-status','حالة المسار'}, -- Used by InternetArchiveBot ['vauthors'] = {'vauthors', 'عدة مؤلفين'}, ['veditors'] = {'veditors', 'عدة محررين' }, ['via'] = {'عبر', 'via'}, ['volume'] = {'المجلد', 'volume', 'الجزء', 'ج'}, ['wikidatacite'] = 'wikidatacite', ['year'] = {'سنة', 'year', 'السنة'}, } local numbered_aliases_t = { ['author#'] = {'مؤلف#', 'الأخير#', 'مؤلف#-الأخير', 'مؤلف# الأخير', 'last#', 'author-last#', 'مؤلف-الأخير#', 'author#-last', 'surname#', 'الأخير#', 'author-surname#', 'author#-surname', 'author#', 'المؤلف#', 'أخير#', 'اخير#', 'مؤلف الأخير#', 'subject#', 'host#', 'الموضوع'}, ['author-link#'] = { 'وصلة مؤلف#', 'مؤلف#-وصلة', 'author#-link', 'subject-link#', 'subject#-link', 'authorlink#', 'author#link', 'مؤلف# وصلة', 'وصلة-مؤلف#', 'وصلة المؤلف#', 'وصلة-المؤلف#' }, ['editor#'] = { 'محرر#', 'محرر#-الأخير', 'محرر# الأخير', 'محرر-الأخير#', 'محرر الأخير#', 'editor-last#', 'editor#-last', 'editor-surname#', 'editor#-surname', 'المحرر#' }, ['editor-link#'] = { 'وصلة محرر#', 'محرر#-وصلة', 'محرر# وصلة', 'وصلة-محرر#', 'editor#-link', 'محرر-وصلة#', 'وصلة-المحرر#', 'وصلة المحرر#' }, ['translator#'] = { 'مترجم#', 'مترجم#-الأخير', 'المترجم#', 'translator-last#', 'المترجم-الأخير#', 'المترجم#-الأخير', 'مترجم الأخير#', 'مترجم# الأخير', 'مترجم-الأخير#', 'translator#-last', 'translator-surname#', 'translator#-surname' }, ['translator-link#'] = { 'وصلة مترجم#', 'مترجم#-وصلة', 'مترجم# وصلة', 'وصلة-مترجم#', 'وصلة المترجم#', 'وصلة-المترجم#', 'translator#-link', 'المترجم#-وصلة' }, ['translator-first#'] = { 'مترجم#-الأول', 'translator#-first', 'مترجم-الأول#', 'مترجم# الأول', 'مترجم-الأول#', 'مترجم الأول#', 'المترجم-الأول#', 'المترجم#-الأول', 'translator-given#', 'translator#-given' }, ['editor-first#'] = { 'محرر#-الأول', 'محرر-الأول#', 'محرر الأول#', 'editor-first#', 'editor#-first', 'editor-given#', 'محرر# الأول', 'محرر# الأول', 'editor#-given' }, ['author-first#'] = { 'الأول#', 'مؤلف#-الأول', 'مؤلف# الأول', 'أول#', 'first#', 'اول#', 'الاول#', 'author-first#', 'مؤلف الأول#', 'author#-first', 'given#', 'author-given#', 'مؤلف-الأول#', 'author#-given' }, } local i18n_t = { ["ar"] = { illustrator = "رسم", composer = "تلحين", animator = "رسم الرسوم المتحركة", director = "إخراج", screenwriter = "سيناريو", signatory = "توقيع", presenter = "تقديم", performer = "أداء", annotator = "تحقيق", reviewer = "مراجعة", unknown = "مجهول", supervisor = "إشراف", ["unknown-publication-date"] = 'nd', ["unknown-publisher"] = '<abbr class="abbr" title="دون ناشر">د.ن.</abbr>', wikidata = "ويكي بيانات" }, ["cite_q"] = "{{استشهاد بويكي بيانات|", ["erratum"] = "(erratum)", ["unknown-author"] = { ar = "غير معروف", en = "Anonymous", fr = "Anonyme", de = "Anonymus", es = "Anónimo" }, ["wdq"] = "[[ويكي بيانات|QID]]", -- go through "WDQ (identifier)" redirect to reduce clutter in "What links here" and improve reverse lookup. Keep in sync with {{QID}}. ["wikisource"] = "arwikisource", ["errors"] = { ["local-article-not-found"] = "المقال المحلي غير موجود.", ["no-label-or-title"] = "لا عنون أو تسمية: ", }, -- Next Line Edited ["unknown-author-trackingcat"] = "", -- [[تصنيف:Cite Q - author unknown]] ["months"] = { "January", "February", "March", "April", "May", "June", "July", "August", "September", "October", "November", "December" }, ["ordinal"] = { [1] = "st", [2] = "nd", [3] = "rd", ["default"] = "th" }, ["trackingcats"] = { ["erratum"] = "[[تصنيف:استشهاد بويكي بيانات:الاستشهاد يعمل مع خطأ]]", ["missing-wikidata"] = "[[تصنيف:مقالات مع بيانات ويكي بيانات نقاصة]]", -- TODO: point to a {{cite q}}-specific category; this one is specific to Module:WikidataIB ['no-label-or-title'] = '[[تصنيف:استشهاد بويكي بيانات:يفتقر إلى عنوان أو تسمية]]', ["retracted"] = "[[تصنيف:الاستشهاد بويكي بيانات: الاستشهاد بعمل مسحوب]]<!-- retracted -->", ["replaced"] = "[[تصنيف:الاستشهاد بويكي بيانات: الاستشهاد بعمل مستبدل]]<!-- replaced -->", ["unknown-author"] = "", -- [[Category:Cite Q - author unknown]] removed as misplaced; someone is welcome to restore it if it is placed outside of the template parameter value ["unknown-template"] = "[[تصنيف:الاستشهاد بويكي بيانات:قالب مجهول]]", }, } local citation_class_map_t = { arxiv = 'استشهاد بأرخايف', ['audio-visual'] = 'استشهاد بوسائط مرئية ومسموعة', ['AV-media-notes'] = 'استشهاد بتسجيلات', biorxiv = 'استشهاد ببيوركسيف', book = 'استشهاد بكتاب', citation = 'استشهاد', citeseerx = 'Cite CiteSeerX', conference = 'استشهاد بمنشورات مؤتمر', document = 'استشهاد بوثيقة', encyclopaedia = 'استشهاد بموسوعة', episode = 'استشهاد بحلقة', interview = 'استشهاد بمقابلة', journal = 'استشهاد بدورية محكمة', magazine = 'استشهاد بمجلة', mailinglist = 'استشهاد بقائمة بريدية', map = 'استشهاد بخريطة', medrxiv = 'cite medRxiv', news = 'استشهاد بخبر', newsgroup = 'استشهاد بمجموعة أخبار', podcast = 'استشهاد بتدوين صوتي', pressrelease = 'استشهاد ببيان صحفي', report = 'استشهاد بتقرير', serial = 'استشهاد بمسلسل', sign = 'استشهاد بإشارة توضيحية', speech = 'استشهاد بخطاب', ssrn = 'cite ssrn', techreport = 'استشهاد بتقرير فني', thesis = 'استشهاد بأطروحة', web = 'استشهاد بويب', } local rejected_t = { sources = { Q465 = true, Q15241312 = true, Q52 = true }, properties = { "P1800", "P1566", "P21"}, instances = { Q5 = true, -- Human is not a direct reference Q184188 = true, Q1250464 = true, -- inatance of refference DBs Q8513 = true, Q17152639 = true, Q7094076 = true, Q5535082 = true, Q856638= true, Q1982918 = true, Q36524 = true, Q36218176 = true, Q526334 = true, Q5227411 = true } } -- Table of simple properties that can be fetched in roughly the same way: -- id = PXXX -- maxvals = maximum number of multiple values (0 for all) -- linked = "no" suppresses linking -- populate_from_journal = true/false determines whether to look in a journal where the source is published -- rank = "best", "preferred", normal, etc. determines how Wikidata ranks are treated -- others = true - the value for the property goes to "others" section local simple_properties_t = { -- Next Line Edited publisher = {id = "P123", maxvals = 3}, oclc = {id = "P243", maxvals = 1}, ['publication-place'] = {id = "P291", maxvals = 0, linked = false}, -- publication place (don't put into |place=; is treated specially in {{citation}} if both are given) doi = {id = "P356", maxvals = 1}, -- take care of |doi-broken-date= (WD "reason for deprecation"/"stated as") and |doi-access= (WD "access status")? issue = {id = "P433", maxvals = 0, populate_from_journal = true}, -- distinguish from |number= ("P1545"?) if both are given (still blocked by {{citation}}, but will be supported in the future) pmid = {id = "P698", maxvals = 1}, -- gbooks = {id = "P675", maxvals = 1}, -- to be added to {{citation}} -- ia = {id = "P724", maxvals = 1}, -- to be added to {{citation}} arxiv = {id = "P818", maxvals = 1}, bibcode = {id = "P819", maxvals = 1}, -- take care of |bibcode-access=? jstor = {id = "P888", maxvals = 1}, -- take care of |jstor-access=? mr = {id = "P889", maxvals = 1}, rfc = {id = "P892", maxvals = 1}, zbl = {id = "P894", maxvals = 1}, ssrn = {id = "P893", maxvals = 1}, place = {id = "P1071", maxvals = 0, linked = false}, -- written-at place -- ['total-pages'] = {id = "P1104", maxvals = 0, linked = 'no'}, -- to be added to {{citation}} / COinS &rft.tpages= -- coden = {id = "P1159", maxvals = 1}, -- to be added to {{citation}} / COinS &rft.coden= s2cid = {id = "P8299", maxvals = 1}, -- take care of |s2cid-access=? pmc = {id = "P932", maxvals = 1}, -- take care of |pmc-embargo-date= (WD "reason for deprecation")? lccn = {id = "P1144", maxvals = 1}, hdl = {id = "P1184", maxvals = 1}, -- take care of |hdl-access=? ismn = {id = "P1208", maxvals = 1}, journal = {id = "P1433", maxvals = 1}, citeseerx = {id = "P3784", maxvals = 1}, osti = {id = "P3894", maxvals = 1}, -- take care of |osti-access=? biorxiv = {id = "P3951", maxvals = 1}, asin = {id = "P5749", maxvals = 1}, -- What about |asin-tld=? (WD examples resolve to .com at present, but may change) -- ['catalog-number'] = {id = "P528", maxvals = 0}, -- to be added to {{citation}} / COinS &rft.artnum= isbn = {id = "P212", maxvals = 1, populate_from_journal = true}, -- ISBN 13 issn = {id = "P236", maxvals = 1, populate_from_journal = true}, -- distinguish from |eissn= for electronic issues? -- jfm = {id = "P?", maxvals = 1}, -- Jahrbuch über die Fortschritte der Mathematik (not Zbl) -- sbn = {id = "P?", maxvals = 1}, -- Standard Book Number (predecessor of ISBN, not ICCU) -- message-id = {id = "P?", maxvals = 1}, -- Usenet message ID chapter = {id = "P792", maxvals = 1, no_vol_qual=true}, ['publication-date'] = {id = "P577", maxvals = 1, populate_from_journal = true}, -- publication date (don't use |date=; is treated specially in {{citation}} if both are given.) series = {id = "P179", maxvals = 1, populate_from_journal = true, qual = "P1545"}, version = {id = "P348", maxvals = 0}, edition = {id = "P393", maxvals = 0}, volume = {id = "P478", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, -- part = {id = "P1545"?, maxvals = 0}, -- to be added to {{citation}} / COinS &rft.part= title = {id = "P1476", rank="p n"}, -- url = {id = "P953", maxvals = 1}, -- deal with this along with archive-url pages = {id = "P304", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, at = {id = "P958", maxvals = 0, populate_from_journal = true, no_vol_qual=true}, -- also incorporate lines (P7421) and columns (P3903) into this (cite map also supports |section=) -- sheets = {id = "P7416", maxvals = 0, populate_from_journal = true}, -- interviewer = {id = "P?", maxvals = 0}, -- does **not** go to "others" section! Multiple interviewers should be n-enumerated illustrator = {id = "P110", maxvals = 10, others = true}, -- goes to "others" section -- foreword and afterword, when contributions to another author's work, are contributions so belong in |contribution=; -- the writer's name goes in |contributor=; requires |title= and |author= -- However, this might need to add support for multiple contributors and their roles to {{citation}}, see Help_talk:Citation_Style_1#Others -- foreword = {id = "P2679", maxvals = 10, others = true}, -- goes to "others" section -- afterword = {id = "P2680", maxvals = 10, others = true}, -- goes to "others" section composer = {id = "P86", maxvals = 10, others = true}, -- goes to "others" section animator = {id = "P6942", maxvals = 10, others = true}, -- goes to "others" section director = {id = "P57", maxvals = 10, others = true}, -- goes to "others" section screenwriter = {id = "P58", maxvals = 10, others = true}, -- goes to "others" section signatory = {id = "P1891", maxvals = 10, others = true}, -- goes to "others" section presenter = {id = "P371", maxvals = 10, others = true}, -- goes to "others" section performer = {id = "P175", maxvals = 10, others = true}, -- goes to "others" section -- next 2 lines added for Arabic Wikipedia annotator = {id ="P11105", maxvals = 10, others = true}, -- goes to "others" section reviewer = {id ="P4032", maxvals = 10, others = true}, -- goes to "others" section supervisor = {id ="P7604", maxvals = 3, others = true}, -- goes to "others" section } local used_args_t = {accessdate=true, language=true, mode=true, ol=true, other=true, ['display-authors']=true, ['display-editors']=true, ['archive-url']=true, ['archive-date']=true, ['orig-date']=true, url=true, title=true, page=true, pages = true, volume=true, } for i = 1, 9 do used_args_t['author' .. i] = true used_args_t['author-link' .. i] = true used_args_t['author-first' .. i] = true used_args_t['editor' .. i] = true used_args_t['editor-link' .. i] = true used_args_t['editor-first' .. i] = true used_args_t['translator' .. i] = true used_args_t['translator-link' .. i] = true used_args_t['translator-first' .. i] = true end -- all_used_properties local used_properties_t = { P1680 = true, P1932 = true, P2699 = true, P407 = true, P2093 = true, -- author string P50 = true, -- author P5769 = true, --editor P655 = true, P856 = true, P953 = true, P98 = true, P854 = true, } for k,v in pairs(simple_properties_t) do used_properties_t[v.id] = { arg = k, maxvals = v.maxvals, linked = v.linked, others = v.others, populate_from_journal = v.populate_from_journal } end simple_properties_t['hijri-date'] = {id = "P577", maxvals = 1, hijri=true} return { aliases_t = aliases_t, numbered_aliases_t = numbered_aliases_t, i18n_t = i18n_t, rejected_t = rejected_t, simple_properties_t = simple_properties_t, used_args_t = used_args_t, used_properties_t = used_properties_t, config_t = config_t, citation_class_map_t=citation_class_map_t } 89a3ngigl1bbmkbtq1fgzscdr73s3y4 828 240854 1158943 1158924 2026-09-10T05:22:36Z ForzaGreen 28665 both parked decisions: p154 (يَمَنَ) recorded corrected, 427 (اِهْبَيَّخَ) extracted after adding its نون assimilation; Template A 83/83, 381 paradigms / 13040 cells 1158943 wikitext text/x-wiki {{صفحة توثيق فرعية}} تُصرِّف هذه الوحدة الأفعال العربية انطلاقا من '''جذر''' و'''وزن صرفي''': الماضي والمضارع بأوجهه، والأمر، مبنيةً للمعلوم وللمجهول، ومعها اسم الفاعل واسم المفعول والمصدر. لا تُستدعى مباشرة في المقالات، بل من خلال [[قالب:تصريف]]: <pre> {{تصريف|جذر=كتب|وزن=فعَل يفعُل}} </pre> == نقاط الدخول == ; <code>show</code> : تعرض جدول التصريف. وهي ما يستدعيه [[قالب:تصريف]]. ; <code>do_generate_forms</code> : تُرجع الصيغ كلها في بنية واحدة، لتُفيد منها وحدة أخرى — وعليها تقوم [[وحدة:ar-verb-form]]. ; <code>infer_radicals</code> : تستخرج حروف الجذر من فعل مكتوب. وللاطلاع على الصيغ في هيئة JSON: <pre> {{تصريف|جذر=كتب|وزن=فعَل يفعُل|json=1}} </pre> == ما تُعالجه == الأوزان الثلاثية والرباعية كلها، مجردةً ومزيدة — وجدولها في [[قالب:تصريف/شرح]]. ومعها المعتل بأنواعه (المثال والأجوف والناقص واللفيف)، والمضعَّف، والمهموز، والأفعال الشاذة مثل [[رَأَى]] و[[أَكَلَ]] و[[أَخَذَ]]. == وسائط أخرى == الوسائط <code>كامل</code> و<code>موسع</code> و<code>بؤرة</code> موصوفة في [[قالب:تصريف/شرح]]. والأخيران وُضعا لصفحات صيغ الأفعال. == الاختبار == الاختبارات موزَّعة على مجموعات، لكلِّ مجموعة صفحةُ نتائجَ تُعرَض عند فتحها: <!-- بداية قائمة مجموعات الاختبار — مولَّدة آليًّا بواسطة tools/make_wiki_testcases.py، لا تُحرَّر يدويًّا --> {| class="wikitable" ! المجموعة !! ما تختبره !! عدد الفحوص !! الصفحات |- | 0 || اختبارات مكتوبة يدويًّا على واجهة الوحدة: استخراج الجذور، وتشكيل المشتقّات، والوسائط والتجاوزات والحواشي، والأفعال الشاذّة || 235 || [[نقاش الوحدة:ar-verb/مختبر|النتائج]] · [[وحدة:ar-verb/مختبر|الوحدة]] |- | 1 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 1394 || [[نقاش الوحدة:ar-verb/testcases/1|النتائج]] · [[وحدة:ar-verb/testcases/1|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/1|البيانات]] |- | 2 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 1423 || [[نقاش الوحدة:ar-verb/testcases/2|النتائج]] · [[وحدة:ar-verb/testcases/2|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/2|البيانات]] |- | 3 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 1190 || [[نقاش الوحدة:ar-verb/testcases/3|النتائج]] · [[وحدة:ar-verb/testcases/3|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/3|البيانات]] |- | 4 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 930 || [[نقاش الوحدة:ar-verb/testcases/4|النتائج]] · [[وحدة:ar-verb/testcases/4|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/4|البيانات]] |- | 5 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 930 || [[نقاش الوحدة:ar-verb/testcases/5|النتائج]] · [[وحدة:ar-verb/testcases/5|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/5|البيانات]] |- | 6 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 930 || [[نقاش الوحدة:ar-verb/testcases/6|النتائج]] · [[وحدة:ar-verb/testcases/6|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/6|البيانات]] |- | 7 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 930 || [[نقاش الوحدة:ar-verb/testcases/7|النتائج]] · [[وحدة:ar-verb/testcases/7|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/7|البيانات]] |- | 8 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 928 || [[نقاش الوحدة:ar-verb/testcases/8|النتائج]] · [[وحدة:ar-verb/testcases/8|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/8|البيانات]] |- | 9 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 926 || [[نقاش الوحدة:ar-verb/testcases/9|النتائج]] · [[وحدة:ar-verb/testcases/9|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/9|البيانات]] |- | 10 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 929 || [[نقاش الوحدة:ar-verb/testcases/10|النتائج]] · [[وحدة:ar-verb/testcases/10|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/10|البيانات]] |- | 11 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 925 || [[نقاش الوحدة:ar-verb/testcases/11|النتائج]] · [[وحدة:ar-verb/testcases/11|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/11|البيانات]] |- | 12 || 30 نموذجًا تصريفيًّا من المرجع المطبوع || 929 || [[نقاش الوحدة:ar-verb/testcases/12|النتائج]] · [[وحدة:ar-verb/testcases/12|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/12|البيانات]] |- | 13 || 21 نموذجًا تصريفيًّا من المرجع المطبوع || 676 || [[نقاش الوحدة:ar-verb/testcases/13|النتائج]] · [[وحدة:ar-verb/testcases/13|الوحدة]] · [[وحدة:ar-verb/testcases/بيانات/13|البيانات]] |- ! المجموع !! 381 نموذجًا تصريفيًّا !! 13275 !! |} <!-- نهاية قائمة مجموعات الاختبار --> والمجموعة 0 مكتوبة يدويًّا وتختبر واجهة الوحدة نفسها؛ وما بعدها مولَّد من [https://github.com/forzagreen/ar-wiktionary-modules المستودع] ومقابَل على «معجم تصريف الأفعال العربية» لأنطوان الدحداح، مكتبة لبنان ناشرون. والتقسيم على مجموعات ليبقى زمن المعالجة في كل صفحة دون الحدّ المسموح به. ولا تُحرَّر الصفحات المولَّدة هنا؛ يُحرَّر المستودع ثم تُولَّد من جديد. ولتجربة استدعاء دون حفظ: [[Special:ExpandTemplates]]. == انظر أيضا == * [[قالب:تصريف]] * [[وحدة:ar-verb-form]] — صفحات الصيغ المصرَّفة، وتقوم على هذه الوحدة * [[قالب:صندوق معلومات فعل]] 44ph2ebaz3wnsdstvladelvzx1vj24o 828 242581 1158944 1158905 2026-09-10T05:22:49Z ForzaGreen 28665 both parked decisions: p154 (يَمَنَ) recorded corrected, 427 (اِهْبَيَّخَ) extracted after adding its نون assimilation; Template A 83/83, 381 paradigms / 13040 cells 1158944 Scribunto text/plain -- بيانات اختبار [[وحدة:ar-verb]] — مولَّدة آليًّا، لا تُحرَّر يدويًّا. -- Auto-generated from the fixtures in github.com/forzagreen/ar-wiktionary-modules. -- Regenerate with tools/make_wiki_testcases.py; edits here are overwritten. -- -- المصدر: أنطوان الدحداح، معجم تصريف الأفعال العربية، مكتبة لبنان ناشرون. -- الصيغ حقائق لغوية مُقابَلة على ذلك المرجع؛ لا يُعاد هنا نصُّ الكتاب ولا ترتيبه. -- -- القيم المنتظرة مكتوبة بإملاء الوحدة لا بإملاء الكتاب حيث اختلفا اختلافًا -- إملائيًّا مقبولًا في الوجهين (ألف التثنية، همزة الوصل، السكون الاختياري). -- الجزء 13 من 13. return { { label = "تَجَفْجَفَ (p418)", root = "جفجف", wazn = "تفعلل", forms = { {"imp_2d", "تَجَفْجَفَا"}, {"imp_2fp", "تَجَفْجَفْنَ"}, {"imp_2fs", "تَجَفْجَفِي"}, {"imp_2mp", "تَجَفْجَفُوا"}, {"imp_2ms", "تَجَفْجَفْ"}, {"ind_1p", "نَتَجَفْجَفُ"}, {"ind_1s", "أَتَجَفْجَفُ"}, {"ind_2d", "تَتَجَفْجَفَانِ"}, {"ind_2fp", "تَتَجَفْجَفْنَ"}, {"ind_2fs", "تَتَجَفْجَفِينَ"}, {"ind_2mp", "تَتَجَفْجَفُونَ"}, {"ind_2ms", "تَتَجَفْجَفُ"}, {"ind_3fd", "تَتَجَفْجَفَانِ"}, {"ind_3fp", "يَتَجَفْجَفْنَ"}, {"ind_3fs", "تَتَجَفْجَفُ"}, {"ind_3md", "يَتَجَفْجَفَانِ"}, {"ind_3mp", "يَتَجَفْجَفُونَ"}, {"ind_3ms", "يَتَجَفْجَفُ"}, {"past_1p", "تَجَفْجَفْنَا"}, {"past_1s", "تَجَفْجَفْتُ"}, {"past_2d", "تَجَفْجَفْتُمَا"}, {"past_2fp", "تَجَفْجَفْتُنَّ"}, {"past_2fs", "تَجَفْجَفْتِ"}, {"past_2mp", "تَجَفْجَفْتُمْ"}, {"past_2ms", "تَجَفْجَفْتَ"}, {"past_3fd", "تَجَفْجَفَتَا"}, {"past_3fp", "تَجَفْجَفْنَ"}, {"past_3fs", "تَجَفْجَفَتْ"}, {"past_3md", "تَجَفْجَفَا"}, {"past_3mp", "تَجَفْجَفُوا"}, {"past_3ms", "تَجَفْجَفَ"}, }, }, { label = "تَرَأْبَلَ (p419)", root = "رأبل", wazn = "تفعلل", forms = { {"imp_2d", "تَرَأْبَلَا"}, {"imp_2fp", "تَرَأْبَلْنَ"}, {"imp_2fs", "تَرَأْبَلِي"}, {"imp_2mp", "تَرَأْبَلُوا"}, {"imp_2ms", "تَرَأْبَلْ"}, {"ind_1p", "نَتَرَأْبَلُ"}, {"ind_1s", "أَتَرَأْبَلُ"}, {"ind_2d", "تَتَرَأْبَلَانِ"}, {"ind_2fp", "تَتَرَأْبَلْنَ"}, {"ind_2fs", "تَتَرَأْبَلِينَ"}, {"ind_2mp", "تَتَرَأْبَلُونَ"}, {"ind_2ms", "تَتَرَأْبَلُ"}, {"ind_3fd", "تَتَرَأْبَلَانِ"}, {"ind_3fp", "يَتَرَأْبَلْنَ"}, {"ind_3fs", "تَتَرَأْبَلُ"}, {"ind_3md", "يَتَرَأْبَلَانِ"}, {"ind_3mp", "يَتَرَأْبَلُونَ"}, {"ind_3ms", "يَتَرَأْبَلُ"}, {"past_1p", "تَرَأْبَلْنَا"}, {"past_1s", "تَرَأْبَلْتُ"}, {"past_2d", "تَرَأْبَلْتُمَا"}, {"past_2fp", "تَرَأْبَلْتُنَّ"}, {"past_2fs", "تَرَأْبَلْتِ"}, {"past_2mp", "تَرَأْبَلْتُمْ"}, {"past_2ms", "تَرَأْبَلْتَ"}, {"past_3fd", "تَرَأْبَلَتَا"}, {"past_3fp", "تَرَأْبَلْنَ"}, {"past_3fs", "تَرَأْبَلَتْ"}, {"past_3md", "تَرَأْبَلَا"}, {"past_3mp", "تَرَأْبَلُوا"}, {"past_3ms", "تَرَأْبَلَ"}, }, }, { label = "تَثَأْلَلَ (p419)", root = "ثألل", wazn = "تفعلل", forms = { {"imp_2d", "تَثَأْلَلَا"}, {"imp_2fp", "تَثَأْلَلْنَ"}, {"imp_2fs", "تَثَأْلَلِي"}, {"imp_2mp", "تَثَأْلَلُوا"}, {"imp_2ms", "تَثَأْلَلْ"}, {"ind_1p", "نَتَثَأْلَلُ"}, {"ind_1s", "أَتَثَأْلَلُ"}, {"ind_2d", "تَتَثَأْلَلَانِ"}, {"ind_2fp", "تَتَثَأْلَلْنَ"}, {"ind_2fs", "تَتَثَأْلَلِينَ"}, {"ind_2mp", "تَتَثَأْلَلُونَ"}, {"ind_2ms", "تَتَثَأْلَلُ"}, {"ind_3fd", "تَتَثَأْلَلَانِ"}, {"ind_3fp", "يَتَثَأْلَلْنَ"}, {"ind_3fs", "تَتَثَأْلَلُ"}, {"ind_3md", "يَتَثَأْلَلَانِ"}, {"ind_3mp", "يَتَثَأْلَلُونَ"}, {"ind_3ms", "يَتَثَأْلَلُ"}, {"past_1p", "تَثَأْلَلْنَا"}, {"past_1s", "تَثَأْلَلْتُ"}, {"past_2d", "تَثَأْلَلْتُمَا"}, {"past_2fp", "تَثَأْلَلْتُنَّ"}, {"past_2fs", "تَثَأْلَلْتِ"}, {"past_2mp", "تَثَأْلَلْتُمْ"}, {"past_2ms", "تَثَأْلَلْتَ"}, {"past_3fd", "تَثَأْلَلَتَا"}, {"past_3fp", "تَثَأْلَلْنَ"}, {"past_3fs", "تَثَأْلَلَتْ"}, {"past_3md", "تَثَأْلَلَا"}, {"past_3mp", "تَثَأْلَلُوا"}, {"past_3ms", "تَثَأْلَلَ"}, }, }, { label = "تَدَرْبَأَ (p420)", root = "دربأ", wazn = "تفعلل", forms = { {"imp_2d", "تَدَرْبَآ"}, {"imp_2fp", "تَدَرْبَأْنَ"}, {"imp_2fs", "تَدَرْبَئِي"}, {"imp_2mp", "تَدَرْبَؤُوا"}, {"imp_2ms", "تَدَرْبَأْ"}, {"ind_1p", "نَتَدَرْبَأُ"}, {"ind_1s", "أَتَدَرْبَأُ"}, {"ind_2d", "تَتَدَرْبَآنِ"}, {"ind_2fp", "تَتَدَرْبَأْنَ"}, {"ind_2fs", "تَتَدَرْبَئِينَ"}, {"ind_2mp", "تَتَدَرْبَؤُونَ"}, {"ind_2ms", "تَتَدَرْبَأُ"}, {"ind_3fd", "تَتَدَرْبَآنِ"}, {"ind_3fp", "يَتَدَرْبَأْنَ"}, {"ind_3fs", "تَتَدَرْبَأُ"}, {"ind_3md", "يَتَدَرْبَآنِ"}, {"ind_3mp", "يَتَدَرْبَؤُونَ"}, {"ind_3ms", "يَتَدَرْبَأُ"}, {"past_1p", "تَدَرْبَأْنَا"}, {"past_1s", "تَدَرْبَأْتُ"}, {"past_2d", "تَدَرْبَأْتُمَا"}, {"past_2fp", "تَدَرْبَأْتُنَّ"}, {"past_2fs", "تَدَرْبَأْتِ"}, {"past_2mp", "تَدَرْبَأْتُمْ"}, {"past_2ms", "تَدَرْبَأْتَ"}, {"past_3fd", "تَدَرْبَأَتَا"}, {"past_3fp", "تَدَرْبَأْنَ"}, {"past_3fs", "تَدَرْبَأَتْ"}, {"past_3md", "تَدَرْبَآ"}, {"past_3mp", "تَدَرْبَؤُوا"}, {"past_3ms", "تَدَرْبَأَ"}, }, }, { label = "تَكَرْثَأَ (p420)", root = "كرثأ", wazn = "تفعلل", forms = { {"imp_2d", "تَكَرْثَآ"}, {"imp_2fp", "تَكَرْثَأْنَ"}, {"imp_2fs", "تَكَرْثَئِي"}, {"imp_2mp", "تَكَرْثَؤُوا"}, {"ind_1p", "نَتَكَرْثَأُ"}, {"ind_1s", "أَتَكَرْثَأُ"}, {"ind_2d", "تَتَكَرْثَآنِ"}, {"ind_2fp", "تَتَكَرْثَأْنَ"}, {"ind_2fs", "تَتَكَرْثَئِينَ"}, {"ind_2mp", "تَتَكَرْثَؤُونَ"}, {"ind_2ms", "تَتَكَرْثَأُ"}, {"ind_3fd", "تَتَكَرْثَآنِ"}, {"ind_3fp", "يَتَكَرْثَأْنَ"}, {"ind_3fs", "تَتَكَرْثَأُ"}, {"ind_3md", "يَتَكَرْثَآنِ"}, {"ind_3mp", "يَتَكَرْثَؤُونَ"}, {"ind_3ms", "يَتَكَرْثَأُ"}, {"past_1p", "تَكَرْثَأْنَا"}, {"past_1s", "تَكَرْثَأْتُ"}, {"past_2d", "تَكَرْثَأْتُمَا"}, {"past_2fp", "تَكَرْثَأْتُنَّ"}, {"past_2fs", "تَكَرْثَأْتِ"}, {"past_2mp", "تَكَرْثَأْتُمْ"}, {"past_2ms", "تَكَرْثَأْتَ"}, {"past_3fd", "تَكَرْثَأَتَا"}, {"past_3fp", "تَكَرْثَأْنَ"}, {"past_3fs", "تَكَرْثَأَتْ"}, {"past_3md", "تَكَرْثَآ"}, {"past_3mp", "تَكَرْثَؤُوا"}, {"past_3ms", "تَكَرْثَأَ"}, }, }, { label = "تَوَكْوَكَ (p421)", root = "وكوك", wazn = "تفعلل", forms = { {"imp_2d", "تَوَكْوَكَا"}, {"imp_2fp", "تَوَكْوَكْنَ"}, {"imp_2fs", "تَوَكْوَكِي"}, {"imp_2mp", "تَوَكْوَكُوا"}, {"imp_2ms", "تَوَكْوَكْ"}, {"ind_1p", "نَتَوَكْوَكُ"}, {"ind_1s", "أَتَوَكْوَكُ"}, {"ind_2d", "تَتَوَكْوَكَانِ"}, {"ind_2fp", "تَتَوَكْوَكْنَ"}, {"ind_2fs", "تَتَوَكْوَكِينَ"}, {"ind_2mp", "تَتَوَكْوَكُونَ"}, {"ind_2ms", "تَتَوَكْوَكُ"}, {"ind_3fd", "تَتَوَكْوَكَانِ"}, {"ind_3fp", "يَتَوَكْوَكْنَ"}, {"ind_3fs", "تَتَوَكْوَكُ"}, {"ind_3md", "يَتَوَكْوَكَانِ"}, {"ind_3mp", "يَتَوَكْوَكُونَ"}, {"ind_3ms", "يَتَوَكْوَكُ"}, {"past_1p", "تَوَكْوَكْنَا"}, {"past_1s", "تَوَكْوَكْتُ"}, {"past_2d", "تَوَكْوَكْتُمَا"}, {"past_2fp", "تَوَكْوَكْتُنَّ"}, {"past_2fs", "تَوَكْوَكْتِ"}, {"past_2mp", "تَوَكْوَكْتُمْ"}, {"past_2ms", "تَوَكْوَكْتَ"}, {"past_3fd", "تَوَكْوَكَتَا"}, {"past_3fp", "تَوَكْوَكْنَ"}, {"past_3fs", "تَوَكْوَكَتْ"}, {"past_3md", "تَوَكْوَكَا"}, {"past_3mp", "تَوَكْوَكُوا"}, {"past_3ms", "تَوَكْوَكَ"}, }, }, { label = "تَوَطْوَطَ (p421)", root = "وطوط", wazn = "تفعلل", forms = { {"imp_2d", "تَوَطْوَطَا"}, {"imp_2fp", "تَوَطْوَطْنَ"}, {"imp_2fs", "تَوَطْوَطِي"}, {"imp_2mp", "تَوَطْوَطُوا"}, {"imp_2ms", "تَوَطْوَطْ"}, {"ind_1p", "نَتَوَطْوَطُ"}, {"ind_1s", "أَتَوَطْوَطُ"}, {"ind_2d", "تَتَوَطْوَطَانِ"}, {"ind_2fp", "تَتَوَطْوَطْنَ"}, {"ind_2fs", "تَتَوَطْوَطِينَ"}, {"ind_2mp", "تَتَوَطْوَطُونَ"}, {"ind_2ms", "تَتَوَطْوَطُ"}, {"ind_3fd", "تَتَوَطْوَطَانِ"}, {"ind_3fp", "يَتَوَطْوَطْنَ"}, {"ind_3fs", "تَتَوَطْوَطُ"}, {"ind_3md", "يَتَوَطْوَطَانِ"}, {"ind_3mp", "يَتَوَطْوَطُونَ"}, {"ind_3ms", "يَتَوَطْوَطُ"}, {"past_1p", "تَوَطْوَطْنَا"}, {"past_1s", "تَوَطْوَطْتُ"}, {"past_2d", "تَوَطْوَطْتُمَا"}, {"past_2fp", "تَوَطْوَطْتُنَّ"}, {"past_2fs", "تَوَطْوَطْتِ"}, {"past_2mp", "تَوَطْوَطْتُمْ"}, {"past_2ms", "تَوَطْوَطْتَ"}, {"past_3fd", "تَوَطْوَطَتَا"}, {"past_3fp", "تَوَطْوَطْنَ"}, {"past_3fs", "تَوَطْوَطَتْ"}, {"past_3md", "تَوَطْوَطَا"}, {"past_3mp", "تَوَطْوَطُوا"}, {"past_3ms", "تَوَطْوَطَ"}, }, }, { label = "تَجَوْرَبَ (p422)", root = "جورب", wazn = "تفعلل", forms = { {"imp_2d", "تَجَوْرَبَا"}, {"imp_2fp", "تَجَوْرَبْنَ"}, {"imp_2fs", "تَجَوْرَبِي"}, {"imp_2mp", "تَجَوْرَبُوا"}, {"imp_2ms", "تَجَوْرَبْ"}, {"ind_1p", "نَتَجَوْرَبُ"}, {"ind_1s", "أَتَجَوْرَبُ"}, {"ind_2d", "تَتَجَوْرَبَانِ"}, {"ind_2fp", "تَتَجَوْرَبْنَ"}, {"ind_2fs", "تَتَجَوْرَبِينَ"}, {"ind_2mp", "تَتَجَوْرَبُونَ"}, {"ind_2ms", "تَتَجَوْرَبُ"}, {"ind_3fd", "تَتَجَوْرَبَانِ"}, {"ind_3fp", "يَتَجَوْرَبْنَ"}, {"ind_3fs", "تَتَجَوْرَبُ"}, {"ind_3md", "يَتَجَوْرَبَانِ"}, {"ind_3mp", "يَتَجَوْرَبُونَ"}, {"ind_3ms", "يَتَجَوْرَبُ"}, {"past_1p", "تَجَوْرَبْنَا"}, {"past_1s", "تَجَوْرَبْتُ"}, {"past_2d", "تَجَوْرَبْتُمَا"}, {"past_2fp", "تَجَوْرَبْتُنَّ"}, {"past_2fs", "تَجَوْرَبْتِ"}, {"past_2mp", "تَجَوْرَبْتُمْ"}, {"past_2ms", "تَجَوْرَبْتَ"}, {"past_3fd", "تَجَوْرَبَتَا"}, {"past_3fp", "تَجَوْرَبْنَ"}, {"past_3fs", "تَجَوْرَبَتْ"}, {"past_3md", "تَجَوْرَبَا"}, {"past_3mp", "تَجَوْرَبُوا"}, {"past_3ms", "تَجَوْرَبَ"}, }, }, { label = "تَسَرْوَلَ (p423)", root = "سرول", wazn = "تفعلل", forms = { {"imp_2d", "تَسَرْوَلَا"}, {"imp_2fp", "تَسَرْوَلْنَ"}, {"imp_2fs", "تَسَرْوَلِي"}, {"imp_2mp", "تَسَرْوَلُوا"}, {"imp_2ms", "تَسَرْوَلْ"}, {"ind_1p", "نَتَسَرْوَلُ"}, {"ind_1s", "أَتَسَرْوَلُ"}, {"ind_2d", "تَتَسَرْوَلَانِ"}, {"ind_2fp", "تَتَسَرْوَلْنَ"}, {"ind_2fs", "تَتَسَرْوَلِينَ"}, {"ind_2mp", "تَتَسَرْوَلُونَ"}, {"ind_2ms", "تَتَسَرْوَلُ"}, {"ind_3fd", "تَتَسَرْوَلَانِ"}, {"ind_3fp", "يَتَسَرْوَلْنَ"}, {"ind_3fs", "تَتَسَرْوَلُ"}, {"ind_3md", "يَتَسَرْوَلَانِ"}, {"ind_3mp", "يَتَسَرْوَلُونَ"}, {"ind_3ms", "يَتَسَرْوَلُ"}, {"past_1p", "تَسَرْوَلْنَا"}, {"past_1s", "تَسَرْوَلْتُ"}, {"past_2d", "تَسَرْوَلْتُمَا"}, {"past_2fp", "تَسَرْوَلْتُنَّ"}, {"past_2fs", "تَسَرْوَلْتِ"}, {"past_2mp", "تَسَرْوَلْتُمْ"}, {"past_2ms", "تَسَرْوَلْتَ"}, {"past_3fd", "تَسَرْوَلَتَا"}, {"past_3fp", "تَسَرْوَلْنَ"}, {"past_3fs", "تَسَرْوَلَتْ"}, {"past_3md", "تَسَرْوَلَا"}, {"past_3mp", "تَسَرْوَلُوا"}, {"past_3ms", "تَسَرْوَلَ"}, }, }, { label = "تَحَمْيَرَ (p423)", root = "حمير", wazn = "تفعلل", forms = { {"imp_2d", "تَحَمْيَرَا"}, {"imp_2fp", "تَحَمْيَرْنَ"}, {"imp_2fs", "تَحَمْيَرِي"}, {"imp_2mp", "تَحَمْيَرُوا"}, {"imp_2ms", "تَحَمْيَرْ"}, {"ind_1p", "نَتَحَمْيَرُ"}, {"ind_1s", "أَتَحَمْيَرُ"}, {"ind_2d", "تَتَحَمْيَرَانِ"}, {"ind_2fp", "تَتَحَمْيَرْنَ"}, {"ind_2fs", "تَتَحَمْيَرِينَ"}, {"ind_2mp", "تَتَحَمْيَرُونَ"}, {"ind_2ms", "تَتَحَمْيَرُ"}, {"ind_3fd", "تَتَحَمْيَرَانِ"}, {"ind_3fp", "يَتَحَمْيَرْنَ"}, {"ind_3fs", "تَتَحَمْيَرُ"}, {"ind_3md", "يَتَحَمْيَرَانِ"}, {"ind_3mp", "يَتَحَمْيَرُونَ"}, {"ind_3ms", "يَتَحَمْيَرُ"}, {"past_1p", "تَحَمْيَرْنَا"}, {"past_1s", "تَحَمْيَرْتُ"}, {"past_2d", "تَحَمْيَرْتُمَا"}, {"past_2fp", "تَحَمْيَرْتُنَّ"}, {"past_2fs", "تَحَمْيَرْتِ"}, {"past_2mp", "تَحَمْيَرْتُمْ"}, {"past_2ms", "تَحَمْيَرْتَ"}, {"past_3fd", "تَحَمْيَرَتَا"}, {"past_3fp", "تَحَمْيَرْنَ"}, {"past_3fs", "تَحَمْيَرَتْ"}, {"past_3md", "تَحَمْيَرَا"}, {"past_3mp", "تَحَمْيَرُوا"}, {"past_3ms", "تَحَمْيَرَ"}, }, }, { label = "اِحْرَنْجَمَ (p426)", root = "حرجم", wazn = "افعنلل", forms = { {"imp_2d", "اِحْرَنْجِمَا"}, {"imp_2fp", "اِحْرَنْجِمْنَ"}, {"imp_2fs", "اِحْرَنْجِمِي"}, {"imp_2mp", "اِحْرَنْجِمُوا"}, {"imp_2ms", "اِحْرَنْجِمْ"}, {"ind_1p", "نَحْرَنْجِمُ"}, {"ind_1s", "أَحْرَنْجِمُ"}, {"ind_2d", "تَحْرَنْجِمَانِ"}, {"ind_2fp", "تَحْرَنْجِمْنَ"}, {"ind_2fs", "تَحْرَنْجِمِينَ"}, {"ind_2mp", "تَحْرَنْجِمُونَ"}, {"ind_2ms", "تَحْرَنْجِمُ"}, {"ind_3fd", "تَحْرَنْجِمَانِ"}, {"ind_3fp", "يَحْرَنْجِمْنَ"}, {"ind_3fs", "تَحْرَنْجِمُ"}, {"ind_3md", "يَحْرَنْجِمَانِ"}, {"ind_3mp", "يَحْرَنْجِمُونَ"}, {"ind_3ms", "يَحْرَنْجِمُ"}, {"past_1p", "اِحْرَنْجَمْنَا"}, {"past_1s", "اِحْرَنْجَمْتُ"}, {"past_2d", "اِحْرَنْجَمْتُمَا"}, {"past_2fp", "اِحْرَنْجَمْتُنَّ"}, {"past_2fs", "اِحْرَنْجَمْتِ"}, {"past_2mp", "اِحْرَنْجَمْتُمْ"}, {"past_2ms", "اِحْرَنْجَمْتَ"}, {"past_3fd", "اِحْرَنْجَمَتَا"}, {"past_3fp", "اِحْرَنْجَمْنَ"}, {"past_3fs", "اِحْرَنْجَمَتْ"}, {"past_3md", "اِحْرَنْجَمَا"}, {"past_3mp", "اِحْرَنْجَمُوا"}, {"past_3ms", "اِحْرَنْجَمَ"}, }, }, { label = "اِفْرَنْقَعَ (p426)", root = "فرقع", wazn = "افعنلل", forms = { {"imp_2d", "اِفْرَنْقِعَا"}, {"imp_2fp", "اِفْرَنْقِعْنَ"}, {"imp_2fs", "اِفْرَنْقِعِي"}, {"imp_2mp", "اِفْرَنْقِعُوا"}, {"imp_2ms", "اِفْرَنْقِعْ"}, {"ind_1p", "نَفْرَنْقِعُ"}, {"ind_1s", "أَفْرَنْقِعُ"}, {"ind_2d", "تَفْرَنْقِعَانِ"}, {"ind_2fp", "تَفْرَنْقِعْنَ"}, {"ind_2fs", "تَفْرَنْقِعِينَ"}, {"ind_2mp", "تَفْرَنْقِعُونَ"}, {"ind_2ms", "تَفْرَنْقِعُ"}, {"ind_3fd", "تَفْرَنْقِعَانِ"}, {"ind_3fp", "يَفْرَنْقِعْنَ"}, {"ind_3fs", "تَفْرَنْقِعُ"}, {"ind_3md", "يَفْرَنْقِعَانِ"}, {"ind_3mp", "يَفْرَنْقِعُونَ"}, {"ind_3ms", "يَفْرَنْقِعُ"}, {"past_1p", "اِفْرَنْقَعْنَا"}, {"past_1s", "اِفْرَنْقَعْتُ"}, {"past_2d", "اِفْرَنْقَعْتُمَا"}, {"past_2fp", "اِفْرَنْقَعْتُنَّ"}, {"past_2fs", "اِفْرَنْقَعْتِ"}, {"past_2mp", "اِفْرَنْقَعْتُمْ"}, {"past_2ms", "اِفْرَنْقَعْتَ"}, {"past_3fd", "اِفْرَنْقَعَتَا"}, {"past_3fp", "اِفْرَنْقَعْنَ"}, {"past_3fs", "اِفْرَنْقَعَتْ"}, {"past_3md", "اِفْرَنْقَعَا"}, {"past_3mp", "اِفْرَنْقَعُوا"}, {"past_3ms", "اِفْرَنْقَعَ"}, }, }, { label = "اِسْلَنْطَأَ (p427)", root = "سلطأ", wazn = "افعنلل", forms = { {"imp_2d", "اِسْلَنْطِئَا"}, {"imp_2fp", "اِسْلَنْطِئْنَ"}, {"imp_2fs", "اِسْلَنْطِئِي"}, {"imp_2mp", "اِسْلَنْطِئُوا"}, {"imp_2ms", "اِسْلَنْطِئْ"}, {"ind_1p", "نَسْلَنْطِئُ"}, {"ind_1s", "أَسْلَنْطِئُ"}, {"ind_2d", "تَسْلَنْطِئَانِ"}, {"ind_2fp", "تَسْلَنْطِئْنَ"}, {"ind_2fs", "تَسْلَنْطِئِينَ"}, {"ind_2mp", "تَسْلَنْطِئُونَ"}, {"ind_2ms", "تَسْلَنْطِئُ"}, {"ind_3fd", "تَسْلَنْطِئَانِ"}, {"ind_3fp", "يَسْلَنْطِئْنَ"}, {"ind_3fs", "تَسْلَنْطِئُ"}, {"ind_3md", "يَسْلَنْطِئَانِ"}, {"ind_3mp", "يَسْلَنْطِئُونَ"}, {"ind_3ms", "يَسْلَنْطِئُ"}, {"past_1p", "اِسْلَنْطَأْنَا"}, {"past_1s", "اِسْلَنْطَأْتُ"}, {"past_2d", "اِسْلَنْطَأْتُمَا"}, {"past_2fp", "اِسْلَنْطَأْتُنَّ"}, {"past_2fs", "اِسْلَنْطَأْتِ"}, {"past_2mp", "اِسْلَنْطَأْتُمْ"}, {"past_2ms", "اِسْلَنْطَأْتَ"}, {"past_3fd", "اِسْلَنْطَأَتَا"}, {"past_3fp", "اِسْلَنْطَأْنَ"}, {"past_3fs", "اِسْلَنْطَأَتْ"}, {"past_3md", "اِسْلَنْطَآ"}, {"past_3mp", "اِسْلَنْطَؤُوا"}, {"past_3ms", "اِسْلَنْطَأَ"}, }, }, { label = "اِطْلَنْفَأَ (p427)", root = "طلفأ", wazn = "افعنلل", forms = { {"imp_2d", "اِطْلَنْفِئَا"}, {"imp_2fp", "اِطْلَنْفِئْنَ"}, {"imp_2fs", "اِطْلَنْفِئِي"}, {"imp_2mp", "اِطْلَنْفِئُوا"}, {"imp_2ms", "اِطْلَنْفِئْ"}, {"ind_1p", "نَطْلَنْفِئُ"}, {"ind_1s", "أَطْلَنْفِئُ"}, {"ind_2d", "تَطْلَنْفِئَانِ"}, {"ind_2fp", "تَطْلَنْفِئْنَ"}, {"ind_2fs", "تَطْلَنْفِئِينَ"}, {"ind_2mp", "تَطْلَنْفِئُونَ"}, {"ind_2ms", "تَطْلَنْفِئُ"}, {"ind_3fd", "تَطْلَنْفِئَانِ"}, {"ind_3fp", "يَطْلَنْفِئْنَ"}, {"ind_3fs", "تَطْلَنْفِئُ"}, {"ind_3md", "يَطْلَنْفِئَانِ"}, {"ind_3mp", "يَطْلَنْفِئُونَ"}, {"ind_3ms", "يَطْلَنْفِئُ"}, {"past_1p", "اِطْلَنْفَأْنَا"}, {"past_1s", "اِطْلَنْفَأْتُ"}, {"past_2d", "اِطْلَنْفَأْتُمَا"}, {"past_2fp", "اِطْلَنْفَأْتُنَّ"}, {"past_2fs", "اِطْلَنْفَأْتِ"}, {"past_2mp", "اِطْلَنْفَأْتُمْ"}, {"past_2ms", "اِطْلَنْفَأْتَ"}, {"past_3fd", "اِطْلَنْفَأَتَا"}, {"past_3fp", "اِطْلَنْفَأْنَ"}, {"past_3fs", "اِطْلَنْفَأَتْ"}, {"past_3md", "اِطْلَنْفَآ"}, {"past_3mp", "اِطْلَنْفَؤُوا"}, {"past_3ms", "اِطْلَنْفَأَ"}, }, }, { label = "اِقْشَعَرَّ (p431)", root = "قشعر", wazn = "افعللّ", forms = { {"imp_2d", "اِقْشَعِرَّا"}, {"imp_2fp", "اِقْشَعْرِرْنَ"}, {"imp_2fs", "اِقْشَعِرِّي"}, {"imp_2mp", "اِقْشَعِرُّوا"}, {"imp_2ms", "اِقْشَعْرِرْ"}, {"ind_1p", "نَقْشَعِرُّ"}, {"ind_1s", "أَقْشَعِرُّ"}, {"ind_2d", "تَقْشَعِرَّانِ"}, {"ind_2fp", "تَقْشَعْرِرْنَ"}, {"ind_2fs", "تَقْشَعِرِّينَ"}, {"ind_2mp", "تَقْشَعِرُّونَ"}, {"ind_2ms", "تَقْشَعِرُّ"}, {"ind_3fd", "تَقْشَعِرَّانِ"}, {"ind_3fp", "يَقْشَعْرِرْنَ"}, {"ind_3fs", "تَقْشَعِرُّ"}, {"ind_3md", "يَقْشَعِرَّانِ"}, {"ind_3mp", "يَقْشَعِرُّونَ"}, {"ind_3ms", "يَقْشَعِرُّ"}, {"past_1p", "اِقْشَعْرَرْنَا"}, {"past_1s", "اِقْشَعْرَرْتُ"}, {"past_2d", "اِقْشَعْرَرْتُمَا"}, {"past_2fp", "اِقْشَعْرَرْتُنَّ"}, {"past_2fs", "اِقْشَعْرَرْتِ"}, {"past_2mp", "اِقْشَعْرَرْتُمْ"}, {"past_2ms", "اِقْشَعْرَرْتَ"}, {"past_3fd", "اِقْشَعَرَّتَا"}, {"past_3fp", "اِقْشَعْرَرْنَ"}, {"past_3fs", "اِقْشَعَرَّتْ"}, {"past_3md", "اِقْشَعَرَّا"}, {"past_3mp", "اِقْشَعَرُّوا"}, {"past_3ms", "اِقْشَعَرَّ"}, }, }, { label = "اِكْفَهَرَّ (p432)", root = "كفهر", wazn = "افعللّ", forms = { {"imp_2d", "اِكْفَهِرَّا"}, {"imp_2fp", "اِكْفَهْرِرْنَ"}, {"imp_2fs", "اِكْفَهِرِّي"}, {"imp_2mp", "اِكْفَهِرُّوا"}, {"imp_2ms", "اِكْفَهْرِرْ"}, {"ind_1p", "نَكْفَهِرُّ"}, {"ind_1s", "أَكْفَهِرُّ"}, {"ind_2d", "تَكْفَهِرَّانِ"}, {"ind_2fp", "تَكْفَهْرِرْنَ"}, {"ind_2fs", "تَكْفَهِرِّينَ"}, {"ind_2mp", "تَكْفَهِرُّونَ"}, {"ind_2ms", "تَكْفَهِرُّ"}, {"ind_3fd", "تَكْفَهِرَّانِ"}, {"ind_3fp", "يَكْفَهْرِرْنَ"}, {"ind_3fs", "تَكْفَهِرُّ"}, {"ind_3md", "يَكْفَهِرَّانِ"}, {"ind_3mp", "يَكْفَهِرُّونَ"}, {"ind_3ms", "يَكْفَهِرُّ"}, {"past_1p", "اِكْفَهْرَرْنَا"}, {"past_1s", "اِكْفَهْرَرْتُ"}, {"past_2d", "اِكْفَهْرَرْتُمَا"}, {"past_2fp", "اِكْفَهْرَرْتُنَّ"}, {"past_2fs", "اِكْفَهْرَرْتِ"}, {"past_2mp", "اِكْفَهْرَرْتُمْ"}, {"past_2ms", "اِكْفَهْرَرْتَ"}, {"past_3fd", "اِكْفَهَرَّتَا"}, {"past_3fp", "اِكْفَهْرَرْنَ"}, {"past_3fs", "اِكْفَهَرَّتْ"}, {"past_3md", "اِكْفَهَرَّا"}, {"past_3mp", "اِكْفَهَرُّوا"}, {"past_3ms", "اِكْفَهَرَّ"}, }, }, { label = "اِضْمَحَلَّ (p432)", root = "ضمحل", wazn = "افعللّ", forms = { {"imp_2d", "اِضْمَحِلَّا"}, {"imp_2fp", "اِضْمَحْلِلْنَ"}, {"imp_2fs", "اِضْمَحِلِّي"}, {"imp_2mp", "اِضْمَحِلُّوا"}, {"imp_2ms", "اِضْمَحْلِلْ"}, {"ind_1p", "نَضْمَحِلُّ"}, {"ind_1s", "أَضْمَحِلُّ"}, {"ind_2d", "تَضْمَحِلَّانِ"}, {"ind_2fp", "تَضْمَحْلِلْنَ"}, {"ind_2fs", "تَضْمَحِلِّينَ"}, {"ind_2mp", "تَضْمَحِلُّونَ"}, {"ind_2ms", "تَضْمَحِلُّ"}, {"ind_3fd", "تَضْمَحِلَّانِ"}, {"ind_3fp", "يَضْمَحْلِلْنَ"}, {"ind_3fs", "تَضْمَحِلُّ"}, {"ind_3md", "يَضْمَحِلَّانِ"}, {"ind_3mp", "يَضْمَحِلُّونَ"}, {"ind_3ms", "يَضْمَحِلُّ"}, {"past_1p", "اِضْمَحْلَلْنَا"}, {"past_1s", "اِضْمَحْلَلْتُ"}, {"past_2d", "اِضْمَحْلَلْتُمَا"}, {"past_2fp", "اِضْمَحْلَلْتُنَّ"}, {"past_2fs", "اِضْمَحْلَلْتِ"}, {"past_2mp", "اِضْمَحْلَلْتُمْ"}, {"past_2ms", "اِضْمَحْلَلْتَ"}, {"past_3fd", "اِضْمَحَلَّتَا"}, {"past_3fp", "اِضْمَحْلَلْنَ"}, {"past_3fs", "اِضْمَحَلَّتْ"}, {"past_3md", "اِضْمَحَلَّا"}, {"past_3mp", "اِضْمَحَلُّوا"}, {"past_3ms", "اِضْمَحَلَّ"}, }, }, { label = "اِزْرَأَمَّ (p433)", root = "زرأم", wazn = "افعللّ", forms = { {"imp_2d", "اِزْرَئِمَّا"}, {"imp_2fp", "اِزْرَأْمِمْنَ"}, {"imp_2fs", "اِزْرَئِمِّي"}, {"imp_2mp", "اِزْرَئِمُّوا"}, {"imp_2ms", "اِزْرَأْمِمْ"}, {"ind_1p", "نَزْرَئِمُّ"}, {"ind_1s", "أَزْرَئِمُّ"}, {"ind_2d", "تَزْرَئِمَّانِ"}, {"ind_2fp", "تَزْرَأْمِمْنَ"}, {"ind_2fs", "تَزْرَئِمِّينَ"}, {"ind_2mp", "تَزْرَئِمُّونَ"}, {"ind_2ms", "تَزْرَئِمُّ"}, {"ind_3fd", "تَزْرَئِمَّانِ"}, {"ind_3fp", "يَزْرَأْمِمْنَ"}, {"ind_3fs", "تَزْرَئِمُّ"}, {"ind_3md", "يَزْرَئِمَّانِ"}, {"ind_3mp", "يَزْرَئِمُّونَ"}, {"ind_3ms", "يَزْرَئِمُّ"}, {"past_1p", "اِزْرَأْمَمْنَا"}, {"past_1s", "اِزْرَأْمَمْتُ"}, {"past_2d", "اِزْرَأْمَمْتُمَا"}, {"past_2fp", "اِزْرَأْمَمْتُنَّ"}, {"past_2fs", "اِزْرَأْمَمْتِ"}, {"past_2mp", "اِزْرَأْمَمْتُمْ"}, {"past_2ms", "اِزْرَأْمَمْتَ"}, {"past_3fd", "اِزْرَأَمَّتَا"}, {"past_3fp", "اِزْرَأْمَمْنَ"}, {"past_3fs", "اِزْرَأَمَّتْ"}, {"past_3md", "اِزْرَأَمَّا"}, {"past_3mp", "اِزْرَأَمُّوا"}, {"past_3ms", "اِزْرَأَمَّ"}, }, }, { label = "اِطْمَأَنَّ (p433)", root = "طمأن", wazn = "افعللّ", forms = { {"imp_2d", "اِطْمَئِنَّا"}, {"imp_2fp", "اِطْمَأْنِنَّ"}, {"imp_2fs", "اِطْمَئِنِّي"}, {"imp_2mp", "اِطْمَئِنُّوا"}, {"imp_2ms", "اِطْمَأْنِنْ"}, {"ind_1p", "نَطْمَئِنُّ"}, {"ind_1s", "أَطْمَئِنُّ"}, {"ind_2d", "تَطْمَئِنَّانِ"}, {"ind_2fp", "تَطْمَأْنِنَّ"}, {"ind_2fs", "تَطْمَئِنِّينَ"}, {"ind_2mp", "تَطْمَئِنُّونَ"}, {"ind_2ms", "تَطْمَئِنُّ"}, {"ind_3fd", "تَطْمَئِنَّانِ"}, {"ind_3fp", "يَطْمَأْنِنَّ"}, {"ind_3fs", "تَطْمَئِنُّ"}, {"ind_3md", "يَطْمَئِنَّانِ"}, {"ind_3mp", "يَطْمَئِنُّونَ"}, {"ind_3ms", "يَطْمَئِنُّ"}, {"past_1p", "اِطْمَأْنَنَّا"}, {"past_1s", "اِطْمَأْنَنْتُ"}, {"past_2d", "اِطْمَأْنَنْتُمَا"}, {"past_2fp", "اِطْمَأْنَنْتُنَّ"}, {"past_2fs", "اِطْمَأْنَنْتِ"}, {"past_2mp", "اِطْمَأْنَنْتُمْ"}, {"past_2ms", "اِطْمَأْنَنْتَ"}, {"past_3fd", "اِطْمَأَنَّتَا"}, {"past_3fp", "اِطْمَأْنَنَّ"}, {"past_3fs", "اِطْمَأَنَّتْ"}, {"past_3md", "اِطْمَأَنَّا"}, {"past_3mp", "اِطْمَأَنُّوا"}, {"past_3ms", "اِطْمَأَنَّ"}, }, }, { label = "يَمَنَ ـُ (p.137، معتل الفاء مثال)", root = "يمن", wazn = "فعَل يفعُل", pass = true, forms = { {"imp_2d", "اُومُنَا"}, {"imp_2fp", "اُومُنَّ"}, {"imp_2fs", "اُومُنِي"}, {"imp_2mp", "اُومُنُوا"}, {"imp_2ms", "اُومُنْ"}, {"ind_1p", "نَيْمُنُ"}, {"ind_1s", "أَيْمُنُ"}, {"ind_2d", "تَيْمُنَانِ"}, {"ind_2fp", "تَيْمُنَّ"}, {"ind_2fs", "تَيْمُنِينَ"}, {"ind_2mp", "تَيْمُنُونَ"}, {"ind_2ms", "تَيْمُنُ"}, {"ind_3fd", "تَيْمُنَانِ"}, {"ind_3fp", "يَيْمُنَّ"}, {"ind_3fs", "تَيْمُنُ"}, {"ind_3md", "يَيْمُنَانِ"}, {"ind_3mp", "يَيْمُنُونَ"}, {"ind_3ms", "يَيْمُنُ"}, {"ind_pass_1p", "نُومَنُ"}, {"ind_pass_1s", "أُومَنُ"}, {"ind_pass_2d", "تُومَنَانِ"}, {"ind_pass_2fp", "تُومَنَّ"}, {"ind_pass_2fs", "تُومَنِينَ"}, {"ind_pass_2mp", "تُومَنُونَ"}, {"ind_pass_2ms", "تُومَنُ"}, {"ind_pass_3fd", "تُومَنَانِ"}, {"ind_pass_3fp", "يُومَنَّ"}, {"ind_pass_3fs", "تُومَنُ"}, {"ind_pass_3md", "يُومَنَانِ"}, {"ind_pass_3mp", "يُومَنُونَ"}, {"ind_pass_3ms", "يُومَنُ"}, {"past_1p", "يَمَنَّا"}, {"past_1s", "يَمَنْتُ"}, {"past_2d", "يَمَنْتُمَا"}, {"past_2fp", "يَمَنْتُنَّ"}, {"past_2fs", "يَمَنْتِ"}, {"past_2mp", "يَمَنْتُمْ"}, {"past_2ms", "يَمَنْتَ"}, {"past_3fd", "يَمَنَتَا"}, {"past_3fp", "يَمَنَّ"}, {"past_3fs", "يَمَنَتْ"}, {"past_3md", "يَمَنَا"}, {"past_3mp", "يَمَنُوا"}, {"past_3ms", "يَمَنَ"}, {"past_pass_1p", "يُمِنَّا"}, {"past_pass_1s", "يُمِنْتُ"}, {"past_pass_2d", "يُمِنْتُمَا"}, {"past_pass_2fp", "يُمِنْتُنَّ"}, {"past_pass_2fs", "يُمِنْتِ"}, {"past_pass_2mp", "يُمِنْتُمْ"}, {"past_pass_2ms", "يُمِنْتَ"}, {"past_pass_3fd", "يُمِنَتَا"}, {"past_pass_3fp", "يُمِنَّ"}, {"past_pass_3fs", "يُمِنَتْ"}, {"past_pass_3md", "يُمِنَا"}, {"past_pass_3mp", "يُمِنُوا"}, {"past_pass_3ms", "يُمِنَ"}, }, }, { label = "اِهْبَيَّخَ (p.415، افعنلل معتل اللام)", root = "هبيخ", wazn = "افعنلل", reduced = true, forms = { {"imp_2d", "اِهْبَيِّخَا"}, {"imp_2fp", "اِهْبَيِّخْنَ"}, {"imp_2fs", "اِهْبَيِّخِي"}, {"imp_2mp", "اِهْبَيِّخُوا"}, {"imp_2ms", "اِهْبَيِّخْ"}, {"ind_1p", "نَهْبَيِّخُ"}, {"ind_1s", "أَهْبَيِّخُ"}, {"ind_2d", "تَهْبَيِّخَانِ"}, {"ind_2fp", "تَهْبَيِّخْنَ"}, {"ind_2fs", "تَهْبَيِّخِينَ"}, {"ind_2mp", "تَهْبَيِّخُونَ"}, {"ind_2ms", "تَهْبَيِّخُ"}, {"ind_3fd", "تَهْبَيِّخَانِ"}, {"ind_3fp", "يَهْبَيِّخْنَ"}, {"ind_3fs", "تَهْبَيِّخُ"}, {"ind_3md", "يَهْبَيِّخَانِ"}, {"ind_3mp", "يَهْبَيِّخُونَ"}, {"ind_3ms", "يَهْبَيِّخُ"}, {"past_1p", "اِهْبَيَّخْنَا"}, {"past_1s", "اِهْبَيَّخْتُ"}, {"past_2d", "اِهْبَيَّخْتُمَا"}, {"past_2fp", "اِهْبَيَّخْتُنَّ"}, {"past_2fs", "اِهْبَيَّخْتِ"}, {"past_2mp", "اِهْبَيَّخْتُمْ"}, {"past_2ms", "اِهْبَيَّخْتَ"}, {"past_3fd", "اِهْبَيَّخَتَا"}, {"past_3fp", "اِهْبَيَّخْنَ"}, {"past_3fs", "اِهْبَيَّخَتْ"}, {"past_3md", "اِهْبَيَّخَا"}, {"past_3mp", "اِهْبَيَّخُوا"}, {"past_3ms", "اِهْبَيَّخَ"}, }, }, } 4vo2t2qiewnn233jrztiiphyh6l3g7e 0 243418 1158928 2026-09-09T15:11:43Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158928 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1498100}} === المعاني === '''جَظَّ''' {{نصر}} {{معجمية|L1498100|معاني}} === النطق === * {{أصد-عر|جَظَّ}} {{-}} === التصريفات === {{تصريف|جذر=جظظ|وزن=فعَل يفعُل}} === المراجع === {{مراجع}} d7w3c52qdnx8alnttk5ooozipnx2au9 0 243419 1158929 2026-09-09T15:16:10Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158929 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1533951}} === المعاني === '''تَجَعَّسَ''' {{تفعّل}} {{معجمية|L1533951|معاني}} === النطق === * {{أصد-عر|تَجَعَّسَ}} {{-}} === التصريفات === {{تصريف|جذر=جعس|وزن=تفعّل}} === المراجع === {{مراجع}} hx72xlvv5d4iptqymldh300j95uwl1g 0 243420 1158930 2026-09-09T15:19:03Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158930 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1533977}} === المعاني === '''جَفْجَفَ''' {{فعلل}} {{معجمية|L1533977|معاني}} === النطق === * {{أصد-عر|جَفْجَفَ}} {{-}} === التصريفات === {{تصريف|جذر=جفجف|وزن=فعلل}} === المراجع === {{مراجع}} 8amnpyv4e2su2xa3c2732fj5twx0ygy 0 243421 1158931 2026-09-09T15:19:26Z نور إسماعيل 37425 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158931 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1536469}} === المعاني === '''زَرَّدَ''' {{فعّل}} {{معجمية|L1536469|معاني}} === النطق === * {{أصد-عر|زَرَّدَ}} {{-}} === التصريفات === {{تصريف|جذر=زرد|وزن=فعّل}} === المراجع === {{مراجع}} 7p6agoqaeav32fm35vm0tkdjg1zknk1 0 243422 1158932 2026-09-09T15:24:45Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158932 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1533993}} === المعاني === '''أَجْفَى''' {{أفعل}} {{معجمية|L1533993|معاني}} === النطق === * {{أصد-عر|أَجْفَى}} {{-}} === التصريفات === {{تصريف|جذر=جفو|وزن=أفعل}} === المراجع === {{مراجع}} gjgafh5hkcxrlpay280x2cbq8u57uqh 0 243423 1158933 2026-09-09T15:30:47Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158933 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1545117}} === المعاني === '''جَلَأَ''' {{فتح}} {{معجمية|L1545117|معاني}} === النطق === * {{أصد-عر|جَلَأَ}} {{-}} === التصريفات === {{تصريف|جذر=جلأ|وزن=فعَل يفعَل}} === المراجع === {{مراجع}} ex62yuj2llmhpnrrrhmr8pledtugl9k 0 243424 1158934 2026-09-09T15:35:55Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158934 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1545668}} === المعاني === '''جَلَتَ''' {{ضرب}} {{معجمية|L1545668|معاني}} === النطق === * {{أصد-عر|جَلَتَ}} {{-}} === التصريفات === {{تصريف|جذر=جلت|وزن=فعَل يفعِل}} === المراجع === {{مراجع}} gal6l5mcz742nd4pj79xs0mkaa7h05m 0 243425 1158935 2026-09-09T15:56:12Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158935 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1534008}} === المعاني === '''اِجْلَخَّ''' {{افعلّ}} {{معجمية|L1534008|معاني}} === النطق === * {{أصد-عر|اِجْلَخَّ}} {{-}} === التصريفات === {{تصريف|جذر=جلخ|وزن=افعلّ}} === المراجع === {{مراجع}} fs3kdhc9dhfw82ujhpjdqvzjsecet43 0 243426 1158936 2026-09-09T16:03:51Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158936 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1534063}} === المعاني === '''جَمَّزَ''' {{فعّل}} {{معجمية|L1534063|معاني}} === النطق === * {{أصد-عر|جَمَّزَ}} {{-}} === التصريفات === {{تصريف|جذر=جمز|وزن=فعّل}} === المراجع === {{مراجع}} m13gt3dxe597wup9ekz5uudbal8usho 0 243427 1158937 2026-09-09T18:48:55Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158937 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1534061}} === المعاني === '''تَجَمَّرَ''' {{تفعّل}} {{معجمية|L1534061|معاني}} === النطق === * {{أصد-عر|تَجَمَّرَ}} {{-}} === التصريفات === {{تصريف|جذر=جمر|وزن=تفعّل}} === المراجع === {{مراجع}} 25uju5emxrxrdnlvsulmyzztebnught 0 243428 1158938 2026-09-09T18:50:39Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158938 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1544255}} === المعاني === '''جَمِئَ''' {{سمع}} {{معجمية|L1544255|معاني}} === النطق === * {{أصد-عر|جَمِئَ}} {{-}} === التصريفات === {{تصريف|جذر=جمأ|وزن=فعِل يفعَل}} === المراجع === {{مراجع}} oj8twgary5wfc1jvwuayh0o9xdb54i1 0 243429 1158939 2026-09-10T00:10:53Z يزن البلوي 37247 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158939 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1486741}} === المعاني === '''أَسَى''' {{ضرب}} {{معجمية|L1486741|معاني}} === النطق === * {{أصد-عر|أَسَى}} {{-}} === التصريفات === {{تصريف|جذر=أسي|وزن=فعَل يفعِل}} === المراجع === {{مراجع}} bqz0iko2iodvgg0mce1fbm7taiuse3q 0 243430 1158940 2026-09-10T00:13:55Z يزن البلوي 37247 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158940 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1485542}} === المعاني === '''اِنْخَنَقَ''' {{انفعل}} {{معجمية|L1485542|معاني}} === النطق === * {{أصد-عر|اِنْخَنَقَ}} {{-}} === التصريفات === {{تصريف|جذر=خنق|وزن=انفعل}} === المراجع === {{مراجع}} 6vyy0hahhwndlyysd851ioq90z9jy4w 0 243431 1158945 2026-09-10T10:11:02Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158945 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1488938}} === المعاني === '''خَشَّبَ''' {{فعّل}} {{معجمية|L1488938|معاني}} === النطق === * {{أصد-عر|خَشَّبَ}} {{-}} === التصريفات === {{تصريف|جذر=خشب|وزن=فعّل}} === المراجع === {{مراجع}} 1ih4i3ufs552retvwqykq6kw8e9rud4 0 243432 1158946 2026-09-10T10:18:02Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158946 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1534997}} === المعاني === '''اِخْتَشَبَ''' {{افتعل}} {{معجمية|L1534997|معاني}} === النطق === * {{أصد-عر|اِخْتَشَبَ}} {{-}} === التصريفات === {{تصريف|جذر=خشب|وزن=افتعل}} === المراجع === {{مراجع}} cuucdo8pawcsge9ua2eo4kd9afahrux 0 243433 1158947 2026-09-10T10:19:00Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158947 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1534995}} === المعاني === '''خَاسَى''' {{فاعل}} {{معجمية|L1534995|معاني}} === النطق === * {{أصد-عر|خَاسَى}} {{-}} === التصريفات === {{تصريف|جذر=خسو|وزن=فاعل}} === المراجع === {{مراجع}} d0hrlzh35q0f6kobx9fycui31eemq94 0 243434 1158949 2026-09-10T10:28:50Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158949 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1485990}} === المعاني === '''اِنْخَسَفَ''' {{انفعل}} {{معجمية|L1485990|معاني}} === النطق === * {{أصد-عر|اِنْخَسَفَ}} {{-}} === التصريفات === {{تصريف|جذر=خسف|وزن=انفعل}} === المراجع === {{مراجع}} a9u6w2hq1zq4r03x899mnj9lvnovpmn 0 243435 1158951 2026-09-10T10:32:23Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158951 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535008}} === المعاني === '''اِنْخَشَفَ''' {{انفعل}} {{معجمية|L1535008|معاني}} === النطق === * {{أصد-عر|اِنْخَشَفَ}} {{-}} === التصريفات === {{تصريف|جذر=خشف|وزن=انفعل}} === المراجع === {{مراجع}} 9t6nj7r0xhotqzrdaq05ydg7f5w357i 0 243436 1158952 2026-09-10T10:40:23Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158952 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535009}} === المعاني === '''خَاشَفَ''' {{فاعل}} {{معجمية|L1535009|معاني}} === النطق === * {{أصد-عر|خَاشَفَ}} {{-}} === التصريفات === {{تصريف|جذر=خشف|وزن=فاعل}} === المراجع === {{مراجع}} h6u16cj129cpf681m47xj3maewvzpxb 0 243437 1158955 2026-09-10T10:44:31Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158955 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1496964}} === المعاني === '''تَخَضَّبَ''' {{تفعّل}} {{معجمية|L1496964|معاني}} === النطق === * {{أصد-عر|تَخَضَّبَ}} {{-}} === التصريفات === {{تصريف|جذر=خضب|وزن=تفعّل}} === المراجع === {{مراجع}} ivqvuq8xop6xujy1e0oeyh0lz57f3b6 0 243438 1158957 2026-09-10T10:47:58Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158957 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1484501}} === المعاني === '''اِخْضَوْضَبَ''' {{افعوعل}} {{معجمية|L1484501|معاني}} === النطق === * {{أصد-عر|اِخْضَوْضَبَ}} {{-}} === التصريفات === {{تصريف|جذر=خضب|وزن=افعوعل}} === المراجع === {{مراجع}} h4qnd5amzg6o58vjfvirs505tgf6wpu 0 243439 1158959 2026-09-10T10:49:17Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158959 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535038}} === المعاني === '''أَخْضَجَ''' {{أفعل}} {{معجمية|L1535038|معاني}} === النطق === * {{أصد-عر|أَخْضَجَ}} {{-}} === التصريفات === {{تصريف|جذر=خضج|وزن=أفعل}} === المراجع === {{مراجع}} 5qru11ymwom2coajxipidwf7is5zxl8 0 243440 1158960 2026-09-10T10:50:24Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158960 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535039}} === المعاني === '''اِنْخَضَجَ''' {{انفعل}} {{معجمية|L1535039|معاني}} === النطق === * {{أصد-عر|اِنْخَضَجَ}} {{-}} === التصريفات === {{تصريف|جذر=خضج|وزن=انفعل}} === المراجع === {{مراجع}} 447rdwqebb3ttkrxgv2qch3apzfnvef 0 243441 1158961 2026-09-10T10:52:34Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158961 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535042}} === المعاني === '''اِخْتَضَدَ''' {{افتعل}} {{معجمية|L1535042|معاني}} === النطق === * {{أصد-عر|اِخْتَضَدَ}} {{-}} === التصريفات === {{تصريف|جذر=خضد|وزن=افتعل}} === المراجع === {{مراجع}} fwgrywg6xa1c0mear7cbp5zpqyukat5 0 243442 1158963 2026-09-10T11:06:55Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158963 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535043}} === المعاني === '''اِنْخَضَدَ''' {{انفعل}} {{معجمية|L1535043|معاني}} === النطق === * {{أصد-عر|اِنْخَضَدَ}} {{-}} === التصريفات === {{تصريف|جذر=خضد|وزن=انفعل}} === المراجع === {{مراجع}} 9jxb0zg53ruf758431fl3ion090zacf 0 243443 1158964 2026-09-10T11:08:50Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158964 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1483714}} === المعاني === '''اِخْضَارَّ''' {{افعالّ}} {{معجمية|L1483714|معاني}} === النطق === * {{أصد-عر|اِخْضَارَّ}} {{-}} === التصريفات === {{تصريف|جذر=خضر|وزن=افعالّ}} === المراجع === {{مراجع}} 86zmml2busotg353z933d0ypyuf5s4m 0 243444 1158965 2026-09-10T11:13:29Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158965 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1474054}} === المعاني === '''اِخْضَوْضَرَ''' {{افعوعل}} {{معجمية|L1474054|معاني}} === النطق === * {{أصد-عر|اِخْضَوْضَرَ}} {{-}} === التصريفات === {{تصريف|جذر=خضر|وزن=افعوعل}} === المراجع === {{مراجع}} jlegsfzljq9tff0iv37ljo9ddfnskv0 0 243445 1158966 2026-09-10T11:14:57Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158966 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535051}} === المعاني === '''خَضْرَعَ''' {{فعلل}} {{معجمية|L1535051|معاني}} === النطق === * {{أصد-عر|خَضْرَعَ}} {{-}} === التصريفات === {{تصريف|جذر=خضرع|وزن=فعلل}} === المراجع === {{مراجع}} 0xqqhxs7dvcq29g49bf2c720klajsxk 0 243446 1158967 2026-09-10T11:17:08Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158967 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535052}} === المعاني === '''تَخَضْرَعَ''' {{تفعلل}} {{معجمية|L1535052|معاني}} === النطق === * {{أصد-عر|تَخَضْرَعَ}} {{-}} === التصريفات === {{تصريف|جذر=خضرع|وزن=تفعلل}} === المراجع === {{مراجع}} tooi5hl0yrdldbttxrnf0m82mu3jsc9 0 243447 1158968 2026-09-10T11:18:49Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158968 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1535058}} === المعاني === '''تَخَضَّعَ''' {{تفعّل}} {{معجمية|L1535058|معاني}} === النطق === * {{أصد-عر|تَخَضَّعَ}} {{-}} === التصريفات === {{تصريف|جذر=خضع|وزن=تفعّل}} === المراجع === {{مراجع}} 7ywec8hos6z1ctbbp35g9ly0857zos2 0 243448 1158969 2026-09-10T11:23:21Z أبو هشام 20788 إنشاء صفحة فعل من بيانات المعاجم (ar-senses tool) 1158969 wikitext text/x-wiki == {{اللغة|عربية}} == {{صندوق معلومات فعل|L1481199}} === المعاني === '''خَاضَعَ''' {{فاعل}} {{معجمية|L1481199|معاني}} === النطق === * {{أصد-عر|خَاضَعَ}} {{-}} === التصريفات === {{تصريف|جذر=خضع|وزن=فاعل}} === المراجع === {{مراجع}} etyal4dyd2nioi4612zpytsns873x7n