ဝိၵ်ႇသျိၼ်ႇၼရီႇ shnwiktionary https://shn.wiktionary.org/wiki/%E1%80%9D%E1%80%AD%E1%81%B5%E1%80%BA%E1%82%87%E1%80%9E%E1%80%BB%E1%80%AD%E1%81%BC%E1%80%BA%E1%82%87%E1%81%BC%E1%80%9B%E1%80%AE%E1%82%87:%E1%81%BC%E1%82%83%E1%82%88%E1%82%81%E1%80%B0%E1%80%9D%E1%80%BA%E1%82%81%E1%82%85%E1%81%B5%E1%80%BA%E1%82%88 MediaWiki 1.47.0-wmf.20 case-sensitive သိုဝ်ႇၶၢဝ်ႇ ၶိုၵ်ႉတွၼ်း ဢုပ်ႇၵုမ် ၽူႈၸႂ်ႉတိုဝ်း ဢုပ်ႇၵုမ် ၽူႈၸႂ်ႉတိုဝ်း ဝိၵ်ႇသျိၼ်ႇၼရီႇ လွင်ႈဢုပ်ႇဢူဝ်း ဝိၵ်ႇသျိၼ်ႇၼရီႇ ၾၢႆႇ ဢုပ်ႇၵုမ် ၾၢႆႇ မီႇတီႇယႃႇဝီႇၶီႇ ဢုပ်ႇၵုမ် မီႇတီႇယႃႇဝီႇၶီႇ ထႅမ်းပလဵတ်ႉ ဢုပ်ႇၵုမ် ထႅမ်းပလဵတ်ႉ လွင်ႈၸွႆႈထႅမ် ဢုပ်ႇၵုမ် လွင်ႈၸွႆႈထႅမ် ပိူင်ထၢၼ်ႈ ဢုပ်ႇၵုမ် ပိူင်ထၢၼ်ႈ ၵိူၼ်ႇတူ ဢုပ်ႇၵုမ် ၵိူၼ်ႇတူ တွၼ်ႈၸပ်းႁၢင် ဢုပ်ႇၵုမ် တွၼ်ႈၸပ်းႁၢင် တူၼ်းၸၢပ်ႈလႅပ်ႈ ဢုပ်ႇၵုမ် တူၼ်းၸၢပ်ႈလႅပ်ႈ ၶိုၼ်းၵေႃႇသၢင်ႈ ဢုပ်ႇၵုမ် ၶိုၼ်းၵေႃႇသၢင်ႈ TimedText TimedText talk မေႃႇၵျူး ဢုပ်ႇၵုမ် မေႃႇၵျူး Event Event talk 828 20443 293602 259895 2026-09-20T13:09:50Z Saimawnkham 9 293602 Scribunto text/plain local export = {} local pos_functions = {} --[==[ Author from 2020 on: mostly Benwing2, with significant contributions from Theknightwho. Based on a prior version by Rua (by now mostly rewritten), with contributions from Erutuon and others (see history for full attribution). ]==] local force_cat = false -- for testing; if true, categories appear in non-mainspace pages local require = require local require_when_needed = require("Module:require when needed") local en_utilities_module = "Module:en-utilities" local headword_utilities_module = "Module:headword utilities" local headword_module = "Module:headword" local inflection_utilities_module = "Module:inflection utilities" local parse_utilities_module = "Module:parse utilities" local JSON_module = "Module:JSON" local labels_module = "Module:labels" local links_module = "Module:links" local parameters_module = "Module:parameters" local string_utilities_module = "Module:string utilities" local table_module = "Module:table" local utilities_module = "Module:utilities" local yesno_module = "Module:yesno" local iut = require_when_needed(inflection_utilities_module) local put = require_when_needed(parse_utilities_module) local m_headword_utilities = require_when_needed(headword_utilities_module) local add_links_to_multiword_term = require_when_needed(headword_utilities_module, "add_links_to_multiword_term") local add_suffix = require_when_needed(en_utilities_module, "add_suffix") local apply_link_modifiers = require_when_needed(headword_utilities_module, "apply_link_modifiers") local concat = table.concat local deepEquals = require_when_needed(table_module, "deepEquals") local dump = mw.dumpObject local format_categories = require_when_needed(utilities_module, "format_categories") local full_headword = require_when_needed(headword_module, "full_headword") local get_label_info = require_when_needed(labels_module, "get_label_info") local get_link_page = require_when_needed(links_module, "get_link_page") local glossary_link = require_when_needed(headword_utilities_module, "glossary_link") local insert = table.insert local insertIfNot = require_when_needed(table_module, "insertIfNot") local ipairs = ipairs local is_regular_plural = require_when_needed(en_utilities_module, "is_regular_plural") local list_to_set = require_when_needed(table_module, "listToSet") local pairs = pairs local process_params = require_when_needed(parameters_module, "process") local remove = table.remove local remove_links = require_when_needed(links_module, "remove_links") local replacement_escape = require_when_needed(string_utilities_module, "replacement_escape") local shallowCopy = require_when_needed(table_module, "shallowCopy") local singularize = require_when_needed(en_utilities_module, "singularize") local split = require_when_needed(string_utilities_module, "split") local toJSON = require_when_needed(JSON_module, "toJSON") local toNFD = mw.ustring.toNFD local type = type local ulen = require_when_needed(string_utilities_module, "len") local ulower = require_when_needed(string_utilities_module, "lower") local umatch = require_when_needed(string_utilities_module, "match") local u = require_when_needed(string_utilities_module, "char") local ugsub = require_when_needed(string_utilities_module, "gsub") local lang = require("Module:languages").getByCode("en") local langname = lang:getCanonicalName() local list_param = {list = true, disallow_holes = true} local list_allow_holes = {list = true, allow_holes = true} local boolean_param = {type = "boolean"} local function ine(val) if val == "" then return nil else return val end end local function track(page) require("Module:debug/track")("en-headword/" .. page) return true end ------------------------------------------- UTILITY FUNCTIONS ------------------------------------------ -- Parse and return an inflection not requiring additional processing. The raw arguments come from `args[field]`, which -- is parsed for inline modifiers. local function parse_inflection(args, field, is_head) local argfield = field if type(argfield) == "table" then argfield = argfield[1] end return m_headword_utilities.parse_term_list_with_modifiers { paramname = field, forms = args[argfield], splitchar = ",", is_head = is_head, } end -- Insert the parsed inflections in `terms` (as parsed by `parse_inflection`) into `data.inflections`, with label -- `label` and optional accelerator spec `accel`. local function insert_inflection(data, terms, label, accel, no_label) for _, termobj in ipairs(terms) do m_headword_utilities.remove_termobj_field_modifiers(termobj) end m_headword_utilities.insert_inflection { headdata = data, terms = terms, label = label, no_label = no_label, accel = accel and {form = accel} or nil, } end -- Insert a fixed label `label` into the inflections for `data`. If `originating_term` is supplied, copy the decorations -- from it into the fixed label. local function insert_fixed_inflection(data, label, originating_term) m_headword_utilities.insert_fixed_inflection { headdata = data, originating_term = originating_term, label = label, } end -- Parse and insert an inflection not requiring additional processing into `data.inflections`. The raw arguments come -- from `args[field]`, which is parsed for inline modifiers. `label` is the label that the inflections are given; -- `accel` is the accelerator form, or nil. local function parse_and_insert_inflection(data, args, field, label, accel) m_headword_utilities.parse_and_insert_inflection { headdata = data, forms = args[field], paramname = field, splitchar = ",", label = label, accel = accel and {form = accel} or nil, } end -- These functions are used directly in the <> format as well as in the utility functions #2 below. local function compute_double_last_cons_stem(term) local last_cons = term:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("Verb stem '" .. term .. "' must end in a consonant to use ++") end return term .. last_cons end local function compute_plusplus_s_form(term, default_s_form) if term:find("[szx]$") then -- regas -> regasses, derez -> derezzes return compute_double_last_cons_stem(term) .. "es" else return default_s_form end end -- The main entry point. -- This is the only function that can be invoked from a template. function export.show(frame) local iparams = { [1] = true, } local iargs = require("Module:parameters").process(frame.args, iparams) local parargs = frame:getParent().args local poscat = iargs[1] local pos_in_1 = not poscat if pos_in_1 then poscat = ine(parargs[1]) or mw.title.getCurrentTitle().fullText == "Template:en-head" and "interjection" or error("Part of speech must be specified in 1=") poscat = require(headword_module).canonicalize_pos(poscat) end local indexing_poscat = pos_in_1 and "head" or poscat local params = { ["head"] = list_param, ["id"] = true, ["json"] = boolean_param, ["sort"] = true, ["splithyph"] = boolean_param, ["nosplithyph"] = boolean_param, ["hyphspace"] = boolean_param, ["nolink"] = boolean_param, ["nolinkhead"] = {type = "boolean_param", alias_of = "nolink"}, ["suffix"] = boolean_param, ["nosuffix"] = boolean_param, ["nomultiwordcat"] = boolean_param, ["abbr"] = list_param, ["the"] = true, ["def"] = {alias_of = "the"}, ["pagename"] = true, -- for testing } if pos_in_1 then params[1] = {required = true} -- required but ignored as already processed above end local pos_data = pos_functions[indexing_poscat] local pos_func if pos_data then local pos_params = pos_data.params if pos_params then for key, val in pairs(pos_params) do params[key] = val end end pos_func = pos_data.func end local args = process_params(parargs, params) -- Account for unsupported titles, e.g. 'C|N>K' instead of 'Unsupported titles/C through N to K'. local pagename = args.pagename or mw.loadData("Module:headword/data").pagename local user_specified_heads = parse_inflection(args, "head", "is_head") local heads = user_specified_heads local autohead if args.nolink or not pagename:find("[ '%-]") then autohead = pagename else local en_no_split_apostrophe_words = list_to_set { "one's", "someone's", "he's", "she's", "it's", } local en_include_hyphen_prefixes = list_to_set { -- We don't include things that are also words even though they are often (perhaps mostly) prefixes, e.g. -- "be", "counter", "cross", "extra", "half", "mid", "over", "pan", "under". "acro", "acousto", "Afro", "agro", "anarcho", "angio", "Anglo", "ante", "anti", "arch", "auto", "bi", "bio", "cis", "co", "cryo", "crypto", "de", "demi", "eco", "electro", "Euro", "ex", "Greco", "hemi", "hydro", "hyper", "hypo", "infra", "Indo", "inter", "intra", "Judeo", "macro", "meta", "micro", "mini", "multi", "neo", "neuro", "non", "para", "peri", "post", "pre", "pro", "proto", "pseudo", "re", "semi", "sub", "super", "trans", "un", "vice", } local function is_english(term) local title = mw.title.new(term) if title and title.exists then local content = title:getContent() if content and content:find("==ဢိင်းၵလဵတ်ႈ==\n") then return true end end return false end local function en_split_hyphen_when_space(word) if not word:find("-", nil, true) then return nil end if args.hyphspace then return "[[" .. word:gsub("%-+", " ") .. "|" .. word .. "]]" end if args.nosplithyph then return "[[" .. word .. "]]" end if not args.splithyph then local space_word = word:gsub("%-+", " ") if is_english(space_word) then return "[[" .. space_word .. "|" .. word .. "]]" end if is_english(word) then return "[[" .. word .. "]]" end end return nil end local function en_split_apostrophe(word) local base = word:match("^(.*)'s$") if base then return "[[" .. base .. "]][[-'s|'s]]" end -- Only treat final apostrophe as possessive if preceded by something that looks like a plural ending in /z/. -- In particular we don't want to do it for words like [[truckin']]. base = word:match("^(.*[sxz])'$") if base then if base:find("s$") then local sg = singularize(base) if is_english(sg) then return "[[" .. sg .. "|" .. base .. "]][[-'|']]" end end return "[[" .. base .. "]][[-'|']]" end return "[[" .. word .. "]]" end autohead = add_links_to_multiword_term(pagename, { split_hyphen_when_space = en_split_hyphen_when_space, split_apostrophe = en_split_apostrophe, no_split_apostrophe_words = en_no_split_apostrophe_words, include_hyphen_prefixes = en_include_hyphen_prefixes, }) end if not heads[1] then heads = {{term = autohead}} else for _, headobj in ipairs(heads) do local head = headobj.term if head:find("^~") then head = apply_link_modifiers(autohead, head:sub(2), lang) headobj.term = head elseif head:find("^[!?]$") then -- If explicit head= just consists of ! or ?, add it to the end of the default head. headobj.term = autohead .. head end if head == autohead then track("redundant-head") end end end -- handle the=/def= if args.the == "~" then local newheads = {} for _, headobj in ipairs(heads) do local barehead = shallowCopy(headobj) insert(newheads, barehead) headobj.term = "the " .. headobj.term insert(newheads, headobj) end heads = newheads elseif args.the then local the = require(yesno_module)(args.the) if the then for _, headobj in ipairs(heads) do headobj.term = "the " .. headobj.term end end end local data = { lang = lang, pos_category = poscat, categories = {}, heads = heads, user_specified_heads = user_specified_heads, -- We use our own splitting algorithm so the redundant head cat will be inaccurate. no_redundant_head_cat = true, inflections = {}, nomultiwordcat = args.nomultiwordcat, sort_key = args.sort, pagename = pagename, id = args.id, force_cat_output = force_cat, } local function inscat(cat) insert(data.categories, langname .. " " .. cat) end local is_suffix = false if args.suffix or not args.nosuffix and pagename:find("^%-") and not pagename:find("^%-%-") and poscat ~= "suffix forms" then is_suffix = true data.pos_category = "suffixes" local singular_poscat = singularize(poscat) inscat(singular_poscat .. "-forming suffixes") insert(data.inflections, {label = singular_poscat .. "-forming suffix"}) end if pos_func then pos_func(args, data, is_suffix) end local extra_categories = {} if pagename:find("[Qq]") then -- Check for q not followed by u. We want to exclude things like [[13q deletion syndrome]] and [[BFOQ]] that -- don't have a lowercase letter on either side, as well as things like [[& seq.]] and [[acq.]] that are -- abbreviations for words containing a following u. -- -- Approximate range of combining diacritics; we want to remove them so the checks below for -- a lowercase letter next to the q aren't tripped up by diacritics on the letter. local u300 = u(0x0300) local u36F = u(0x036F) local pagename_no_diacritics = ugsub(toNFD(pagename), "[" .. u300 .. "-" .. u36F .. "]", "") if pagename_no_diacritics:find("[Qq][a-tv-z]") or pagename_no_diacritics:find("[a-z]q[^u.]") or pagename_no_diacritics:find("[a-z]q$") then inscat("words containing Q not followed by U") end end -- toNFD performs decomposition, so letters that decompose to an ASCII -- vowel and a diacritic, such as é, are counted as vowels and do not do not -- need to be included in the pattern. if not umatch(ulower(toNFD(pagename)), "[aeiouyæœøəªºαεηιουω]") then inscat("words spelled without vowels") end if pagename:find("yre$") then inscat('ဢၼ်သုတ်းတင်း "-yre"') end if not pagename:find(" ") and ulen(pagename) >= 25 then insert(extra_categories, "Long " .. langname .. " words") end if pagename:find("^[^aeiou ]*a[^aeiou ]*e[^aeiou ]*i[^aeiou ]*o[^aeiou ]*u[^aeiou ]*$") then inscat("words that use all vowels in alphabetical order") end parse_and_insert_inflection(data, args, "abbr", "abbreviation") if args.json then return toJSON(data) end return full_headword(data) .. (extra_categories[1] and format_categories(extra_categories, lang, args.sort) or "") end local function make_default_comparative(word) if word == "good" or word == "well" then return {"better"} elseif word == "bad" or word == "badly" then return {"worse"} elseif word == "far" then return {"further", "farther"} else return {add_suffix(word, "r")} end end local function make_default_superlative(word) if word == "good" or word == "well" then return {"best"} elseif word == "bad" or word == "badly" then return {"worst"} elseif word == "far" then return {"furthest", "farthest"} else return {add_suffix(word, "st.superlative")} end end -- This function does the common work between adjectives and adverbs. local function process_comparative_args(data, args, plpos) local pagename = data.pagename local comps = parse_inflection(args, 1) local sups = parse_inflection(args, "sup") local outcomps, outsups if args.componly then if comps[1] then error("Can't specify comparatives of comparative-only " .. plpos) end insert(data.inflections, {label = glossary_link("comparative") .. " form only"}) insert(data.categories, langname .. " comparative-only " .. plpos) -- Set to empty list so we don't get any comparatives output, but process superlatives if specified. outcomps = {} if not sups[1] then -- Set to empty list so we don't get any superlatives output unless explicitly given. outsups = {} end elseif args.suponly then if comps[1] or sups[1] then error("Can't specify comparatives or superlatives of or superlative-only " .. plpos) end insert(data.inflections, {label = glossary_link("superlative") .. " form only"}) insert(data.categories, langname .. " superlative-only " .. plpos) return end -- If the first parameter is ?, then don't show anything, just return. if comps[1] and comps[1].term == "?" then if comps[2] then error("Can't specify additional comparatives along with '?'") end if sups[1] then error("Can't specify superlatives along with '?' for the comparative") end return end if comps[1] and comps[1].term == "-" then local hyphencomp = remove(comps, 1) -- Remove the "-" but retain for decorations. -- Not (generally) comparable; may occasionally have a comparative if comps[1] then insert_fixed_inflection(data, "not generally <<comparable>>", hyphencomp) elseif not sups[1] then insert_fixed_inflection(data, "not <<comparable>>", hyphencomp) insert(data.categories, plpos .. langname .. " ၸိူဝ်းတႅၵ်ႈၼိူင်းဢမ်ႇလႆႈ ") return else -- No comparative, but a superlative. insert_inflection() will correctly generate 'no comparative' if we -- pass in "-" as the value. outcomps = {hyphencomp} end elseif not comps[1] then comps = {{term = "more"}} end if not outcomps then -- not if we set `outcomps` to "-" above or processed a comparative-only term outcomps = {} -- Go over each parameter given and create a comparative and superlative form. for _, compobj in ipairs(comps) do local comp = compobj.term if comp == "-" then error("Comparative of '-' only allowed as first comparative") end if comp == "+" then comp = "+more" elseif comp == "more" and pagename ~= "many" and pagename ~= "much" then comp = "+more" elseif comp == "further" and pagename ~= "far" then comp = "+further" elseif comp == "better" and pagename ~= "good" and pagename ~= "well" then comp = "+better" elseif comp:find("~") then comp = comp:gsub("~", replacement_escape(pagename)) end compobj.origterm = comp if comp == "+more" then comp = "more [[" .. pagename .. "]]" elseif comp == "+further" then comp = {"further [[" .. pagename .. "]]", "farther [[" .. pagename .. "]]"} elseif comp == "+better" then comp = "better [[" .. pagename .. "]]" elseif comp == "er" then -- Add -er. comp = add_suffix(pagename, "r") elseif comp == "ier" then if pagename:sub(-1) ~= "y" then error("Can't specify 'ier' comparative unless the term ends with 'y': " .. pagename) end comp = pagename:gsub("e?y$", "ier") elseif comp:find("^%+") then local special = m_headword_utilities.get_special_indicator(comp, "noerror") if special then comp = m_headword_utilities.handle_multiword(pagename, special, make_default_comparative) end end if type(comp) == "table" and not comp[2] then comp = comp[1] end if type(comp) == "table" then for i = 1, #comp - 1 do local outobj = shallowCopy(compobj) outobj.term = comp[i] insert(outcomps, outobj) end compobj.term = comp[#comp] insert(outcomps, compobj) else compobj.term = comp insert(outcomps, compobj) end end end if sups[1] and sups[1].term == "-" then if sups[2] then error("Can't specify '-' as superlative followed by further values") end -- No superlative. insert_inflection() will correctly generate 'no superlative' if we pass in "-" as the value. outsups = sups else if not sups[1] then sups = {{term = "+"}} end end -- `outsups` will be set if we set `outsups` to "-" above or processed a comparative-only term without superlatives. if not outsups then outsups = {} local function process_sup(sup, special, supobj, compobj) if special then sup = m_headword_utilities.handle_multiword(pagename, special, make_default_superlative) elseif sup == "-" or sup == "+" then error(("Internal error: Superlative value of '%s' should have been handled earlier"):format(sup)) elseif sup == "+most" then sup = "most [[" .. pagename .. "]]" elseif sup == "+furthest" then sup = {"furthest [[" .. pagename .. "]]", "farthest [[" .. pagename .. "]]"} elseif sup == "+best" then sup = "best [[" .. pagename .. "]]" elseif sup == "est" then -- Add -est. sup = add_suffix(pagename, "st.superlative") elseif sup == "iest" then if pagename:sub(-1) ~= "y" then error("Can't specify 'iest' superlative unless the term ends with 'y': " .. pagename) end sup = pagename:gsub("e?y$", "iest") end if type(sup) == "table" and not sup[2] then sup = sup[1] end if compobj then supobj = shallowCopy(supobj) supobj = m_headword_utilities.combine_termobj_decorations(supobj, compobj) end if type(sup) == "table" then for i = 1, #sup - 1 do local outobj = shallowCopy(supobj) outobj.term = sup[i] insert(outsups, outobj) end supobj.term = sup[#sup] insert(outsups, supobj) else supobj.term = sup insert(outsups, supobj) end end for _, supobj in ipairs(sups) do local sup = supobj.term if sup == "-" then error("Superlative of '-' only allowed as first superlative") end if sup == "+" then if not comps[1] then error("Superlative of '+' can't be specified when there are no comparatives") end for _, compobj in ipairs(comps) do local comp = compobj.origterm local special if comp == "+more" then sup = "+most" elseif comp == "+further" then sup = "+furthest" elseif comp == "+better" then sup = "+best" elseif comp == "er" then sup = "est" elseif comp == "ier" then sup = "iest" else if comp:find("^%+") then special = m_headword_utilities.get_special_indicator(comp, "noerror") end if not special then -- If the full comparative was given, then derive the superlative by replacing -er with -- -est. if comp:sub(-2) == "er" then sup = comp:sub(1, -3) .. "est" else error(("The superlative cannot be derived automatically from comparative '%s' because it doesn't end in -er"):format(comp)) end end end process_sup(sup, special, supobj, compobj) end else local special = m_headword_utilities.get_special_indicator(sup, "noerror") -- Do some work here rather than in process_sup() so we don't end up double-processing a term with a '~' -- in it or a term that happens to be 'most' or similar after substitution of ~ in the comparative. if not special then if sup == "most" and pagename ~= "many" and pagename ~= "much" then sup = "+most" elseif sup == "furthest" and pagename ~= "far" then sup = "+furthest" elseif sup == "best" and pagename ~= "good" and pagename ~= "well" then sup = "+best" elseif sup:find("~") then sup = sup:gsub("~", replacement_escape(pagename)) end end process_sup(sup, special, supobj) end end end insert_inflection(data, outcomps, "<<comparative>>", "comparative") insert_inflection(data, outsups, "<<superlative>>", "superlative") end pos_functions["ၵမ်ႉႁိၵ်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉႁိၵ်ႈ") end, } pos_functions["ၵမ်ႉသၢင်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉသၢင်ႈ") end, } local function escape(str) return (str:gsub("\\([:#])", "\\\\%1") :gsub("[:#]", "\\%0")) end local function canonicalize_plural(pl, pagename, pos) if pl == "+" then return escape(add_suffix(pagename, "s.plural", pos)) elseif pl == "++" then return escape(compute_plusplus_s_form(pagename, add_suffix(pagename, "s.plural", pos))) elseif pl == "*" then return escape(pagename) elseif pl == "ies" then if pagename:sub(-1) == "y" then return escape(pagename:gsub("e?y$", pl)) end error("Can't specify 'ies' plural unless the term ends with 'y'.") elseif pl == "s" or pl == "es" or pl == "'s" then return escape(pagename .. pl) end end local function do_nouns(args, data, pos) local pagename = data.pagename pos = pos or "ႁိၵ်ႈ" local plurals = parse_inflection(args, 1) local function insert_plurale_tantum_inflections(is_plural_only, originating_label) if args.sg[1] then insert_fixed_inflection(data, "normally plural", originating_label) parse_and_insert_inflection(data, args, "sg", "singular") elseif is_plural_only then insert_fixed_inflection(data, "plural only", originating_label) end if args.attr[1] then parse_and_insert_inflection(data, args, "attr", "attributive") end end local function first_pl_term() return plurals[1] and plurals[1].term or nil end if first_pl_term() == "p" then -- plurale tantum if plurals[2] then error("With plurale tantum noun, can't specify more than one plural") end data.genders = {"p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections("plural only", plurals[1]) return end local function inscat(cat) insert(data.categories,"ႁိၵ်ႈ".. langname .. " " .. cat) end local need_default_plural = pos == "ႁိၵ်ႈ" if first_pl_term() == "sp" then -- construed as singular or plural sp = remove(plurals, 1) -- Remove the "sp" but retain it for its decorations. inscat("ဢၼ်ဢဝ်တီႈပွင်ႇဝႃႈ ပဵၼ်တူဝ်လဵဝ် ဢမ်ႇၼၼ် တူဝ်ၼမ်") data.genders = {"s", "p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections(nil, sp) need_default_plural = false elseif first_pl_term() == "-" then -- Uncountable noun; may occasionally have a plural local hyphpl = remove(plurals, 1) -- Remove the "-" but retain for decorations. inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") -- If plural forms were given explicitly, then show "usually" if plurals[1] then insert_fixed_inflection(data, "usually <<uncountable>>", hyphpl) else insert_fixed_inflection(data, "<<uncountable>>", hyphpl) end need_default_plural = false elseif first_pl_term() == "#" then -- Usually countable (e.g., "grilled cheese") local hashpl = remove(plurals, 1) -- Remove the "#" but retain for decorations. insert_fixed_inflection(data, "usually <<countable>>", hashpl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end elseif first_pl_term() == "~" then -- Mixed countable/uncountable noun, always has a plural local tildepl = remove(plurals, 1) -- Remove the "~" but retain for decorations. insert_fixed_inflection(data, "<<countable>> and <<uncountable>>", tildepl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end end -- Plural is unknown if first_pl_term() == "?" then local questionpl = remove(plurals, 1) -- Remove the "?" but retain for decorations. -- Not desired; see [[Wiktionary:Tea_room/2021/August#"Plural unknown or uncertain"]] -- insert_fixed_inflection(data, "plural unknown or uncertain", questionpl) inscat("ဢၼ်ဢမ်ႇႁူႉ ဢမ်ႇၼၼ် ဢမ်ႇတႅတ်ႈၼႅတ်ႈ တူဝ်ၼမ်") if plurals[1] then error("Can't specify explicit plurals along with '?' for unknown/uncertain plural") end return end -- Plural is not attested if first_pl_term() == "!" then local exclampl = remove(plurals, 1) -- Remove the "!" but retain for decorations. insert_fixed_inflection(data, "plural not attested", exclampl) inscat("nouns with unattested plurals") if plurals[1] then error("Can't specify explicit plurals along with '!' for unattested plural") end return end -- If no plural was given, maybe add a default one, otherwise (when "-" was given or proper noun) return. if not plurals[1] then if not need_default_plural then inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") return end plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end -- There are plural forms to show, so show them. inscat("countable nouns") local irregular, indeclinable for i, pl in ipairs(plurals) do local canon_pl = canonicalize_plural(pl.term, pagename, pos) if canon_pl then pl.term = canon_pl end local pl_term = get_link_page(pl.term, lang) if not (pagename:find(" ") or is_regular_plural(pl_term, pagename)) then irregular = true if pl_term == pagename then indeclinable = true end end end if irregular then inscat("ဢၼ်ၵိုၵ်းပႃး တူဝ်ၼမ် ဢၼ်ဢမ်ႇၸႂ်ႈပၵ်းၵဝ်ႇပိူင်ၵဝ်ႇ") end if indeclinable then inscat("ဢၼ်ဢမ်ႇပိၼ်ႇၸွမ်းပိူင်မၼ်း") end insert_inflection(data, plurals, "plural", "p") end -- Return the parameters to be used for nouns and proper nouns. Currently the same. local noun_params = { [1] = list_param, ["pl\1qual"] = {list = true, allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the plural", }, -- The following four only used for pluralia tantum (1=p) ["sg"] = list_param, ["attr"] = list_param, } pos_functions["ႁိၵ်ႈ"] = { params = noun_params, func = do_nouns, } pos_functions["ႁိၵ်ႈသုၼ်ႇလဵဝ်"] = { params = noun_params, func = function(args, data) return do_nouns(args, data, "ႁိၵ်ႈသုၼ်ႇလဵဝ်") end, } local function base_default_verb_forms(verb) return escape(add_suffix(verb, "s.verb")), escape(add_suffix(verb, "ing")), escape(add_suffix(verb, "d")) end local function default_verb_forms(verb) local full_s_form, full_ing_form, full_ed_form = base_default_verb_forms(verb) if verb:find(" ") then local first, rest = verb:match("^(.-)( .*)$") local first_s_form, first_ing_form, first_ed_form = base_default_verb_forms(first) return full_s_form, full_ing_form, full_ed_form, first_s_form .. rest, first_ing_form .. rest, first_ed_form .. rest, first, rest else return full_s_form, full_ing_form, full_ed_form, nil, nil, nil, nil, nil end end local function compute_double_last_cons_stem_of_split_verb(verb, ending) local first, rest = verb:match("^(.-)( .*)$") if not first then error("Verb '" .. verb .. "' must have a space in it to use **") end local last_cons = first:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("First word '" .. first .. "' must end in a consonant to use **") end return first .. last_cons .. ending .. rest end local function check_non_nil_star_form(form, pagename) if form == nil then error("Verb '" .. pagename .. "' must have a space in it to use *, **, *l, *! or *'") end return form end local function sub_tilde(form, pagename) if not form then return nil end if form:find("~") then form = form:gsub("~", replacement_escape(pagename)) end return form end local deprecated_qual_replaced_by_inline_modifier = { list = true, allow_holes = true, replaced_by = false, instead = "use an inline modifier <q:...> or <l:...> on the value" } pos_functions["သၢင်ႈ"] = { params = { [1] = {list = "pres_3sg", disallow_holes = true}, ["pres_3sg\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [2] = {list = "pres_ptc", disallow_holes = true}, ["pres_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [3] = {list = "past", disallow_holes = true}, ["past\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [4] = {list = "past_ptc", allow_holes = true}, ["past_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, ["noautolinkverb"] = boolean_param, ["angle_bracket"] = boolean_param, }, func = function(args, data) -- Get parameters local par1s local par2s = parse_inflection(args, {2, "pres_ptc"}) local par3s = parse_inflection(args, {3, "past"}) local par4s = parse_inflection(args, {4, "past_ptc"}) local pres_3sgs, pres_ptcs, pasts, past_ptcs local pagename = data.pagename ------------------------------------------- UTILITY FUNCTIONS #2 ------------------------------------------ -- These functions are used in both in the separate-parameter format and in the override params such as past_ptc2=. local full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed local lemma local function set_lemma_and_default_forms(the_lemma) lemma = the_lemma full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed, lemma_first, lemma_rest = default_verb_forms(the_lemma) end local function canonicalize_s_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_s elseif form == "*" then return check_non_nil_star_form(split_default_s, lemma) elseif form == "++" then return compute_plusplus_s_form(lemma, full_default_s) elseif form == "**" then if lemma:find("^[^ ]*[szx] ") then return compute_double_last_cons_stem_of_split_verb(lemma, "es") else return check_non_nil_star_form(split_default_s, lemma) end elseif form == "+!" then return lemma .. "s" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "s" .. lemma_rest elseif form == "+'" then return lemma .. "'s" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'s" .. lemma_rest elseif form == "+l" then if lemma:find("[szx]$") then return {{term = full_default_s, l = {"US"}}, {term = compute_plusplus_s_form(lemma, full_default_s), l = {"UK"}}} else return compute_plusplus_s_form(lemma, full_default_s) end elseif form == "*l" then if lemma:find("^[^ ]*[szx] ") then return {{term = check_non_nil_star_form(split_default_s, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "es"), l = {"UK"}}} else return check_non_nil_star_form(split_default_s, lemma) end else return sub_tilde(form, lemma) end end local function canonicalize_ing_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ing elseif form == "*" then return check_non_nil_star_form(split_default_ing, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ing" elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ing") elseif form == "+!" then return lemma .. "ing" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ing" .. lemma_rest elseif form == "+'" then return lemma .. "'ing" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'ing" .. lemma_rest elseif form == "+l" then return {{term = full_default_ing, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ing", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ing, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ing"), l = {"UK"}}} else return sub_tilde(form, lemma) end end local function canonicalize_ed_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ed elseif form == "*" then return check_non_nil_star_form(split_default_ed, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ed" elseif form == "+!" then return lemma .. "ed" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ed" .. lemma_rest elseif form == "+'" then return {{term = lemma .. "'d"}, {term = lemma .. "'ed"}} elseif form == "*'" then return {{term = check_non_nil_star_form(lemma_first) .. "'d" .. lemma_rest}, {term = check_non_nil_star_form(lemma_first) .. "'ed" .. lemma_rest}} elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ed") elseif form == "+l" then return {{term = full_default_ed, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ed", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ed, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ed"), l = {"UK"}}} else return sub_tilde(form, lemma) end end -- FIXME: options should be "+", "*", "++", "**", "+n", "*n", "++n" and "**n", but not "n" local function canonicalize_en_form(form) if form == "n" then track("n4") return add_suffix(lemma, "n") end return canonicalize_ed_form(form) end --------------------------------- MAIN PARSING/CONJUGATING CODE -------------------------------- local is_angle_bracket = args.angle_bracket if is_angle_bracket then if par2s[1] or par3s[1] or par4s[1] then error("Can't specify explicit values for 2=, 3= or 4= along with the angle-bracket format") end elseif is_angle_bracket == nil and not par2s[1] and not par3s[1] and not par4s[1] and not args[1][2] and args[1][1] and args[1][1]:find("<") then if put.term_contains_top_level_html(args[1][1]) then -- Often, term_contains_top_level_html() returns true on the angle-bracket format, which would -- make the pcall() below succeed but leave the angle brackets as-is. Check for this and only do the -- pcall() if term_contains_top_level_html() returns false. is_angle_bracket = true else -- If it's ambiguous whether it's an angle-bracket format or separate params with an inline modifier, -- try to parse as the latter. If an error occurs, treat as the former. local ok ok, par1s = pcall(parse_inflection, args, {1, "pres_3sg"}) if not ok then par1s = nil is_angle_bracket = true end end end if is_angle_bracket then -------------------------- ANGLE-BRACKET FORMAT -------------------------- -- (0) Expand multiword term with angle brackets just on the first word. local arg11 = args[1][1] if arg11:find("^<.*>$") and pagename:find(" ") then local first, rest = pagename:match("^(.-)( .*)$") arg11 = first .. arg11 .. rest end -- (1) Parse the indicator specs inside of angle brackets. local function parse_indicator_spec(angle_bracket_spec) local inside = angle_bracket_spec:match("^<(.*)>$") assert(inside) local segments = put.parse_balanced_segment_run(inside, "[", "]") local comma_separated_groups = put.split_alternating_runs(segments, ",") if #comma_separated_groups > 4 then error("Too many comma-separated parts in indicator spec, expected at most 4: " .. 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 error("Extraneous text after bracketed footnotes: '" .. concat(separated_group) .. "'") end if not footnotes then footnotes = {} end insert(footnotes, separated_group[j]) end return footnotes end local function fetch_specs(comma_separated_group) if not comma_separated_group then return {{term = "+"}} end local specs = {} local colon_separated_groups = put.split_alternating_runs(comma_separated_group, ":") for _, colon_separated_group in ipairs(colon_separated_groups) do local form = colon_separated_group[1] if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error("*, **, *l, *! and *' not allowed inside of indicator specs: " .. angle_bracket_spec) end if form == "" then form = "+" end local termobj = { term = form } local footnotes = fetch_footnotes(colon_separated_group) if footnotes then for _, footnote in ipairs(footnotes) do m_headword_utilities.add_footnote_to_termobj(termobj, footnote) end end insert(specs, termobj) end return specs end local s_specs = fetch_specs(comma_separated_groups[1]) local ing_specs = fetch_specs(comma_separated_groups[2]) local ed_specs = fetch_specs(comma_separated_groups[3]) local en_specs = fetch_specs(comma_separated_groups[4]) return { forms = {}, s_specs = s_specs, ing_specs = ing_specs, ed_specs = ed_specs, en_specs = en_specs, } end local parse_props = { parse_indicator_spec = parse_indicator_spec, } local alternant_multiword_spec = iut.parse_inflected_text(arg11, parse_props) -- (2) Check for user-specified brackets; remove any links from the lemma, but remember the original -- form so we can use it below in the 'lemma_linked' form. -- Check to see if there are brackets in the pre-text or post-text. If so, use the linked lemma (with the -- verb autolinked unless noautolinkverb is given). Otherwise, use the default headword algorithm. local function check_bracket(val) if val:find("%[%[") then alternant_multiword_spec.saw_bracket = true end end for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do check_bracket(alternant_or_word_spec.before_text) if alternant_or_word_spec.alternants then for _, multiword_spec in ipairs(alternant_or_word_spec.alternants) do for _, word_spec in ipairs(multiword_spec.word_specs) do check_bracket(word_spec.before_text) end check_bracket(multiword_spec.post_text) end end end check_bracket(alternant_multiword_spec.post_text) iut.map_word_specs(alternant_multiword_spec, function(base) if base.lemma == "" then base.lemma = pagename end base.orig_lemma = base.lemma base.lemma = remove_links(base.lemma) if args.noautolinkverb or base.orig_lemma:find("%[%[") then base.linked_lemma = base.orig_lemma else base.linked_lemma = "[[" .. base.orig_lemma .. "]]" end end) -- (3) Conjugate the verbs according to the indicator specs parsed above. local all_verb_slots = { lemma = "infinitive", lemma_linked = "infinitive", s_form = "3|s|pres", ing_form = "pres|ptcp", ed_form = "past", en_form = "past|ptcp", } local function conjugate_verb(base) local function process_specs(slot, specs, canon_func, default_values, default_already_formobj) local function insert_termobj_into_slot(termobj) local formobj = m_headword_utilities.convert_termobj_to_formobj(termobj) -- If the form is -, don't insert any forms, which will result in there being no overall forms -- (in fact it will be nil). We check for that down below and substitute a single "-" as the -- form, which in turn gets turned into special labels like "no present participle". if formobj.form == "-" then if formobj.footnotes then error("Unable to preserve footnotes specified on missing form '-': FIXME: " .. dump(formobj.footnotes)) end else iut.insert_form(base.forms, slot, formobj) end end local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg insert_termobj_into_slot(arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_decorations(canon, arg) insert_termobj_into_slot(canon) end end end for _, arg in ipairs(specs) do if arg.term == "+" then if default_values then -- will be nil if past tense specified as - and no past ptc given for _, val in ipairs(default_values) do val = shallowCopy(val) if default_already_formobj then local argformobj = m_headword_utilities.convert_termobj_to_formobj(arg) val.footnotes = iut.combine_footnotes(val.footnotes, argformobj.footnotes) iut.insert_form(base.forms, slot, val) else m_headword_utilities.combine_termobj_decorations(val, arg) canonicalize_and_insert(val) end end end else canonicalize_and_insert(arg) end end end set_lemma_and_default_forms(base.lemma) local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error(("Internal error: '%s' should have already thrown an error"):format(form)) end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_specs("s_form", base.s_specs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_specs("ing_form", base.ing_specs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_specs("ed_form", base.ed_specs, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_specs("en_form", base.en_specs, function(arg) return canonicalize_en_form(arg.term) end, base.forms.ed_form, "default already formobj") iut.insert_form(base.forms, "lemma", {form = base.lemma}) -- Add linked version of lemma for use in head=. We write this in a general fashion in case -- there are multiple lemma forms (which isn't possible currently at this level, although it's -- possible overall using the ((...,...)) notation). iut.insert_forms(base.forms, "lemma_linked", iut.map_forms(base.forms.lemma, function(form) if form == base.lemma and base.linked_lemma:find("%[%[") then return base.linked_lemma else return form end end)) end local inflect_props = { slot_table = all_verb_slots, inflect_word_spec = conjugate_verb, } iut.inflect_multiword_or_alternant_multiword_spec(alternant_multiword_spec, inflect_props) -- (4) Fetch the forms and put the conjugated lemmas in data.heads if not explicitly given. local function fetch_termobjs(slot) local forms = alternant_multiword_spec.forms[slot] -- See above. This should only occur if the user explicitly used - for a spec. if not forms or not forms[1] then return {{term = "-"}} end local termobjs = {} for _, formobj in ipairs(forms) do insert(termobjs, m_headword_utilities.convert_formobj_to_termobj(formobj)) end return termobjs end pres_3sgs = fetch_termobjs("s_form") pres_ptcs = fetch_termobjs("ing_form") pasts = fetch_termobjs("ed_form") past_ptcs = fetch_termobjs("en_form") -- Use the "linked" form of the lemma as the head if no head= explicitly given and the user specified -- brackets in one of the lemmas. Otherwise we use the default headword-linking algorithm. if not data.user_specified_heads[1] and alternant_multiword_spec.saw_bracket then data.heads = {} for _, lemma_obj in ipairs(alternant_multiword_spec.forms.lemma_linked) do insert(data.heads, m_headword_utilities.convert_formobj_to_termobj(lemma_obj)) end end else -------------------------- SEPARATE-PARAM FORMAT -------------------------- set_lemma_and_default_forms(pagename) par1s = par1s or parse_inflection(args, {1, "pres_3sg"}) pres_3sgs = {} pres_ptcs = {} pasts = {} past_ptcs = {} if not par1s[1] then par1s = {{term = "+"}} end if not par2s[1] then par2s = {{term = "+"}} end if not par3s[1] then par3s = {{term = "+"}} end if not par4s[1] then par4s = {{term = "+"}} end local function process_argument(args, dest, canon_func, default_values, default_already_canonicalized) local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg m_headword_utilities.insert_termobj_combining_duplicates(dest, arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_decorations(canon, arg) m_headword_utilities.insert_termobj_combining_duplicates(dest, canon) end end end for _, arg in ipairs(args) do if arg.term == "+" then for _, val in ipairs(default_values) do val = shallowCopy(val) m_headword_utilities.combine_termobj_decorations(val, arg) if default_already_canonicalized then m_headword_utilities.insert_termobj_combining_duplicates(dest, val) else canonicalize_and_insert(val) end end else canonicalize_and_insert(arg) end end end local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" or form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_argument(par1s, pres_3sgs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_argument(par2s, pres_ptcs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_argument(par3s, pasts, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_argument(par4s, past_ptcs, function(arg) return canonicalize_en_form(arg.term) end, pasts, "default already canonicalized") end ------------------------------------------- INSERT INFLECTIONS ------------------------------------------ insert_inflection(data, pres_3sgs, "third-person singular simple present", "s-verb-form") insert_inflection(data, pres_ptcs, "present participle", "ing-form") if deepEquals(pasts, past_ptcs) then insert_inflection(data, pasts, "simple past and past participle", "ed-form", "no simple past or past participle") else insert_inflection(data, pasts, "simple past", "spast") insert_inflection(data, past_ptcs, "past participle", "past|part") end if pagename:find(" ") then -- Check for placeholder "it" local words = split(pagename, " ") for _, word in ipairs(words) do if word == "it" or word == "its" or word == "it's" then insert(data.categories, langname .. ' terms with placeholder "it"') break end end -- Check for phrasal verbs local phrasal_adverbs = list_to_set{ -- NOTE: This should only contain common phrasal adverbs, not random words like [[low]], -- [[adrift]], etc. "aback", "about", "above", "across", "after", "against", "ahead", "along", "apart", "around", "as", "aside", "at", "away", "back", "before", "behind", "below", "between", "beyond", "by", "down", "for", "forth", "from", "in", "into", "of", "off", "on", "onto", "out", "over", "past", "round", "through", "to", "together", "towards", "under", "up", "upon", "with", "without", } local allowed_non_adverb_words = list_to_set{ "it", "one", "oneself", "someone", } local base = pagename local seen_adverbs = {} -- Only consider a verb to be phrasal if it consists of a single base verb followed exclusively by either -- adverbs from `phrasal_adverbs` or placeholder words from `allowed_non_adverb_words`, where at -- least one following word is from `phrasal_adverbs` (hence [[can it]] is not a phrasal verb). while true do local prev, word = base:match("^(.+) (.-)$") if not prev then break end if phrasal_adverbs[word] then insert(seen_adverbs, word) elseif allowed_non_adverb_words[word] then -- do nothing else break end base = prev end if not base:find(" ") and seen_adverbs[1] then insert(data.categories, langname .. " phrasal verbs") for i = #seen_adverbs, 1, -1 do insert(data.categories,'ပွင်ႈၵႂၢမ်းသၢင်ႈ' .. langname .. ' ဢၼ်ၽွတ်ႈႁႅၼ်းဝႆႉတင်း "' .. seen_adverbs[i] .. '"') end end end end, } ----------------------------------------------------------------------------------------- -- Suffix forms -- ----------------------------------------------------------------------------------------- pos_functions["suffix forms"] = { params = { [1] = {required = true, list = true, disallow_holes = true}, }, func = function(args, data, is_suffix) local suffix_type = {} for _, typ in ipairs(args[1]) do insert(suffix_type, typ .. "-forming suffix") end insert(data.inflections, {label = "non-lemma form of " .. m_table.serialCommaJoin(suffix_type, {conj = "or"})}) end, } return export l4rjmsahe5cnjlvrl6ckc72pcpzhvxk 293603 293602 2026-09-20T13:10:59Z Saimawnkham 9 293603 Scribunto text/plain local export = {} local pos_functions = {} --[==[ Author from 2020 on: mostly Benwing2, with significant contributions from Theknightwho. Based on a prior version by Rua (by now mostly rewritten), with contributions from Erutuon and others (see history for full attribution). ]==] local force_cat = false -- for testing; if true, categories appear in non-mainspace pages local require = require local require_when_needed = require("Module:require when needed") local en_utilities_module = "Module:en-utilities" local headword_utilities_module = "Module:headword utilities" local headword_module = "Module:headword" local inflection_utilities_module = "Module:inflection utilities" local parse_utilities_module = "Module:parse utilities" local JSON_module = "Module:JSON" local labels_module = "Module:labels" local links_module = "Module:links" local parameters_module = "Module:parameters" local string_utilities_module = "Module:string utilities" local table_module = "Module:table" local utilities_module = "Module:utilities" local yesno_module = "Module:yesno" local iut = require_when_needed(inflection_utilities_module) local put = require_when_needed(parse_utilities_module) local m_headword_utilities = require_when_needed(headword_utilities_module) local add_links_to_multiword_term = require_when_needed(headword_utilities_module, "add_links_to_multiword_term") local add_suffix = require_when_needed(en_utilities_module, "add_suffix") local apply_link_modifiers = require_when_needed(headword_utilities_module, "apply_link_modifiers") local concat = table.concat local deepEquals = require_when_needed(table_module, "deepEquals") local dump = mw.dumpObject local format_categories = require_when_needed(utilities_module, "format_categories") local full_headword = require_when_needed(headword_module, "full_headword") local get_label_info = require_when_needed(labels_module, "get_label_info") local get_link_page = require_when_needed(links_module, "get_link_page") local glossary_link = require_when_needed(headword_utilities_module, "glossary_link") local insert = table.insert local insertIfNot = require_when_needed(table_module, "insertIfNot") local ipairs = ipairs local is_regular_plural = require_when_needed(en_utilities_module, "is_regular_plural") local list_to_set = require_when_needed(table_module, "listToSet") local pairs = pairs local process_params = require_when_needed(parameters_module, "process") local remove = table.remove local remove_links = require_when_needed(links_module, "remove_links") local replacement_escape = require_when_needed(string_utilities_module, "replacement_escape") local shallowCopy = require_when_needed(table_module, "shallowCopy") local singularize = require_when_needed(en_utilities_module, "singularize") local split = require_when_needed(string_utilities_module, "split") local toJSON = require_when_needed(JSON_module, "toJSON") local toNFD = mw.ustring.toNFD local type = type local ulen = require_when_needed(string_utilities_module, "len") local ulower = require_when_needed(string_utilities_module, "lower") local umatch = require_when_needed(string_utilities_module, "match") local u = require_when_needed(string_utilities_module, "char") local ugsub = require_when_needed(string_utilities_module, "gsub") local lang = require("Module:languages").getByCode("en") local langname = lang:getCanonicalName() local list_param = {list = true, disallow_holes = true} local list_allow_holes = {list = true, allow_holes = true} local boolean_param = {type = "boolean"} local function ine(val) if val == "" then return nil else return val end end local function track(page) require("Module:debug/track")("en-headword/" .. page) return true end ------------------------------------------- UTILITY FUNCTIONS ------------------------------------------ -- Parse and return an inflection not requiring additional processing. The raw arguments come from `args[field]`, which -- is parsed for inline modifiers. local function parse_inflection(args, field, is_head) local argfield = field if type(argfield) == "table" then argfield = argfield[1] end return m_headword_utilities.parse_term_list_with_modifiers { paramname = field, forms = args[argfield], splitchar = ",", is_head = is_head, } end -- Insert the parsed inflections in `terms` (as parsed by `parse_inflection`) into `data.inflections`, with label -- `label` and optional accelerator spec `accel`. local function insert_inflection(data, terms, label, accel, no_label) for _, termobj in ipairs(terms) do m_headword_utilities.remove_termobj_field_modifiers(termobj) end m_headword_utilities.insert_inflection { headdata = data, terms = terms, label = label, no_label = no_label, accel = accel and {form = accel} or nil, } end -- Insert a fixed label `label` into the inflections for `data`. If `originating_term` is supplied, copy the decorations -- from it into the fixed label. local function insert_fixed_inflection(data, label, originating_term) m_headword_utilities.insert_fixed_inflection { headdata = data, originating_term = originating_term, label = label, } end -- Parse and insert an inflection not requiring additional processing into `data.inflections`. The raw arguments come -- from `args[field]`, which is parsed for inline modifiers. `label` is the label that the inflections are given; -- `accel` is the accelerator form, or nil. local function parse_and_insert_inflection(data, args, field, label, accel) m_headword_utilities.parse_and_insert_inflection { headdata = data, forms = args[field], paramname = field, splitchar = ",", label = label, accel = accel and {form = accel} or nil, } end -- These functions are used directly in the <> format as well as in the utility functions #2 below. local function compute_double_last_cons_stem(term) local last_cons = term:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("Verb stem '" .. term .. "' must end in a consonant to use ++") end return term .. last_cons end local function compute_plusplus_s_form(term, default_s_form) if term:find("[szx]$") then -- regas -> regasses, derez -> derezzes return compute_double_last_cons_stem(term) .. "es" else return default_s_form end end -- The main entry point. -- This is the only function that can be invoked from a template. function export.show(frame) local iparams = { [1] = true, } local iargs = require("Module:parameters").process(frame.args, iparams) local parargs = frame:getParent().args local poscat = iargs[1] local pos_in_1 = not poscat if pos_in_1 then poscat = ine(parargs[1]) or mw.title.getCurrentTitle().fullText == "Template:en-head" and "interjection" or error("Part of speech must be specified in 1=") poscat = require(headword_module).canonicalize_pos(poscat) end local indexing_poscat = pos_in_1 and "head" or poscat local params = { ["head"] = list_param, ["id"] = true, ["json"] = boolean_param, ["sort"] = true, ["splithyph"] = boolean_param, ["nosplithyph"] = boolean_param, ["hyphspace"] = boolean_param, ["nolink"] = boolean_param, ["nolinkhead"] = {type = "boolean_param", alias_of = "nolink"}, ["suffix"] = boolean_param, ["nosuffix"] = boolean_param, ["nomultiwordcat"] = boolean_param, ["abbr"] = list_param, ["the"] = true, ["def"] = {alias_of = "the"}, ["pagename"] = true, -- for testing } if pos_in_1 then params[1] = {required = true} -- required but ignored as already processed above end local pos_data = pos_functions[indexing_poscat] local pos_func if pos_data then local pos_params = pos_data.params if pos_params then for key, val in pairs(pos_params) do params[key] = val end end pos_func = pos_data.func end local args = process_params(parargs, params) -- Account for unsupported titles, e.g. 'C|N>K' instead of 'Unsupported titles/C through N to K'. local pagename = args.pagename or mw.loadData("Module:headword/data").pagename local user_specified_heads = parse_inflection(args, "head", "is_head") local heads = user_specified_heads local autohead if args.nolink or not pagename:find("[ '%-]") then autohead = pagename else local en_no_split_apostrophe_words = list_to_set { "one's", "someone's", "he's", "she's", "it's", } local en_include_hyphen_prefixes = list_to_set { -- We don't include things that are also words even though they are often (perhaps mostly) prefixes, e.g. -- "be", "counter", "cross", "extra", "half", "mid", "over", "pan", "under". "acro", "acousto", "Afro", "agro", "anarcho", "angio", "Anglo", "ante", "anti", "arch", "auto", "bi", "bio", "cis", "co", "cryo", "crypto", "de", "demi", "eco", "electro", "Euro", "ex", "Greco", "hemi", "hydro", "hyper", "hypo", "infra", "Indo", "inter", "intra", "Judeo", "macro", "meta", "micro", "mini", "multi", "neo", "neuro", "non", "para", "peri", "post", "pre", "pro", "proto", "pseudo", "re", "semi", "sub", "super", "trans", "un", "vice", } local function is_english(term) local title = mw.title.new(term) if title and title.exists then local content = title:getContent() if content and content:find("==ဢိင်းၵလဵတ်ႈ==\n") then return true end end return false end local function en_split_hyphen_when_space(word) if not word:find("-", nil, true) then return nil end if args.hyphspace then return "[[" .. word:gsub("%-+", " ") .. "|" .. word .. "]]" end if args.nosplithyph then return "[[" .. word .. "]]" end if not args.splithyph then local space_word = word:gsub("%-+", " ") if is_english(space_word) then return "[[" .. space_word .. "|" .. word .. "]]" end if is_english(word) then return "[[" .. word .. "]]" end end return nil end local function en_split_apostrophe(word) local base = word:match("^(.*)'s$") if base then return "[[" .. base .. "]][[-'s|'s]]" end -- Only treat final apostrophe as possessive if preceded by something that looks like a plural ending in /z/. -- In particular we don't want to do it for words like [[truckin']]. base = word:match("^(.*[sxz])'$") if base then if base:find("s$") then local sg = singularize(base) if is_english(sg) then return "[[" .. sg .. "|" .. base .. "]][[-'|']]" end end return "[[" .. base .. "]][[-'|']]" end return "[[" .. word .. "]]" end autohead = add_links_to_multiword_term(pagename, { split_hyphen_when_space = en_split_hyphen_when_space, split_apostrophe = en_split_apostrophe, no_split_apostrophe_words = en_no_split_apostrophe_words, include_hyphen_prefixes = en_include_hyphen_prefixes, }) end if not heads[1] then heads = {{term = autohead}} else for _, headobj in ipairs(heads) do local head = headobj.term if head:find("^~") then head = apply_link_modifiers(autohead, head:sub(2), lang) headobj.term = head elseif head:find("^[!?]$") then -- If explicit head= just consists of ! or ?, add it to the end of the default head. headobj.term = autohead .. head end if head == autohead then track("redundant-head") end end end -- handle the=/def= if args.the == "~" then local newheads = {} for _, headobj in ipairs(heads) do local barehead = shallowCopy(headobj) insert(newheads, barehead) headobj.term = "the " .. headobj.term insert(newheads, headobj) end heads = newheads elseif args.the then local the = require(yesno_module)(args.the) if the then for _, headobj in ipairs(heads) do headobj.term = "the " .. headobj.term end end end local data = { lang = lang, pos_category = poscat, categories = {}, heads = heads, user_specified_heads = user_specified_heads, -- We use our own splitting algorithm so the redundant head cat will be inaccurate. no_redundant_head_cat = true, inflections = {}, nomultiwordcat = args.nomultiwordcat, sort_key = args.sort, pagename = pagename, id = args.id, force_cat_output = force_cat, } local function inscat(cat) insert(data.categories, langname .. " " .. cat) end local is_suffix = false if args.suffix or not args.nosuffix and pagename:find("^%-") and not pagename:find("^%-%-") and poscat ~= "suffix forms" then is_suffix = true data.pos_category = "suffixes" local singular_poscat = singularize(poscat) inscat(singular_poscat .. "-forming suffixes") insert(data.inflections, {label = singular_poscat .. "-forming suffix"}) end if pos_func then pos_func(args, data, is_suffix) end local extra_categories = {} if pagename:find("[Qq]") then -- Check for q not followed by u. We want to exclude things like [[13q deletion syndrome]] and [[BFOQ]] that -- don't have a lowercase letter on either side, as well as things like [[& seq.]] and [[acq.]] that are -- abbreviations for words containing a following u. -- -- Approximate range of combining diacritics; we want to remove them so the checks below for -- a lowercase letter next to the q aren't tripped up by diacritics on the letter. local u300 = u(0x0300) local u36F = u(0x036F) local pagename_no_diacritics = ugsub(toNFD(pagename), "[" .. u300 .. "-" .. u36F .. "]", "") if pagename_no_diacritics:find("[Qq][a-tv-z]") or pagename_no_diacritics:find("[a-z]q[^u.]") or pagename_no_diacritics:find("[a-z]q$") then inscat("words containing Q not followed by U") end end -- toNFD performs decomposition, so letters that decompose to an ASCII -- vowel and a diacritic, such as é, are counted as vowels and do not do not -- need to be included in the pattern. if not umatch(ulower(toNFD(pagename)), "[aeiouyæœøəªºαεηιουω]") then inscat("words spelled without vowels") end if pagename:find("yre$") then inscat('ဢၼ်သုတ်းတင်း "-yre"') end if not pagename:find(" ") and ulen(pagename) >= 25 then insert(extra_categories, "Long " .. langname .. " words") end if pagename:find("^[^aeiou ]*a[^aeiou ]*e[^aeiou ]*i[^aeiou ]*o[^aeiou ]*u[^aeiou ]*$") then inscat("words that use all vowels in alphabetical order") end parse_and_insert_inflection(data, args, "abbr", "abbreviation") if args.json then return toJSON(data) end return full_headword(data) .. (extra_categories[1] and format_categories(extra_categories, lang, args.sort) or "") end local function make_default_comparative(word) if word == "good" or word == "well" then return {"better"} elseif word == "bad" or word == "badly" then return {"worse"} elseif word == "far" then return {"further", "farther"} else return {add_suffix(word, "r")} end end local function make_default_superlative(word) if word == "good" or word == "well" then return {"best"} elseif word == "bad" or word == "badly" then return {"worst"} elseif word == "far" then return {"furthest", "farthest"} else return {add_suffix(word, "st.superlative")} end end -- This function does the common work between adjectives and adverbs. local function process_comparative_args(data, args, plpos) local pagename = data.pagename local comps = parse_inflection(args, 1) local sups = parse_inflection(args, "sup") local outcomps, outsups if args.componly then if comps[1] then error("Can't specify comparatives of comparative-only " .. plpos) end insert(data.inflections, {label = glossary_link("comparative") .. " form only"}) insert(data.categories, langname .. " comparative-only " .. plpos) -- Set to empty list so we don't get any comparatives output, but process superlatives if specified. outcomps = {} if not sups[1] then -- Set to empty list so we don't get any superlatives output unless explicitly given. outsups = {} end elseif args.suponly then if comps[1] or sups[1] then error("Can't specify comparatives or superlatives of or superlative-only " .. plpos) end insert(data.inflections, {label = glossary_link("superlative") .. " form only"}) insert(data.categories, langname .. " superlative-only " .. plpos) return end -- If the first parameter is ?, then don't show anything, just return. if comps[1] and comps[1].term == "?" then if comps[2] then error("Can't specify additional comparatives along with '?'") end if sups[1] then error("Can't specify superlatives along with '?' for the comparative") end return end if comps[1] and comps[1].term == "-" then local hyphencomp = remove(comps, 1) -- Remove the "-" but retain for decorations. -- Not (generally) comparable; may occasionally have a comparative if comps[1] then insert_fixed_inflection(data, "not generally <<comparable>>", hyphencomp) elseif not sups[1] then insert_fixed_inflection(data, "not <<comparable>>", hyphencomp) insert(data.categories, plpos .. langname .. " ၸိူဝ်းတႅၵ်ႈၼိူင်းဢမ်ႇလႆႈ ") return else -- No comparative, but a superlative. insert_inflection() will correctly generate 'no comparative' if we -- pass in "-" as the value. outcomps = {hyphencomp} end elseif not comps[1] then comps = {{term = "more"}} end if not outcomps then -- not if we set `outcomps` to "-" above or processed a comparative-only term outcomps = {} -- Go over each parameter given and create a comparative and superlative form. for _, compobj in ipairs(comps) do local comp = compobj.term if comp == "-" then error("Comparative of '-' only allowed as first comparative") end if comp == "+" then comp = "+more" elseif comp == "more" and pagename ~= "many" and pagename ~= "much" then comp = "+more" elseif comp == "further" and pagename ~= "far" then comp = "+further" elseif comp == "better" and pagename ~= "good" and pagename ~= "well" then comp = "+better" elseif comp:find("~") then comp = comp:gsub("~", replacement_escape(pagename)) end compobj.origterm = comp if comp == "+more" then comp = "more [[" .. pagename .. "]]" elseif comp == "+further" then comp = {"further [[" .. pagename .. "]]", "farther [[" .. pagename .. "]]"} elseif comp == "+better" then comp = "better [[" .. pagename .. "]]" elseif comp == "er" then -- Add -er. comp = add_suffix(pagename, "r") elseif comp == "ier" then if pagename:sub(-1) ~= "y" then error("Can't specify 'ier' comparative unless the term ends with 'y': " .. pagename) end comp = pagename:gsub("e?y$", "ier") elseif comp:find("^%+") then local special = m_headword_utilities.get_special_indicator(comp, "noerror") if special then comp = m_headword_utilities.handle_multiword(pagename, special, make_default_comparative) end end if type(comp) == "table" and not comp[2] then comp = comp[1] end if type(comp) == "table" then for i = 1, #comp - 1 do local outobj = shallowCopy(compobj) outobj.term = comp[i] insert(outcomps, outobj) end compobj.term = comp[#comp] insert(outcomps, compobj) else compobj.term = comp insert(outcomps, compobj) end end end if sups[1] and sups[1].term == "-" then if sups[2] then error("Can't specify '-' as superlative followed by further values") end -- No superlative. insert_inflection() will correctly generate 'no superlative' if we pass in "-" as the value. outsups = sups else if not sups[1] then sups = {{term = "+"}} end end -- `outsups` will be set if we set `outsups` to "-" above or processed a comparative-only term without superlatives. if not outsups then outsups = {} local function process_sup(sup, special, supobj, compobj) if special then sup = m_headword_utilities.handle_multiword(pagename, special, make_default_superlative) elseif sup == "-" or sup == "+" then error(("Internal error: Superlative value of '%s' should have been handled earlier"):format(sup)) elseif sup == "+most" then sup = "most [[" .. pagename .. "]]" elseif sup == "+furthest" then sup = {"furthest [[" .. pagename .. "]]", "farthest [[" .. pagename .. "]]"} elseif sup == "+best" then sup = "best [[" .. pagename .. "]]" elseif sup == "est" then -- Add -est. sup = add_suffix(pagename, "st.superlative") elseif sup == "iest" then if pagename:sub(-1) ~= "y" then error("Can't specify 'iest' superlative unless the term ends with 'y': " .. pagename) end sup = pagename:gsub("e?y$", "iest") end if type(sup) == "table" and not sup[2] then sup = sup[1] end if compobj then supobj = shallowCopy(supobj) supobj = m_headword_utilities.combine_termobj_decorations(supobj, compobj) end if type(sup) == "table" then for i = 1, #sup - 1 do local outobj = shallowCopy(supobj) outobj.term = sup[i] insert(outsups, outobj) end supobj.term = sup[#sup] insert(outsups, supobj) else supobj.term = sup insert(outsups, supobj) end end for _, supobj in ipairs(sups) do local sup = supobj.term if sup == "-" then error("Superlative of '-' only allowed as first superlative") end if sup == "+" then if not comps[1] then error("Superlative of '+' can't be specified when there are no comparatives") end for _, compobj in ipairs(comps) do local comp = compobj.origterm local special if comp == "+more" then sup = "+most" elseif comp == "+further" then sup = "+furthest" elseif comp == "+better" then sup = "+best" elseif comp == "er" then sup = "est" elseif comp == "ier" then sup = "iest" else if comp:find("^%+") then special = m_headword_utilities.get_special_indicator(comp, "noerror") end if not special then -- If the full comparative was given, then derive the superlative by replacing -er with -- -est. if comp:sub(-2) == "er" then sup = comp:sub(1, -3) .. "est" else error(("The superlative cannot be derived automatically from comparative '%s' because it doesn't end in -er"):format(comp)) end end end process_sup(sup, special, supobj, compobj) end else local special = m_headword_utilities.get_special_indicator(sup, "noerror") -- Do some work here rather than in process_sup() so we don't end up double-processing a term with a '~' -- in it or a term that happens to be 'most' or similar after substitution of ~ in the comparative. if not special then if sup == "most" and pagename ~= "many" and pagename ~= "much" then sup = "+most" elseif sup == "furthest" and pagename ~= "far" then sup = "+furthest" elseif sup == "best" and pagename ~= "good" and pagename ~= "well" then sup = "+best" elseif sup:find("~") then sup = sup:gsub("~", replacement_escape(pagename)) end end process_sup(sup, special, supobj) end end end insert_inflection(data, outcomps, "<<comparative>>", "comparative") insert_inflection(data, outsups, "<<superlative>>", "superlative") end pos_functions["ၵမ်ႉႁိၵ်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉႁိၵ်ႈ") end, } pos_functions["ၵမ်ႉသၢင်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉသၢင်ႈ") end, } local function escape(str) return (str:gsub("\\([:#])", "\\\\%1") :gsub("[:#]", "\\%0")) end local function canonicalize_plural(pl, pagename, pos) if pl == "+" then return escape(add_suffix(pagename, "s.plural", pos)) elseif pl == "++" then return escape(compute_plusplus_s_form(pagename, add_suffix(pagename, "s.plural", pos))) elseif pl == "*" then return escape(pagename) elseif pl == "ies" then if pagename:sub(-1) == "y" then return escape(pagename:gsub("e?y$", pl)) end error("Can't specify 'ies' plural unless the term ends with 'y'.") elseif pl == "s" or pl == "es" or pl == "'s" then return escape(pagename .. pl) end end local function do_nouns(args, data, pos) local pagename = data.pagename pos = pos or "ႁိၵ်ႈ" local plurals = parse_inflection(args, 1) local function insert_plurale_tantum_inflections(is_plural_only, originating_label) if args.sg[1] then insert_fixed_inflection(data, "normally plural", originating_label) parse_and_insert_inflection(data, args, "sg", "singular") elseif is_plural_only then insert_fixed_inflection(data, "plural only", originating_label) end if args.attr[1] then parse_and_insert_inflection(data, args, "attr", "attributive") end end local function first_pl_term() return plurals[1] and plurals[1].term or nil end if first_pl_term() == "p" then -- plurale tantum if plurals[2] then error("With plurale tantum noun, can't specify more than one plural") end data.genders = {"p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections("plural only", plurals[1]) return end local function inscat(cat) insert(data.categories,"ႁိၵ်ႈ".. langname .. " " .. cat) end local need_default_plural = pos == "ႁိၵ်ႈ" if first_pl_term() == "sp" then -- construed as singular or plural sp = remove(plurals, 1) -- Remove the "sp" but retain it for its decorations. inscat("ဢၼ်ဢဝ်တီႈပွင်ႇဝႃႈ ပဵၼ်တူဝ်လဵဝ် ဢမ်ႇၼၼ် တူဝ်ၼမ်") data.genders = {"s", "p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections(nil, sp) need_default_plural = false elseif first_pl_term() == "-" then -- Uncountable noun; may occasionally have a plural local hyphpl = remove(plurals, 1) -- Remove the "-" but retain for decorations. inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") -- If plural forms were given explicitly, then show "usually" if plurals[1] then insert_fixed_inflection(data, "usually <<uncountable>>", hyphpl) else insert_fixed_inflection(data, "<<uncountable>>", hyphpl) end need_default_plural = false elseif first_pl_term() == "#" then -- Usually countable (e.g., "grilled cheese") local hashpl = remove(plurals, 1) -- Remove the "#" but retain for decorations. insert_fixed_inflection(data, "usually <<countable>>", hashpl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end elseif first_pl_term() == "~" then -- Mixed countable/uncountable noun, always has a plural local tildepl = remove(plurals, 1) -- Remove the "~" but retain for decorations. insert_fixed_inflection(data, "<<countable>> and <<uncountable>>", tildepl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end end -- Plural is unknown if first_pl_term() == "?" then local questionpl = remove(plurals, 1) -- Remove the "?" but retain for decorations. -- Not desired; see [[Wiktionary:Tea_room/2021/August#"Plural unknown or uncertain"]] -- insert_fixed_inflection(data, "plural unknown or uncertain", questionpl) inscat("ဢၼ်ဢမ်ႇႁူႉ ဢမ်ႇၼၼ် ဢမ်ႇတႅတ်ႈၼႅတ်ႈ တူဝ်ၼမ်") if plurals[1] then error("Can't specify explicit plurals along with '?' for unknown/uncertain plural") end return end -- Plural is not attested if first_pl_term() == "!" then local exclampl = remove(plurals, 1) -- Remove the "!" but retain for decorations. insert_fixed_inflection(data, "plural not attested", exclampl) inscat("nouns with unattested plurals") if plurals[1] then error("Can't specify explicit plurals along with '!' for unattested plural") end return end -- If no plural was given, maybe add a default one, otherwise (when "-" was given or proper noun) return. if not plurals[1] then if not need_default_plural then inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") return end plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end -- There are plural forms to show, so show them. inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") local irregular, indeclinable for i, pl in ipairs(plurals) do local canon_pl = canonicalize_plural(pl.term, pagename, pos) if canon_pl then pl.term = canon_pl end local pl_term = get_link_page(pl.term, lang) if not (pagename:find(" ") or is_regular_plural(pl_term, pagename)) then irregular = true if pl_term == pagename then indeclinable = true end end end if irregular then inscat("ဢၼ်ၵိုၵ်းပႃး တူဝ်ၼမ် ဢၼ်ဢမ်ႇၸႂ်ႈပၵ်းၵဝ်ႇပိူင်ၵဝ်ႇ") end if indeclinable then inscat("ဢၼ်ဢမ်ႇပိၼ်ႇၸွမ်းပိူင်မၼ်း") end insert_inflection(data, plurals, "plural", "p") end -- Return the parameters to be used for nouns and proper nouns. Currently the same. local noun_params = { [1] = list_param, ["pl\1qual"] = {list = true, allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the plural", }, -- The following four only used for pluralia tantum (1=p) ["sg"] = list_param, ["attr"] = list_param, } pos_functions["ႁိၵ်ႈ"] = { params = noun_params, func = do_nouns, } pos_functions["ႁိၵ်ႈသုၼ်ႇလဵဝ်"] = { params = noun_params, func = function(args, data) return do_nouns(args, data, "ႁိၵ်ႈသုၼ်ႇလဵဝ်") end, } local function base_default_verb_forms(verb) return escape(add_suffix(verb, "s.verb")), escape(add_suffix(verb, "ing")), escape(add_suffix(verb, "d")) end local function default_verb_forms(verb) local full_s_form, full_ing_form, full_ed_form = base_default_verb_forms(verb) if verb:find(" ") then local first, rest = verb:match("^(.-)( .*)$") local first_s_form, first_ing_form, first_ed_form = base_default_verb_forms(first) return full_s_form, full_ing_form, full_ed_form, first_s_form .. rest, first_ing_form .. rest, first_ed_form .. rest, first, rest else return full_s_form, full_ing_form, full_ed_form, nil, nil, nil, nil, nil end end local function compute_double_last_cons_stem_of_split_verb(verb, ending) local first, rest = verb:match("^(.-)( .*)$") if not first then error("Verb '" .. verb .. "' must have a space in it to use **") end local last_cons = first:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("First word '" .. first .. "' must end in a consonant to use **") end return first .. last_cons .. ending .. rest end local function check_non_nil_star_form(form, pagename) if form == nil then error("Verb '" .. pagename .. "' must have a space in it to use *, **, *l, *! or *'") end return form end local function sub_tilde(form, pagename) if not form then return nil end if form:find("~") then form = form:gsub("~", replacement_escape(pagename)) end return form end local deprecated_qual_replaced_by_inline_modifier = { list = true, allow_holes = true, replaced_by = false, instead = "use an inline modifier <q:...> or <l:...> on the value" } pos_functions["သၢင်ႈ"] = { params = { [1] = {list = "pres_3sg", disallow_holes = true}, ["pres_3sg\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [2] = {list = "pres_ptc", disallow_holes = true}, ["pres_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [3] = {list = "past", disallow_holes = true}, ["past\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [4] = {list = "past_ptc", allow_holes = true}, ["past_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, ["noautolinkverb"] = boolean_param, ["angle_bracket"] = boolean_param, }, func = function(args, data) -- Get parameters local par1s local par2s = parse_inflection(args, {2, "pres_ptc"}) local par3s = parse_inflection(args, {3, "past"}) local par4s = parse_inflection(args, {4, "past_ptc"}) local pres_3sgs, pres_ptcs, pasts, past_ptcs local pagename = data.pagename ------------------------------------------- UTILITY FUNCTIONS #2 ------------------------------------------ -- These functions are used in both in the separate-parameter format and in the override params such as past_ptc2=. local full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed local lemma local function set_lemma_and_default_forms(the_lemma) lemma = the_lemma full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed, lemma_first, lemma_rest = default_verb_forms(the_lemma) end local function canonicalize_s_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_s elseif form == "*" then return check_non_nil_star_form(split_default_s, lemma) elseif form == "++" then return compute_plusplus_s_form(lemma, full_default_s) elseif form == "**" then if lemma:find("^[^ ]*[szx] ") then return compute_double_last_cons_stem_of_split_verb(lemma, "es") else return check_non_nil_star_form(split_default_s, lemma) end elseif form == "+!" then return lemma .. "s" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "s" .. lemma_rest elseif form == "+'" then return lemma .. "'s" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'s" .. lemma_rest elseif form == "+l" then if lemma:find("[szx]$") then return {{term = full_default_s, l = {"US"}}, {term = compute_plusplus_s_form(lemma, full_default_s), l = {"UK"}}} else return compute_plusplus_s_form(lemma, full_default_s) end elseif form == "*l" then if lemma:find("^[^ ]*[szx] ") then return {{term = check_non_nil_star_form(split_default_s, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "es"), l = {"UK"}}} else return check_non_nil_star_form(split_default_s, lemma) end else return sub_tilde(form, lemma) end end local function canonicalize_ing_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ing elseif form == "*" then return check_non_nil_star_form(split_default_ing, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ing" elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ing") elseif form == "+!" then return lemma .. "ing" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ing" .. lemma_rest elseif form == "+'" then return lemma .. "'ing" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'ing" .. lemma_rest elseif form == "+l" then return {{term = full_default_ing, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ing", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ing, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ing"), l = {"UK"}}} else return sub_tilde(form, lemma) end end local function canonicalize_ed_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ed elseif form == "*" then return check_non_nil_star_form(split_default_ed, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ed" elseif form == "+!" then return lemma .. "ed" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ed" .. lemma_rest elseif form == "+'" then return {{term = lemma .. "'d"}, {term = lemma .. "'ed"}} elseif form == "*'" then return {{term = check_non_nil_star_form(lemma_first) .. "'d" .. lemma_rest}, {term = check_non_nil_star_form(lemma_first) .. "'ed" .. lemma_rest}} elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ed") elseif form == "+l" then return {{term = full_default_ed, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ed", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ed, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ed"), l = {"UK"}}} else return sub_tilde(form, lemma) end end -- FIXME: options should be "+", "*", "++", "**", "+n", "*n", "++n" and "**n", but not "n" local function canonicalize_en_form(form) if form == "n" then track("n4") return add_suffix(lemma, "n") end return canonicalize_ed_form(form) end --------------------------------- MAIN PARSING/CONJUGATING CODE -------------------------------- local is_angle_bracket = args.angle_bracket if is_angle_bracket then if par2s[1] or par3s[1] or par4s[1] then error("Can't specify explicit values for 2=, 3= or 4= along with the angle-bracket format") end elseif is_angle_bracket == nil and not par2s[1] and not par3s[1] and not par4s[1] and not args[1][2] and args[1][1] and args[1][1]:find("<") then if put.term_contains_top_level_html(args[1][1]) then -- Often, term_contains_top_level_html() returns true on the angle-bracket format, which would -- make the pcall() below succeed but leave the angle brackets as-is. Check for this and only do the -- pcall() if term_contains_top_level_html() returns false. is_angle_bracket = true else -- If it's ambiguous whether it's an angle-bracket format or separate params with an inline modifier, -- try to parse as the latter. If an error occurs, treat as the former. local ok ok, par1s = pcall(parse_inflection, args, {1, "pres_3sg"}) if not ok then par1s = nil is_angle_bracket = true end end end if is_angle_bracket then -------------------------- ANGLE-BRACKET FORMAT -------------------------- -- (0) Expand multiword term with angle brackets just on the first word. local arg11 = args[1][1] if arg11:find("^<.*>$") and pagename:find(" ") then local first, rest = pagename:match("^(.-)( .*)$") arg11 = first .. arg11 .. rest end -- (1) Parse the indicator specs inside of angle brackets. local function parse_indicator_spec(angle_bracket_spec) local inside = angle_bracket_spec:match("^<(.*)>$") assert(inside) local segments = put.parse_balanced_segment_run(inside, "[", "]") local comma_separated_groups = put.split_alternating_runs(segments, ",") if #comma_separated_groups > 4 then error("Too many comma-separated parts in indicator spec, expected at most 4: " .. 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 error("Extraneous text after bracketed footnotes: '" .. concat(separated_group) .. "'") end if not footnotes then footnotes = {} end insert(footnotes, separated_group[j]) end return footnotes end local function fetch_specs(comma_separated_group) if not comma_separated_group then return {{term = "+"}} end local specs = {} local colon_separated_groups = put.split_alternating_runs(comma_separated_group, ":") for _, colon_separated_group in ipairs(colon_separated_groups) do local form = colon_separated_group[1] if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error("*, **, *l, *! and *' not allowed inside of indicator specs: " .. angle_bracket_spec) end if form == "" then form = "+" end local termobj = { term = form } local footnotes = fetch_footnotes(colon_separated_group) if footnotes then for _, footnote in ipairs(footnotes) do m_headword_utilities.add_footnote_to_termobj(termobj, footnote) end end insert(specs, termobj) end return specs end local s_specs = fetch_specs(comma_separated_groups[1]) local ing_specs = fetch_specs(comma_separated_groups[2]) local ed_specs = fetch_specs(comma_separated_groups[3]) local en_specs = fetch_specs(comma_separated_groups[4]) return { forms = {}, s_specs = s_specs, ing_specs = ing_specs, ed_specs = ed_specs, en_specs = en_specs, } end local parse_props = { parse_indicator_spec = parse_indicator_spec, } local alternant_multiword_spec = iut.parse_inflected_text(arg11, parse_props) -- (2) Check for user-specified brackets; remove any links from the lemma, but remember the original -- form so we can use it below in the 'lemma_linked' form. -- Check to see if there are brackets in the pre-text or post-text. If so, use the linked lemma (with the -- verb autolinked unless noautolinkverb is given). Otherwise, use the default headword algorithm. local function check_bracket(val) if val:find("%[%[") then alternant_multiword_spec.saw_bracket = true end end for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do check_bracket(alternant_or_word_spec.before_text) if alternant_or_word_spec.alternants then for _, multiword_spec in ipairs(alternant_or_word_spec.alternants) do for _, word_spec in ipairs(multiword_spec.word_specs) do check_bracket(word_spec.before_text) end check_bracket(multiword_spec.post_text) end end end check_bracket(alternant_multiword_spec.post_text) iut.map_word_specs(alternant_multiword_spec, function(base) if base.lemma == "" then base.lemma = pagename end base.orig_lemma = base.lemma base.lemma = remove_links(base.lemma) if args.noautolinkverb or base.orig_lemma:find("%[%[") then base.linked_lemma = base.orig_lemma else base.linked_lemma = "[[" .. base.orig_lemma .. "]]" end end) -- (3) Conjugate the verbs according to the indicator specs parsed above. local all_verb_slots = { lemma = "infinitive", lemma_linked = "infinitive", s_form = "3|s|pres", ing_form = "pres|ptcp", ed_form = "past", en_form = "past|ptcp", } local function conjugate_verb(base) local function process_specs(slot, specs, canon_func, default_values, default_already_formobj) local function insert_termobj_into_slot(termobj) local formobj = m_headword_utilities.convert_termobj_to_formobj(termobj) -- If the form is -, don't insert any forms, which will result in there being no overall forms -- (in fact it will be nil). We check for that down below and substitute a single "-" as the -- form, which in turn gets turned into special labels like "no present participle". if formobj.form == "-" then if formobj.footnotes then error("Unable to preserve footnotes specified on missing form '-': FIXME: " .. dump(formobj.footnotes)) end else iut.insert_form(base.forms, slot, formobj) end end local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg insert_termobj_into_slot(arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_decorations(canon, arg) insert_termobj_into_slot(canon) end end end for _, arg in ipairs(specs) do if arg.term == "+" then if default_values then -- will be nil if past tense specified as - and no past ptc given for _, val in ipairs(default_values) do val = shallowCopy(val) if default_already_formobj then local argformobj = m_headword_utilities.convert_termobj_to_formobj(arg) val.footnotes = iut.combine_footnotes(val.footnotes, argformobj.footnotes) iut.insert_form(base.forms, slot, val) else m_headword_utilities.combine_termobj_decorations(val, arg) canonicalize_and_insert(val) end end end else canonicalize_and_insert(arg) end end end set_lemma_and_default_forms(base.lemma) local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error(("Internal error: '%s' should have already thrown an error"):format(form)) end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_specs("s_form", base.s_specs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_specs("ing_form", base.ing_specs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_specs("ed_form", base.ed_specs, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_specs("en_form", base.en_specs, function(arg) return canonicalize_en_form(arg.term) end, base.forms.ed_form, "default already formobj") iut.insert_form(base.forms, "lemma", {form = base.lemma}) -- Add linked version of lemma for use in head=. We write this in a general fashion in case -- there are multiple lemma forms (which isn't possible currently at this level, although it's -- possible overall using the ((...,...)) notation). iut.insert_forms(base.forms, "lemma_linked", iut.map_forms(base.forms.lemma, function(form) if form == base.lemma and base.linked_lemma:find("%[%[") then return base.linked_lemma else return form end end)) end local inflect_props = { slot_table = all_verb_slots, inflect_word_spec = conjugate_verb, } iut.inflect_multiword_or_alternant_multiword_spec(alternant_multiword_spec, inflect_props) -- (4) Fetch the forms and put the conjugated lemmas in data.heads if not explicitly given. local function fetch_termobjs(slot) local forms = alternant_multiword_spec.forms[slot] -- See above. This should only occur if the user explicitly used - for a spec. if not forms or not forms[1] then return {{term = "-"}} end local termobjs = {} for _, formobj in ipairs(forms) do insert(termobjs, m_headword_utilities.convert_formobj_to_termobj(formobj)) end return termobjs end pres_3sgs = fetch_termobjs("s_form") pres_ptcs = fetch_termobjs("ing_form") pasts = fetch_termobjs("ed_form") past_ptcs = fetch_termobjs("en_form") -- Use the "linked" form of the lemma as the head if no head= explicitly given and the user specified -- brackets in one of the lemmas. Otherwise we use the default headword-linking algorithm. if not data.user_specified_heads[1] and alternant_multiword_spec.saw_bracket then data.heads = {} for _, lemma_obj in ipairs(alternant_multiword_spec.forms.lemma_linked) do insert(data.heads, m_headword_utilities.convert_formobj_to_termobj(lemma_obj)) end end else -------------------------- SEPARATE-PARAM FORMAT -------------------------- set_lemma_and_default_forms(pagename) par1s = par1s or parse_inflection(args, {1, "pres_3sg"}) pres_3sgs = {} pres_ptcs = {} pasts = {} past_ptcs = {} if not par1s[1] then par1s = {{term = "+"}} end if not par2s[1] then par2s = {{term = "+"}} end if not par3s[1] then par3s = {{term = "+"}} end if not par4s[1] then par4s = {{term = "+"}} end local function process_argument(args, dest, canon_func, default_values, default_already_canonicalized) local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg m_headword_utilities.insert_termobj_combining_duplicates(dest, arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_decorations(canon, arg) m_headword_utilities.insert_termobj_combining_duplicates(dest, canon) end end end for _, arg in ipairs(args) do if arg.term == "+" then for _, val in ipairs(default_values) do val = shallowCopy(val) m_headword_utilities.combine_termobj_decorations(val, arg) if default_already_canonicalized then m_headword_utilities.insert_termobj_combining_duplicates(dest, val) else canonicalize_and_insert(val) end end else canonicalize_and_insert(arg) end end end local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" or form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_argument(par1s, pres_3sgs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_argument(par2s, pres_ptcs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_argument(par3s, pasts, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_argument(par4s, past_ptcs, function(arg) return canonicalize_en_form(arg.term) end, pasts, "default already canonicalized") end ------------------------------------------- INSERT INFLECTIONS ------------------------------------------ insert_inflection(data, pres_3sgs, "third-person singular simple present", "s-verb-form") insert_inflection(data, pres_ptcs, "present participle", "ing-form") if deepEquals(pasts, past_ptcs) then insert_inflection(data, pasts, "simple past and past participle", "ed-form", "no simple past or past participle") else insert_inflection(data, pasts, "simple past", "spast") insert_inflection(data, past_ptcs, "past participle", "past|part") end if pagename:find(" ") then -- Check for placeholder "it" local words = split(pagename, " ") for _, word in ipairs(words) do if word == "it" or word == "its" or word == "it's" then insert(data.categories, langname .. ' terms with placeholder "it"') break end end -- Check for phrasal verbs local phrasal_adverbs = list_to_set{ -- NOTE: This should only contain common phrasal adverbs, not random words like [[low]], -- [[adrift]], etc. "aback", "about", "above", "across", "after", "against", "ahead", "along", "apart", "around", "as", "aside", "at", "away", "back", "before", "behind", "below", "between", "beyond", "by", "down", "for", "forth", "from", "in", "into", "of", "off", "on", "onto", "out", "over", "past", "round", "through", "to", "together", "towards", "under", "up", "upon", "with", "without", } local allowed_non_adverb_words = list_to_set{ "it", "one", "oneself", "someone", } local base = pagename local seen_adverbs = {} -- Only consider a verb to be phrasal if it consists of a single base verb followed exclusively by either -- adverbs from `phrasal_adverbs` or placeholder words from `allowed_non_adverb_words`, where at -- least one following word is from `phrasal_adverbs` (hence [[can it]] is not a phrasal verb). while true do local prev, word = base:match("^(.+) (.-)$") if not prev then break end if phrasal_adverbs[word] then insert(seen_adverbs, word) elseif allowed_non_adverb_words[word] then -- do nothing else break end base = prev end if not base:find(" ") and seen_adverbs[1] then insert(data.categories, langname .. " phrasal verbs") for i = #seen_adverbs, 1, -1 do insert(data.categories,'ပွင်ႈၵႂၢမ်းသၢင်ႈ' .. langname .. ' ဢၼ်ၽွတ်ႈႁႅၼ်းဝႆႉတင်း "' .. seen_adverbs[i] .. '"') end end end end, } ----------------------------------------------------------------------------------------- -- Suffix forms -- ----------------------------------------------------------------------------------------- pos_functions["suffix forms"] = { params = { [1] = {required = true, list = true, disallow_holes = true}, }, func = function(args, data, is_suffix) local suffix_type = {} for _, typ in ipairs(args[1]) do insert(suffix_type, typ .. "-forming suffix") end insert(data.inflections, {label = "non-lemma form of " .. m_table.serialCommaJoin(suffix_type, {conj = "or"})}) end, } return export d1v7psjvdfwl4zy9qkivs3gqq8u5ok8 293605 293603 2026-09-20T13:15:45Z Saimawnkham 9 293605 Scribunto text/plain local export = {} local pos_functions = {} --[==[ Author from 2020 on: mostly Benwing2, with significant contributions from Theknightwho. Based on a prior version by Rua (by now mostly rewritten), with contributions from Erutuon and others (see history for full attribution). ]==] local force_cat = false -- for testing; if true, categories appear in non-mainspace pages local require = require local require_when_needed = require("Module:require when needed") local en_utilities_module = "Module:en-utilities" local headword_utilities_module = "Module:headword utilities" local headword_module = "Module:headword" local inflection_utilities_module = "Module:inflection utilities" local parse_utilities_module = "Module:parse utilities" local JSON_module = "Module:JSON" local labels_module = "Module:labels" local links_module = "Module:links" local parameters_module = "Module:parameters" local string_utilities_module = "Module:string utilities" local table_module = "Module:table" local utilities_module = "Module:utilities" local yesno_module = "Module:yesno" local iut = require_when_needed(inflection_utilities_module) local put = require_when_needed(parse_utilities_module) local m_headword_utilities = require_when_needed(headword_utilities_module) local add_links_to_multiword_term = require_when_needed(headword_utilities_module, "add_links_to_multiword_term") local add_suffix = require_when_needed(en_utilities_module, "add_suffix") local apply_link_modifiers = require_when_needed(headword_utilities_module, "apply_link_modifiers") local concat = table.concat local deepEquals = require_when_needed(table_module, "deepEquals") local dump = mw.dumpObject local format_categories = require_when_needed(utilities_module, "format_categories") local full_headword = require_when_needed(headword_module, "full_headword") local get_label_info = require_when_needed(labels_module, "get_label_info") local get_link_page = require_when_needed(links_module, "get_link_page") local glossary_link = require_when_needed(headword_utilities_module, "glossary_link") local insert = table.insert local insertIfNot = require_when_needed(table_module, "insertIfNot") local ipairs = ipairs local is_regular_plural = require_when_needed(en_utilities_module, "is_regular_plural") local list_to_set = require_when_needed(table_module, "listToSet") local pairs = pairs local process_params = require_when_needed(parameters_module, "process") local remove = table.remove local remove_links = require_when_needed(links_module, "remove_links") local replacement_escape = require_when_needed(string_utilities_module, "replacement_escape") local shallowCopy = require_when_needed(table_module, "shallowCopy") local singularize = require_when_needed(en_utilities_module, "singularize") local split = require_when_needed(string_utilities_module, "split") local toJSON = require_when_needed(JSON_module, "toJSON") local toNFD = mw.ustring.toNFD local type = type local ulen = require_when_needed(string_utilities_module, "len") local ulower = require_when_needed(string_utilities_module, "lower") local umatch = require_when_needed(string_utilities_module, "match") local u = require_when_needed(string_utilities_module, "char") local ugsub = require_when_needed(string_utilities_module, "gsub") local lang = require("Module:languages").getByCode("en") local langname = lang:getCanonicalName() local list_param = {list = true, disallow_holes = true} local list_allow_holes = {list = true, allow_holes = true} local boolean_param = {type = "boolean"} local function ine(val) if val == "" then return nil else return val end end local function track(page) require("Module:debug/track")("en-headword/" .. page) return true end ------------------------------------------- UTILITY FUNCTIONS ------------------------------------------ -- Parse and return an inflection not requiring additional processing. The raw arguments come from `args[field]`, which -- is parsed for inline modifiers. local function parse_inflection(args, field, is_head) local argfield = field if type(argfield) == "table" then argfield = argfield[1] end return m_headword_utilities.parse_term_list_with_modifiers { paramname = field, forms = args[argfield], splitchar = ",", is_head = is_head, } end -- Insert the parsed inflections in `terms` (as parsed by `parse_inflection`) into `data.inflections`, with label -- `label` and optional accelerator spec `accel`. local function insert_inflection(data, terms, label, accel, no_label) for _, termobj in ipairs(terms) do m_headword_utilities.remove_termobj_field_modifiers(termobj) end m_headword_utilities.insert_inflection { headdata = data, terms = terms, label = label, no_label = no_label, accel = accel and {form = accel} or nil, } end -- Insert a fixed label `label` into the inflections for `data`. If `originating_term` is supplied, copy the qualifiers, -- labels and references from it into the fixed label. local function insert_fixed_inflection(data, label, originating_term) m_headword_utilities.insert_fixed_inflection { headdata = data, originating_term = originating_term, label = label, } end -- Parse and insert an inflection not requiring additional processing into `data.inflections`. The raw arguments come -- from `args[field]`, which is parsed for inline modifiers. `label` is the label that the inflections are given; -- `accel` is the accelerator form, or nil. local function parse_and_insert_inflection(data, args, field, label, accel) m_headword_utilities.parse_and_insert_inflection { headdata = data, forms = args[field], paramname = field, splitchar = ",", label = label, accel = accel and {form = accel} or nil, -- If we want check_missing support, we need to supply the following: -- check_missing = true, -- lang = lang, -- plpos = plpos, } end -- These functions are used directly in the <> format as well as in the utility functions #2 below. local function compute_double_last_cons_stem(term) local last_cons = term:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("Verb stem '" .. term .. "' must end in a consonant to use ++") end return term .. last_cons end local function compute_plusplus_s_form(term, default_s_form) if term:find("[szx]$") then -- regas -> regasses, derez -> derezzes return compute_double_last_cons_stem(term) .. "es" else return default_s_form end end -- The main entry point. -- This is the only function that can be invoked from a template. function export.show(frame) local iparams = { [1] = true, } local iargs = require("Module:parameters").process(frame.args, iparams) local parargs = frame:getParent().args local poscat = iargs[1] local pos_in_1 = not poscat if pos_in_1 then poscat = ine(parargs[1]) or mw.title.getCurrentTitle().fullText == "Template:en-head" and "interjection" or error("Part of speech must be specified in 1=") poscat = require(headword_module).canonicalize_pos(poscat) end local indexing_poscat = pos_in_1 and "head" or poscat local params = { ["head"] = list_param, ["id"] = true, ["json"] = boolean_param, ["sort"] = true, ["splithyph"] = boolean_param, ["nosplithyph"] = boolean_param, ["hyphspace"] = boolean_param, ["nolink"] = boolean_param, ["nolinkhead"] = {type = "boolean_param", alias_of = "nolink"}, ["suffix"] = boolean_param, ["nosuffix"] = boolean_param, ["nomultiwordcat"] = boolean_param, ["abbr"] = list_param, ["the"] = true, ["def"] = {alias_of = "the"}, ["pagename"] = true, -- for testing } if pos_in_1 then params[1] = {required = true} -- required but ignored as already processed above end local pos_data = pos_functions[indexing_poscat] local pos_func if pos_data then local pos_params = pos_data.params if pos_params then for key, val in pairs(pos_params) do params[key] = val end end pos_func = pos_data.func end local args = process_params(parargs, params) -- Account for unsupported titles, e.g. 'C|N>K' instead of 'Unsupported titles/C through N to K'. local pagename = args.pagename or mw.loadData("Module:headword/data").pagename local user_specified_heads = parse_inflection(args, "head", "is_head") local heads = user_specified_heads local autohead if args.nolink or not pagename:find("[ '%-]") then autohead = pagename else local en_no_split_apostrophe_words = list_to_set { "one's", "someone's", "he's", "she's", "it's", } local en_include_hyphen_prefixes = list_to_set { -- We don't include things that are also words even though they are often (perhaps mostly) prefixes, e.g. -- "be", "counter", "cross", "extra", "half", "mid", "over", "pan", "under". "acro", "acousto", "Afro", "agro", "anarcho", "angio", "Anglo", "ante", "anti", "arch", "auto", "bi", "bio", "cis", "co", "cryo", "crypto", "de", "demi", "eco", "electro", "Euro", "ex", "Greco", "hemi", "hydro", "hyper", "hypo", "infra", "Indo", "inter", "intra", "Judeo", "macro", "meta", "micro", "mini", "multi", "neo", "neuro", "non", "para", "peri", "post", "pre", "pro", "proto", "pseudo", "re", "semi", "sub", "super", "trans", "un", "vice", } local function is_english(term) local title = mw.title.new(term) if title and title.exists then local content = title:getContent() if content and content:find("==ဢိင်းၵလဵတ်ႈ==\n") then return true end end return false end local function en_split_hyphen_when_space(word) if not word:find("-", nil, true) then return nil end if args.hyphspace then return "[[" .. word:gsub("%-+", " ") .. "|" .. word .. "]]" end if args.nosplithyph then return "[[" .. word .. "]]" end if not args.splithyph then local space_word = word:gsub("%-+", " ") if is_english(space_word) then return "[[" .. space_word .. "|" .. word .. "]]" end if is_english(word) then return "[[" .. word .. "]]" end end return nil end local function en_split_apostrophe(word) local base = word:match("^(.*)'s$") if base then return "[[" .. base .. "]][[-'s|'s]]" end -- Only treat final apostrophe as possessive if preceded by something that looks like a plural ending in /z/. -- In particular we don't want to do it for words like [[truckin']]. base = word:match("^(.*[sxz])'$") if base then if base:find("s$") then local sg = singularize(base) if is_english(sg) then return "[[" .. sg .. "|" .. base .. "]][[-'|']]" end end return "[[" .. base .. "]][[-'|']]" end return "[[" .. word .. "]]" end autohead = add_links_to_multiword_term(pagename, { split_hyphen_when_space = en_split_hyphen_when_space, split_apostrophe = en_split_apostrophe, no_split_apostrophe_words = en_no_split_apostrophe_words, include_hyphen_prefixes = en_include_hyphen_prefixes, }) end if not heads[1] then heads = {{term = autohead}} else for _, headobj in ipairs(heads) do local head = headobj.term if head:find("^~") then head = apply_link_modifiers(autohead, head:sub(2), lang) headobj.term = head elseif head:find("^[!?]$") then -- If explicit head= just consists of ! or ?, add it to the end of the default head. headobj.term = autohead .. head end if head == autohead then track("redundant-head") end end end -- handle the=/def= if args.the == "~" then local newheads = {} for _, headobj in ipairs(heads) do local barehead = shallowCopy(headobj) insert(newheads, barehead) headobj.term = "the " .. headobj.term insert(newheads, headobj) end heads = newheads elseif args.the then local the = require(yesno_module)(args.the) if the then for _, headobj in ipairs(heads) do headobj.term = "the " .. headobj.term end end end local data = { lang = lang, pos_category = poscat, categories = {}, heads = heads, user_specified_heads = user_specified_heads, -- We use our own splitting algorithm so the redundant head cat will be inaccurate. no_redundant_head_cat = true, inflections = {}, nomultiwordcat = args.nomultiwordcat, sort_key = args.sort, pagename = pagename, id = args.id, force_cat_output = force_cat, } local function inscat(cat) insert(data.categories, langname .. " " .. cat) end local is_suffix = false if args.suffix or not args.nosuffix and pagename:find("^%-") and not pagename:find("^%-%-") and poscat ~= "suffix forms" then is_suffix = true data.pos_category = "suffixes" local singular_poscat = singularize(poscat) inscat(singular_poscat .. "-forming suffixes") insert(data.inflections, {label = singular_poscat .. "-forming suffix"}) end if pos_func then pos_func(args, data, is_suffix) end local extra_categories = {} if pagename:find("[Qq]") then -- Check for q not followed by u. We want to exclude things like [[13q deletion syndrome]] and [[BFOQ]] that -- don't have a lowercase letter on either side, as well as things like [[& seq.]] and [[acq.]] that are -- abbreviations for words containing a following u. -- -- Approximate range of combining diacritics; we want to remove them so the checks below for -- a lowercase letter next to the q aren't tripped up by diacritics on the letter. local u300 = u(0x0300) local u36F = u(0x036F) local pagename_no_diacritics = ugsub(toNFD(pagename), "[" .. u300 .. "-" .. u36F .. "]", "") if pagename_no_diacritics:find("[Qq][a-tv-z]") or pagename_no_diacritics:find("[a-z]q[^u.]") or pagename_no_diacritics:find("[a-z]q$") then inscat("words containing Q not followed by U") end end -- toNFD performs decomposition, so letters that decompose to an ASCII -- vowel and a diacritic, such as é, are counted as vowels and do not do not -- need to be included in the pattern. if not umatch(ulower(toNFD(pagename)), "[aeiouyæœøəªºαεηιουω]") then inscat("words spelled without vowels") end if pagename:find("yre$") then inscat('ဢၼ်သုတ်းတင်း "-yre"') end if not pagename:find(" ") and ulen(pagename) >= 25 then insert(extra_categories, "Long " .. langname .. " words") end if pagename:find("^[^aeiou ]*a[^aeiou ]*e[^aeiou ]*i[^aeiou ]*o[^aeiou ]*u[^aeiou ]*$") then inscat("words that use all vowels in alphabetical order") end parse_and_insert_inflection(data, args, "abbr", "abbreviation") if args.json then return toJSON(data) end return full_headword(data) .. (extra_categories[1] and format_categories(extra_categories, lang, args.sort) or "") end local function make_default_comparative(word) if word == "good" or word == "well" then return {"better"} elseif word == "bad" or word == "badly" then return {"worse"} elseif word == "far" then return {"further", "farther"} else return {add_suffix(word, "r")} end end local function make_default_superlative(word) if word == "good" or word == "well" then return {"best"} elseif word == "bad" or word == "badly" then return {"worst"} elseif word == "far" then return {"furthest", "farthest"} else return {add_suffix(word, "st.superlative")} end end -- This function does the common work between adjectives and adverbs. local function process_comparative_args(data, args, plpos) local pagename = data.pagename local comps = parse_inflection(args, 1) local sups = parse_inflection(args, "sup") local outcomps, outsups if args.componly then if comps[1] then error("Can't specify comparatives of comparative-only " .. plpos) end insert(data.inflections, {label = glossary_link("comparative") .. " form only"}) insert(data.categories, langname .. " comparative-only " .. plpos) -- Set to empty list so we don't get any comparatives output, but process superlatives if specified. outcomps = {} if not sups[1] then -- Set to empty list so we don't get any superlatives output unless explicitly given. outsups = {} end elseif args.suponly then if comps[1] or sups[1] then error("Can't specify comparatives or superlatives of or superlative-only " .. plpos) end insert(data.inflections, {label = glossary_link("superlative") .. " form only"}) insert(data.categories, langname .. " superlative-only " .. plpos) return end -- If the first parameter is ?, then don't show anything, just return. if comps[1] and comps[1].term == "?" then if comps[2] then error("Can't specify additional comparatives along with '?'") end if sups[1] then error("Can't specify superlatives along with '?' for the comparative") end return end if comps[1] and comps[1].term == "-" then local hyphencomp = remove(comps, 1) -- Remove the "-" but retain for qualifiers, labels, references -- Not (generally) comparable; may occasionally have a comparative if comps[1] then insert_fixed_inflection(data, "not generally <<comparable>>", hyphencomp) elseif not sups[1] then insert_fixed_inflection(data, "not <<comparable>>", hyphencomp) insert(data.categories, plpos .. langname .. " ၸိူဝ်းတႅၵ်ႈၼိူင်းဢမ်ႇလႆႈ ") return else -- No comparative, but a superlative. insert_inflection() will correctly generate 'no comparative' if we -- pass in "-" as the value. outcomps = {hyphencomp} end elseif not comps[1] then comps = {{term = "more"}} end if not outcomps then -- not if we set `outcomps` to "-" above or processed a comparative-only term outcomps = {} -- Go over each parameter given and create a comparative and superlative form. for _, compobj in ipairs(comps) do local comp = compobj.term if comp == "-" then error("Comparative of '-' only allowed as first comparative") end if comp == "+" then comp = "+more" elseif comp == "more" and pagename ~= "many" and pagename ~= "much" then comp = "+more" elseif comp == "further" and pagename ~= "far" then comp = "+further" elseif comp == "better" and pagename ~= "good" and pagename ~= "well" then comp = "+better" elseif comp:find("~") then comp = comp:gsub("~", replacement_escape(pagename)) end compobj.origterm = comp if comp == "+more" then comp = "more [[" .. pagename .. "]]" elseif comp == "+further" then comp = {"further [[" .. pagename .. "]]", "farther [[" .. pagename .. "]]"} elseif comp == "+better" then comp = "better [[" .. pagename .. "]]" elseif comp == "er" then -- Add -er. comp = add_suffix(pagename, "r") elseif comp == "ier" then if pagename:sub(-1) ~= "y" then error("Can't specify 'ier' comparative unless the term ends with 'y': " .. pagename) end comp = pagename:gsub("e?y$", "ier") elseif comp:find("^%+") then local special = m_headword_utilities.get_special_indicator(comp, "noerror") if special then comp = m_headword_utilities.handle_multiword(pagename, special, make_default_comparative) end end if type(comp) == "table" and not comp[2] then comp = comp[1] end if type(comp) == "table" then for i = 1, #comp - 1 do local outobj = shallowCopy(compobj) outobj.term = comp[i] insert(outcomps, outobj) end compobj.term = comp[#comp] insert(outcomps, compobj) else compobj.term = comp insert(outcomps, compobj) end end end if sups[1] and sups[1].term == "-" then if sups[2] then error("Can't specify '-' as superlative followed by further values") end -- No superlative. insert_inflection() will correctly generate 'no superlative' if we pass in "-" as the value. outsups = sups else if not sups[1] then sups = {{term = "+"}} end end -- `outsups` will be set if we set `outsups` to "-" above or processed a comparative-only term without superlatives. if not outsups then outsups = {} local function process_sup(sup, special, supobj, compobj) if special then sup = m_headword_utilities.handle_multiword(pagename, special, make_default_superlative) elseif sup == "-" or sup == "+" then error(("Internal error: Superlative value of '%s' should have been handled earlier"):format(sup)) elseif sup == "+most" then sup = "most [[" .. pagename .. "]]" elseif sup == "+furthest" then sup = {"furthest [[" .. pagename .. "]]", "farthest [[" .. pagename .. "]]"} elseif sup == "+best" then sup = "best [[" .. pagename .. "]]" elseif sup == "est" then -- Add -est. sup = add_suffix(pagename, "st.superlative") elseif sup == "iest" then if pagename:sub(-1) ~= "y" then error("Can't specify 'iest' superlative unless the term ends with 'y': " .. pagename) end sup = pagename:gsub("e?y$", "iest") end if type(sup) == "table" and not sup[2] then sup = sup[1] end if compobj then supobj = shallowCopy(supobj) supobj = m_headword_utilities.combine_termobj_qualifiers_labels(supobj, compobj) end if type(sup) == "table" then for i = 1, #sup - 1 do local outobj = shallowCopy(supobj) outobj.term = sup[i] insert(outsups, outobj) end supobj.term = sup[#sup] insert(outsups, supobj) else supobj.term = sup insert(outsups, supobj) end end for _, supobj in ipairs(sups) do local sup = supobj.term if sup == "-" then error("Superlative of '-' only allowed as first superlative") end if sup == "+" then if not comps[1] then error("Superlative of '+' can't be specified when there are no comparatives") end for _, compobj in ipairs(comps) do local comp = compobj.origterm local special if comp == "+more" then sup = "+most" elseif comp == "+further" then sup = "+furthest" elseif comp == "+better" then sup = "+best" elseif comp == "er" then sup = "est" elseif comp == "ier" then sup = "iest" else if comp:find("^%+") then special = m_headword_utilities.get_special_indicator(comp, "noerror") end if not special then -- If the full comparative was given, then derive the superlative by replacing -er with -- -est. if comp:sub(-2) == "er" then sup = comp:sub(1, -3) .. "est" else error(("The superlative cannot be derived automatically from comparative '%s' because it doesn't end in -er"):format(comp)) end end end process_sup(sup, special, supobj, compobj) end else local special = m_headword_utilities.get_special_indicator(sup, "noerror") -- Do some work here rather than in process_sup() so we don't end up double-processing a term with a '~' -- in it or a term that happens to be 'most' or similar after substitution of ~ in the comparative. if not special then if sup == "most" and pagename ~= "many" and pagename ~= "much" then sup = "+most" elseif sup == "furthest" and pagename ~= "far" then sup = "+furthest" elseif sup == "best" and pagename ~= "good" and pagename ~= "well" then sup = "+best" elseif sup:find("~") then sup = sup:gsub("~", replacement_escape(pagename)) end end process_sup(sup, special, supobj) end end end insert_inflection(data, outcomps, "<<comparative>>", "comparative") insert_inflection(data, outsups, "<<superlative>>", "superlative") end pos_functions["ၵမ်ႉႁိၵ်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉႁိၵ်ႈ") end, } pos_functions["ၵမ်ႉသၢင်ႈ"] = { params = { [1] = list_param, ["comp_qual"] = {list = "comp\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the comparative value", }, ["sup"] = list_param, ["sup_qual"] = {list = "sup\1_qual", allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the superlative value", }, ["componly"] = boolean_param, ["suponly"] = boolean_param, }, func = function(args, data) -- Process the comparatives and superlatives. process_comparative_args(data, args, "ၵမ်ႉသၢင်ႈ") end, } local function escape(str) return (str:gsub("\\([:#])", "\\\\%1") :gsub("[:#]", "\\%0")) end local function canonicalize_plural(pl, pagename, pos) if pl == "+" then return escape(add_suffix(pagename, "s.plural", pos)) elseif pl == "++" then return escape(compute_plusplus_s_form(pagename, add_suffix(pagename, "s.plural", pos))) elseif pl == "*" then return escape(pagename) elseif pl == "ies" then if pagename:sub(-1) == "y" then return escape(pagename:gsub("e?y$", pl)) end error("Can't specify 'ies' plural unless the term ends with 'y'.") elseif pl == "s" or pl == "es" or pl == "'s" then return escape(pagename .. pl) end end local function do_nouns(args, data, pos) local pagename = data.pagename pos = pos or "ႁိၵ်ႈ" local plurals = parse_inflection(args, 1) local function insert_plurale_tantum_inflections(is_plural_only, originating_label) if args.sg[1] then insert_fixed_inflection(data, "normally plural", originating_label) parse_and_insert_inflection(data, args, "sg", "singular") elseif is_plural_only then insert_fixed_inflection(data, "plural only", originating_label) end if args.attr[1] then parse_and_insert_inflection(data, args, "attr", "attributive") end end local function first_pl_term() return plurals[1] and plurals[1].term or nil end if first_pl_term() == "p" then -- plurale tantum if plurals[2] then error("With plurale tantum noun, can't specify more than one plural") end data.genders = {"p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections("plural only", plurals[1]) return end local function inscat(cat) insert(data.categories,"ႁိၵ်ႈ".. langname .. " " .. cat) end local need_default_plural = pos == "ႁိၵ်ႈ" if first_pl_term() == "sp" then -- construed as singular or plural sp = remove(plurals, 1) -- Remove the "sp" but retain it for its qualifiers, labels, references inscat("ဢၼ်ဢဝ်တီႈပွင်ႇဝႃႈ ပဵၼ်တူဝ်လဵဝ် ဢမ်ႇၼၼ် တူဝ်ၼမ်") data.genders = {"s", "p"} -- this should auto-insert the correct 'pluralia tantum' category insert_plurale_tantum_inflections(nil, sp) need_default_plural = false elseif first_pl_term() == "-" then -- Uncountable noun; may occasionally have a plural local hyphpl = remove(plurals, 1) -- Remove the "-" but retain for qualifiers, labels, references inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") -- If plural forms were given explicitly, then show "usually" if plurals[1] then insert_fixed_inflection(data, "usually <<uncountable>>", hyphpl) else insert_fixed_inflection(data, "<<uncountable>>", hyphpl) end need_default_plural = false elseif first_pl_term() == "#" then -- Usually countable (e.g., "grilled cheese") local hashpl = remove(plurals, 1) -- Remove the "#" but retain for qualifiers, labels, references insert_fixed_inflection(data, "usually <<countable>>", hashpl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end elseif first_pl_term() == "~" then -- Mixed countable/uncountable noun, always has a plural local tildepl = remove(plurals, 1) -- Remove the "~" but retain for qualifiers, labels, references insert_fixed_inflection(data, "<<countable>> and <<uncountable>>", tildepl) inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") -- If no plural was given, add a default one now if not plurals[1] then plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end end -- Plural is unknown if first_pl_term() == "?" then local questionpl = remove(plurals, 1) -- Remove the "?" but retain for qualifiers, labels, references -- Not desired; see [[Wiktionary:Tea_room/2021/August#"Plural unknown or uncertain"]] -- insert_fixed_inflection(data, "plural unknown or uncertain", questionpl) inscat("ဢၼ်ဢမ်ႇႁူႉ ဢမ်ႇၼၼ် ဢမ်ႇတႅတ်ႈၼႅတ်ႈ တူဝ်ၼမ်") if plurals[1] then error("Can't specify explicit plurals along with '?' for unknown/uncertain plural") end return end -- Plural is not attested if first_pl_term() == "!" then local exclampl = remove(plurals, 1) -- Remove the "!" but retain for qualifiers, labels, references insert_fixed_inflection(data, "plural not attested", exclampl) inscat("nouns with unattested plurals") if plurals[1] then error("Can't specify explicit plurals along with '!' for unattested plural") end return end -- If no plural was given, maybe add a default one, otherwise (when "-" was given or proper noun) return. if not plurals[1] then if not need_default_plural then inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇဢမ်ႇလႆႈ") return end plurals[1] = {term = escape(add_suffix(pagename, "s.plural", pos))} end -- There are plural forms to show, so show them. inscat("ၸိူဝ်းၼပ်ႉဢၢၼ်ႇလႆႈ") local irregular, indeclinable for i, pl in ipairs(plurals) do local canon_pl = canonicalize_plural(pl.term, pagename, pos) if canon_pl then pl.term = canon_pl end local pl_term = get_link_page(pl.term, lang) if not (pagename:find(" ") or is_regular_plural(pl_term, pagename)) then irregular = true if pl_term == pagename then indeclinable = true end end end if irregular then inscat("ဢၼ်ၵိုၵ်းပႃး တူဝ်ၼမ် ဢၼ်ဢမ်ႇၸႂ်ႈပၵ်းၵဝ်ႇပိူင်ၵဝ်ႇ") end if indeclinable then inscat("ဢၼ်ဢမ်ႇပိၼ်ႇၸွမ်းပိူင်မၼ်း") end insert_inflection(data, plurals, "plural", "p") end -- Return the parameters to be used for nouns and proper nouns. Currently the same. local noun_params = { [1] = list_param, ["pl\1qual"] = {list = true, allow_holes = true, replaced_by = false, instead = "use <l:...> or <q:...> inline modifier on the plural", }, -- The following four only used for pluralia tantum (1=p) ["sg"] = list_param, ["attr"] = list_param, } pos_functions["ႁိၵ်ႈ"] = { params = noun_params, func = do_nouns, } pos_functions["ႁိၵ်ႈသုၼ်ႇလဵဝ်"] = { params = noun_params, func = function(args, data) return do_nouns(args, data, "ႁိၵ်ႈသုၼ်ႇလဵဝ်") end, } local function base_default_verb_forms(verb) return escape(add_suffix(verb, "s.verb")), escape(add_suffix(verb, "ing")), escape(add_suffix(verb, "d")) end local function default_verb_forms(verb) local full_s_form, full_ing_form, full_ed_form = base_default_verb_forms(verb) if verb:find(" ") then local first, rest = verb:match("^(.-)( .*)$") local first_s_form, first_ing_form, first_ed_form = base_default_verb_forms(first) return full_s_form, full_ing_form, full_ed_form, first_s_form .. rest, first_ing_form .. rest, first_ed_form .. rest, first, rest else return full_s_form, full_ing_form, full_ed_form, nil, nil, nil, nil, nil end end local function compute_double_last_cons_stem_of_split_verb(verb, ending) local first, rest = verb:match("^(.-)( .*)$") if not first then error("Verb '" .. verb .. "' must have a space in it to use **") end local last_cons = first:match("([bcdfghjklmnpqrstvwxyzBCDFGHJKLMNPQRSTVWXYZ])$") if not last_cons then error("First word '" .. first .. "' must end in a consonant to use **") end return first .. last_cons .. ending .. rest end local function check_non_nil_star_form(form, pagename) if form == nil then error("Verb '" .. pagename .. "' must have a space in it to use *, **, *l, *! or *'") end return form end local function sub_tilde(form, pagename) if not form then return nil end if form:find("~") then form = form:gsub("~", replacement_escape(pagename)) end return form end local deprecated_qual_replaced_by_inline_modifier = { list = true, allow_holes = true, replaced_by = false, instead = "use an inline modifier <q:...> or <l:...> on the value" } pos_functions["သၢင်ႈ"] = { params = { [1] = {list = "pres_3sg", disallow_holes = true}, ["pres_3sg\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [2] = {list = "pres_ptc", disallow_holes = true}, ["pres_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [3] = {list = "past", disallow_holes = true}, ["past\1_qual"] = deprecated_qual_replaced_by_inline_modifier, [4] = {list = "past_ptc", allow_holes = true}, ["past_ptc\1_qual"] = deprecated_qual_replaced_by_inline_modifier, ["noautolinkverb"] = boolean_param, ["angle_bracket"] = boolean_param, }, func = function(args, data) -- Get parameters local par1s local par2s = parse_inflection(args, {2, "pres_ptc"}) local par3s = parse_inflection(args, {3, "past"}) local par4s = parse_inflection(args, {4, "past_ptc"}) local pres_3sgs, pres_ptcs, pasts, past_ptcs local pagename = data.pagename ------------------------------------------- UTILITY FUNCTIONS #2 ------------------------------------------ -- These functions are used in both in the separate-parameter format and in the override params such as past_ptc2=. local full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed local lemma local function set_lemma_and_default_forms(the_lemma) lemma = the_lemma full_default_s, full_default_ing, full_default_ed, split_default_s, split_default_ing, split_default_ed, lemma_first, lemma_rest = default_verb_forms(the_lemma) end local function canonicalize_s_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_s elseif form == "*" then return check_non_nil_star_form(split_default_s, lemma) elseif form == "++" then return compute_plusplus_s_form(lemma, full_default_s) elseif form == "**" then if lemma:find("^[^ ]*[szx] ") then return compute_double_last_cons_stem_of_split_verb(lemma, "es") else return check_non_nil_star_form(split_default_s, lemma) end elseif form == "+!" then return lemma .. "s" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "s" .. lemma_rest elseif form == "+'" then return lemma .. "'s" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'s" .. lemma_rest elseif form == "+l" then if lemma:find("[szx]$") then return {{term = full_default_s, l = {"US"}}, {term = compute_plusplus_s_form(lemma, full_default_s), l = {"UK"}}} else return compute_plusplus_s_form(lemma, full_default_s) end elseif form == "*l" then if lemma:find("^[^ ]*[szx] ") then return {{term = check_non_nil_star_form(split_default_s, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "es"), l = {"UK"}}} else return check_non_nil_star_form(split_default_s, lemma) end else return sub_tilde(form, lemma) end end local function canonicalize_ing_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ing elseif form == "*" then return check_non_nil_star_form(split_default_ing, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ing" elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ing") elseif form == "+!" then return lemma .. "ing" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ing" .. lemma_rest elseif form == "+'" then return lemma .. "'ing" elseif form == "*'" then return check_non_nil_star_form(lemma_first) .. "'ing" .. lemma_rest elseif form == "+l" then return {{term = full_default_ing, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ing", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ing, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ing"), l = {"UK"}}} else return sub_tilde(form, lemma) end end local function canonicalize_ed_form(form) if form == "+" then error("Internal error: Should not see '+' here") elseif form == "^" then return full_default_ed elseif form == "*" then return check_non_nil_star_form(split_default_ed, lemma) elseif form == "++" then return compute_double_last_cons_stem(lemma) .. "ed" elseif form == "+!" then return lemma .. "ed" elseif form == "*!" then return check_non_nil_star_form(lemma_first) .. "ed" .. lemma_rest elseif form == "+'" then return {{term = lemma .. "'d"}, {term = lemma .. "'ed"}} elseif form == "*'" then return {{term = check_non_nil_star_form(lemma_first) .. "'d" .. lemma_rest}, {term = check_non_nil_star_form(lemma_first) .. "'ed" .. lemma_rest}} elseif form == "**" then return compute_double_last_cons_stem_of_split_verb(lemma, "ed") elseif form == "+l" then return {{term = full_default_ed, l = {"US"}}, {term = compute_double_last_cons_stem(lemma) .. "ed", l = {"UK"}}} elseif form == "*l" then return {{term = check_non_nil_star_form(split_default_ed, lemma), l = {"US"}}, {term = compute_double_last_cons_stem_of_split_verb(lemma, "ed"), l = {"UK"}}} else return sub_tilde(form, lemma) end end -- FIXME: options should be "+", "*", "++", "**", "+n", "*n", "++n" and "**n", but not "n" local function canonicalize_en_form(form) if form == "n" then track("n4") return add_suffix(lemma, "n") end return canonicalize_ed_form(form) end --------------------------------- MAIN PARSING/CONJUGATING CODE -------------------------------- local is_angle_bracket = args.angle_bracket if is_angle_bracket then if par2s[1] or par3s[1] or par4s[1] then error("Can't specify explicit values for 2=, 3= or 4= along with the angle-bracket format") end elseif is_angle_bracket == nil and not par2s[1] and not par3s[1] and not par4s[1] and not args[1][2] and args[1][1] and args[1][1]:find("<") then if put.term_contains_top_level_html(args[1][1]) then -- Often, term_contains_top_level_html() returns true on the angle-bracket format, which would -- make the pcall() below succeed but leave the angle brackets as-is. Check for this and only do the -- pcall() if term_contains_top_level_html() returns false. is_angle_bracket = true else -- If it's ambiguous whether it's an angle-bracket format or separate params with an inline modifier, -- try to parse as the latter. If an error occurs, treat as the former. local ok ok, par1s = pcall(parse_inflection, args, {1, "pres_3sg"}) if not ok then par1s = nil is_angle_bracket = true end end end if is_angle_bracket then -------------------------- ANGLE-BRACKET FORMAT -------------------------- -- (0) Expand multiword term with angle brackets just on the first word. local arg11 = args[1][1] if arg11:find("^<.*>$") and pagename:find(" ") then local first, rest = pagename:match("^(.-)( .*)$") arg11 = first .. arg11 .. rest end -- (1) Parse the indicator specs inside of angle brackets. local function parse_indicator_spec(angle_bracket_spec) local inside = angle_bracket_spec:match("^<(.*)>$") assert(inside) local segments = put.parse_balanced_segment_run(inside, "[", "]") local comma_separated_groups = put.split_alternating_runs(segments, ",") if #comma_separated_groups > 4 then error("Too many comma-separated parts in indicator spec, expected at most 4: " .. 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 error("Extraneous text after bracketed footnotes: '" .. concat(separated_group) .. "'") end if not footnotes then footnotes = {} end insert(footnotes, separated_group[j]) end return footnotes end local function fetch_specs(comma_separated_group) if not comma_separated_group then return {{term = "+"}} end local specs = {} local colon_separated_groups = put.split_alternating_runs(comma_separated_group, ":") for _, colon_separated_group in ipairs(colon_separated_groups) do local form = colon_separated_group[1] if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error("*, **, *l, *! and *' not allowed inside of indicator specs: " .. angle_bracket_spec) end if form == "" then form = "+" end local termobj = { term = form } local footnotes = fetch_footnotes(colon_separated_group) if footnotes then for _, footnote in ipairs(footnotes) do m_headword_utilities.add_footnote_to_termobj(termobj, footnote) end end insert(specs, termobj) end return specs end local s_specs = fetch_specs(comma_separated_groups[1]) local ing_specs = fetch_specs(comma_separated_groups[2]) local ed_specs = fetch_specs(comma_separated_groups[3]) local en_specs = fetch_specs(comma_separated_groups[4]) return { forms = {}, s_specs = s_specs, ing_specs = ing_specs, ed_specs = ed_specs, en_specs = en_specs, } end local parse_props = { parse_indicator_spec = parse_indicator_spec, } local alternant_multiword_spec = iut.parse_inflected_text(arg11, parse_props) -- (2) Check for user-specified brackets; remove any links from the lemma, but remember the original -- form so we can use it below in the 'lemma_linked' form. -- Check to see if there are brackets in the pre-text or post-text. If so, use the linked lemma (with the -- verb autolinked unless noautolinkverb is given). Otherwise, use the default headword algorithm. local function check_bracket(val) if val:find("%[%[") then alternant_multiword_spec.saw_bracket = true end end for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do check_bracket(alternant_or_word_spec.before_text) if alternant_or_word_spec.alternants then for _, multiword_spec in ipairs(alternant_or_word_spec.alternants) do for _, word_spec in ipairs(multiword_spec.word_specs) do check_bracket(word_spec.before_text) end check_bracket(multiword_spec.post_text) end end end check_bracket(alternant_multiword_spec.post_text) iut.map_word_specs(alternant_multiword_spec, function(base) if base.lemma == "" then base.lemma = pagename end base.orig_lemma = base.lemma base.lemma = remove_links(base.lemma) if args.noautolinkverb or base.orig_lemma:find("%[%[") then base.linked_lemma = base.orig_lemma else base.linked_lemma = "[[" .. base.orig_lemma .. "]]" end end) -- (3) Conjugate the verbs according to the indicator specs parsed above. local all_verb_slots = { lemma = "infinitive", lemma_linked = "infinitive", s_form = "3|s|pres", ing_form = "pres|ptcp", ed_form = "past", en_form = "past|ptcp", } local function conjugate_verb(base) local function process_specs(slot, specs, canon_func, default_values, default_already_formobj) local function insert_termobj_into_slot(termobj) local formobj = m_headword_utilities.convert_termobj_to_formobj(termobj) -- If the form is -, don't insert any forms, which will result in there being no overall forms -- (in fact it will be nil). We check for that down below and substitute a single "-" as the -- form, which in turn gets turned into special labels like "no present participle". if formobj.form == "-" then if formobj.footnotes then error("Unable to preserve footnotes specified on missing form '-': FIXME: " .. dump(formobj.footnotes)) end else iut.insert_form(base.forms, slot, formobj) end end local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg insert_termobj_into_slot(arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_qualifiers_labels(canon, arg) insert_termobj_into_slot(canon) end end end for _, arg in ipairs(specs) do if arg.term == "+" then if default_values then -- will be nil if past tense specified as - and no past ptc given for _, val in ipairs(default_values) do val = shallowCopy(val) if default_already_formobj then local argformobj = m_headword_utilities.convert_termobj_to_formobj(arg) val.footnotes = iut.combine_footnotes(val.footnotes, argformobj.footnotes) iut.insert_form(base.forms, slot, val) else m_headword_utilities.combine_termobj_qualifiers_labels(val, arg) canonicalize_and_insert(val) end end end else canonicalize_and_insert(arg) end end end set_lemma_and_default_forms(base.lemma) local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then error(("Internal error: '%s' should have already thrown an error"):format(form)) end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_specs("s_form", base.s_specs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_specs("ing_form", base.ing_specs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_specs("ed_form", base.ed_specs, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_specs("en_form", base.en_specs, function(arg) return canonicalize_en_form(arg.term) end, base.forms.ed_form, "default already formobj") iut.insert_form(base.forms, "lemma", {form = base.lemma}) -- Add linked version of lemma for use in head=. We write this in a general fashion in case -- there are multiple lemma forms (which isn't possible currently at this level, although it's -- possible overall using the ((...,...)) notation). iut.insert_forms(base.forms, "lemma_linked", iut.map_forms(base.forms.lemma, function(form) if form == base.lemma and base.linked_lemma:find("%[%[") then return base.linked_lemma else return form end end)) end local inflect_props = { slot_table = all_verb_slots, inflect_word_spec = conjugate_verb, } iut.inflect_multiword_or_alternant_multiword_spec(alternant_multiword_spec, inflect_props) -- (4) Fetch the forms and put the conjugated lemmas in data.heads if not explicitly given. local function fetch_termobjs(slot) local forms = alternant_multiword_spec.forms[slot] -- See above. This should only occur if the user explicitly used - for a spec. if not forms or not forms[1] then return {{term = "-"}} end local termobjs = {} for _, formobj in ipairs(forms) do insert(termobjs, m_headword_utilities.convert_formobj_to_termobj(formobj)) end return termobjs end pres_3sgs = fetch_termobjs("s_form") pres_ptcs = fetch_termobjs("ing_form") pasts = fetch_termobjs("ed_form") past_ptcs = fetch_termobjs("en_form") -- Use the "linked" form of the lemma as the head if no head= explicitly given and the user specified -- brackets in one of the lemmas. Otherwise we use the default headword-linking algorithm. if not data.user_specified_heads[1] and alternant_multiword_spec.saw_bracket then data.heads = {} for _, lemma_obj in ipairs(alternant_multiword_spec.forms.lemma_linked) do insert(data.heads, m_headword_utilities.convert_formobj_to_termobj(lemma_obj)) end end else -------------------------- SEPARATE-PARAM FORMAT -------------------------- set_lemma_and_default_forms(pagename) par1s = par1s or parse_inflection(args, {1, "pres_3sg"}) pres_3sgs = {} pres_ptcs = {} pasts = {} past_ptcs = {} if not par1s[1] then par1s = {{term = "+"}} end if not par2s[1] then par2s = {{term = "+"}} end if not par3s[1] then par3s = {{term = "+"}} end if not par4s[1] then par4s = {{term = "+"}} end local function process_argument(args, dest, canon_func, default_values, default_already_canonicalized) local function canonicalize_and_insert(arg) local canon_arg = canon_func(arg) if type(canon_arg) == "string" then arg.term = canon_arg m_headword_utilities.insert_termobj_combining_duplicates(dest, arg) else for _, canon in ipairs(canon_arg) do m_headword_utilities.combine_termobj_qualifiers_labels(canon, arg) m_headword_utilities.insert_termobj_combining_duplicates(dest, canon) end end end for _, arg in ipairs(args) do if arg.term == "+" then for _, val in ipairs(default_values) do val = shallowCopy(val) m_headword_utilities.combine_termobj_qualifiers_labels(val, arg) if default_already_canonicalized then m_headword_utilities.insert_termobj_combining_duplicates(dest, val) else canonicalize_and_insert(val) end end else canonicalize_and_insert(arg) end end end local all_part_default_specs = {} local function process_and_canonicalize_s_form(arg) local form = arg.term if form == "+" then error("Internal error: '+' should have been converted to '^' by now") end if form == "^" or form == "++" or form == "+l" or form == "+!" or form == "+'" or form == "*" or form == "**" or form == "*l" or form == "*!" or form == "*'" then insert(all_part_default_specs, shallowCopy(arg)) end return canonicalize_s_form(form) end process_argument(par1s, pres_3sgs, process_and_canonicalize_s_form, {{term = "^"}}) if not all_part_default_specs[1] then all_part_default_specs[1] = {term = "^"} end process_argument(par2s, pres_ptcs, function(arg) return canonicalize_ing_form(arg.term) end, all_part_default_specs) process_argument(par3s, pasts, function(arg) return canonicalize_ed_form(arg.term) end, all_part_default_specs) process_argument(par4s, past_ptcs, function(arg) return canonicalize_en_form(arg.term) end, pasts, "default already canonicalized") end ------------------------------------------- INSERT INFLECTIONS ------------------------------------------ insert_inflection(data, pres_3sgs, "third-person singular simple present", "s-verb-form") insert_inflection(data, pres_ptcs, "present participle", "ing-form") if deepEquals(pasts, past_ptcs) then insert_inflection(data, pasts, "simple past and past participle", "ed-form", "no simple past or past participle") else insert_inflection(data, pasts, "simple past", "spast") insert_inflection(data, past_ptcs, "past participle", "past|part") end if pagename:find(" ") then -- Check for placeholder "it" local words = split(pagename, " ") for _, word in ipairs(words) do if word == "it" or word == "its" or word == "it's" then insert(data.categories, langname .. ' terms with placeholder "it"') break end end -- Check for phrasal verbs local phrasal_adverbs = list_to_set{ -- NOTE: This should only contain common phrasal adverbs, not random words like [[low]], -- [[adrift]], etc. "aback", "about", "above", "across", "after", "against", "ahead", "along", "apart", "around", "as", "aside", "at", "away", "back", "before", "behind", "below", "between", "beyond", "by", "down", "for", "forth", "from", "in", "into", "of", "off", "on", "onto", "out", "over", "past", "round", "through", "to", "together", "towards", "under", "up", "upon", "with", "without", } local allowed_non_adverb_words = list_to_set{ "it", "one", "oneself", "someone", } local base = pagename local seen_adverbs = {} -- Only consider a verb to be phrasal if it consists of a single base verb followed exclusively by either -- adverbs from `phrasal_adverbs` or placeholder words from `allowed_non_adverb_words`, where at -- least one following word is from `phrasal_adverbs` (hence [[can it]] is not a phrasal verb). while true do local prev, word = base:match("^(.+) (.-)$") if not prev then break end if phrasal_adverbs[word] then insert(seen_adverbs, word) elseif allowed_non_adverb_words[word] then -- do nothing else break end base = prev end if not base:find(" ") and seen_adverbs[1] then insert(data.categories, langname .. " phrasal verbs") for i = #seen_adverbs, 1, -1 do insert(data.categories,'ပွင်ႈၵႂၢမ်းသၢင်ႈ' .. langname .. ' ဢၼ်ၽွတ်ႈႁႅၼ်းဝႆႉတင်း "' .. seen_adverbs[i] .. '"') end end end end, } ----------------------------------------------------------------------------------------- -- Suffix forms -- ----------------------------------------------------------------------------------------- pos_functions["suffix forms"] = { params = { [1] = {required = true, list = true, disallow_holes = true}, }, func = function(args, data, is_suffix) local suffix_type = {} for _, typ in ipairs(args[1]) do insert(suffix_type, typ .. "-forming suffix") end insert(data.inflections, {label = "non-lemma form of " .. m_table.serialCommaJoin(suffix_type, {conj = "or"})}) end, } return export i2yhhs04i90v5aovskm77p6i1q6sc1n 828 41815 293604 255748 2026-09-20T13:13:10Z Saimawnkham 9 293604 Scribunto text/plain local export = {} local m_links = require("Module:links") local m_str_utils = require("Module:string utilities") local m_table = require("Module:table") local put = require("Module:parse utilities") local headword_data_module = "Module:headword/data" local script_utilities_module = "Module:script utilities" local pages_module = "Module:pages" local template_parser_module = "Module:template parser" local is_callable = require("Module:fun").is_callable local split = m_str_utils.split local ufind = m_str_utils.find local usub = m_str_utils.sub local ucfirst = m_str_utils.ucfirst local unpack = unpack or table.unpack -- Lua 5.2 compatibility local dump = mw.dumpObject local new_title = mw.title.new local insert = table.insert local concat = table.concat local insert_if_not = m_table.insertIfNot local shallow_copy = m_table.shallowCopy local function track(page) require("Module:debug/track")("inflection utilities/" .. page) return true end local footnote_abbrevs = { ["a"] = "archaic", ["c"] = "colloquial", ["d"] = "dialectal", ["fp"] = "folk-poetic", ["l"] = "literary", ["lc"] = "low colloquial", ["p"] = "poetic", ["pej"] = "pejorative", ["r"] = "rare", } --[==[ intro: The following code is used in building up the inflection of terms in inflected languages, where a term can potentially consist of several inflected words, each surrounded by fixed text, and a given slot (e.g. accusative singular) of a given word can potentially consist of multiple possible inflected forms. In addition, each form may be associated with a manual transliteration and/or a list of footnotes (or qualifiers, in the case of headword lines). The following terminology is helpful to understand: * A '''term''' is a word or multiword expression that can be inflected. A multiword term may in turn consist of several single-word inflected terms with surrounding fixed text. A term belongs to a particular '''part of speech''' (e.g. noun, verb, adjective, etc.). * An '''inflection dimension''' is a particular dimension over which a term may be inflected, such as case, number, gender, person, tense, mood, voice, aspect, etc. * The '''lemma''' is the particular form of a term under which the term is entered into a dictionary. For example, for verbs, it is most commonly the infinitive, but this differs for some languages: e.g. Latin, Greek and Bulgarian use the first-person singular present indicative (active voice in the case of Latin and Greek); Sanskrit and Macedonian use the third-person singular present indicative (active voice in the case of Sanskrit); Hebrew and Arabic use the third-person singular masculine past (aka "perfect"); etc. For nouns, the lemma form is most commonly the nominative singular, but e.g. for Old French it is the objective singular and for Sanskrit it is the root. * A '''slot''' is a particular combination of inflection dimensions. An example might be "accusative plural" for a noun, or "first-person singular present indicative" for a verb. Slots are named in a language-specific fashion. For example, the slot "accusative plural" might have a name `accpl`, while "first-person singular present indicative" might be variously named `pres1s`, `pres_ind_1_sg`, etc. Each slot is filled with zero or more '''forms'''. * A '''form''' is a particular inflection of a slot for a particular term. Note that a given slot may (and often does) have more than one associated form; these different forms are termed '''variants'''. An example is {{m+|de|Bug||bow (of a ship)}}, which has two genitive singular forms ''Buges'' and ''Bugs''; two plural forms in all cases, e.g. nominative plural ''Buge'' and ''Büge''; and two dative singular forms ''Bug'' and rare/archaic ''Buge''. The form variants for a given slot are ordered, and generally should have the more common and/or preferred variants first, along with rare, archaic or obsolete variants last (if they are included at all). * Forms are described using '''form objects''', which are Lua objects taking the form `{form="``form_value``", translit="``manual_translit``", footnotes={"``footnote``", "``footnote``", ...}}`. (Additional '''metadata''' may be present in a form object, although the support for preserving such metadata when transformations are applied to form objects isn't yet complete.) ``form_value`` is a '''form value''' specifying the value of the form itself in the term's script. ``manual_translit`` specifies optional manual transliteration for the form, in case (a) the form value is in a different script; and (b) either the form's automatic transliteration is incorrect and needs to be overridden, or the language of the term has no automatic transliteration (e.g. in the case of Persian and Hebrew). ``footnote`` is a footnote to be attached to the form in question, and should be e.g. {"[archaic]"} or {"[only in the meaning 'to succeed (an officeholder)']"}, i.e. the string must be surrounded by brackets and should begin with a lowercase letter and not end in a period/full stop. When such footnotes are converted to actual footnotes in a table of inflected forms, the brackets will be removed, the first letter will be capitalized and a period/full stop will be added to the end. (However, when such footnotes are used as qualifiers in headword lines, only the brackets will be removed, with no capitalization or final period.) Note that only ``form_value`` is mandatory. * A list of zero or more form objects is termed a '''form object list''', or usually just a '''form list'''. Such lists are ordered and go into form tables (see below). * A '''form table''' is a Lua table (i.e. a dictionary) describing all the possible inflections of a given term. The keys in such a table are slots (strings) and the values are form lists. '''NOTE:''' All inflection code assumes and maintains the invariant that no two slots, and no two forms in a single slot, share the same form object (by reference, i.e. the Lua object describing a form object should never be shared in two places). This allows for safely side-effecting form objects in certain sorts of operations. This same invariant necessarily applies to the Lua list objects containing the form objects, but does '''NOT''' apply to metadata inside of form objects. In particular, a list of footnotes may well be shared among different form objects. This means it is '''NOT''' safe to side-effect such lists, and in fact no code in this module that manipulates footnote lists will ever side-effect such lists; they are treated as immutable. * Some functions, to save memory, accept and work with abbreviated forms of form objects and/or form lists. Specifically, an '''abbreviated form object''' is either a form object or a string, the latter corresponding to a form object whose form value is the string and all other properties are nil. Similarly, an '''abbreviated form list''' is either a single abbreviated form object or a list of such objects, i.e. any of a string, form object or list of strings and/or form objects. Functions that do not accept such abbreviated structures may be said to insist on being passed form objects in '''full form''', or form lists in '''full list form'''. * Each slot is associated with an '''accelerator tag set''', which is a list of inflection tags that are used when generating an accelerator entry for the forms in the slot (see [[WT:ACCEL]]). For example, the first singular present indicative of a verb might have slot name `pres_1sg` and corresponding accelerator tag set `1|s|pres|ind`. As shown, the accelerator tag set is a string consisting of inflection tags (as used in {{tl|inflection of}}) separated by `|`. Despite the terminology ''tag set'', the tags in a tag set are ordered, although the same tag should never occur twice. * Some inflected terms are '''multiword''', i.e. they consist of multiple '''words''', where each word is generally separated by spaces or sometimes hyphens. In such a term, some of the words inflect, while others remain fixed. Words that inflect are termed '''inflecting words''' (or more correctly '''inflecting parts''', since in some circumstances, parts of a word can inflect). The '''fixed text''' is all the parts of a multiword term that do not inflect. * The descriptor that describes how a given term inflects is called an '''inflection spec''', and consists of the lemma form of the term itself, annotated with an '''angle bracket spec''' after each inflecting word. As the name implies, an angle bracket spec is surrounded by angle brackets (`<...>`). A simple example is {{m+|de|Feder||feather}}, whose inflection spec looks like `Feder<f>`, where `f` specifies the feminine gender. In this case, although there are several properties that could be specified between angle brackets, all except the gender are optional and have been left out, indicating that defaults should be used. Another example is {{m+|de|Baske|Basque person}}, whose inflection spec looks like `Baske<m.weak>`, where `m` specifies the masculine gender and `weak` specifies the weak inflection. Note that individual components of an angle bracket spec like `m` and `weak` are termed '''indicators''' and are separated by periods/full stops. A slightly more complex example is {{m+|de|Zeitgeist||zeitgeist}}, whose inflection spec looks like `Zeitgeist<m,es:s,er>` and which specifies three things in a single '''compound indicator''': `m` (the masculine gender); `es:s` (the genitive singular, which can end in either ''-es'' or ''-s''); and `er` (the nomininative plural, which ends in ''-er''). * If there are several inflecting words in a term, each one will be followed by its own angle bracket spec. An example is {{m+|de|schwarzes Loch||black hole}}, whose inflection spec looks like `schwarzes<+> Loch<n,es:s,^er>`. Here, the adjective ''schwarzes'' (the nominative neuter singular of {{m|de|schwarz||black}}) is followed by the angle bracket spec `<+>` specifying that it inflects as an adjective, and the noun ''Loch'' has the angle bracket spec `<n,es:s,^er>`, indicating (similarly to the above example) that it is neuter, has a genitive singular in either ''-es'' or ''-s'', and has a nominative plural in ''-er'' with umlaut, hence ''Löcher'' (the `^` specifies that the form requires umlaut). * Sometimes a given term has multiple ways of inflecting that differ in ways that can't be specified using a single angle bracket spec. This is supported using '''alternants''', which are specified using double parentheses. (This is so that terms that themselves contain parentheses can be specified without interference.) An example is {{m+|uk|русин||Rusyn}}, which can be stressed either as ''ру́син'' (stress on the first syllable and following accent paradigm ''a'', hence genitive singular ''ру́сина'') or ''руси́н'' (stress on the second syllable and following accent paradigm ''b'', hence genitive singular ''русина́''; note how the stress moves onto the ending, in accordance with the accent paradigm). This is specified using `((ру́син<pr>,руси́н<b.pr>))`, i.e. each separate the alternants with a comma and surround them with double parentheses. (Here, `pr` means that the terms belong to the personal animacy class, and `b` specifies the accent paradigm; paradigm ''a'' is the default and hence is omitted.) * Note that occasionally, parts of a single space-delimited word can inflect separately. An example is {{m+|la|rōsmarīnus||rosemary}}, which is a compound of {{m+|la|rōs||dew}} and {{m|la|marīnus||marine, of the sea}}. In this compound, both parts of the compound can inflect separately; hence genitive singular ''rōrismarīnī'', accusative singular ''rōremmarīnum'', etc. Alternatively, only the second part inflects; hence genitive singular ''rōsmarīnī'', accusative singular ''rōsmarīnum'', etc. This is specified as `((rōs/rōr<3.M>marīnus<2>,rōsmarīnus<2>))`. Here, the term {{m|la|rōs}} by itself would have inflection spec `rōs/rōr<3.M>` (indicating that it is third declension masculine with a non-nominative-singular stem ''rōr-'') and the term {{m|la|marīnus}} would have inflection spec `<2>` (indicating that it is second declension; the masculine gender is inferred from the ''-us'' ending). When combined in a single inflection spec, the doubly-inflecting alternant is written `rōs/rōr<3.M>marīnus<2>`, with each inflecting part followed by its corresponding angle bracket spec, and the singly-inflecting alternant is written `rōsmarīnus<2>`. As this example shows, the two alternants need not correspond in how many inflecting parts there are. It should also be noted that fixed text can surround an alternant and it is even possible to supply multiple alternants in a single inflection spec (e.g. if the term has two words in it and each word requires an alternant to inflect). * The result of parsing a single angle bracket spec is stored into a '''word spec'''. The structure of a word spec is fairly arbitrary and is determined by the user-written `parse_indicator_spec` function, but always contains a form table under the `forms` key that is populated during inflection (see below). A parameter or local variable that holds a word spec is conventionally named `base` for historical reasons. Word specs are grouped together into a structure termed a '''multiword spec''', which describes one or more word specs along with the fixed text in between and around the inflected words. Multiword specs are in turn grouped into structures termed '''alternant specs''', indicating the distinct alternants and the words in each alternant. Finally, multiword specs and alternant specs are grouped into an '''alternant multiword spec''', which is the top-level object describing an inflection spec. Each of these different specs has a form table in it stored in the `forms` key that is populated during the inflection process and contains the form objects that specify the inflections of this part of the full multiword term. (It should be noted that the term '''spec''' is overloaded to mean two different things: the user-specified descriptor that specifies the lemma form of the term and associated inflection, and the associated internal Lua object that encapsulates all information derived from the descriptor, along with later-generated information on how to inflect the term(s) being described.) * Among these various "spec" structures, the two most important are the top-level alternant multiword spec and the bottom-level word spec or "base". You will rarely find it necessary to manipulate the intermediate structures or concern yourself with the details of their formation. * The term ''form'' is unfortunately overloaded in various modules to mean several things. In particular, for historical reasons, the form value inside of a form object is stored using the key `form`; the form table inside of an alternant multiword spec, a word spec (or "base") and the intermediate structures is stored using the key `forms`; and the accelerator tag set is internally referred to in [[WT:ACCEL]] as a "form". To avoid confusion, the following conventions are followed in code in this module, and should be followed for code in invoking modules as well: *# Functions that accept form objects often name the relevant parameter `form` (if a single form object is required) or `forms` (if a list of form objects, aka form list, is required). *# Functions that accept abbreviated form objects should (but don't always) indicate this by naming the parameter `abform` (for a single abbreviated form object) or `abforms` (for an abbreviated form list). *# Functions that accept a form value (the native-script string portion of a form object, stored for historical reasons in the `.form` property) should '''not''' call such a parameter `form`, but instead use something that makes clear that a form value is required, such as `formval` or sometimes just `val`. *# Similarly, functions that accept a form table should '''not''' call such a parameter `forms` (although for historical reasons the form table in an alternant multiword spec is stored in the field `forms`). Instead, use `formtable` or `formtab`, or similar name that makes clear that the value is a form table (i.e. a map from slot to form list). ====Footnote handling==== Each form can have one or more attached footnotes. The form of a footnote as specified by the user and stored in form values is e.g. {"[archaic]"} or {"[only in the meaning 'to succeed (an officeholder)']"}, i.e. the string must be surrounded by brackets and should begin with a lowercase letter and not end in a period/full stop. When such footnotes are converted to actual footnotes in a table of inflected forms, the brackets will be removed, the first letter will be capitalized and a period/full stop will be added to the end. (However, when such footnotes are used as qualifiers in headword lines, only the brackets will be removed, with no capitalization or final period.) When merging two forms into one, such as when concatenating the form objects of two inflected words in a multiword term or deduplicating form objects sharing the same form value during `show_forms()`, the footnotes are generally combined as well. This means that if one form object has footnotes and the other doesn't, the resulting form object inherits the footnotes of the object that has them, and if both form objects have footnotes, the resulting form object gets all footnotes from both source form objects, with duplicates removed. However, when inserting a form into a form table slot that already has a form whose form value and translit are identical to the new form, the behavior is different. In under normal circumstances the footnotes of the new form are ''not'' incorporated into those of the existing form (if any), but are simply dropped. To understand why this makes sense, consider a term that has two possible forms of its lemma (e.g. two forms differing in stress or in vowel length), where the second form is archaic, rare, colloquial or the like, and has an attached footnote indicating this. An example of this is {{m+|ru|кожух||sheepskin coat; bullet shell}}, where the form ''кожу́х'' with accent pattern ''b'' is more common overall but the form ''ко́жух'' with accent pattern ''c(1)'' is more common among professionals. On first glance, this could be indicated using `((кожу́х&lt;b>,ко́жух<c(1).[professional usage only]>))`. But some forms of these two declensions are the same (in particular, the genitive, dative, instrumental and prepositional plural). If for these slots, the footnotes of the duplicate forms were combined (i.e. the footnotes of the second declension pattern were added to the already-existing form taken from the first declension pattern), these forms would wrongly be labeled as ''professional usage only''. For this reason, it makes more sense to drop the footnotes of the second form when deduplicating. The same sort of behavior makes sense when a single lemma can have two different declensions, the second of which requires a footnote and where some forms in the two declensions are shared. An example of this is {{m+|uk|окови́та||strong, high-quality liquor}}, which can be inflected adjectivally or (rarely) nominally. This would be indicated as `((окови́та<sg.+>,окови́та<sg.[rare]>))` where the `+` indicates adjectival declension and the `sg` indicates that this term only exists in the singular. Here, the two declensions differ in the genitive, dative/locative and vocative (respectively, adjectival ''окови́тої'', ''окови́тій'', ''окови́та'' vs. nominal ''окови́ти'', ''окови́ті'', ''окови́то'') but are the same in the accusative (''окови́ту'') and instrumental (''окови́тою''). Again, dropping the footnotes of the second form when deduplicating is correct and including them would be wrong. This behavior can be changed by attaching a '''footnote modifier''' to the footnote associated the second form. A footnote modifier is a symbol attached to the beginning of a footnote, directly following the opening bracket. The following modifiers are currently recognized: * `!` or `+`: If placed on a footnote of the second form, combine that footnote with those of the first form (if any) rather than dropping it. * `*`: If placed on a footnote of the first form, drop that footnote when merging a second form with any footnotes. An example where the `*` modifier makes sense is a modification of the above example with {{m+|ru|кожух}}. If we notated it as `((кожу́х<b.[more common among laymen]>,ко́жух<c(1).[more common among professionals]>))`, the shared forms would wrongly have the footnote ''more common among laymen'' when in fact they are the only possible forms. If instead we used `((кожу́х<b.[*more common among laymen]>,ко́жух<c(1).[more common among professionals]>))`, the shared forms would correctly have no footnote. ]==] local function extract_footnote_modifiers(footnote) local footnote_mods, footnote_without_mods = footnote:match("^%[([!*+]?)(.*)%]$") if not footnote_mods then error("Saw footnote '" .. footnote .. "' not surrounded by brackets") end return footnote_mods, footnote_without_mods end function export.deduplicate_footnotes(existing, new) if not new then return existing end -- Check to see if there are existing footnotes with *; if so, remove them. if existing then local any_footnotes_with_asterisk = false for _, footnote in ipairs(existing) do local footnote_mods, _ = extract_footnote_modifiers(footnote) if footnote_mods:find("%*") then any_footnotes_with_asterisk = true break end end if any_footnotes_with_asterisk then local filtered_footnotes = {} for _, footnote in ipairs(existing) do local footnote_mods, _ = extract_footnote_modifiers(footnote) if not footnote_mods:find("%*") then insert(filtered_footnotes, footnote) end end if #filtered_footnotes > 0 then existing = filtered_footnotes else existing = nil end end end -- The behavior here has changed; track cases where the old behavior might -- be needed by adding ! to the footnote. track("combining-footnotes") local any_footnotes_with_bang = false for _, footnote in ipairs(new) do local footnote_mods, _ = extract_footnote_modifiers(footnote) if footnote_mods:find("[!+]") then any_footnotes_with_bang = true break end end if any_footnotes_with_bang then if not existing then existing = {} else existing = shallow_copy(existing) end for _, footnote in ipairs(new) do local already_seen = false local footnote_mods, footnote_without_mods = extract_footnote_modifiers(footnote) if footnote_mods:find("[!+]") then for _, existing_footnote in ipairs(existing) do local _, existing_footnote_without_mods = extract_footnote_modifiers(existing_footnote) if existing_footnote_without_mods == footnote_without_mods then already_seen = true break end end if not already_seen then insert(existing, footnote) end end end end return existing end --[==[ Combine two sets of footnotes. If either is {nil}, just return the other, and if both are {nil}, return {nil}. ]==] function export.combine_footnotes(notes1, notes2) if not notes1 and not notes2 then return nil end if not notes1 then return notes2 end if not notes2 then return notes1 end local combined = shallow_copy(notes1) for _, note in ipairs(notes2) do insert_if_not(combined, note) end return combined end function export.combine_formobj_lists(list1, list2, metadata) if not list1 and not list2 then return nil end if not list1 then return list2 end if not list2 then return list1 end return export.insert_formobjs_into_list { existing_list = list1, formobjs = list2, metadata = metadata, } end function export.looks_like_formobj_list(list) if list == nil then return true end if type(list) ~= "table" then return false end for k, v in pairs(list) do if type(k) ~= "number" then return false end if type(v) ~= "table" then return false end if type(v.form) ~= "string" then return false end end return true end function export.looks_like_boolean(object) return object == nil or type(object) == "boolean" end function export.default_fallback_deduplicate_handler(_prop, existing, _new) return existing, true end function export.default_fallback_combine_handler(_prop, existing, new) if export.looks_like_formobj_list(existing) and export.looks_like_formobj_list(new) then return export.combine_formobj_lists(existing, new), true end if export.looks_like_boolean(existing) and export.looks_like_boolean(new) then return existing or new, true end if existing ~= nil then return existing, true end return new, true end --[=[ Default handler functions for working with form object metadata. The user-specific metadata structure has the same format. The key is a specific metadata field and the value is an object with fields specifying handlers and other properties of the metadata. The recognized per-metadata fields are: * `deduplicate`: Handler to deduplicate form objects when inserting a form object into an existing list (e.g. in a form table) and two form objects have the same form value and translit. It is passed two values, respectively the metadata property from the existing list object and the new object being inserted into the list. The default and most common behavior here is to preserve the property from the existing object, even if {nil}, and discard the value from the new object. This is what is done in footnotes, for example, unless the existing footnote is preceded by a * (meaning to discard both values) or the new one is preceded by a + (meaning to preserve and combine both values, which is not currrently used anywhere and may not make sense). This ensures, for example, that if a noun can be declined according to two paradigms and the second one is rare, the {"rare"} footnote on the second one gets discarded whenever the two paradigms coincide. * `combine`: Handler to combine form objects when deriving a new object from an existing one. This is conceptually very different from deduplication. In this situation, the existing and new objects may or may not match in form value and translit, and since the new form object is a derivative of the existing one, it should combine the metadata properties of both, preserving as much information as possible and not discarding anything. Combining footnotes, for example, means taking the footnotes from either one, and if both exising and new objects have footnotes, merging the two lists and removing duplicates. (The only exception is if a new footnote is preceded by !, which causes existing footnotes to be discarded. This is termed an ''override'', and allows e.g. for one form of a paradigm to override the footnote associated with the paradigm as a whole, as for example if a paradigm is archaic except for one of the forms.) The default `combine` strategy is as follows: *# If both the existing and new objects are {nil} or "look like" form object lists (i.e. are tables where all keys are numeric and all values are tables with a `form` key whose value is a string), combine them using `insert_formobjs_into_list`. *# If both the existing and new objects are {nil} or boolean, combine them using `A or B`. *# Otherwise, take the existing value if it is non-{nil}, and if not, take the new value. Note that this means that the existing value will be kept even if it is {false} (which would not happen if the new value were {true}); this allows for metadata where {false} is a special signal and keys are otherwise strings or objects. However, in general if the metadata objects are not boolean and not form object lists, they will need a custom combine handler. Some keys in the metadata handler object are special: * `*`: This provides "fallback" handlers. These handlers are called for all metadata fields without dedicated handlers. These handlers work similar to normal `deduplicate` and `combine` handlers except (a) they are passed three values, first the metadata property name and then the existing and new values; (b) they should return two values, first the appropriate merged value and second a boolean indicating whether the property was handled. (If {false} is returned, the first value is ignored and should be {nil}.) Dedicated and fallback handlers are called in the following order: (1) dedicated user-specified handlers; (2) dedicated default handlers; (3) fallback user-specified handlers; (4) fallback default handlers. This ensures that standard dedicated default handlers (particularly for footnotes) get run prior to the user-specified fallback handlers; otherwise, practically all such fallback handlers will need to check for and ignore footnotes. * `_side_effect_formobjs`: If true, existing form objects can be side-effected during deduplication or combining. This requires that two logically distinct form objects in a given form table never have the same reference. (Generally two such objects should be in different slots. If two references to the same object occur in different positions in the list of form objects under a single slot, it indicates an internal error in the deduplication mechanism.) This means that user code that generates form objects for insertion into a table must take care never to reuse existing form objects (e.g. when setting one slot equal to another slot), but must always shallow-copy the form objects. This is not the default for this reason, and in fact it's not clear that implementing this results in any efficiency gain (it may actually be the reverse, since form objects have to be copied that could otherwise be shared). ]=] local default_metadata_handlers = { footnotes = { deduplicate = export.deduplicate_footnotes, combine = export.combine_footnotes, }, ["*"] = { deduplicate = export.default_fallback_deduplicate_handler, combine = export.default_fallback_combine_handler, } } local function is_non_metadata_property(k) return k == "form" or k == "translit" end -- Call the appropriate handler for merging the metadata for property `prop` with values `existing_val` and `new_val`, -- using method `method` (either "deduplicate" or "combine"). `metadata` is the structure describing all properties of -- all metadata in a given form object. This looks for, in order: -- 1. A handler for the specific property in `metadata`. -- 2. A default handler for the specific property. -- 3. A fallback handler in `metadata`. -- 4. A default fallback handler. -- Return the new value. local function call_merge_metadata_handler(metadata, prop, method, existing_val, new_val) local function try_specific(handler_struct, prop) if handler_struct and handler_struct[prop] and handler_struct[prop][method] then return (handler_struct[prop][method](existing_val, new_val)), true end return nil, false end local function try_fallback(handler_struct, prop) if handler_struct and handler_struct["*"] and handler_struct["*"][method] then -- Two values are returned from the handler, the new value and whether the property was handled. return handler_struct["*"][method](prop, existing_val, new_val) end return nil, false end local retval, matched = try_specific(metadata, prop) if matched then return retval end retval, matched = try_specific(default_metadata_handlers, prop) if matched then return retval end retval, matched = try_fallback(metadata, prop) if matched then return retval end retval, matched = try_fallback(default_metadata_handlers, prop) return retval end -- Merge the metadata of two form objects using `method` (either "deduplicate" or "combine"). `metadata` is the -- structure describing all properties of all metadata in a given form object. Metadata is merged between -- `existing_formobj` (on the left) and `new_formobj` (on the right), and is normally stored into `existing_formobj`, -- but will be stored into `new_formobj` if `merge_into_new` is given. If `metadata` allows for side-effecting, the -- changes will be made by side-effecting the object to be changed; otherwise, a shallow copy of the object will be -- made if necessary (only if the new value is not equal to the existing value). Returns two values, the changed form -- object (which may be the same as one of the passed-in objects) and a flag which is true if a copy was made of a form -- object. local function merge_metadata(metadata, method, existing_formobj, new_formobj, merge_into_new) local props_of_either = nil for k, _ in pairs(existing_formobj) do if not is_non_metadata_property(k) then props_of_either = props_of_either or {} props_of_either[k] = true end end for k, _ in pairs(new_formobj) do if not is_non_metadata_property(k) then props_of_either = props_of_either or {} props_of_either[k] = true end end local store_object = merge_into_new and new_formobj or existing_formobj if props_of_either then if metadata._side_effect_formobjs then for k, _ in pairs(props_of_either) do store_object[k] = call_merge_metadata_handler(metadata, k, method, existing_formobj[k], new_formobj[k]) end else local need_to_copy = false local new_values = {} for k, _ in pairs(props_of_either) do local newval = call_merge_metadata_handler(metadata, k, method, existing_formobj[k], new_formobj[k]) if newval ~= store_object[k] and not m_table.deepEquals(newval, store_object[k]) then need_to_copy = true end new_values[k] = newval end if need_to_copy then store_object = shallow_copy(store_object) for k, _ in pairs(props_of_either) do store_object[k] = new_values[k] end end return store_object, need_to_copy end else return store_object, false end end --[==[ Insert a form object (see above) into a list of such objects. If the form is already present (i.e. both the form value and translit, if any, match), the footnotes of the existing and new form might be combined (specifically, footnotes in the new form beginning with `+` will be combined). ]==] function export.insert_formobj_into_list(data) local existing_list = data.existing_list local formobj = data.formobj -- Don't do anything if the form object is nil. if not formobj then return existing_list end for i, existing_formobj in ipairs(existing_list) do if existing_formobj.form == formobj.form and existing_formobj.translit == formobj.translit then local new_existing, copied = merge_metadata(data.metadata, "deduplicate", existing_formobj, formobj) if copied then local list_copy = shallow_copy(existing_list) list_copy[i] = new_existing return list_copy end return existing_list end end -- Form not found. insert(existing_list, formobj) return existing_list end --[==[ Insert a list of form objects (see above) into an existing list of such objects. ``forms`` can be {nil}, in which case nothing happens. This is equivalent to calling `insert_form_into_list` for each form object. ]==] function export.insert_formobjs_into_list(data) local existing_list = data.existing_list local formobjs = data.formobjs if not formobjs then return existing_list end for _, formobj in ipairs(formobjs) do existing_list = export.insert_formobj_into_list { existing_list = existing_list, formobj = formobj, metadata = data.metadata, } end return existing_list end --[==[ Insert a form object (see above) into the given slot in the given form table. On input, `data` is an object with the following fields: * `formtable`: A form table (see above). * `slot`: A slot in the form table, i.e. a key whose value is a form object list. * `formobj`: A form object to be inserted into the slot. Can be {nil}, in which case nothing happens. * `metadata`: The metadata structure describing how to combine user-specified metadata. Can be {nil} if there is no user-specified metadata or the default fallback metadata handlers are sufficient (which is usually the case if the all user-specified metadata is boolean). ]==] function export.insert_formobj_into_form_table(data) local formtable, slot, formobj, metadata = data.formtable, data.slot, data.formobj, data.metadata -- Don't do anything if the form object is nil. if not formobj then return end if not formtable[slot] then formtable[slot] = {} end formtable[slot] = export.insert_formobj_into_list { existing_list = formtable[slot], formobj = formobj, metadata = metadata, } end --[==[ Insert a list of form objects (see above) into the given slot in the given form table. On input, `data` is an object with the following fields: * `formtable`: A form table (see above). * `slot`: A slot in the form table, i.e. a key whose value is a form object list. * `formobjs`: A list of form objects (in "full list" form) to be inserted into the slot. Can be {nil} or an empty list, in which case nothing happens. * `metadata`: The metadata structure describing how to combine user-specified metadata. Can be {nil} if there is no user-specified metadata or the default fallback metadata handlers are sufficient (which is usually the case if the all user-specified metadata is boolean). ]==] function export.insert_formobjs_into_form_table(data) local formtable, slot, formobjs, metadata = data.formtable, data.slot, data.formobjs, data.metadata if not formobjs or not formobjs[1] then return end if not formtable[slot] then formtable[slot] = {} end formtable[slot] = export.insert_formobjs_into_list { existing_list = formtable[slot], formobjs = formobjs, metadata = metadata, } end --[==[ Flatmap a function over a list of form objects and return the (potentially flattened) results. Flatmapping is like a normal mapping operation except that the function being called may return a list, and all values returned are concatenated into a single flattened list. On input, `data` is an object with the following fields: * `formobjs`: A list of form objects to map over, in "full list form" (see above). * `fn`: The function to map over each form object. If a form object's value is {"?"}, the form object is preserved on output and the function is not called. Otherwise, the function is called with one argument, the form object, and should {nil} (equivalent to returning an empty list) or an abbreviated form list (i.e. anything that is convertible into a full list form, such as a single form value, a list of form values, a form object or a list of form objects). For each form object in the return value, the metadata of the object mapped over is combined with the metadata of the returned object, unless `no_combine_metadata` is given. The resulting form objects from all function calls are flattened into a list and deduplicated in the same fashion as `insert_formobj_into_list()` (in case two different form objects map to the same form value). * `metadata`: The metadata structure describing how to combine user-specified metadata. Can be {nil} if there is no user-specified metadata or the default fallback metadata handlers are sufficient (which is usually the case if the all user-specified metadata is boolean). * `no_combine_metadata`: See `fn` above. If given, metadata from objects mapped over will not be combined into the returned objects. NOTE: There is no `map_formobjs` function provided because `flatmap_formobjs` will correctly handle functions that return single form objects. ]==] function export.flatmap_formobjs(data) local formobjs, fn, metadata, no_combine_metadata = data.formobjs, data.fn, data.metadata, data.no_combine_metadata if not formobjs then return nil end local retval = {} for _, formobj in ipairs(formobjs) do local funret if formobj.form == "?" then if metadata and metadata._side_effect_formobjs then funret = shallow_copy(formobj) else funret = formobj end else funret = fn(formobj) end if funret then funret = export.convert_to_full_list_form(funret) for _, fr in ipairs(funret) do if not no_combine_metadata then fr = merge_metadata(metadata, "combine", formobj, fr, "merge_into_new") end retval = export.insert_formobj_into_list { existing_list = retval, formobj = fr, metadata = metadata, } end end end return retval end --[==[ Insert a form object (see above) into a list of such objects. If the form is already present (i.e. both the form value and translit, if any, match), the footnotes of the existing and new form might be combined (specifically, footnotes in the new form beginning with `!` will be combined). ]==] function export.insert_form_into_list(list, form) -- Don't do anything if the form object or the form inside it is nil. This simplifies -- form insertion in the presence of inflection generating functions that may return nil, -- such as generate_noun_vocative() and generate_noun_count_form(). if not form or not form.form then return end for _, listform in ipairs(list) do if listform.form == form.form and listform.translit == form.translit then listform.footnotes = export.deduplicate_footnotes(listform.footnotes, form.footnotes) return end end -- Form not found. insert(list, form) end --[==[ Insert a list of form objects (see above) into an existing list of such objects. ``forms`` can be {nil}, in which case nothing happens. This is equivalent to calling `insert_form_into_list` for each form object. ]==] function export.insert_forms_into_list(list, forms) if not forms then return end for _, form in ipairs(forms) do export.insert_form_into_list(list, form) end end --[==[ Insert a form object (see above) into the given slot in the given form table. ``form`` can be {nil}, in which case nothing happens. ]==] function export.insert_form(formtable, slot, form) -- Don't do anything if the form object or the form inside it is nil. This simplifies -- form insertion in the presence of inflection generating functions that may return nil, -- such as generate_noun_vocative() and generate_noun_count_form(). if not form or not form.form then return end if not formtable[slot] then formtable[slot] = {} end export.insert_form_into_list(formtable[slot], form) end --[==[ Insert a list of form objects (see above) into the given slot in the given form table. ``forms`` can be {nil}, in which case nothing happens. ]==] function export.insert_forms(formtable, slot, forms) if not forms then return end for _, form in ipairs(forms) do export.insert_form(formtable, slot, form) end end --[==[ Identity mapping function. ]==] function export.identity(formval, translit) return formval, translit end local function form_value_transliterable(formval) return formval ~= "?" and formval ~= "—" end local function call_map_function_str(str, fun) if str == "?" then return "?" end local newformval, newtranslit = fun(str) if newtranslit then return {form=newformval, translit=newtranslit} else return newformval end end -- FIXME: This doesn't correctly handle metadata. local function call_map_function_obj(form, fun) if form.form == "?" then return {form = "?", footnotes = form.footnotes} end local newformval, newtranslit = fun(form.form, form.translit) return {form = newformval, translit = newtranslit, footnotes = form.footnotes} end --[==[ Map a function over the form values in ``forms`` (a list of form objects in "full list form"; see above). If an input form value is {"?"}, it is preserved on output and the function is not called. Otherwise, the function is called with two arguments, the original form and manual translit; if manual translit isn't relevant, it's fine to declare the function with only one argument. The return value is either a single value (the new form) or two values (the new form and new manual translit). The footnotes (if any) from the input form objects are preserved on output. Uses `insert_form_into_list()` to insert the resulting form objects into the returned list in case two different forms map to the same thing. FIXME: Expand this to correctly handle metadata, or create a variant that correctly handles metadata. ]==] function export.map_forms(forms, fun) if not forms then return nil end local retval = {} for _, form in ipairs(forms) do export.insert_form_into_list(retval, call_map_function_obj(form, fun)) end return retval end --[==[ Map a list-returning function over the form values in ``forms`` (a list of form objects in "full list form"; see above). If an input form value is {"?"}, it is preserved on output and the function is not called. Otherwise, the function is called with two arguments, the original form and manual translit; if manual translit isn't relevant, it's fine to declare the function with only one argument. The return value of the function can be {nil} or an abbreviated form list (i.e. anything that is convertible into a full list form, such as a single form value, a list of form values, a form object or a list of form objects). For each form object in the return value, the footnotes of that form object (if any) are combined with any footnotes from the input form object, and the result inserted into the returned list using `insert_form_into_list()` in case two different forms map to the same thing. ]==] function export.flatmap_forms(forms, fun) if not forms then return nil end local retval = {} for _, form in ipairs(forms) do local funret = form.form == "?" and form or fun(form.form, form.translit) if funret then funret = export.convert_to_general_list_form(funret) for _, fr in ipairs(funret) do local newform = { form = fr.form, translit = fr.translit, footnotes = export.combine_footnotes(form.footnotes, fr.footnotes) } export.insert_form_into_list(retval, newform) end end end return retval end --[==[ Map a function over the form values in ``abforms`` (an abbreviated form list). If the input form value is {"?"}, it is preserved on output and the function is not called. If ``first_only`` is given and ``abforms`` is a list, only map over the first element. Return value is of the same form as ``abforms``, unless ``abforms`` is a string and the function returns both form value and manual translit (in which case the return value is a form object). The function is called with two arguments, the original form value and manual translit; if manual translit isn't relevant, it's fine to declare the function with only one argument. The return value is either a single value (the new form value) or two values (the new form value and new manual translit). The footnotes (if any) from the input form objects are preserved on output. FIXME: This function is used only in [[Module:bg-verb]] and should be moved into that module. ]==] function export.map_form_or_forms(abforms, fun, first_only) if not abforms then return nil elseif type(abforms) == "string" then return call_map_function_str(abforms, fun) elseif abforms.form then return call_map_function_obj(abforms, fun) else local retval = {} for i, abform in ipairs(abforms) do if first_only then return export.map_form_or_forms(abform, fun) end insert(retval, export.map_form_or_forms(abform, fun)) end return retval end end --[==[ Expand a given footnote (as specified by the user, including the surrounding brackets) into the form to be inserted into the final generated table. If ``no_parse_refs`` is not given and the footnote is a reference (of the form {"[ref:...]"}), parse and return the specified reference(s). Two values are returned, `footnote_string` (the expanded footnote, or nil if the second value is present) and `references` (a list of objects of the form `{text = ``text``, name = ``name``, group = ``group``}` if the footnote is a reference and ``no_parse_refs`` is not given, otherwise {nil}). Unless ``return_raw`` is given, the returned footnote string is capitalized and has a final period added. ]==] function export.expand_footnote_or_references(note, return_raw, no_parse_refs) local _, notetext = extract_footnote_modifiers(note) if not no_parse_refs and notetext:find("^ref:") then -- a reference notetext = notetext:gsub("^ref:", "") local parsed_refs = require("Module:references").parse_references(notetext) for i, ref in ipairs(parsed_refs) do if type(ref) == "string" then parsed_refs[i] = {text = ref} end end return nil, parsed_refs end if footnote_abbrevs[notetext] then notetext = footnote_abbrevs[notetext] track("footnote-whole-abbrev") else local split_notes = split(notetext, "<(.-)>") for i, split_note in ipairs(split_notes) do if i % 2 == 0 then split_notes[i] = footnote_abbrevs[split_note] track("footnote-angle-bracket-abbrev") if not split_notes[i] then -- Don't error for now, because HTML might be in the footnote. -- Instead we should switch the syntax here to e.g. <<a>> to avoid -- conflicting with HTML. split_notes[i] = "<" .. split_note .. ">" track("footnote-unrecognized-angle-bracket-abbrev") --error("Unrecognized footnote abbrev: <" .. split_note .. ">") else track("footnote-recognized-angle-bracket-abbrev") end end end notetext = concat(split_notes) end return return_raw and notetext or ucfirst(notetext) .. "." end --[==[ Convert a list of foonotes to qualifiers and references for use in [[Module:headword]] or similar. Returns two values, a list of qualifiers (possibly {nil}) and a list of reference structures (possibly {nil}), following the structure defined in [[Module:references]]). ]==] function export.convert_footnotes_to_qualifiers_and_references(footnotes) if not footnotes then return nil end local quals, refs for _, qualifier in ipairs(footnotes) do local this_footnote, this_refs = export.expand_footnote_or_references(qualifier, "return raw") if this_refs then if not refs then refs = this_refs else for _, ref in ipairs(this_refs) do insert(refs, ref) end end else if not quals then quals = {this_footnote} else insert(quals, this_footnote) end end end return quals, refs end --[==[ Combine an abbreviated form object (either a string or a table) with additional footnotes, possibly replacing the form value and/or translit in the process. Normally called in one of two ways: (1) `combine_form_and_footnotes(``form_obj``, ``addl_footnotes``, ``new_form``, ``new_translit``)` where ``form_obj`` is an existing abbreviated form object; ``addl_footnotes`` is either {nil}, a single string (a footnote) or a list of footnotes; ``new_formval`` is either {nil} or the new form value to substitute; and ``new_translit`` is either {nil} or the new translit string to substitute. (2) `combine_form_and_footnotes(``form_value``, ``footnotes``)`, where ``form_value`` is a form value (a string) and ``footnotes`` is either {nil}, a single string (a footnote) or a list of footnotes. In either case, a form object is returned, preserving as many properties as possible from any existing form object in ``abform``. Do the minimal amount of work; e.g. if ``abform`` is a form object and ``addl_footnotes``, ``new_formval`` and ``new_translit`` are all {nil}, the same object as passed in is returned. Under no circumstances is the existing form object side-effected. '''FIXME:''' This does not correctly preserve metadata. ]==] function export.combine_form_and_footnotes(abform, addl_footnotes, new_formval, new_translit) if type(addl_footnotes) == "string" then addl_footnotes = {addl_footnotes} end if not addl_footnotes and not new_formval and not new_translit then return abform end if type(abform) == "string" then new_formval = new_formval or abform return {form = new_formval, translit = new_translit, footnotes = addl_footnotes} end abform = shallow_copy(abform) if new_formval then abform.form = new_formval end if new_translit then abform.translit = new_translit end if addl_footnotes then abform.footnotes = export.combine_footnotes(abform.footnotes, addl_footnotes) end return abform end --[==[ Convert an abbreviated form list (either a string, form object, or list of either) into full list form. This function does not side-effect any of the objects passed into ``abforms``, but will return ``abforms`` unchanged if already in full list form. ]==] function export.convert_to_full_list_form(abforms) if type(abforms) == "string" then return {{form = abforms}} elseif abforms.form then return {abforms} else -- Check if already in full list form and return directly if so. local must_convert = false for _, form in ipairs(abforms) do if type(form) == "string" then must_convert = true break end end if not must_convert then return abforms end end local retval = {} for _, form in ipairs(abforms) do if type(form) == "string" then insert(retval, {form = form}) else insert(retval, form) end end return retval end --[==[ Convert an abbreviated form list (either a string, form object, or list of either) into full list form. If ``footnotes`` is supplied, then for each form in the form list, combine the form's footnotes with ``footnotes``. This function does not side-effect any of the objects passed into ``abforms``, but will return ``abforms`` unchanged if already in full list form and ``footnotes`` is {nil}. '''FIXME:''' This does not correctly preserve metadata. ]==] function export.convert_to_general_list_form(abforms, footnotes) if type(footnotes) == "string" then footnotes = {footnotes} end if type(abforms) == "string" then return {{form = abforms, footnotes = footnotes}} elseif abforms.form then return {export.combine_form_and_footnotes(abforms, footnotes)} elseif not footnotes then -- Check if already in full list form and return directly if so. local must_convert = false for _, form in ipairs(abforms) do if type(form) == "string" then must_convert = true break end end if not must_convert then return abforms end end local retval = {} for _, form in ipairs(abforms) do if type(form) == "string" then insert(retval, {form = form, footnotes = footnotes}) else insert(retval, export.combine_form_and_footnotes(form, footnotes)) end end return retval end local function is_table_of_strings(forms) for k, v in pairs(forms) do if type(k) ~= "number" or type(v) ~= "string" then return false end end return true end local function lang_or_func_transliterate(func, lang, text) local retval if func then retval = func(text) else retval = (lang:transliterate(text)) end -- FIXME! Hack to work around bug in ...:transliterate(). Remove me as soon as this bug is fixed. if not retval and (text == " " or text == "-" or text == "-́" or text == "?" or text == "-∅") then retval = text end if not retval then error(("Unable to transliterate text '%s'"):format(text)) end return retval end --[==[ Combine ``stems`` and ``endings`` and store into slot ``slot`` of form table ``formtable``. Either of ``stems`` and ``endings`` can be {nil} or an abbreviated form list. The combination of a given stem and ending happens using ``combine_stem_ending``, which takes two parameters (stem and ending, each a string) and returns one value (a string). If manual transliteration is present in either ``stems`` or ``endings``, ``lang`` (a language object or a function of one argument to transliterate a string) along with ``combine_stem_ending_tr`` (a function for combining manual transliterations that works much like ``combine_stem_ending``) must be given. ``footnotes``, if specified, is a list of additional footnotes to attach to the resulting inflections (stem+ending combinations). The resulting inflections are inserted into the form table using `insert_form()`, in case of duplication. ]==] function export.add_forms(formtable, slot, stems, endings, combine_stem_ending, lang, combine_stem_ending_tr, footnotes) if stems == nil or endings == nil then return end local function combine(stem, ending) if stem == "?" or ending == "?" then return "?" end return combine_stem_ending(stem, ending) end local function transliterate(text) return lang_or_func_transliterate(is_callable(lang) and lang or nil, lang, text) end if type(stems) == "string" and type(endings) == "string" then export.insert_form(formtable, slot, {form = combine(stems, endings), footnotes = footnotes}) elseif type(stems) == "string" and is_table_of_strings(endings) then for _, ending in ipairs(endings) do export.insert_form(formtable, slot, {form = combine(stems, ending), footnotes = footnotes}) end else stems = export.convert_to_general_list_form(stems) endings = export.convert_to_general_list_form(endings, footnotes) for _, stem in ipairs(stems) do for _, ending in ipairs(endings) do local new_footnotes = nil if stem.footnotes and ending.footnotes then new_footnotes = shallow_copy(stem.footnotes) for _, footnote in ipairs(ending.footnotes) do insert_if_not(new_footnotes, footnote) end elseif stem.footnotes then new_footnotes = stem.footnotes elseif ending.footnotes then new_footnotes = ending.footnotes end local new_form = combine(stem.form, ending.form) local new_translit if new_form ~= "?" and (stem.translit or ending.translit) then if not lang or not combine_stem_ending_tr then error("Internal error: With manual translit, 'lang' and 'combine_stem_ending_tr' must be passed to 'add_forms'") end local stem_tr = stem.translit or transliterate(m_links.remove_links(stem.form)) local ending_tr = ending.translit or transliterate(m_links.remove_links(ending.form)) new_translit = combine_stem_ending_tr(stem_tr, ending_tr) end export.insert_form(formtable, slot, {form = new_form, translit = new_translit, footnotes = new_footnotes}) end end end end --[==[ Combine any number of form components and store into slot ``slot`` of form table ``formtable``. ``components`` is a list of abbreviated form lists which should be concatenated similarly to how `add_forms()` does it, and stored in ``slot`` along with any footnotes in ``footnotes``. More specifically: # If there are no components, nothing happens. # If there is one component, it is converted to full list form and `insert_forms()` called. # If there are two components, they are treated as stems and endings respectively and `add_forms()` is called. # If there are three or more components, they are concatenated left-to-right in the manner of a `reduce()` operation: the first two components are combined using `add_forms()` and stored into a temporary table, then the next component is combined with the result of the previous operation, etc. In the last combination, footnotes in `footnotes` are combined in, and the result stored into `formtable`. This should generally be used when you are likely to have three or more components, as in [[Module:ar-verb]] (prefixes, stems and endings) and [[Module:de-verb]] (which in some situations has five components combined together). ``combine_stem_ending``, ``lang``, ``combine_stem_ending_tr`` and ``footnotes`` are as in `add_forms()`. ]==] function export.add_multiple_forms(formtable, slot, components, combine_stem_ending, lang, combine_stem_ending_tr, footnotes) if #components == 0 then return elseif #components == 1 then local forms = export.convert_to_general_list_form(components[1], footnotes) export.insert_forms(formtable, slot, forms) elseif #components == 2 then local stems = components[1] local endings = components[2] export.add_forms(formtable, slot, stems, endings, combine_stem_ending, lang, combine_stem_ending_tr, footnotes) else local prev = components[1] for i=2, #components do local temptable = {} export.add_forms(temptable, slot, prev, components[i], combine_stem_ending, lang, combine_stem_ending_tr, i == #components and footnotes or nil) prev = temptable[slot] end export.insert_forms(formtable, slot, prev) end end local function iterate_slot_list_or_table(props, do_slot) if props.slot_list then for _, slot_and_accel_tag_set in ipairs(props.slot_list) do local slot, accel_tag_set = unpack(slot_and_accel_tag_set) do_slot(slot, accel_tag_set) end else for slot, accel_tag_set in pairs(props.slot_table) do do_slot(slot, accel_tag_set) end end end function export.default_split_bracketed_runs_into_words(bracketed_runs, data) -- If the text begins with a hyphen, include the hyphen in the set of allowed characters -- for an inflected segment. This way, e.g. conjugating "-ir" is treated as a regular -- -ir verb rather than a hyphen + irregular [[ir]]. local is_suffix = (not data or data.text_index == 1) and bracketed_runs[1]:find("^%-") local split_pattern = is_suffix and " " or "[ %-]" return put.split_alternating_runs(bracketed_runs, split_pattern, "preserve splitchar") end local function props_transliterate(props, text) return lang_or_func_transliterate(props.transliterate, props.lang, text) end local function parse_before_or_post_text(data) -- NOTE: data.text_index is also used (in `default_split_bracketed_runs_into_words`). local props, text, segments, lemma_is_last = data.props, data.text, data.segments, data.lemma_is_last -- Call parse_balanced_segment_run() to keep multiword links together. local bracketed_runs = put.parse_balanced_segment_run(text, "[", "]") -- Split normally on space or hyphen (but customizable). Use preserve_splitchar so we know whether the separator was -- a space or hyphen. local space_separated_groups if props.split_bracketed_runs_into_words then space_separated_groups = props.split_bracketed_runs_into_words(bracketed_runs) end if not space_separated_groups then space_separated_groups = export.default_split_bracketed_runs_into_words(bracketed_runs, data) end local parsed_components = {} local parsed_components_translit = {} local saw_manual_translit = false local lemma for j, space_separated_group in ipairs(space_separated_groups) do local component = concat(space_separated_group) if lemma_is_last and j == #space_separated_groups then lemma = component if lemma == "" and not props.allow_blank_lemma then error("Word is blank: '" .. concat(segments) .. "'") end elseif component:find("//") then -- Manual translit or respelling specified. if not props.lang then error("Manual translit not allowed for this language; if this is incorrect, 'props.lang' must be set internally") end saw_manual_translit = true local split_components = split(component, "//", "plain") if #split_components ~= 2 then error("Term with translit or respelling should have only one // in it: " .. component) end local translit component, translit = unpack(split_components) if props.transliterate_respelling then translit = props.transliterate_respelling(translit) end insert(parsed_components, component) insert(parsed_components_translit, translit) else insert(parsed_components, component) insert(parsed_components_translit, false) -- signal that it may need later transliteration end end if saw_manual_translit then for j, parsed_component in ipairs(parsed_components) do if not parsed_components_translit[j] then parsed_components_translit[j] = props_transliterate(props, m_links.remove_links(parsed_component)) end end end text = concat(parsed_components) local translit if saw_manual_translit then translit = concat(parsed_components_translit) end return text, translit, lemma end --[=[ Parse a segmented multiword spec such as "[[медичний|меди́чна]]<+> [[сестра́]]<*,*#.pr>" (in Ukrainian). "Segmented" here means it is broken up on <...> segments using parse_balanced_segment_run(text, "<", ">"), e.g. the above text would be passed in as {"[[медичний|меди́чна]]", "<+>", " [[сестра́]]", "<*,*#.pr>", ""}. The return value is a table of the form { word_specs = {``word_spec``, ``word_spec``, ...}, post_text = "``text-at-end``", post_text_no_links = "``text-at-end-no-links``", post_text_translit = "``manual-translit-of-text-at-end``" or nil (if no manual translit or respelling was specified in the post-text) } where ``word_spec`` describes an individual inflected word and "``text-at-end``" is any raw text that may occur after all inflected words. Individual words or linked text (including multiword text) may be given manual transliteration or respelling in languages that support this using ``text``//``translit`` or ``text``//``respelling``. Each ``word_spec`` is of the form returned by parse_indicator_spec(): { lemma = "``lemma``", before_text = "``text-before-word``", before_text_no_links = "``text-before-word-no-links``", before_text_translit = "``manual-translit-of-text-before-word``" or nil (if no manual translit or respelling was specified in the before-text) -- Fields as described in parse_indicator_spec() ... } For example, the return value for "[[медичний|меди́чна]]<+> [[сестра́]]<*,*#.pr>" is { word_specs = { { lemma = "[[медичний|меди́чна]]", overrides = {}, adj = true, before_text = "", before_text_no_links = "", forms = {}, }, { lemma = "[[сестра́]]", overrides = {}, stresses = { { reducible = true, genpl_reversed = false, }, { reducible = true, genpl_reversed = true, }, }, animacy = "pr", before_text = " ", before_text_no_links = " ", forms = {}, }, }, post_text = "", post_text_no_links = "", } ]=] local function parse_multiword_spec(segments, props, disable_allow_default_indicator) local multiword_spec = { word_specs = {} } if not disable_allow_default_indicator then if #segments == 1 then if props.allow_default_indicator then insert(segments, "<>") insert(segments, "") elseif props.angle_brackets_omittable then segments[1] = "<" .. segments[1] .. ">" insert(segments, 1, "") insert(segments, "") end end end -- Loop over every other segment. The even-numbered segments are angle-bracket specs while -- the odd-numbered segments are the text between them. for i = 2, #segments - 1, 2 do local before_text, before_text_translit, lemma = parse_before_or_post_text { props = props, text = segments[i - 1], text_index = i - 1, segments = segments, lemma_is_last = true } local base = props.parse_indicator_spec(segments[i], lemma) base.before_text = before_text base.before_text_no_links = m_links.remove_links(base.before_text) base.before_text_translit = before_text_translit base.lemma = base.lemma or lemma insert(multiword_spec.word_specs, base) end multiword_spec.post_text, multiword_spec.post_text_translit = parse_before_or_post_text { props = props, text = segments[#segments], text_index = #segments, segments = segments, lemma_is_last = false } multiword_spec.post_text_no_links = m_links.remove_links(multiword_spec.post_text) return multiword_spec end --[=[ Parse an alternant, e.g. "((родо́вий,родови́й))" or "((ру́син<pr>,руси́н<b.pr>))" (both in Ukrainian). The return value is a table of the form { alternants = {``multiword_spec``, ``multiword_spec``, ...} } where ``multiword_spec`` describes a given alternant and is as returned by parse_multiword_spec(). ]=] local function parse_alternant(alternant, props) local alternant_text = alternant:match("^%(%((.*)%)%)$") local segments = put.parse_balanced_segment_run(alternant_text, "<", ">") local comma_separated_groups = put.split_alternating_runs(segments, "%s*,%s*") local alternant_spec = {alternants = {}} for _, comma_separated_group in ipairs(comma_separated_groups) do insert(alternant_spec.alternants, parse_multiword_spec(comma_separated_group, props)) end return alternant_spec end --[==[ Top-level parsing function. Parse text describing one or more inflected words. `text` is the inflected text to parse, which generally has `<...>` specs following words to be inflected, and may have alternants indicated using double parens. Examples: * {"[[медичний|меди́чна]]<+> [[сестра́]]<*,*#.pr>"} (Ukrainian, for {{m|uk|меди́чна сестра́||nurse|lit=medical sister}}); * {"((ру́син<pr>,руси́н<b.pr>))"} (Ukrainian, for {{m|uk|русин||Rusyn}}, with two possible stress patterns); * {"पंचायती//पंचाय*ती राज<M>"} (Hindi, for {{m|hi|पंचायती राज||village council}}, with phonetic respelling in the before-text component); * {"((<M>,<M.plstem:फ़तूह.dirpl:फ़तूह>))"} (Hindi, for {{m|hi|फ़तह||win, victory}} when used on that page, where the lemma is omitted and taken from the pagename); * {""} (for any number of Hindi adjectives, where the lemma is omitted and taken from the pagename, and the angle bracket spec <> is assumed); * {"काला<+>धन<M>"} (Hindi, for {{m|hi|कालाधन||black money}}, showing that closed compounds where each part is declined can be correctly handled). `props` is an object specifying properties used during parsing, as follows: ```{ parse_indicator_spec = __function__(``angle_bracket_spec``, ``lemma``) `''(required)''`, lang = __lang object__, transliterate_respelling = __function__(``respelling_or_translit``) `''(optional)''`, split_bracketed_runs_into_words = __function__(``bracket_split_runs``) `''(optional)''`, allow_default_indicator = __boolean__, angle_brackets_omittable = __boolean__, allow_blank_lemma = __boolean__, }``` `parse_indicator_spec` is a required function that takes two arguments, a string surrounded by angle brackets and the lemma, and should return an arbitrary object containing properties describing the indicators inside of the angle brackets). This object is often called a '''base''' and given the argument name `base` in inflection code. `lang` is the language object for the language in question; only needed if manual translit or respelling may be present using `//`. `transliterate_respelling` is a function that is only needed if respelling is allowed in place of manual translit after `//`. It takes one argument, the respelling or translit, and should return the transliteration of any respelling but return any translit unchanged. `split_bracketed_runs_into_words` is an optional function to split the passed-in text into words. It is used, for example, to determine what text constitutes a word when followed by an angle-bracket spec, i.e. what the lemma to be inflected is vs. surrounding fixed text. It takes one argument, the result of splitting the original text on brackets, and should return alternating runs of words and split characters, or nil to apply the default algorithm. Specifically, the value passed in is the result of calling `parse_balanced_segment_run(``text``, "[", "]")` from [[Module:parse utilities]] on the original text, and the default version of this function calls `split_alternating_runs(``bracketed_runs``, ``pattern``, "preserve splitchar")`, where ``bracketed_runs`` is the value passed in and ``pattern`` splits on either spaces or hyphens (unless the text begins with a hyphen, in which case splitting is only on spaces, so that suffixes can be inflected). `allow_default_indicator` should be {true} if an empty indicator in angle brackets `<>` can be omitted and should be automatically added at the end of the multiword text (if no alternants) or at the end of each alternant (if alternants present). `angle_brackets_omittable` should be {true} if angle brackets can be omitted around a non-empty indicator in the presence of a blank lemma. In this case, if the combined indicator spec has no angle brackets, they will be added around the indicator (or around all indicators, if alternants are present). This only makes sense when `allow_blank_lemma` is specified. `allow_blank_lemma` should be {true} of if a blank lemma is allowed; in such a case, the calling function should substitute a default lemma, typically taken from the pagename. The return value is a table referred to as an '''alternant multiword spec''', and is of the form ```{ alternant_or_word_specs = {``alternant_or_word_spec``, ``alternant_or_word_spec``, ...}, post_text = "``text_at_end``", post_text_no_links = "``text_at_end_no_links``", post_text_translit = "``translit_of_text_at_end``" `(or nil)`, }``` where `alternant_or_word_spec` is either an '''alternant spec''' as returned by `parse_alternant()` or a '''multiword spec''' as described in the comment above `parse_multiword_spec()`. An alternant spec looks as follows: ```{ alternants = {``multiword_spec``, ``multiword_spec``, ...}, before_text = "``text_before_alternant``", before_text_no_links = "``text_before_alternant``", before_text_translit = "``translit_of_text_before_alternant``" `(or nil)`, }``` i.e. it is like what is returned by `parse_alternant()` but has extra `before_text` and `before_text_no_links` fields. ]==] function export.parse_inflected_text(text, props) if props.angle_brackets_omittable and not props.allow_blank_lemma then error("If 'angle_brackets_omittable' is specified, so should 'allow_blank_lemma'") end local alternant_multiword_spec = {alternant_or_word_specs = {}} local alternant_segments = split(text, "(%(%(.-%)%))") local last_post_text, last_post_text_no_links, last_post_text_translit for i = 1, #alternant_segments do if i % 2 == 1 then local segments = put.parse_balanced_segment_run(alternant_segments[i], "<", ">") -- Disable allow_default_indicator if alternants are present and we're processing -- the non-alternant text. Otherwise we will try to treat the non-alternant text -- surrounding the alternants as an inflected word rather than as raw text. local multiword_spec = parse_multiword_spec(segments, props, #alternant_segments ~= 1) for _, word_spec in ipairs(multiword_spec.word_specs) do insert(alternant_multiword_spec.alternant_or_word_specs, word_spec) end last_post_text = multiword_spec.post_text last_post_text_no_links = multiword_spec.post_text_no_links last_post_text_translit = multiword_spec.post_text_translit else local alternant_spec = parse_alternant(alternant_segments[i], props) alternant_spec.before_text = last_post_text alternant_spec.before_text_no_links = last_post_text_no_links alternant_spec.before_text_translit = last_post_text_translit insert(alternant_multiword_spec.alternant_or_word_specs, alternant_spec) end end alternant_multiword_spec.post_text = last_post_text alternant_multiword_spec.post_text_no_links = last_post_text_no_links alternant_multiword_spec.post_text_translit = last_post_text_translit -- Save boolean properties from `props`. We need at least `allow_default_indicator` when implementing -- `reconstruct_original_spec()`. alternant_multiword_spec.allow_default_indicator = props.allow_default_indicator alternant_multiword_spec.angle_brackets_omittable = props.angle_brackets_omittable alternant_multiword_spec.allow_blank_lemma = props.allow_blank_lemma return alternant_multiword_spec end -- Inflect alternants in ``alternant_spec`` (an object as returned by parse_alternant()). -- This sets the form values in ```alternant_spec``.forms` for all slots. -- (If a given slot has no values, it will not be present in ```alternant_spec``.forms`). local function inflect_alternants(alternant_spec, props) alternant_spec.forms = {} for _, multiword_spec in ipairs(alternant_spec.alternants) do export.inflect_multiword_or_alternant_multiword_spec(multiword_spec, props) iterate_slot_list_or_table(props, function(slot) if not props.skip_slot or not props.skip_slot(slot) then export.insert_forms(alternant_spec.forms, slot, multiword_spec.forms[slot]) end end) end end --[=[ Subfunction of `inflect_multiword_or_alternant_multiword_spec()`. This is used in building up the inflections of multiword expressions. The basic purpose of this function is to append a set of forms representing the inflections of a given inflected term in a given slot onto the existing forms for that slot. Given a multiword expression potentially consisting of several inflected terms along with fixed text in between, we work iteratively from left to right, adding the new forms onto the existing ones. Normally, all combinations of new and existing forms are created, meaning if there are M existing forms and N new ones, we will end up with M*N forms. However, some of these combinations can be rejected using the variant mechanism (see the description of get_variants below). Specifically, `formtable` is a table of per-slot forms, where the key is a slot and the value is a list of form objects (objects of the form {form=``form``, translit=``manual_translit``, footnotes=``footnotes``}). `slot` is the slot in question. `forms` specifies the forms to be appended onto the existing forms, and is likewise a list of form objects. `props` is the same as in `inflect_multiword_or_alternant_multiword_spec()`. `before_text` is the fixed text that goes before the forms to be added. `before_text_no_links` is the same as `before_text` but with any links (i.e. hyperlinks of the form [[``term``]] or [[``term``|``display``]]) converted into raw terms using remove_links() in [[Module:links]], and `before_text_translit` is optional manual translit of `before_text_no_links`. Note that the value "?" in a form is "infectious" in that if either the existing or new form has the value "?", the resulting combination will also be "?". This allows "?" to be used to mean "unknown". ]=] local function append_forms(props, formtable, slot, forms, before_text, before_text_no_links, before_text_translit) if not forms then return end local old_forms = formtable[slot] or {{form = ""}} local ret_forms = {} for _, old_form in ipairs(old_forms) do for _, form in ipairs(forms) do local old_form_vars = props.get_variants and props.get_variants(old_form.form) or "" local form_vars = props.get_variants and props.get_variants(form.form) or "" if old_form_vars ~= "" and form_vars ~= "" and old_form_vars ~= form_vars then -- Reject combination due to non-matching variant codes. else local new_formval local new_translit if old_form.form == "?" or form.from == "?" then new_formval = "?" else new_formval = old_form.form .. before_text .. form.form if old_form.translit or before_text_translit or form.translit then if not props.lang then error("Internal error: If manual translit is given, 'props.lang' must be set") end if not before_text_translit then before_text_translit = props_transliterate(props, before_text_no_links) or "" end local old_translit = old_form.translit or props_transliterate(props, m_links.remove_links(old_form.form)) or "" local translit = form.translit or props_transliterate(props, m_links.remove_links(form.form)) or "" new_translit = old_translit .. before_text_translit .. translit end end local new_formobj local new_footnotes = export.combine_footnotes(old_form.footnotes, form.footnotes) if new_formval == form.form and new_translit == form.translit then -- Automatically preserve metadata when possible. new_formobj = shallow_copy(form) new_formobj.footnotes = new_footnotes else new_formobj = {form=new_formval, translit=new_translit, footnotes=new_footnotes} if props.combine_metadata then props.combine_metadata { slot = slot, dest_form = new_formobj, form1 = old_form, form2 = form, between_text = before_text, between_text_no_links = before_text_no_links, between_text_translit = before_text_translit, } end end insert(ret_forms, new_formobj) end end end formtable[slot] = ret_forms end --[==[ Top-level inflection function. Create the inflections of a noun, verb, adjective or similar. `alternant_multiword_spec` is as returned by `parse_inflected_text` and describes the properties of the term to be inflected, including all the user-provided inflection specifications (e.g. the number, gender, conjugation/declension/etc. of each word) and the surrounding text. `props` indicates how to do the actual inflection (see below). The resulting inflected forms are stored into the `.forms` property of `multiword_spec`. This property holds a table whose keys are slots (i.e. ID's of individual inflected forms, such as "pres_1sg" for the first-person singular present indicative tense of a verb) and whose values are lists of the form `{ form = ``form``, translit = ``manual_translit_or_nil``, footnotes = ``footnote_list_or_nil``}`, where ``form`` is a string specifying the value of the form (e.g. "ouço" for the first-person singular present indicative of the Portuguese verb [[ouvir]]); ``manual_translit_or_nil`` is the corresponding manual transliteration if needed (i.e. if the form is in a non-Latin script and the automatic transliteration is incorrect or unavailable), otherwise nil; and ``footnote_list_or_nil`` is a list of footnotes to be attached to the form, or nil for no footnotes. Note that currently footnotes must be surrounded by brackets, e.g "[archaic]", and should not begin with a capital letter or end with a period. (Conversion from "[archaic]" to "Archaic." happens automatically.) This function has no return value, but modifies `multiword_spec` in-place, adding the `forms` table as described above. After calling this function, call show_forms() on the `forms` table to convert the forms and footnotes given in this table to strings suitable for display. `props` is an object specifying properties used during inflection, as follows: ```{ slot_list = {{"``slot``", "``accel``"}, {"``slot``", "``accel``"}, ...}, slot_table = {``slot`` = "``accel``", ``slot`` = "``accel``", ...}, skip_slot = nil `or` __function__(slot), lang = nil `or` __lang_object__, inflect_word_spec = __function__(base), get_variants = nil 'or` __function__(formval), combine_metadata = nil `or` __function__(data), include_user_specified_links = __boolean__, }``` `slot_list` is a list of two-element lists of slots and associated accelerator tags. ``slot`` is arbitrary but should correspond with slot names as generated by `inflect_word_spec`. ``accel`` is the corresponding accelerator tags; e.g. if ``slot`` is "pres_1sg", ``accel`` might be "1|s|pres|ind". ``accel`` is actually unused during inflection, but is used during `show_forms()`, which takes the same `slot_list` as a property upon input. `slot_table` is a table mapping slots to associated accelerator tags and serves the same function as `slot_list`. Only one of `slot_list` or `slot_table` must be given. For new code it is preferable to use `slot_list` because this allows you to control the order of processing slots, which may occasionally be important. `skip_slot` is a function of one argument, a slot name, and should return a boolean indicating whether to skip the given slot during inflection. It can be used, for example, to skip singular slots if the overall term being inflected is plural-only, and vice-versa. `lang` is a language object. This is only used to generate manual transliteration. If the language is written in the Latin script or manual transliteration cannot be specified in the input to parse_inflected_text(), this can be omitted. (Manual transliteration is allowed if the `lang` object is set in the `props` passed to parse_inflected_text().) `inflect_word_spec` is the function to do the actual inflection. It is passed a single argument, which is a ``word_spec`` object describing the word to be inflected and the user-provided inflection specifications. It is exactly the same as was returned by the `parse_indicator_spec` function provided in the `props` sent on input to `parse_inflected_text`, but has additional fields describing the word to be inflected and the surrounding text, as follows: ```{ lemma = "``lemma``", before_text = "``text-before-word``", before_text_no_links = "``text-before-word-no-links``", before_text_translit = "``manual-translit-of-text-before-word``" or nil (if no manual translit or respelling was specified in the before-text) -- Fields as described in parse_indicator_spec() ... }``` Here ``lemma`` is the word to be inflected as specified by the user (including any links if so given), and the `before_text*` fields describe the raw text preceding the word to be inflected. Any other fields in this object are as set by `parse_inflected_text`, and describe things like the gender, number, conjugation/declension, etc. as specified by the user in the <...> spec following the word to be inflected. `inflect_word_spec` should initialize the `.forms` property of the passed-in ``word_spec`` object to the inflected forms of the word in question. The value of this property is a table of the same format as the `.forms` property that is ultimately generated by inflect_multiword_or_alternant_multiword_spec() and described above near the top of this documentation: i.e. a table whose keys are slots and whose values are lists of the form `{ form = ``form``, translit = ``manual_translit_or_nil``, footnotes = ``footnote_list_or_nil``}`. `get_variants` is either {nil} or a function of one argument (a string, a form value). The purpose of this function is to ensure that in a multiword term where a given slot has more than one possible variant, the final output has only parallel variants in it. For example, feminine nouns and adjectives in Russian have two possible endings, one typically in -ой (-oj) and the other in -ою (-oju). If we have a feminine adjective-noun combination (or a hyphenated feminine noun-noun combination, or similar), and we don't specify `get_variants`, we'll end up with four values for the instrumental singular: one where both adjective and noun end in -ой, one where both end in -ою, and two where one of the words ends in -ой and the other in -ою. In general if we have N words each with K variants, we'll end up with an explosion of N^K possibilities. `get_variants` avoids this by returning a variant code (an arbitary string) for each variant. If two words each have a non-empty variant code, and the variant codes disagree, the combination will be rejected. If `get_variants` is not provided, or either variant code is an empty string, or the variant codes agree, the combination is allowed. The recommended way to use `get_variants` is as follows: 1. During inflection in `inflect_word_spec`, add a special character or string to each of the variants generated for a given slot when there is more than one. (As an optimization, do this only when there is more than one word being inflected.) Special Unicode characters can be used for this purpose, e.g. U+FFF0, U+FFF1, ..., U+FFFD, which have no meaning in Unicode. 2. Specify `get_variants` as a function that pulls out and returns the special character(s) or string included in the variant forms. 3. When calling show_forms(), specify a `canonicalize` function that removes the variant code character(s) or string from each form before converting to the display form. See [[Module:hi-verb]] and [[Module:hi-common]] for an example of doing this in a generalized fashion. (Look for add_variant_codes(), get_variants() and remove_variant_codes().) `combine_metadata` is a function that is invoked when combining two form objects along along with in-between text and storing into a destination form object. When this happens, if the the form value and translit in the first form object is empty and the in-between text is likewise empty (which regularly happens when appending the form object describing the first word in a multiword expression to empty base text), the second form object is simply shallow-copied along with all of its metadata, and any footnotes are combined appropriately (normally the first form object is such a case won't have footnotes). Otherwise, a new form object is constructed by combining the form values, translit and footnotes from the two objects and in-between text, and calling `combine_metadata` to combine any other metadata. Leave this unspecified if there is no additional metadata or if you don't want any metadata carried over. (Examples of metadata that should generally not be carried over are glosses of individual words, sense ID's and similar word-level properties that can't easily be combined to generate a multiword equivalent. Examples of metadata that should be carried over and combined are qualifiers, labels and certain boolean properties such as an uncertainty flag indicating that a given form is uncertain. For some metadata, it is more complex; for example, if both source words have the same gender or part of speech, the destination should keep that value, but if they differ, it may be safest to leave the field blank.) This function, if specified, is called with a single argument as follows: ```{ slot = "__string__", dest_form = __formobj__, form1 = __formobj__, form2 = __formobj__, between_text = "__string__", between_text_no_links = "__string__", between_text_translit = "__string__" `or` nil }``` Here, `slot` is the slot whose forms are being constructed. `dest_form` is the destination form object into which the combined metadata should be written, and is pre-populated with appropriate `form`, `translit` and `footnotes` fields. `form1` and `form2` are the two source forms being combined, and `between_text` is the text to be inserted between the two source forms. `between_text_no_links` is the same as `between_text` but with double-bracket links removed, and `between_text_translit` is the manual transliteration of `between_text_no_links`, if specified. The function should return nothing, but should side-effect `dest_form` as appropriate. `include_user_specified_links`, if given, ensures that user-specified links in the raw text surrounding a given word are preserved in the output. If omitted or set to false, such links will be removed and the whole multiword expression will be linked. ]==] function export.inflect_multiword_or_alternant_multiword_spec(multiword_spec, props) multiword_spec.forms = {} local is_alternant_multiword = not not multiword_spec.alternant_or_word_specs for _, word_spec in ipairs(is_alternant_multiword and multiword_spec.alternant_or_word_specs or multiword_spec.word_specs) do if word_spec.alternants then inflect_alternants(word_spec, props) else props.inflect_word_spec(word_spec) end iterate_slot_list_or_table(props, function(slot) if not props.skip_slot or not props.skip_slot(slot) then append_forms(props, multiword_spec.forms, slot, word_spec.forms[slot], (slot:find("linked") or props.include_user_specified_links) and word_spec.before_text or word_spec.before_text_no_links, word_spec.before_text_no_links, word_spec.before_text_translit ) end end) end if multiword_spec.post_text ~= "" then local pseudoform = {{form=""}} iterate_slot_list_or_table(props, function(slot) -- If slot is empty or should be skipped, don't try to append post-text. if (not props.skip_slot or not props.skip_slot(slot)) and multiword_spec.forms[slot] then append_forms(props, multiword_spec.forms, slot, pseudoform, (slot:find("linked") or props.include_user_specified_links) and multiword_spec.post_text or multiword_spec.post_text_no_links, multiword_spec.post_text_no_links, multiword_spec.post_text_translit ) end end) end end function export.map_word_specs(alternant_multiword_spec, fun) for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do if alternant_or_word_spec.alternants then for _, multiword_spec in ipairs(alternant_or_word_spec.alternants) do for _, word_spec in ipairs(multiword_spec.word_specs) do fun(word_spec) end end else fun(alternant_or_word_spec) end end end function export.create_footnote_obj() return { notes = {}, seen_notes = {}, noteindex = 1, seen_refs = {}, } end function export.get_footnote_text(footnotes, footnote_obj) if not footnotes then return "" end local link_indices = {} local all_refs = {} for _, footnote in ipairs(footnotes) do local refs footnote, refs = export.expand_footnote_or_references(footnote) if footnote then local this_noteindex = footnote_obj.seen_notes[footnote] if not this_noteindex then -- Generate a footnote index. this_noteindex = footnote_obj.noteindex footnote_obj.noteindex = footnote_obj.noteindex + 1 insert(footnote_obj.notes, '<sup style="color: var(--wikt-palette-red, red)">' .. this_noteindex .. '</sup>' .. footnote) footnote_obj.seen_notes[footnote] = this_noteindex end insert_if_not(link_indices, this_noteindex) end if refs then for _, ref in ipairs(refs) do if not ref.name then local this_refhash = footnote_obj.seen_refs[ref.text] if not this_refhash then -- Different text needs to have different auto-generated names, globally across the entire page, -- including across different invocations of {{it-verb}} or {{it-conj}}. The easiest way to accomplish -- this is to use a message-digest hashing function. It does not have to be cryptographically secure -- (MD5 is insecure); it just needs to have low probability of collisions. this_refhash = mw.hash.hashValue("md5", ref.text) footnote_obj.seen_refs[ref.text] = this_refhash end ref.autoname = this_refhash end -- I considered using "n" as the default group rather than nothing, to more clearly distinguish regular -- footnotes from references, but this requires referencing group "n" as <references group="n"> below, -- which is non-obvious. insert_if_not(all_refs, ref) end end end table.sort(link_indices) local function sort_refs(r1, r2) -- FIXME, we are now sorting on an arbitrary hash. Should we keep track of the order we -- saw the autonamed references and sort on that? if r1.autoname and r2.name then return true elseif r1.name and r2.autoname then return false elseif r1.name and r2.name then return r1.name < r2.name else return r1.autoname < r2.autoname end end table.sort(all_refs, sort_refs) for i, ref in ipairs(all_refs) do local refargs = {name = ref.name or ref.autoname, group = ref.group} all_refs[i] = mw.getCurrentFrame():extensionTag("ref", ref.text, refargs) end local link_text if #link_indices > 0 then link_text = '<sup style="color: var(--wikt-palette-red, red)">' .. concat(link_indices, ",") .. '</sup>' else link_text = "" end local ref_text = concat(all_refs) if link_text ~= "" and ref_text ~= "" then return link_text .. "<sup>,</sup>" .. ref_text else return link_text .. ref_text end end --[==[ Add links around words in a term. If multiword_only, do it only in multiword terms. ]==] function export.add_links(form, multiword_only) if form == "" or form == " " then return form end if not form:find("%[%[") then if ufind(form, "[%s%p]") then --optimization to avoid loading [[Module:headword]] on single-word forms local m_headword = require("Module:headword") if m_headword.head_is_multiword(form) then form = m_headword.add_multiword_links(form) end end if not multiword_only and not form:find("%[%[") then form = "[[" .. form .. "]]" end end return form end --[==[ Remove redundant link surrounding entire term. ]==] function export.remove_redundant_links(term) return (term:gsub("^%[%[([^%[%]|]*)%]%]$", "%1")) end --[==[ Add links to all before and after text; for use in inflection modules that preserve links in multiword lemmas and include links in non-lemma forms rather than allowing the entire form to be a link. If `remember_original`, remember the original user-specified before/after text so we can reconstruct the original spec later. `add_links` is a function of one argument to add links to a given piece of text; if unspecified, it defaults to `export.add_links`. ]==] function export.add_links_to_before_and_after_text(alternant_multiword_spec, remember_original, add_links) add_links = add_links or export.add_links local function add_links_remember_original(object, field) if remember_original then object["user_specified_" .. field] = object[field] end object[field] = add_links(object[field]) end for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do add_links_remember_original(alternant_or_word_spec, "before_text") if alternant_or_word_spec.alternants then for _, multiword_spec in ipairs(alternant_or_word_spec.alternants) do for _, word_spec in ipairs(multiword_spec.word_specs) do add_links_remember_original(word_spec, "before_text") end add_links_remember_original(multiword_spec, "post_text") end end end add_links_remember_original(alternant_multiword_spec, "post_text") end --[==[ Reconstruct the original overall spec from the output of parse_inflected_text(), so we can use it in the language-specific acceleration module in the implementation of {{tl|pt-verb form of}} and the like. `props` is an optional table of properties. Currently only `preprocess_angle_bracket_spec` is recognized, and is an optional function of one argument that is called to process an angle-bracket spec before inserting into the reconstructed spec. ]==] function export.reconstruct_original_spec(alternant_multiword_spec, props) local parts = {} props = props or {} local function ins(txt) insert(parts, txt) end local function insert_angle_bracket_spec(spec) if props.preprocess_angle_bracket_spec then spec = props.preprocess_angle_bracket_spec(spec) end ins(spec) end for _, alternant_or_word_spec in ipairs(alternant_multiword_spec.alternant_or_word_specs) do ins(alternant_or_word_spec.user_specified_before_text) if alternant_or_word_spec.alternants then ins("((") for i, multiword_spec in ipairs(alternant_or_word_spec.alternants) do if i > 1 then ins(",") end for _, word_spec in ipairs(multiword_spec.word_specs) do ins(word_spec.user_specified_before_text) ins(word_spec.user_specified_lemma) insert_angle_bracket_spec(word_spec.angle_bracket_spec) end ins(multiword_spec.user_specified_post_text) end ins("))") else ins(alternant_or_word_spec.user_specified_lemma) insert_angle_bracket_spec(alternant_or_word_spec.angle_bracket_spec) end end ins(alternant_multiword_spec.user_specified_post_text) local retval = concat(parts) if alternant_multiword_spec.allow_default_indicator then -- As a special case, if we see e.g. "amar<>", remove the <>. Don't do this if there are spaces or alternants. if not retval:find(" ") and not retval:find("%(%(") then local retval_no_angle_brackets = retval:match("^(.*)<>$") if retval_no_angle_brackets then return retval_no_angle_brackets end end end return retval end --[==[ Convert the forms in ``formtable`` (a form table, whose keys are slots and whose values are lists of form objects, each of which is a table of the form `form = ``form``, translit = ``manual_translit_or_nil``, footnotes = ``footnote_list_or_nil``, no_accel = ``true_to_suppress_accelerators``, ... `) into strings. The form table is side-effected. Each form list turns into a string consisting of a comma-separated list of linked forms, with accelerators (unless `no_accel` is set in a given form object). If `include_translit` is specified, each string consists of a comma-separated list of form values (each formatted as a link), an HTML `&lt;br/>`, and a comma-separated list of transliterations. `props` is a table used in generating the strings, as follows: ```{ lang = __lang_object__, lemmas = {"``lemma``", "``lemma``", ...}, slot_list = {{"``slot``", "``accel``"}, {"``slot``", "``accel``"}, ...}, slot_table = {``slot`` = "``accel``", ``slot`` = "``accel``", ...}, include_translit = __boolean__, create_footnote_obj = nil `or` __function__(), canonicalize = nil or __function__(formval), preprocess_forms = nil `or` __function__(data), no_deduplicate_forms = __boolean__, combine_metadata_during_dedup = nil `or` __function__(data), transform_accel_obj = nil `or` __function__(slot, form, accel_obj), format_forms = nil `or` __function__(data), generate_link = nil `or` __function__(data), format_tr = nil `or` __function__(data), join_spans = nil `or` __function__(data), footnotes = nil or {"``extra_footnote``", "``extra_footnote``", ...}, }``` `lemmas` is the list of lemmas, used in the accelerators. `slot_list` is a list of two-element lists of slots and associated accelerator tag sets. ``slot`` should correspond to slots generated during `inflect_multiword_or_alternant_multiword_spec()`. ``accel`` is the corresponding accelerator tag set; e.g. if ``slot`` is "pres_1sg", ``accel`` might be "1|s|pres|ind". ``accel`` is used in generating entries for accelerator support (see [[WT:ACCEL]]). `slot_table` is a table mapping slots to associated accelerator tag sets and serves the same function as `slot_list`. Only one of `slot_list` or `slot_table` must be given. For new code it is preferable to use `slot_list` because this allows you to control the order of processing slots, which may occasionally be important. `include_translit`, if given, causes transliteration to be included in the generated strings. The function works as follows: # Create an object to hold footnotes (customizable using `create_footnote_obj`). # Generate the comma-separated lemma form values and store in `.lemma` in the form table. # Loop over the slots specified using `slot_list` or `slot_table`. For each slot: ## Canonicalize the form values (customizable using `canonicalize`; by default does nothing). ## Preprocess the forms (customizable using `preprocess_forms`; by default does nothing). ## Unless `no_deduplicate_forms` is set, deduplicate forms in a slot sharing the same form value but possibly different transliteration. (This happens e.g. in Russian, where it is relatively common for a given form to have two possible transliterations, one reflecting a more nativized pronunciation where Cyrillic е triggers palatalization of the preceding consonant, and one reflecting a more "foreign" pronunciation where this palatalization does not happen. In such a case, the automatic transliteration would normally suffice for the more nativized pronunciation but the more "foreign" pronunciation will need manual transliteration.) As part of deduplication, footnotes will be combined using `combine_footnotes`; distinct manual transliterations will be combined into a list (meaning the `translit` field of form objects in some subsequent `props` functions may hold a list; this will be noted when possible); and any remaining metadata will be combined using the `combine_metadata_during_dedup` method, if provided. ## Add acceleration to all forms. The acceleration tag set associated with a given form comes from `slot_list` or `slot_table`, i.e. all forms in a given slot have the same tag set. However, different forms will have different associated transliterations stored into the accelerator object associated with the form, as well as possibly different lemmas. In particular, when there are multiple lemma forms, this is often due to alternative ways to pronounce the lemma (e.g. alternative stress positions or vowel lengths), and there are often associated non-lemma forms that match each lemma. An example given in the introduction is {{m+|uk|русин||Rusyn}}, stressed in the lemma as ''ру́син'' or ''руси́н'' with associated genitive singulars ''ру́сина'' and ''русина́''. We would like the auto-generated accelerator entry for {{m|uk|русина}} to show the variant ''ру́сина'' as having lemma ''ру́син'' and the variant ''русина́'' as having the lemma ''руси́н'', rather than showing both variants as having both lemmas, which is less accurate. As a result, the code that generates acceleration objects for forms matches up forms and lemmas one-to-one if possible. If this is not possible, the matching is usually one lemma to many forms, as in {{m+|uk|міст||bridge}} with genitive singular ''мо́сту'' or ''моста́'' (in which case all forms get the same lemma), or many lemmas to one form, as in {{m+|uk|черга||turn, queue}} stressed either ''че́рга'' or ''черга́'' with nominative singular only ''че́рги'' (in which case the single form gets assicated all lemmas). If there are multiple lemmas and multiple forms, the algorithm attempts to align them as evenly as possible (e.g. two lemma variants to four forms means the first two forms get assigned the first lemma variant and the last two forms get assigned the second lemma variant); this is often going to be incorrect, but (a) there's unlikely to be a single algorithm that works in all such circumstances, and (b) these cases are very rare. Finally, note the following: ##* No acceleration is assigned to a form if any of the following apply: (a) there are no lemmas given in `props.lemmas`; (b) the `no_accel` key in the form object has a non-falsy value; (c) the form value of the form is {"?"} or an em-dash ({"—"}); (d) the accelerator tag set is given as a hyphen {"-"}); or (e) the form value contains an internal link. ##* The accelerator code sets the `formval_for_link` key in each form object to the version of the form value that should be passed to `full_link()` in [[Module:links]]. This is usually the same as the passed-in form value, but differs when the entire form value is surrounded with a redundant internal link (which is removed). ##* The resulting accelerator object can be modified (or replaced entirely) by the `transform_accel_obj` function. This is used, for example, in [[Module:es-verb]], [[Module:pt-verb]] and other Romance-language verb conjugation modules (likewise [[Module:ar-verb]]) to replace the tag set with the original verb spec used to generate the verb, so that the accelerator code can generate the appropriate call to {{tl|es-verb form of}}, {{tl|pt-verb form of}} or the like, which computes the inflections, instead of directly listing the inflections. ## Format the forms into strings. The entire default process can be replaced using `format_forms`; otherwise the default algorithm works as follows: ### Generate the '''form value spans''', with one entry (a linked HTML-ized version of the form value) per form. This can be customized using `generate_link`. (Various modules do this. For example, the Arabic verb module includes qualifiers, labels, ID's and the like that can be specified by the user; the Portuguese and reintegrated Galician verb modules italicize certain superseded or otherwise less-desirable forms instead of linking them normally; the German verb module adds {{m|de|dass}} to subjunctive forms and optional pronouns to imperative forms; and the German adjective module adds articles to adjective forms normally accompanied by articles and the equivalent of "he/she is" etc. to predicate forms.) The default uses `full_link()` in [[Module:links]] (with transliteration generation disabled) concatenated with the appropriate footnote symbol(s) (if any). ### Generate the '''transliteration spans''', with one entry per distinct translit, auto-generated if manual translit isn't available. Note that, due to the earlier form value deduplication step, there may be multiple translits per form object. These translits are themselves deduplicated to get the list of spans. (Such duplication can happen, for example, in Arabic with terms containing a glottal stop in them; there may be multiple ways of spelling the glottal stop or ''hamza'' in Arabic, but only one way of transliterating it.) Each span consists of an object specifying the translit and the list of footnotes. These objects are then converted to formatted strings, either using `format_tr` if supplied or else calling `tag_translit()` in [[Module:script utilities]] and concatenating the appropriate footnote symbol(s) (if any). ### Combine the form value and transliteration spans. If `join_spans` is supplied, use it; otherwise, concatenate the form value spans (comma-separated) and (if available) transliteration spans (comma-separated), and (if appropriate) combine them using {<br />}. `create_footnote_obj` is an optional function of no arguments to create the footnote object used to track footnotes; see `create_footnote_obj()`. Customizing it is useful to prepopulate the footnote table using `get_footnote_text()`. `canonicalize` is an optional function of one argument (a form value) to canonicalize each form before processing; it can return nil for no change. The most common purpose of this function is to remove variant codes from the form value. See the documentation for `inflect_multiword_or_alternant_multiword_spec()` for a description of variant codes and their purpose. `preprocess_forms` is an optional function of one argument (a table of properties) to preprocess the form objects as a whole. It runs after `canonicalize` (meaning that the form values passed in are canonicalized) and before deduplication and the addition of acceleration info. The property table passed in has the following properties: * `slot`: The slot being processed. * `forms`: The list of form objects for this slot. * `accel_tag_set`: The accelerator tag set for this slot, taken from `slot_list` or `slot_table`. * `footnote_obj`: The footnote object returned by the `create_footnote_obj` property or the default `create_footnote_obj()` function. `preprocess_forms` should return a list of preprocessed form objects, or {nil} to use the passed-in `forms`. If this function does deduplication, you should set `no_deduplicate_forms` to disable the default deduplication process. `no_deduplicate_forms`, if set, disables the deduplication step (see above). `combine_metadata_during_dedup` is an optional function of one argument (a table of properties) to combine the metadata of deduplicated form objects. The property table passed in has the following properties: * `slot`: The slot being processed. * `existing_form`: The existing form object into which a duplicated form is being combined. * `dup_form`: The duplicated form being combined into `existing_form`. * `existing_form_pos`: The one-based position of the existing form in the deduplicated form list (not necessarily its original position). * `dup_form_pos`: The one-based position of the duplicated form in its original list. The following should be noted about the form objects passed in: # The form values in `.form` have been canonicalized using `.canonicalize`, if provided. # The form values in `existing_form` and `dup_form` are always the same. # The footnotes in `existing_form` have already been combined with those in `dup_form`. # If there was manual translit either in `existing_form` (prior to deduplication) or in `dup_form`, there will be manual translit in `existing_form.translit` that is a list and combines any previous accumulated translits in `existing_form` as well as the translit in `dup_form` (even if one of them was specified as {nil} indicating an automatic translit). This means that the translit in `existing_form.translit` is always either {nil} or a list of strings (and the same applies to `dup_form.translit`). `transform_accel_obj` is an optional function of three arguments (``slot``, ``formobj``, ``accel_obj``) to transform the default constructed accelerator object in ``accel_obj`` into an object that should be passed to `full_link()` in [[Module:links]]. It should return the new accelerator object, or {nil} for no acceleration. (If {nil} is returned, the corresponding form has no acceleration; this is unlike most customization functions, where returning {nil} causes the default algorithm to be invoked.) The function can destructively modify the accelerator object passed in. '''NOTE''': This is called even when the passed-in ``accel_obj`` is {nil} (see the (a) through (e) reasons above why no acceleration may be assigned to a form). Thus, your code needs to do something sensible in this case. The description above of how `show_forms()` works inclues various examples of modules that supply a `transform_accel_obj` function and the reasons for doing so. `format_forms`, if supplied, is a function that entirely replaces the formatting portion of `show_forms()`. An example of why you might want to do this is to get a different layout than the default, e.g. one where translit is displayed next to each form value instead of the form values and translits grouped and displayed on separate lines. Under normal circumstances, you should not do this, but instead customize the functions that replace specific parts of the default formatting algorithm (see below). This function is passed one argument (a table of properties) and should return a string (the formatted forms, ready to store into the slot in the form table) or {nil} to proceed with the default algorithm (see above). The property table passed in has the following properties: * `slot`: The slot being processed. * `forms`: The list of form objects, deduplicated and with accelerator info added. * `footnote_obj`: The footnote object returned by the `create_footnote_obj` property or the default `create_footnote_obj()` function. The following should be noted about the form objects in `forms`: # There are extra fields `formval_for_link` and `accel_obj`. The first one is as described above under the paragraph beginning "Add acceleration to all forms" under "The function works as follows". The second one is the accelerator object in the format expected by [[Module:links]]. # The `translit` field, if non-{nil}, is a list of transliterations rather than a single transliteration; this is due to the form value deduplication step. `generate_link` is an optional function to generate the link text for a given form value. It is passed a single argument (a table of properties) and should return a string, the formatted link. If it returns {nil}, the default algorithm (see above) is invoked. The property table passed in has the following properties: * `slot`: The slot being processed. * `form`: The form to be converted to a formatted link. As with the `format_forms` function described above, the form objects passed in contain extra fields `formval_for_link` and `accel_obj` (both of which will normally be used), and the `translit` field, if non-{nil}, is a list. * `pos`: The one-based position of the form being processed, in the list of form value spans. Rarely used. * `footnote_obj`: The footnote object returned by the `create_footnote_obj` property or the default `create_footnote_obj()` function. Normally used in order to get the footnote symbol associated with any footnotes in `footnotes`. The description above of how `show_forms()` works inclues various examples of modules that supply a `generate_link` function and the reasons for doing so. `format_tr` is an optional function to generate the formatted text for a given transliteration. It is passed a single argument (a table of properties) and should return a string, the formatted transliteration text. If it returns {nil}, the default algorithm (see above) is invoked. The property table passed in has the following properties: * `slot`: The slot being processed. * `tr_for_tag`: The transliteration to process. * `pos`: The one-based position of the transliteration being processed, in the list of transliteration spans. Rarely used. * `footnotes`: The list of footnotes associated with all form objects with this transliteration. (If there were multiple form objects with the same transliteration, the list of footnotes will have been generated using `combine_footnotes()`.) * `footnote_obj`: The footnote object returned by the `create_footnote_obj` property or the default `create_footnote_obj()` function. Normally used in order to get the footnote symbol associated with any footnotes in `footnotes`. `join_spans` is an optional function to join the processed form value and transliteration spans into a formatted string. It is passed a single argument (a table of properties) and should return the final string to store into the form table slot. If it returns {nil}, the default algorithm (see above) is invoked. The property table passed in has the following properties: * `slot`: The slot being processed. * `formval_spans`: A list of strings, the formatted form value spans. * `tr_spans`: A list of strings, the formatted transliteration spans. If there is no transliteration, this will be an empty list. A custom `join_spans` is provided by [[Module:de-verb]], which concatenates the form value spans vertically (using {"<br />"}) instead of horizontally using a comma, as is normal; this is because there is no translit and the form values are often long, containing extra words attached during `generate_link()`. The only exception is the `aux` slot holding the auxiliaries, which is concatenated horizontally using {" or "}. [[Module:de-adjective]] similarly provides a custom `join_spans` function that concatenates the form value spans vertically. ]==] function export.show_forms(formtable, props) local footnote_obj = props.create_footnote_obj and props.create_footnote_obj() or export.create_footnote_obj() local function fetch_formval_and_translit(entry, remove_links) local formval, translit if type(entry) == "table" then formval, translit = entry.form, entry.translit else formval = entry end if remove_links then formval = m_links.remove_links(formval) end return formval, translit end local lemma_formvals = {} for _, lemma in ipairs(props.lemmas) do local lemma_formval, _ = fetch_formval_and_translit(lemma) insert_if_not(lemma_formvals, lemma_formval) end formtable.lemma = #lemma_formvals > 0 and concat(lemma_formvals, ", ") or mw.loadData(headword_data_module).pagename -- For safety, since we in-place modify `lemmas` usually before processing a given slot, make a copy. local props_lemmas = shallow_copy(props.lemmas) for i, lemma in ipairs(props_lemmas) do props_lemmas[i] = shallow_copy(lemma) end local function do_slot(slot, accel_tag_set) local formobjs = formtable[slot] if formobjs then if type(formobjs) ~= "table" then error("Internal error: For slot '" .. slot .. "', expected table but saw " .. dump(formobjs)) end -- Maybe canonicalize the form values (e.g. remove variant codes and monosyllabic accents). if props.canonicalize then for _, form in ipairs(formobjs) do form.form = props.canonicalize(form.form) or form.form end end -- Preprocess the forms as a whole if called for. if props.preprocess_forms then formobjs = props.preprocess_forms { slot = slot, forms = formobjs, accel_tag_set = accel_tag_set, footnote_obj = footnote_obj, } or formobjs end -- Maybe deduplicate form values (happens e.g. in Russian with two terms with the same Russian form but -- different translits). if not props.no_deduplicate_forms then local deduped_formobjs = {} for i, form in ipairs(formobjs) do local function combine_forms(existing_form, dup_form, pos) assert(existing_form.form == dup_form.form) -- Combine footnotes. existing_form.footnotes = export.combine_footnotes(existing_form.footnotes, dup_form.footnotes) -- If translit is being generated, and there's manual translit associated with either form, we -- need to generate any missing translits and combine them, taking into account the fact that a -- translit value may actually be a list of translits (particularly with the existing form if we -- already combined an item with manual translit into it). if props.include_translit and form_value_transliterable(existing_form.form) and ( existing_form.translit or dup_form.translit) then local combined_translit if not existing_form.translit then combined_translit = { props_transliterate(props, m_links.remove_links(existing_form.form)) } elseif type(existing_form.translit) == "string" then combined_translit = {existing_form.translit} else combined_translit = existing_form.translit end local dup_form_translit = dup_form.translit if not dup_form_translit then -- dup_form.form is the same as existing_form.form (see assert above), but this is -- defensive programming in case that changes dup_form_translit = {props_transliterate(props, m_links.remove_links(dup_form.form))} elseif type(dup_form_translit) == "string" then dup_form_translit = {dup_form_translit} end for _, translit in ipairs(dup_form_translit) do insert_if_not(combined_translit, translit) end existing_form.translit = combined_translit end if props.combine_metadata_during_dedup then props.combine_metadata_during_dedup { slot = slot, existing_form = existing_form, existing_form_pos = pos, dup_form = dup_form, dup_form_pos = i, } end end insert_if_not(deduped_formobjs, form, { key = function(formobj) return formobj.form end, combine = combine_forms, }) end formobjs = deduped_formobjs end -- Add acceleration info to form objects. for i, form in ipairs(formobjs) do local formval = form.form if not form_value_transliterable(formval) then form.formval_for_link = formval else local formval_for_link formval_for_link = formval -- remove redundant link surrounding entire form formval_for_link = export.remove_redundant_links(formval_for_link) form.formval_for_link = formval_for_link -------------------- Compute the accelerator object. ----------------- local accel_obj -- Check if form still has links; if so, don't add accelerators because the resulting entries will -- be wrong. if props_lemmas[1] and not form.no_accel and accel_tag_set ~= "-" and not formval_for_link:find("%[%[") then -- If there is more than one form or more than one lemma, things get tricky. Often, there are -- the same number of forms as lemmas, e.g. for Ukrainian [[зимовий]] "wintry; winter (rel.)", -- which can be stressed зимо́вий or зимови́й with corresponding masculine/neuter genitive -- singulars зимо́вого or зимово́го etc. In this case, usually the forms and lemmas match up so -- we do this. If there are different numbers of forms than lemmas, it's usually one lemma -- against several forms e.g. Ukrainian [[міст]] "bridge" with genitive singular мо́сту or моста́ -- (accent patterns b or c) or [[ложка|ло́жка]] "spoon" with nominative plural ло́жки or ложки́ -- (accent patterns a or c). Here, we should assign the same lemma to both forms. The opposite -- can happen, e.g. [[черга]] "turn, queue" stressed че́рга or черга́ with nominative plural only -- че́рги (accent patterns a or d). Here we should assign both lemmas to the same form. In more -- complicated cases, with more than one lemma and form and different numbers of each, we try -- to align them as much as possible, e.g. if there are somehow eight forms and three lemmas, -- we assign lemma 1 to forms 1-3, lemma 2 to forms 4-6 and lemma 3 to forms 7 and 8, and -- conversely if there are somehow three forms and eight lemmas. This is likely to be wrong, but -- (a) there's unlikely to be a single algorithm that works in all such circumstances, and (b) -- these cases are vanishingly rare or nonexistent. Properly we should try to remember which -- form was generated by which lemma, but that is significant extra work for little gain. local first_lemma, last_lemma if #formobjs >= #props_lemmas then -- More forms than lemmas. Try to even out the forms assigned per lemma. local forms_per_lemma = math.ceil(#formobjs / #props_lemmas) first_lemma = math.floor((i - 1) / forms_per_lemma) + 1 last_lemma = first_lemma else -- More lemmas than forms. Try to even out the lemmas assigned per form. local lemmas_per_form = math.ceil(#props_lemmas / #formobjs) first_lemma = (i - 1) * lemmas_per_form + 1 last_lemma = math.min(first_lemma + lemmas_per_form - 1, #props_lemmas) end local accel_lemma, accel_lemma_translit if first_lemma == last_lemma then accel_lemma, accel_lemma_translit = fetch_formval_and_translit(props_lemmas[first_lemma], "remove links") else accel_lemma = {} accel_lemma_translit = {} for j=first_lemma, last_lemma do local this_accel_lemma, this_accel_lemma_translit = fetch_formval_and_translit(props_lemmas[j], "remove links") -- Do not use insert() especially for the translit because it may be nil and in -- that case we want gaps in the array. accel_lemma[j - first_lemma + 1] = this_accel_lemma accel_lemma_translit[j - first_lemma + 1] = this_accel_lemma_translit end end local accel_translit if props.include_translit and form.translit then if type(form.translit) == "table" then accel_translit = concat(form.translit, ", ") elseif type(form.translit) == "string" then accel_translit = form.translit else error(("Internal error: For slot '%s', form translit is not a table or string: %s"): format(slot, dump(form.translit))) end end accel_obj = { form = accel_tag_set, translit = accel_translit, lemma = accel_lemma, lemma_translit = props.include_translit and accel_lemma_translit or nil, } end -- Postprocess if requested. if props.transform_accel_obj then accel_obj = props.transform_accel_obj(slot, form, accel_obj) end form.accel_obj = accel_obj end end -- Format the form objects into a string for insertion into the table. local formatted_forms if props.format_forms then formatted_forms = props.format_forms { slot = slot, forms = formobjs, footnote_obj = footnote_obj, } end if not formatted_forms then -- Default algorithm: Separate form values and translits and concatenate on separate lines. -- Form values have already been deduplicated but we may need to deduplicate translits (this happens -- e.g. in Arabic where there may be multiple ways of spelling a hamza in the Arabic script but only -- one way in transliteration). local formval_spans = {} local tr_spans = {} for i, form in ipairs(formobjs) do local link if props.generate_link then link = props.generate_link { slot = slot, pos = i, form = form, footnote_obj = footnote_obj, } end if not link then link = m_links.full_link { lang = props.lang, term = form.formval_for_link, tr = "-", accel = form.accel_obj } .. export.get_footnote_text(form.footnotes, footnote_obj) end formval_spans[i] = link if props.include_translit then local translits = form.translit or props_transliterate(props, m_links.remove_links(form.form)) if type(translits) == "string" then translits = {translits} end for _, tr in ipairs(translits) do local tr_for_tag = tr -- formerly may have differed due to old-style footnote symbols insert_if_not(tr_spans, { tr_for_tag = tr_for_tag, footnotes = form.footnotes, }, { key = function(trobj) return trobj.tr_for_tag end, combine = function(trobj, newtrobj) -- Combine footnotes. trobj.footnotes = export.combine_footnotes(trobj.footnotes, newtrobj.footnotes) end, }) end end end for i, tr_span in ipairs(tr_spans) do local formatted_tr if props.format_tr then formatted_tr = props.format_tr { slot = slot, pos = i, tr_for_tag = tr_span.tr_for_tag, footnotes = tr_span.footnotes, footnote_obj = footnote_obj, } end if not formatted_tr then formatted_tr = require(script_utilities_module).tag_translit(tr_span.tr_for_tag, props.lang, "default", " style=\"color: var(--wikt-palette-grey-8,#888);\"") .. export.get_footnote_text(tr_span.footnotes, footnote_obj) end tr_spans[i] = formatted_tr end if props.join_spans then formatted_forms = props.join_spans { slot = slot, formval_spans = formval_spans, tr_spans = tr_spans, } end if not formatted_forms then local formval_span = concat(formval_spans, ", ") local tr_span if #tr_spans > 0 then tr_span = concat(tr_spans, ", ") end if tr_span then formatted_forms = formval_span .. "<br />" .. tr_span else formatted_forms = formval_span end end end formtable[slot] = formatted_forms else formtable[slot] = "—" end end iterate_slot_list_or_table(props, do_slot) formtable.footnote = concat(footnote_obj.notes, "<br />") end --[==[ Find the inflection spec for a given lemma by scraping the contents of the section for the lemma's language, looking for the appropriate inflection template (which may in some cases be a headword template with inflections specified in the headword, as for {{tl|en-verb}}). An inflection ID can be given; in that case, the value of the appropriate ID param (usually `id=`) must match the specified ID; otherwise, any inflection template call of the right name will work. If anything goes wrong in the process, a string is returned describing the error message; otherwise an object is returned, where `template` is the template found (an object of the form returned by `find_templates` in [[Module:template parser]]). There is one param, an object with the following fields: * `lemma`: The page to scrape for inflection calls. * `langname`: The name of the language of the lemma (e.g. {"English"} or {"Icelandic"}). Only that section is scraped. * `infltemp`: A string specifying the name of the template containing the inflections, or a function of one argument, the template object (an object of the form returned by `find_templates` in [[Module:template parser]]), which should return true if the given object specifies an appropriate inflection template. * `idparam`: The name of the parameter holding the inflection ID used to disambiguate lemmas with multiple inflections. * `inflid`: The inflection ID that must match the value of the `idparam` parameter specified by the inflection template. If not specified, there must be exactly one matching inflection template present, and it must not have a value given for the `idparam` parameter. Otherwise, all matching inflection templates must have a value specified for the `idparam` parameter of the template, and there must be exactly one template whose `idparam` parameter value is the same as the value of `inflid`. Find the inflection spec by scraping the contents of the Icelandic section of `lemma`, looking for `infltemp` calls (where template is e.g. "is-ndecl", "is-adecl" or "is-conj") If `inflid` is given, it must match the value of the |id= param specified to the inflection template; otherwise, any inflection template call will work. If anything goes wrong in the process, a string is returned describing the error message; otherwise a table of inflections is returned, each containing a field `infl` with the inflection spec. The inflection spec comes from the |deriv=, |deriv2=, etc. params in the inflection template if specified and `is_deriv` is given; otherwise from 1=, 2=, etc. If `allow_empty_infl` is given, a missing inflection spec in 1= is allowed and converted to an empty string; otherwise, an error is signaled. ]==] function export.scrape_inflection(data) local langname, lemma, infltemp, inflid, idparam = data.langname, data.lemma, data.infltemp, data.inflid, data.idparam local lang_lemma = ("%s base lemma '[[%s]]'"):format(langname, lemma) local title = new_title(lemma) if title then local content = title:getContent() if content then local langtext = require(pages_module).get_section(content, langname) if langtext then local inflobjs_by_id = {} local inflobjs_without_id = {} local ordered_seen_ids = {} for template in require(template_parser_module).find_templates(langtext) do local matches if type(infltemp) == "string" then matches = template:get_name() == infltemp else matches = infltemp(template) end if matches then local inflobj = {template = template} local args = template:get_arguments() local tid = args[idparam] if tid then if inflobjs_by_id[tid] then return ("For %s base lemma '[[%s]]', saw id='%s' twice"):format( langname, lemma, tid ) end inflobjs_by_id[tid] = inflobj insert_if_not(ordered_seen_ids, tid) else insert(inflobjs_without_id, inflobj) end end end local function concat_ordered_seen_ids() local quoted_seen_ids = {} for _, seen_id in ipairs(ordered_seen_ids) do insert(quoted_seen_ids, "'" .. seen_id .. "'") end return m_table.serialCommaJoin(quoted_seen_ids, {dontTag = true}) end local formatted_template = type(infltemp) == "string" and ("[[Template:%s]]"):format(infltemp) or "inflection template" local formatted_template_sing_article = type(infltemp) == "string" and ("[[Template:%s]]"):format(infltemp) or "an inflection template" local formatted_template_plural = type(infltemp) == "string" and "'s" or "s" if ordered_seen_ids[1] and inflobjs_without_id[1] then return ("For %s, saw %s %s%s with id=%s as well as %s %s%s without ID; this is not allowed; " .. "with multiple %s calls, all must have id= params"):format(lang_lemma, #ordered_seen_ids, formatted_template, ordered_seen_ids[2] and formatted_template_plural or "", concat_ordered_seen_ids(), #inflobjs_without_id, formatted_template, inflobjs_without_id[2] and formatted_template_plural or "", formatted_template) elseif not ordered_seen_ids[1] and not inflobjs_without_id[1] then return ("For %s, found %s section but couldn't find any calls to %s"):format( lang_lemma, langname, formatted_template_sing_article) elseif #inflobjs_without_id > 1 then return ("For %s, found %s %s%s without ID's; this is not allowed; with multiple %s calls, " .. "all must have id= params"):format(lang_lemma, #inflobjs_without_id, formatted_template, formatted_template_plural, formatted_template) elseif inflid then if inflobjs_by_id[inflid] then return inflobjs_by_id[inflid] elseif ordered_seen_ids[1] then return ("For %s, found %s section but couldn't find any inflections matching ID '%s'; " .. "instead found ID's %s; you may have misspelled the ID"):format(lang_lemma, langname, inflid, concat_ordered_seen_ids()) else return ("For %s, found %s section with a single %s without ID, but ID requirement '%s' " .. "specified; consider removing the ID requirement"):format(lang_lemma, langname, formatted_template, inflid) end elseif inflobjs_without_id[1] then -- only one infl template, and it doesn't have an id=; return it return inflobjs_without_id[1] elseif ordered_seen_ids[2] then return ("For %s, saw %s %s%s with id=%s but with a request to return the inflection without " .. "ID; consider adding an ID restriction to the scrape request, e.g. '@@:%s' for " .. "self-scraping or '@%s:%s' for scraping from another page"):format(lang_lemma, #ordered_seen_ids, formatted_template, formatted_template_plural, concat_ordered_seen_ids(), ordered_seen_ids[1], usub(lemma, 1, 1), ordered_seen_ids[1]) else return ("For %s, found %s section with a single %s with id=%s, but the scrape request didn't " .. "specify an ID; consider removing the id= param from the %s call unless you expect to add " .. "more such calls to the page in the future (in which case add the ID to the scrape request, " .. "e.g. '@@:%s' for self-scraping or '@%s:%s' for scraping from another page"):format(lang_lemma, langname, formatted_template, concat_ordered_seen_ids(), formatted_template, ordered_seen_ids[1], usub(lemma, 1, 1), ordered_seen_ids[1]) end else return ("For %s, page exists but has no %s section"):format(lang_lemma, langname) end else return ("For %s, couldn't fetch contents for page; page may not exist"):format(lang_lemma) end else return ("Bad %s; couldn't create title object"):format(lang_lemma) end end return export nv8di3n6mnnnmvmo7fjstb2viyp4j22 0 76366 293620 293380 2026-09-20T13:30:56Z Saimawnkham 9 293620 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== [[File:Northern Goshawk ad M2.jpg|thumb]] ===သဵင်ဢွၵ်ႇ=== * {{a|en|UK}} ** {{IPA|en|/ˈhɔːk/|a=RP,Northern England,Wales}} ** {{IPA|en|/ˈhoːk/|a=SSB}} ** {{IPA|en|/ˈhɔk/|a=Scotland}} * {{a|en|US}}: ** {{IPA|en|/ˈhɔk/|a=non-cot-caught}} ** {{IPA|en|/ˈhɑk/|a=cot-caught}} ** {{audio|en|en-us-hawk.ogg|a=US}} * {{IPA|en|/ˈhɒk/|a=CA}} * {{IPA|en|/ˈhoːk/|a=AU,NZ}} * {{rhymes|en|ɔːk|s=1}} * {{hyph|en|hawk}} * {{hmp|en|Hauke|Hawk|hawke|Hawke|;|hock<q:cot-caught>|hok}} ===ႁိၵ်ႈ=== {{en-noun}} # ၼူၵ်ႉႁုင်ႉ။ # ၵေႃႉၶႂ်ႈတိုၵ်းတေႃး။ ===သၢင်ႈ=== {{en-verb}} # ႁႃၶႆၢတၢင်းၶၢႆ ၵိဝ်ႇၼိုင်ႈယဝ်ႉၵိဝ်ႇၼိုင်ႈ။ # ၶီႈၶၢၵ်ႇၶမ်ၶေႃး။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} q1i6cvbm0o1h4v3i1qmprts99kqo1f9 0 76376 293621 293386 2026-09-20T13:31:29Z Saimawnkham 9 /* ဢိင်းၵလဵတ်ႈ */ 293621 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== [[File:Hawser rope of the Rio Tagus (ship, 1979), Sète cf01.jpg|thumb|a typical hawser]] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/hɔːzə/|a=RP}} * {{IPA|en|/hɔzɚ/|a=GA}} * {{audio|en|LL-Q1860 (eng)-Vealhurl-hawser.wav|a=Southern England}} * {{rhymes|en|ɔːzə(ɹ)|s=2}} ===ႁိၵ်ႈ=== {{en-noun}} # ၸိူၵ်ႈလဵၵ်းၸႂ်ႉၸွမ်းႁိူဝ်းၼမ်ႉ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} k0wmnsnt3rbqb2lueurvihx5wiy91x9 0 76427 293584 293439 2026-09-20T12:45:37Z Saimawnkham 9 /* ႁိၵ်ႈ */ 293584 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{enPR|hē}}, {{IPA|en|/ˈhiː/}} ** {{IPA|en|[hiː]|a=RP,CA}} *** {{audio|en|En-uk-he.ogg|a=UK}} ** {{IPA|en|[hɪj]|a=SSB,AU,Southern US}} ** {{IPA|en|[hi(ː)]|[çi(ː)]|a=US}} *** {{audio|en|en-us-he.ogg|a=US}} * {{IPA|en|/(h)i/|/(h)ɪ/|a=unstressed}} * {{rhymes|en|iː|s=1}} ===ႁိၵ်ႈ=== {{en-pron}} # (pron) မၼ်း။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} jra8bp221b0v67moakdyhz62r6n71re 293585 293584 2026-09-20T12:45:59Z Saimawnkham 9 /* ႁိၵ်ႈ */ 293585 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{enPR|hē}}, {{IPA|en|/ˈhiː/}} ** {{IPA|en|[hiː]|a=RP,CA}} *** {{audio|en|En-uk-he.ogg|a=UK}} ** {{IPA|en|[hɪj]|a=SSB,AU,Southern US}} ** {{IPA|en|[hi(ː)]|[çi(ː)]|a=US}} *** {{audio|en|en-us-he.ogg|a=US}} * {{IPA|en|/(h)i/|/(h)ɪ/|a=unstressed}} * {{rhymes|en|iː|s=1}} ===ႁိၵ်ႈ=== {{en-pronoun}} # (pron) မၼ်း။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} 7piuhcamtsqnldcpqikudwcwc0y8s04 293586 293585 2026-09-20T12:46:09Z Saimawnkham 9 /* ႁိၵ်ႈ */ 293586 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{enPR|hē}}, {{IPA|en|/ˈhiː/}} ** {{IPA|en|[hiː]|a=RP,CA}} *** {{audio|en|En-uk-he.ogg|a=UK}} ** {{IPA|en|[hɪj]|a=SSB,AU,Southern US}} ** {{IPA|en|[hi(ː)]|[çi(ː)]|a=US}} *** {{audio|en|en-us-he.ogg|a=US}} * {{IPA|en|/(h)i/|/(h)ɪ/|a=unstressed}} * {{rhymes|en|iː|s=1}} ===ႁိၵ်ႈ=== {{en-pronoun}} # မၼ်း။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} e95h8fndxwfiyzua87jxbzrx3xy9d49 0 76434 293589 293447 2026-09-20T12:51:10Z Saimawnkham 9 293589 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-prepphr}} # ႁႂ်ႈလိုမ်းပႅတ်ႈ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} hapoiqrs4506kw8ry9h86ztikg2xpah 293590 293589 2026-09-20T12:51:24Z Saimawnkham 9 Saimawnkham ၶၢႆႉယဝ်ႉ ၼႃႈလိၵ်ႈ [[Put something out of someone's head]] တေႃႇ [[put something out of someone's head]] 293589 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-prepphr}} # ႁႂ်ႈလိုမ်းပႅတ်ႈ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} hapoiqrs4506kw8ry9h86ztikg2xpah 0 76441 293593 293454 2026-09-20T12:55:16Z Saimawnkham 9 /* သၢင်ႈ */ 293593 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-verb|<,,stood>}} # ပူၵ်းမိုဝ်းလႆႈ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} baddzwid3qdgatrs3h5424f7vg61sxw 0 76457 293597 293470 2026-09-20T12:57:07Z Saimawnkham 9 293597 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-verb|<,,made>}} # ၶိုင်သၢင်ႈဢဝ်ႁင်းၵူၺ်း။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} lfz3e4enjjb64k6oz46pkn62f50ksfc 0 76458 293601 293471 2026-09-20T12:58:58Z Saimawnkham 9 /* သၢင်ႈ */ 293601 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-verb|<,,bitten>}} # ၸႂ်လမ်သေလႃႇ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} rz3xfm2rrv3a4ow8w59p89i8cre0yd1 293606 293601 2026-09-20T13:16:06Z Saimawnkham 9 293606 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{en-verb|<,,bitten>}} # ၸႂ်လမ်သေလႃႇ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} m01bzr7szclddxjanwlpopzsggmd6si 0 76502 293616 293518 2026-09-20T13:27:47Z Saimawnkham 9 293616 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== [[File:Audio-Technica ATH-A500 (2).jpg|thumb|right|a pair of circumaural '''headphones''']] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɛdfəʊnz/|a=RP}} ** {{audio|en|LL-Q1860 (eng)-Vealhurl-headphones.wav|a=Southern England}} * {{IPA|en|/ˈhɛdfoʊnz/|a=US}} ** {{audio|en|LL-Q1860 (eng)-Aadenboy-headphones.wav|a=US}} * {{IPA|en|/ˈhedfəʉnz/|a=AU}} * {{IPA|en|/ˈhedfɐʉnz/|a=NZ}} * {{IPA|en|/ˈhɛdfonz/|a=Scotland}} * {{IPA|en|/ˈhɛdfoːnz/|a=Wales,non-toe-tow}} ===ႁိၵ်ႈ=== {{en-noun|p|sg=headphone}} # ၶိူင်ႈထွမ်ႇႁူ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} eumbj81sby989widbf0esks3jxxjzwr 0 76503 293618 293519 2026-09-20T13:28:45Z Saimawnkham 9 /* ဢိင်းၵလဵတ်ႈ */ 293618 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== [[File:Shirt of Mail MET DP359868.jpg|thumb]] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɛdˌpis/|a=GA}} ===ႁိၵ်ႈ=== {{en-noun}} # ၵုပ်း။ # ဢွၵ်းဢေႃ။ # ပႆၢးဝူၼ်ႉ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} bguejs09ss2p9sy8z8bdyxjej6nbe3i 293619 293618 2026-09-20T13:29:03Z Saimawnkham 9 /* သဵင်ဢွၵ်ႇ */ 293619 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== [[File:Shirt of Mail MET DP359868.jpg|thumb]] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɔːbɜː(ɹ)k/}} * {{rhymes|en|ɔːbɜː(ɹ)k|s=2}} * {{hyph|en|hau|berk}} ===ႁိၵ်ႈ=== {{en-noun}} # ၵုပ်း။ # ဢွၵ်းဢေႃ။ # ပႆၢးဝူၼ်ႉ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} 29ypp9cly38fsukhzibugz6ftsem654 0 76557 293578 2026-09-20T12:41:29Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[hateful]] ([[WT:ACCEL|Accelerated]]) 293578 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===ၵမ်ႉႁိၵ်ႈ=== {{head|en|ၶေႃႈတႅၵ်ႈၼိူင်း ၵမ်ႉႁိၵ်ႈ}} # {{en-comparative of|hateful}} hbk08su5smhl84c2o8cx3ymi9wa9l37 0 76558 293579 2026-09-20T12:41:40Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[hateful]] ([[WT:ACCEL|Accelerated]]) 293579 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===ၵမ်ႉႁိၵ်ႈ=== {{head|en|ၶေႃႈသုင်သုတ်း ၵမ်ႉႁိၵ်ႈ}} # {{en-superlative of|hateful}} 0aasg9bbqm7w1yhu6wcg5xrp8fduoj9 0 76559 293580 2026-09-20T12:42:55Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[have]] ([[WT:ACCEL|Accelerated]]) 293580 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ====သဵင်ဢွၵ်ႇ==== * {{enPR|hăz|a=stressed}}, {{IPA|en|/hæz/}} ** {{audio|en|LL-Q1860 (eng)-Pvanp7-has.wav|a=UK}} ** {{audio|en|en-us-has.ogg|a=US}} * {{enPR|həz|əz|a=unstressed}}, {{IPA|en|/həz/|/əz/}} * {{sense|has to}} {{enPR|hăs}}, {{IPA|en|/hæs/}} * {{rhymes|en|æz|s=1}} ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ}} # {{infl of|en|have||s-verb-form}} 1dcxq4oo5vo7z0f8fdy1phawr376lq6 14 76560 293581 2026-09-20T12:43:12Z Saimawnkham 9 ၵေႃႇသၢင်ႈၼႃႈလိၵ်ႈဝႆႉ တင်း "[[ပိူင်ထၢၼ်ႈ:တူၼ်းၸၢပ်ႈလႅပ်ႈ:ဢိင်းၵလဵတ်ႈ/æz]]" 293581 wikitext text/x-wiki [[ပိူင်ထၢၼ်ႈ:တူၼ်းၸၢပ်ႈလႅပ်ႈ:ဢိင်းၵလဵတ်ႈ/æz]] 29oqvn08lmxoz818ca6faq9s7ppf6ib 14 76561 293582 2026-09-20T12:44:00Z Saimawnkham 9 ၵေႃႇသၢင်ႈၼႃႈလိၵ်ႈဝႆႉ တင်း "__HIDDENCAT__" 293582 wikitext text/x-wiki __HIDDENCAT__ 2twjmejn56ditxo46hqinfh52nh6flb 0 76562 293583 2026-09-20T12:44:43Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[make hay while the sun shines]] ([[WT:ACCEL|Accelerated]]) 293583 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=made hay while the sun shone}} # {{infl of|en|make hay while the sun shines||ed-form}} kqlikggczkevme88632gbd7dtkz4u7v 0 76563 293587 2026-09-20T12:49:23Z Saimawnkham 9 ၵေႃႇသၢင်ႈၼႃႈလိၵ်ႈဝႆႉ တင်း "==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/hɛdz/}} * {{audio|en|en-us-heads.ogg|a=US}} * {{rhymes|en|ɛdz|s=1}} ===ႁိၵ်ႈ=== {{head|en|ပိူင်ၽၢင်ႁိၵ်ႈ}} # {{plural of|en|head}}." 293587 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/hɛdz/}} * {{audio|en|en-us-heads.ogg|a=US}} * {{rhymes|en|ɛdz|s=1}} ===ႁိၵ်ႈ=== {{head|en|ပိူင်ၽၢင်ႁိၵ်ႈ}} # {{plural of|en|head}}. o2zbnv3eypwjaj8kud16t9jjvzb4y42 0 76564 293588 2026-09-20T12:50:00Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[head]] ([[WT:ACCEL|Accelerated]]) 293588 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɛdɪd/}} * {{audio|en|LL-Q1860 (eng)-Vealhurl-headed.wav|a=Southern England}} * {{rhymes|en|ɛdɪd|s=2}} ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ}} # {{infl of|en|head||ed-form}} jnf54vus3hgwd5l9vq9qxwtqcx52rqb 0 76565 293591 2026-09-20T12:51:24Z Saimawnkham 9 Saimawnkham ၶၢႆႉယဝ်ႉ ၼႃႈလိၵ်ႈ [[Put something out of someone's head]] တေႃႇ [[put something out of someone's head]] 293591 wikitext text/x-wiki #REDIRECT [[put something out of someone's head]] 9afpo2av0obuiztn5o89ynjve6jm9ft 0 76566 293592 2026-09-20T12:54:44Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[keep one's head above water]] ([[WT:ACCEL|Accelerated]]) 293592 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=kept one's head above water}} # {{infl of|en|keep one's head above water||ed-form}} f7m37bb0h675cebb7slfkywvbbqgtlr 0 76567 293594 2026-09-20T12:55:25Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[stand on one’s head]] ([[WT:ACCEL|Accelerated]]) 293594 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=stands on one’s head}} # {{infl of|en|stand on one’s head||s-verb-form}} pei79h0cooq4fz6g10dhuuvbo4j1h0b 0 76568 293595 2026-09-20T12:55:35Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[stand on one’s head]] ([[WT:ACCEL|Accelerated]]) 293595 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=standing on one’s head}} # {{infl of|en|stand on one’s head||ing-form}} c4k1nyc346s8zgw3j7v20ox5bhxcv7e 0 76569 293596 2026-09-20T12:55:41Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[stand on one’s head]] ([[WT:ACCEL|Accelerated]]) 293596 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=stood on one’s head}} # {{infl of|en|stand on one’s head||ed-form}} qgfyl0vckwftd600ksfn996ok4pxhup 0 76570 293598 2026-09-20T12:57:18Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[make something up out of one’s head]] ([[WT:ACCEL|Accelerated]]) 293598 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=makes something up out of one’s head}} # {{infl of|en|make something up out of one’s head||s-verb-form}} 8p6uywjt6dqgukxrou86v8o928h43z0 0 76571 293599 2026-09-20T12:57:23Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[make something up out of one’s head]] ([[WT:ACCEL|Accelerated]]) 293599 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=making something up out of one’s head}} # {{infl of|en|make something up out of one’s head||ing-form}} iz4uqp7blj71p7qwp8hxuae8msc34nf 0 76572 293600 2026-09-20T12:57:27Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[make something up out of one’s head]] ([[WT:ACCEL|Accelerated]]) 293600 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=made something up out of one’s head}} # {{infl of|en|make something up out of one’s head||ed-form}} 1xukoz8boir6jowmelnsbp95w0pb55m 0 76573 293607 2026-09-20T13:16:15Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[bite one’s head off]] ([[WT:ACCEL|Accelerated]]) 293607 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=bites one’s head off}} # {{infl of|en|bite one’s head off||s-verb-form}} npuhqy4yqmzbh6msprzzkplyoccnyv6 0 76574 293608 2026-09-20T13:16:37Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[bite one’s head off]] ([[WT:ACCEL|Accelerated]]) 293608 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=biting one’s head off}} # {{infl of|en|bite one’s head off||ing-form}} 7jiwy0r0ydz68ltf043yztzna6ooqmk 0 76575 293609 2026-09-20T13:16:48Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[bite one’s head off]] ([[WT:ACCEL|Accelerated]]) 293609 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=bitten one’s head off}} # {{infl of|en|bite one’s head off||ed-form}} ni3dcch1hbi66o415e6zqxj4zisru25 0 76576 293610 2026-09-20T13:23:52Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[give someone their head]] ([[WT:ACCEL|Accelerated]]) 293610 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=gives someone their head}} # {{infl of|en|give someone their head||s-verb-form}} 5v2pnyghlryq9nifpsk09jdkrpthkjw 0 76577 293611 2026-09-20T13:23:57Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[give someone their head]] ([[WT:ACCEL|Accelerated]]) 293611 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=giving someone their head}} # {{infl of|en|give someone their head||ing-form}} ov5gwxbpdkhickawx2xql7b5pajeole 0 76578 293612 2026-09-20T13:24:03Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[give someone their head]] ([[WT:ACCEL|Accelerated]]) 293612 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=gave someone their head}} # {{infl of|en|give someone their head||spast}} 9gklxdgc8z8r7qie4ryy5on221h2oxl 0 76579 293613 2026-09-20T13:24:08Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[give someone their head]] ([[WT:ACCEL|Accelerated]]) 293613 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=given someone their head}} # {{infl of|en|give someone their head||past|part}} m56572uqpq1fj6yi5aebqp01ux7r8qr 0 76580 293614 2026-09-20T13:25:00Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[lose one's head]] ([[WT:ACCEL|Accelerated]]) 293614 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=lost one's head}} # {{infl of|en|lose one's head||ed-form}} t2sbvhwcsgjzymk5w3mgh8l35xfjw4e 0 76581 293615 2026-09-20T13:25:16Z Saimawnkham 9 Creating forms of ဢိင်းၵလဵတ်ႈ [[keep one's head]] ([[WT:ACCEL|Accelerated]]) 293615 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== ===သၢင်ႈ=== {{head|en|ပိူင်ၽၢင်သၢင်ႈ|head=kept one's head}} # {{infl of|en|keep one's head||ed-form}} bxoa93ek8kt697z58u1rrwpkmdo7zln 0 76582 293617 2026-09-20T13:27:48Z Saimawnkham 9 ၵေႃႇသၢင်ႈၼႃႈလိၵ်ႈဝႆႉ တင်း "==ဢိင်းၵလဵတ်ႈ== {{wp}} [[File:Audio-Technica ATH-A500 (2).jpg|thumb|right|a pair of circumaural '''headphones''']] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɛdfəʊn/|a=RP}} ** {{audio|en|LL-Q1860 (eng)-Vealhurl-headphone.wav|a=Southern England}} * {{IPA|en|/ˈhɛdfoʊn/|a=US}} ** {{audio|en|En-us-headphone.oga|a=GA}} * {{IPA|en|/ˈhedfəʉn/|a=AU}} * {{IPA|en|/ˈhedfɐʉn/|a=NZ}} * {{..." 293617 wikitext text/x-wiki ==ဢိင်းၵလဵတ်ႈ== {{wp}} [[File:Audio-Technica ATH-A500 (2).jpg|thumb|right|a pair of circumaural '''headphones''']] ===သဵင်ဢွၵ်ႇ=== * {{IPA|en|/ˈhɛdfəʊn/|a=RP}} ** {{audio|en|LL-Q1860 (eng)-Vealhurl-headphone.wav|a=Southern England}} * {{IPA|en|/ˈhɛdfoʊn/|a=US}} ** {{audio|en|En-us-headphone.oga|a=GA}} * {{IPA|en|/ˈhedfəʉn/|a=AU}} * {{IPA|en|/ˈhedfɐʉn/|a=NZ}} * {{IPA|en|/ˈhɛdfon/|a=Scotland}} * {{IPA|en|/ˈhɛdfoːn/|a=Wales,non-toe-tow}} ===ႁိၵ်ႈ=== {{en-noun}} # ၶိူင်ႈထွမ်ႇႁူ။ ===ၽိုၼ်ဢိင်=== # {{VPS Ref}} d6e6hl5fllm6ofx45ysvy08n2q2a87f