Vikisözlük
trwiktionary
https://tr.wiktionary.org/wiki/Vikis%C3%B6zl%C3%BCk:Anasayfa
MediaWiki 1.47.0-wmf.7
case-sensitive
Ortam
Özel
Tartışma
Kullanıcı
Kullanıcı mesaj
Vikisözlük
Vikisözlük tartışma
Dosya
Dosya tartışma
MediaWiki
MediaWiki tartışma
Şablon
Şablon tartışma
Yardım
Yardım tartışma
Kategori
Kategori tartışma
Portal
Portal tartışma
YeniKurum
YeniKurum tartışma
Ek
Ek tartışma
Alıntılar
Alıntılar tartışma
TimedText
TimedText talk
Modül
Modül tartışma
Event
Event talk
misafir
0
53753
5669941
5257723
2026-06-23T10:22:02Z
Huntersenpai2
68080
/* Çeviriler */ Yazım hatası düzeltildi
5669941
wikitext
text/x-wiki
{{bakınız|misafirhane}}
==Türkçe==
[mi·sa:·fir]
===Söyleniş===
* {{IPA|dil=tr|/mi.saː.ˈfiɾ/}}
===Köken===
{{k|dil=tr|ota}} {{z|ota|مسافر}}, [[Arapça]] {{z|ar|مِسَافِر}}
===Ad===
{{tr-ad}}
{{Vikipedi}}
# {{t|dil=tr|sosyoloji}} [[bir]] [[yer|yere]] [[veya]] [[biri|birinin]] [[ev|evine]] [[kısa]] [[bir]] [[süre]] [[kalmak]] [[için]] [[gelmek|gelen]] [[kişi]], [[konuk]], [[mihman]]
====Çekimleme====
{{tr-ad-tablo}}
====Çeviriler====
{{Üst|tip=çeviriler}}
*Almanca: {{ç|de|Besucher|m}}, {{ç|de|Besucherin|f}}
*Arapça: {{ç|ar|مُسَافِر|m}}
*Ersya dili: {{ç|myv|инже}}
*Fince: {{ç|fi|vieras}}
*Friuli dili: {{ç|fur|forest}}
*Hausaca: {{ç|ha|baƙo}}
*İngilizce: {{ç|en|guest}}
*Lehçe: {{ç|pl|gość}}
*Macarca: {{ç|hu|vendég}}
*Malgaşça: {{ç|mg|vahiny}}
*Mari dili: {{ç|chm|уна}}
*Mokşa dili: {{ç|mdf|конак}}
*Osmanlı Türkçesi: {{ç|ota|مسافر}}
*Pencapça: {{ç|pa|ਅਤਿਥਿ}}
*Rumence: {{ç|ro|musafir}}
*Telugu dili: {{ç|te|అతిథి}}
*Yunanca: {{ç|el|μουσαφίρης}}
{{Alt}}
===Kaynakça===
* {{Kaynak-TDK}}
====Atasözleri====
* [[Ahmak misafir ev sahibini ağırlar]]
* [[Köylü, misafir kabul etmeyiz demez, konacak konak yoktur der]]
* [[Misafir balığa benzer, bekledi mi kokar]]
* [[Misafir ev sahibinin kuzusudur]]
* [[Misafir kısmeti ile gelir]]
* [[Misafir misafiri istemez, ev sahibi ikisini de]]
* [[Misafir on kısmetle gelir; birini yer, dokuzunu bırakır]]
* [[Misafir umduğunu değil bulduğunu yer]]
qm4sxddz4v84fjb4qkawa2l81fw9row
batırık
0
127583
5669903
4354579
2026-06-22T20:19:02Z
~2026-36427-17
68146
/* Ad */ Yazım hatası düzeltildi, açıklama detaylandirildi
5669903
wikitext
text/x-wiki
==Türkçe==
===Ad===
{{tr-ad}}
# {{terim|dil=tr|halk ağzı}} [[Kısır]]ın su ve taze sebzeler ile karıştırılmasıyla yapılan yöresel bir gıda
23hvum881xg8n4sf2yhrfd2juicyxnf
შიში
0
148291
5669906
4541794
2026-06-22T21:35:32Z
Turgut46
26665
/* Gürcüce */
5669906
wikitext
text/x-wiki
== Gürcüce ==
===Ad===
{{başlık başı|ka|ad}}
# [[korku]]
n4j16tttogoi3djms6ofvj7luh9rnpz
грейпфрут
0
178379
5669945
4537736
2026-06-23T11:20:59Z
MustafaCavlak
59368
/* Rusça */
5669945
wikitext
text/x-wiki
==Rusça==
===Ad===
{{başlık başı|ru|ad|baş=грейпфру́т|c=e}}
# [[greyfurt]]
====Çekimleme====
{{ru-ad-cansız-s-E-a|грейпфру́т}}
2j75lkdepndkgf7wgm3n84jfpfo26gm
Vikisözlük:Değişiklik sayılarına göre Vikisözlükçüler listesi
4
466356
5669904
5669610
2026-06-22T21:05:16Z
YBot
37147
Güncelleme
5669904
wikitext
text/x-wiki
{{/begin}}
<center>
{| class="wikitable"
! #
! Kullanıcı
! Değişiklik sayısı
! Kullanıcı grupları
|-
| 1
| [[Kullanıcı:HastaLaVi2|HastaLaVi2]]
| align="center" | 1.264.519
| arayüz yöneticisi, hizmetli
|-
| 2
| [[Kullanıcı:Victor Trevor|Victor Trevor]]
| align="center" | 144.832
|
|-
| 3
| [[Kullanıcı:MustafaCavlak|MustafaCavlak]]
| align="center" | 102.273
| hizmetli, geçici hesap IP görüntüleyici
|-
| 4
| [[Kullanıcı:Viki|Viki]]
| align="center" | 45.152
|
|-
| 5
| [[Kullanıcı:By erdo can|By erdo can]]
| align="center" | 42.866
|
|-
| 6
| [[Kullanıcı:Satirdan kahraman|Satirdan kahraman]]
| align="center" | 37.378
|
|-
| 7
| [[Kullanıcı:Sae1962|Sae1962]]
| align="center" | 34.149
|
|-
| 8
| [[Kullanıcı:Saltinbas|Saltinbas]]
| align="center" | 26.434
|
|-
| 9
| [[Kullanıcı:MüjdeA|MüjdeA]]
| align="center" | 25.502
|
|-
| 10
| [[Kullanıcı:Sabri76|Sabri76]]
| align="center" | 20.349
|
|-
| 11
| [[Kullanıcı:Katakay|Katakay]]
| align="center" | 19.415
|
|-
| 12
| [[Kullanıcı:Turgut46|Turgut46]]
| align="center" | 17.527
| hizmetli
|-
| 13
| [[Kullanıcı:Vuzorg|Vuzorg]]
| align="center" | 15.932
|
|-
| 14
| [[Kullanıcı:Ender|Ender]]
| align="center" | 14.945
|
|-
| 15
| [[Kullanıcı:Kök|Kök]]
| align="center" | 8.437
|
|-
| 16
| [[Kullanıcı:Curious|Curious]]
| align="center" | 7.438
|
|-
| 17
| [[Kullanıcı:Bobcats|Bobcats]]
| align="center" | 6.526
|
|-
| 18
| [[Kullanıcı:Nosferatü|Nosferatü]]
| align="center" | 6.408
|
|-
| 19
| [[Kullanıcı:Uncitoyen|Uncitoyen]]
| align="center" | 5.685
| hizmetli
|-
| 20
| [[Kullanıcı:Vikiyılmaz|Vikiyılmaz]]
| align="center" | 5.369
|
|-
| 21
| [[Kullanıcı:Sinek~trwiktionary|Sinek~trwiktionary]]
| align="center" | 5.221
|
|-
| 22
| [[Kullanıcı:YZCTEK|YZCTEK]]
| align="center" | 4.880
|
|-
| 23
| [[Kullanıcı:Vorhon|Vorhon]]
| align="center" | 4.699
|
|-
| 24
| [[Kullanıcı:LA2|LA2]]
| align="center" | 4.433
|
|-
| 25
| [[Kullanıcı:Apisite|Apisite]]
| align="center" | 4.428
|
|-
| 26
| [[Kullanıcı:B5.101.222.187|B5.101.222.187]]
| align="center" | 4.259
|
|-
| 27
| [[Kullanıcı:Kavurt|Kavurt]]
| align="center" | 4.088
|
|-
| 28
| [[Kullanıcı:Vitruvian|Vitruvian]]
| align="center" | 4.054
|
|-
| 29
| [[Kullanıcı:123snake45|123snake45]]
| align="center" | 3.971
|
|-
| 30
| [[Kullanıcı:Alikiroglu61|Alikiroglu61]]
| align="center" | 3.773
|
|-
| 31
| [[Kullanıcı:3210|3210]]
| align="center" | 3.716
|
|-
| 32
| [[Kullanıcı:Goreno|Goreno]]
| align="center" | 3.474
|
|-
| 33
| [[Kullanıcı:Runningfridgesrule|Runningfridgesrule]]
| align="center" | 3.119
|
|-
| 34
| [[Kullanıcı:Reality006|Reality006]]
| align="center" | 3.099
|
|-
| 35
| [[Kullanıcı:HâlitM|HâlitM]]
| align="center" | 3.052
|
|-
| 36
| [[Kullanıcı:Ahmed922229|Ahmed922229]]
| align="center" | 3.009
|
|-
| 37
| [[Kullanıcı:Barishan|Barishan]]
| align="center" | 2.971
|
|-
| 38
| [[Kullanıcı:Ekrem691|Ekrem691]]
| align="center" | 2.398
|
|-
| 39
| [[Kullanıcı:Esege|Esege]]
| align="center" | 2.321
|
|-
| 40
| [[Kullanıcı:Ugur Basak|Ugur Basak]]
| align="center" | 2.267
|
|-
| 41
| [[Kullanıcı:Gamlıbaykuş|Gamlıbaykuş]]
| align="center" | 2.098
|
|-
| 42
| [[Kullanıcı:AML ZİYA AKYASAN|AML ZİYA AKYASAN]]
| align="center" | 1.989
|
|-
| 43
| [[Kullanıcı:Entry62|Entry62]]
| align="center" | 1.959
|
|-
| 44
| [[Kullanıcı:Magurale|Magurale]]
| align="center" | 1.854
|
|-
| 45
| [[Kullanıcı:Sihl fatma|Sihl fatma]]
| align="center" | 1.835
|
|-
| 46
| [[Kullanıcı:Don Alessandro|Don Alessandro]]
| align="center" | 1.815
|
|-
| 47
| [[Kullanıcı:Burudet88|Burudet88]]
| align="center" | 1.791
|
|-
| 48
| [[Kullanıcı:Şêr|Şêr]]
| align="center" | 1.668
|
|-
| 49
| [[Kullanıcı:Abuk78|Abuk78]]
| align="center" | 1.534
|
|-
| 50
| [[Kullanıcı:Göktürker|Göktürker]]
| align="center" | 1.508
|
|-
| 51
| [[Kullanıcı:Flanoz|Flanoz]]
| align="center" | 1.463
|
|-
| 52
| [[Kullanıcı:Sa'y|Sa'y]]
| align="center" | 1.345
|
|-
| 53
| [[Kullanıcı:Mzekiu|Mzekiu]]
| align="center" | 1.344
|
|-
| 54
| [[Kullanıcı:Samleml enes bilgin|Samleml enes bilgin]]
| align="center" | 1.315
|
|-
| 55
| [[Kullanıcı:İbrahim alican|İbrahim alican]]
| align="center" | 1.292
|
|-
| 56
| [[Kullanıcı:SİHL Gökhan|SİHL Gökhan]]
| align="center" | 1.278
|
|-
| 57
| [[Kullanıcı:Nazendegül|Nazendegül]]
| align="center" | 1.260
|
|-
| 58
| [[Kullanıcı:ליאון12|ליאון12]]
| align="center" | 1.178
|
|-
| 59
| [[Kullanıcı:Semlaml110|Semlaml110]]
| align="center" | 1.174
|
|-
| 60
| [[Kullanıcı:Ahmetaslan|Ahmetaslan]]
| align="center" | 1.141
|
|-
| 61
| [[Kullanıcı:Kibele|Kibele]]
| align="center" | 1.136
|
|-
| 62
| [[Kullanıcı:Wwww8888|Wwww8888]]
| align="center" | 1.125
|
|-
| 63
| [[Kullanıcı:Kmlblg 332|Kmlblg 332]]
| align="center" | 1.124
|
|-
| 64
| [[Kullanıcı:Lagrium|Lagrium]]
| align="center" | 1.094
|
|-
| 65
| [[Kullanıcı:SİHL ALİ MUSA|SİHL ALİ MUSA]]
| align="center" | 1.045
|
|-
| 66
| [[Kullanıcı:Ömer faruk çakmak|Ömer faruk çakmak]]
| align="center" | 996
|
|-
| 67
| [[Kullanıcı:Aybeg|Aybeg]]
| align="center" | 968
|
|-
| 68
| [[Kullanıcı:ВМНС|ВМНС]]
| align="center" | 962
|
|-
| 69
| [[Kullanıcı:Almanbet Janışev|Almanbet Janışev]]
| align="center" | 954
|
|-
| 70
| [[Kullanıcı:Seksen iki yüz kırk beş|Seksen iki yüz kırk beş]]
| align="center" | 931
|
|-
| 71
| [[Kullanıcı:Maxa|Maxa]]
| align="center" | 912
|
|-
| 72
| [[Kullanıcı:Wiamboo|Wiamboo]]
| align="center" | 880
|
|-
| 73
| [[Kullanıcı:SİHL Menzulekumaş|SİHL Menzulekumaş]]
| align="center" | 846
|
|-
| 74
| [[Kullanıcı:Melkorettin|Melkorettin]]
| align="center" | 812
|
|-
| 75
| [[Kullanıcı:Yasemin|Yasemin]]
| align="center" | 802
|
|-
| 76
| [[Kullanıcı:Peirep|Peirep]]
| align="center" | 791
|
|-
| 77
| [[Kullanıcı:Jeuwre|Jeuwre]]
| align="center" | 786
|
|-
| 78
| [[Kullanıcı:S.ihl.selmaaksoy|S.ihl.selmaaksoy]]
| align="center" | 781
|
|-
| 79
| [[Kullanıcı:Gibranist|Gibranist]]
| align="center" | 773
|
|-
| 80
| [[Kullanıcı:Kmlblgfatmanur|Kmlblgfatmanur]]
| align="center" | 771
|
|-
| 81
| [[Kullanıcı:Asım Hoca|Asım Hoca]]
| align="center" | 765
|
|-
| 82
| [[Kullanıcı:Malafaya|Malafaya]]
| align="center" | 765
|
|-
| 83
| [[Kullanıcı:Hagvoda|Hagvoda]]
| align="center" | 760
|
|-
| 84
| [[Kullanıcı:YosefHan7|YosefHan7]]
| align="center" | 756
|
|-
| 85
| [[Kullanıcı:Mustafahoca|Mustafahoca]]
| align="center" | 745
|
|-
| 86
| [[Kullanıcı:ZazaLejyoner|ZazaLejyoner]]
| align="center" | 740
|
|-
| 87
| [[Kullanıcı:Sürmene aml ve eml yusufcakir|Sürmene aml ve eml yusufcakir]]
| align="center" | 721
|
|-
| 88
| [[Kullanıcı:Mehmet-k61|Mehmet-k61]]
| align="center" | 670
|
|-
| 89
| [[Kullanıcı:Sinus46|Sinus46]]
| align="center" | 670
|
|-
| 90
| [[Kullanıcı:Kml serhan|Kml serhan]]
| align="center" | 668
|
|-
| 91
| [[Kullanıcı:TrDAsh|TrDAsh]]
| align="center" | 639
|
|-
| 92
| [[Kullanıcı:Itidal|Itidal]]
| align="center" | 635
|
|-
| 93
| [[Kullanıcı:Ömerbysl|Ömerbysl]]
| align="center" | 635
|
|-
| 94
| [[Kullanıcı:Chelik99|Chelik99]]
| align="center" | 587
|
|-
| 95
| [[Kullanıcı:Mereyü|Mereyü]]
| align="center" | 556
|
|-
| 96
| [[Kullanıcı:Octahedron80|Octahedron80]]
| align="center" | 547
|
|-
| 97
| [[Kullanıcı:ALİ İHSAN KARAMAN|ALİ İHSAN KARAMAN]]
| align="center" | 533
|
|-
| 98
| [[Kullanıcı:Gullerdiyari|Gullerdiyari]]
| align="center" | 520
|
|-
| 99
| [[Kullanıcı:Yeniler|Yeniler]]
| align="center" | 515
|
|-
| 100
| [[Kullanıcı:Andac20|Andac20]]
| align="center" | 509
|
|-
| 101
| [[Kullanıcı:Dilyâresi|Dilyâresi]]
| align="center" | 485
|
|-
| 102
| [[Kullanıcı:Nerepla Keskin|Nerepla Keskin]]
| align="center" | 483
|
|-
| 103
| [[Kullanıcı:Berkucar0|Berkucar0]]
| align="center" | 473
|
|-
| 104
| [[Kullanıcı:İsmetaydin|İsmetaydin]]
| align="center" | 465
|
|-
| 105
| [[Kullanıcı:Sadmleml enes|Sadmleml enes]]
| align="center" | 464
|
|-
| 106
| [[Kullanıcı:Condmatstrel|Condmatstrel]]
| align="center" | 461
|
|-
| 107
| [[Kullanıcı:Wutsje|Wutsje]]
| align="center" | 440
|
|-
| 108
| [[Kullanıcı:Lo Ximiendo|Lo Ximiendo]]
| align="center" | 435
|
|-
| 109
| [[Kullanıcı:Kml znnt|Kml znnt]]
| align="center" | 430
|
|-
| 110
| [[Kullanıcı:CommonsDelinker|CommonsDelinker]]
| align="center" | 423
|
|-
| 111
| [[Kullanıcı:Kendim|Kendim]]
| align="center" | 418
|
|-
| 112
| [[Kullanıcı:Pathoschild|Pathoschild]]
| align="center" | 413
|
|-
| 113
| [[Kullanıcı:Kemal.KURBAN|Kemal.KURBAN]]
| align="center" | 413
|
|-
| 114
| [[Kullanıcı:Merbeg|Merbeg]]
| align="center" | 410
|
|-
| 115
| [[Kullanıcı:Hamzahoca|Hamzahoca]]
| align="center" | 405
|
|-
| 116
| [[Kullanıcı:AHMET AYSOY|AHMET AYSOY]]
| align="center" | 393
|
|-
| 117
| [[Kullanıcı:Sihl ayhan|Sihl ayhan]]
| align="center" | 385
|
|-
| 118
| [[Kullanıcı:A.d.m.l.ve.e.m.lmuhammet alican|A.d.m.l.ve.e.m.lmuhammet alican]]
| align="center" | 383
|
|-
| 119
| [[Kullanıcı:Kvazimodo|Kvazimodo]]
| align="center" | 381
|
|-
| 120
| [[Kullanıcı:Muhammet yardım|Muhammet yardım]]
| align="center" | 378
|
|-
| 121
| [[Kullanıcı:Egemensen~trwiktionary|Egemensen~trwiktionary]]
| align="center" | 376
|
|-
| 122
| [[Kullanıcı:Johnny281993|Johnny281993]]
| align="center" | 376
|
|-
| 123
| [[Kullanıcı:Türkçe|Türkçe]]
| align="center" | 368
|
|-
| 124
| [[Kullanıcı:Vito Genovese|Vito Genovese]]
| align="center" | 365
|
|-
| 125
| [[Kullanıcı:Ata Kemal|Ata Kemal]]
| align="center" | 365
|
|-
| 126
| [[Kullanıcı:ToprakM|ToprakM]]
| align="center" | 365
|
|-
| 127
| [[Kullanıcı:Yunus Emre Biçer|Yunus Emre Biçer]]
| align="center" | 364
|
|-
| 128
| [[Kullanıcı:Bm|Bm]]
| align="center" | 347
|
|-
| 129
| [[Kullanıcı:Ampouble|Ampouble]]
| align="center" | 343
|
|-
| 130
| [[Kullanıcı:SLÖzlem|SLÖzlem]]
| align="center" | 343
|
|-
| 131
| [[Kullanıcı:S.orsev.ayşe kibaroğlu|S.orsev.ayşe kibaroğlu]]
| align="center" | 341
|
|-
| 132
| [[Kullanıcı:MelekArı|MelekArı]]
| align="center" | 336
|
|-
| 133
| [[Kullanıcı:Tegel|Tegel]]
| align="center" | 313
|
|-
| 134
| [[Kullanıcı:The cat|The cat]]
| align="center" | 311
|
|-
| 135
| [[Kullanıcı:SAL 378|SAL 378]]
| align="center" | 305
|
|-
| 136
| [[Kullanıcı:Saml livess 61|Saml livess 61]]
| align="center" | 297
|
|-
| 137
| [[Kullanıcı:Guljan|Guljan]]
| align="center" | 296
|
|-
| 138
| [[Kullanıcı:Kmlblg272|Kmlblg272]]
| align="center" | 292
|
|-
| 139
| [[Kullanıcı:SLDerya|SLDerya]]
| align="center" | 292
|
|-
| 140
| [[Kullanıcı:Srhat|Srhat]]
| align="center" | 285
|
|-
| 141
| [[Kullanıcı:2004onuralp|2004onuralp]]
| align="center" | 285
|
|-
| 142
| [[Kullanıcı:Swam pl|Swam pl]]
| align="center" | 284
|
|-
| 143
| [[Kullanıcı:Kml ibrahim|Kml ibrahim]]
| align="center" | 275
|
|-
| 144
| [[Kullanıcı:MSchoentgen|MSchoentgen]]
| align="center" | 274
|
|-
| 145
| [[Kullanıcı:Underfell Flowey|Underfell Flowey]]
| align="center" | 270
|
|-
| 146
| [[Kullanıcı:Dbl2010|Dbl2010]]
| align="center" | 264
|
|-
| 147
| [[Kullanıcı:Funda|Funda]]
| align="center" | 260
|
|-
| 148
| [[Kullanıcı:Kebab in my butt|Kebab in my butt]]
| align="center" | 260
|
|-
| 149
| [[Kullanıcı:Biyolojiyabikurdi|Biyolojiyabikurdi]]
| align="center" | 258
|
|-
| 150
| [[Kullanıcı:Arhancelo|Arhancelo]]
| align="center" | 256
|
|-
| 151
| [[Kullanıcı:SLnuray|SLnuray]]
| align="center" | 251
|
|-
| 152
| [[Kullanıcı:Gilthuanas|Gilthuanas]]
| align="center" | 250
|
|-
| 153
| [[Kullanıcı:Yorınçga573|Yorınçga573]]
| align="center" | 249
|
|-
| 154
| [[Kullanıcı:SL.resul akyıldız|SL.resul akyıldız]]
| align="center" | 247
|
|-
| 155
| [[Kullanıcı:SLözgül|SLözgül]]
| align="center" | 246
|
|-
| 156
| [[Kullanıcı:Azxpetro|Azxpetro]]
| align="center" | 244
|
|-
| 157
| [[Kullanıcı:Determinator|Determinator]]
| align="center" | 243
|
|-
| 158
| [[Kullanıcı:Серк.123|Серк.123]]
| align="center" | 243
|
|-
| 159
| [[Kullanıcı:Ceaeven|Ceaeven]]
| align="center" | 240
|
|-
| 160
| [[Kullanıcı:Sihl gülten deveci|Sihl gülten deveci]]
| align="center" | 237
|
|-
| 161
| [[Kullanıcı:Axbyc|Axbyc]]
| align="center" | 236
|
|-
| 162
| [[Kullanıcı:Superyetkin|Superyetkin]]
| align="center" | 235
| arayüz yöneticisi
|-
| 163
| [[Kullanıcı:Wikihez|Wikihez]]
| align="center" | 228
|
|-
| 164
| [[Kullanıcı:SLFatma|SLFatma]]
| align="center" | 227
|
|-
| 165
| [[Kullanıcı:Kmlblg240|Kmlblg240]]
| align="center" | 224
|
|-
| 166
| [[Kullanıcı:Erdemaslancan|Erdemaslancan]]
| align="center" | 224
|
|-
| 167
| [[Kullanıcı:Stml 172|Stml 172]]
| align="center" | 220
|
|-
| 168
| [[Kullanıcı:SLUğurcan|SLUğurcan]]
| align="center" | 219
|
|-
| 169
| [[Kullanıcı:Deknuydt|Deknuydt]]
| align="center" | 217
|
|-
| 170
| [[Kullanıcı:Kwamikagami|Kwamikagami]]
| align="center" | 212
|
|-
| 171
| [[Kullanıcı:Axis09|Axis09]]
| align="center" | 210
|
|-
| 172
| [[Kullanıcı:OSMAN ÇAVUŞLU|OSMAN ÇAVUŞLU]]
| align="center" | 209
|
|-
| 173
| [[Kullanıcı:Public class|Public class]]
| align="center" | 208
|
|-
| 174
| [[Kullanıcı:Aivasedo|Aivasedo]]
| align="center" | 205
|
|-
| 175
| [[Kullanıcı:Sal 390|Sal 390]]
| align="center" | 203
|
|-
| 176
| [[Kullanıcı:Onuralpateş2004|Onuralpateş2004]]
| align="center" | 202
|
|-
| 177
| [[Kullanıcı:Aslan yakup|Aslan yakup]]
| align="center" | 197
|
|-
| 178
| [[Kullanıcı:SLebru|SLebru]]
| align="center" | 197
|
|-
| 179
| [[Kullanıcı:Sal-zeynep|Sal-zeynep]]
| align="center" | 196
|
|-
| 180
| [[Kullanıcı:SLRahman|SLRahman]]
| align="center" | 194
|
|-
| 181
| [[Kullanıcı:Chapultepec|Chapultepec]]
| align="center" | 193
|
|-
| 182
| [[Kullanıcı:Trong Dang|Trong Dang]]
| align="center" | 192
|
|-
| 183
| [[Kullanıcı:Syum90|Syum90]]
| align="center" | 185
|
|-
| 184
| [[Kullanıcı:NURULLAH GÜLDAL|NURULLAH GÜLDAL]]
| align="center" | 183
|
|-
| 185
| [[Kullanıcı:Vikiçizer|Vikiçizer]]
| align="center" | 182
|
|-
| 186
| [[Kullanıcı:MehmedKM|MehmedKM]]
| align="center" | 182
|
|-
| 187
| [[Kullanıcı:Bikarhêner|Bikarhêner]]
| align="center" | 180
|
|-
| 188
| [[Kullanıcı:Şaşılık|Şaşılık]]
| align="center" | 179
|
|-
| 189
| [[Kullanıcı:Felecita|Felecita]]
| align="center" | 176
|
|-
| 190
| [[Kullanıcı:Ddenkel~trwiktionary|Ddenkel~trwiktionary]]
| align="center" | 175
|
|-
| 191
| [[Kullanıcı:Sezinumaykucuk|Sezinumaykucuk]]
| align="center" | 175
|
|-
| 192
| [[Kullanıcı:Sallapattias|Sallapattias]]
| align="center" | 171
|
|-
| 193
| [[Kullanıcı:SLburhantuncer|SLburhantuncer]]
| align="center" | 168
|
|-
| 194
| [[Kullanıcı:Bahattin|Bahattin]]
| align="center" | 168
|
|-
| 195
| [[Kullanıcı:Kml suleakguc|Kml suleakguc]]
| align="center" | 167
|
|-
| 196
| [[Kullanıcı:Sihl ASİYE|Sihl ASİYE]]
| align="center" | 167
|
|-
| 197
| [[Kullanıcı:Mehmetkasımay|Mehmetkasımay]]
| align="center" | 165
|
|-
| 198
| [[Kullanıcı:SLfatma254|SLfatma254]]
| align="center" | 163
|
|-
| 199
| [[Kullanıcı:Yko 235|Yko 235]]
| align="center" | 162
|
|-
| 200
| [[Kullanıcı:Stml 158|Stml 158]]
| align="center" | 159
|
|-
| 201
| [[Kullanıcı:Stml 121|Stml 121]]
| align="center" | 156
|
|-
| 202
| [[Kullanıcı:SLEngin|SLEngin]]
| align="center" | 154
|
|-
| 203
| [[Kullanıcı:Serdar62|Serdar62]]
| align="center" | 154
|
|-
| 204
| [[Kullanıcı:Terryclient|Terryclient]]
| align="center" | 151
|
|-
| 205
| [[Kullanıcı:Czeski Wratek|Czeski Wratek]]
| align="center" | 151
|
|-
| 206
| [[Kullanıcı:Calq|Calq]]
| align="center" | 148
|
|-
| 207
| [[Kullanıcı:Sihl Ramazan Erdöl|Sihl Ramazan Erdöl]]
| align="center" | 146
|
|-
| 208
| [[Kullanıcı:Drabdullayev17|Drabdullayev17]]
| align="center" | 140
|
|-
| 209
| [[Kullanıcı:Lionel Cristiano|Lionel Cristiano]]
| align="center" | 139
|
|-
| 210
| [[Kullanıcı:Hasan Sami|Hasan Sami]]
| align="center" | 138
|
|-
| 211
| [[Kullanıcı:Abuk SABUK|Abuk SABUK]]
| align="center" | 138
|
|-
| 212
| [[Kullanıcı:Dilci|Dilci]]
| align="center" | 137
|
|-
| 213
| [[Kullanıcı:Matiia|Matiia]]
| align="center" | 137
|
|-
| 214
| [[Kullanıcı:SİHL İSA ELER|SİHL İSA ELER]]
| align="center" | 135
|
|-
| 215
| [[Kullanıcı:Ata Kemal Ayaser|Ata Kemal Ayaser]]
| align="center" | 135
|
|-
| 216
| [[Kullanıcı:Emperyan|Emperyan]]
| align="center" | 135
|
|-
| 217
| [[Kullanıcı:Lou|Lou]]
| align="center" | 133
|
|-
| 218
| [[Kullanıcı:Rolby|Rolby]]
| align="center" | 133
|
|-
| 219
| [[Kullanıcı:Peter Berbe|Peter Berbe]]
| align="center" | 133
|
|-
| 220
| [[Kullanıcı:Eminovič|Eminovič]]
| align="center" | 132
|
|-
| 221
| [[Kullanıcı:Hdayetozsoy|Hdayetozsoy]]
| align="center" | 131
|
|-
| 222
| [[Kullanıcı:Fankibiber|Fankibiber]]
| align="center" | 129
|
|-
| 223
| [[Kullanıcı:GerardM|GerardM]]
| align="center" | 125
|
|-
| 224
| [[Kullanıcı:Ka dir 641|Ka dir 641]]
| align="center" | 125
|
|-
| 225
| [[Kullanıcı:Mesutbolukbas|Mesutbolukbas]]
| align="center" | 124
|
|-
| 226
| [[Kullanıcı:SLemine607 10D|SLemine607 10D]]
| align="center" | 122
|
|-
| 227
| [[Kullanıcı:Tarihçi Selo|Tarihçi Selo]]
| align="center" | 120
|
|-
| 228
| [[Kullanıcı:Lambiam|Lambiam]]
| align="center" | 120
|
|-
| 229
| [[Kullanıcı:Dodecaplex|Dodecaplex]]
| align="center" | 120
|
|-
| 230
| [[Kullanıcı:Vay~trwiktionary|Vay~trwiktionary]]
| align="center" | 119
|
|-
| 231
| [[Kullanıcı:Yavuz Atasever|Yavuz Atasever]]
| align="center" | 118
|
|-
| 232
| [[Kullanıcı:Metal Militia|Metal Militia]]
| align="center" | 117
|
|-
| 233
| [[Kullanıcı:Aarp65|Aarp65]]
| align="center" | 117
|
|-
| 234
| [[Kullanıcı:Eyüp Türküm|Eyüp Türküm]]
| align="center" | 117
|
|-
| 235
| [[Kullanıcı:Kalem.metin|Kalem.metin]]
| align="center" | 117
|
|-
| 236
| [[Kullanıcı:Mario The PS2 Guy|Mario The PS2 Guy]]
| align="center" | 117
|
|-
| 237
| [[Kullanıcı:Meyegon|Meyegon]]
| align="center" | 115
|
|-
| 238
| [[Kullanıcı:Cûndûllah el-Kurdî|Cûndûllah el-Kurdî]]
| align="center" | 115
|
|-
| 239
| [[Kullanıcı:Minorax|Minorax]]
| align="center" | 115
|
|-
| 240
| [[Kullanıcı:Baris6161TURK|Baris6161TURK]]
| align="center" | 113
|
|-
| 241
| [[Kullanıcı:Kml 207aysegül|Kml 207aysegül]]
| align="center" | 112
|
|-
| 242
| [[Kullanıcı:SLseda|SLseda]]
| align="center" | 111
|
|-
| 243
| [[Kullanıcı:Kmoksy|Kmoksy]]
| align="center" | 111
|
|-
| 244
| [[Kullanıcı:LesChloroformistes|LesChloroformistes]]
| align="center" | 110
|
|-
| 245
| [[Kullanıcı:Ladsgroup|Ladsgroup]]
| align="center" | 109
|
|-
| 246
| [[Kullanıcı:Müntesib|Müntesib]]
| align="center" | 109
|
|-
| 247
| [[Kullanıcı:Kmlrbabalik|Kmlrbabalik]]
| align="center" | 108
|
|-
| 248
| [[Kullanıcı:Kmlblg86|Kmlblg86]]
| align="center" | 108
|
|-
| 249
| [[Kullanıcı:S htk 117|S htk 117]]
| align="center" | 105
|
|-
| 250
| [[Kullanıcı:SLGamze|SLGamze]]
| align="center" | 105
|
|-
| 251
| [[Kullanıcı:Doğru Tercüman|Doğru Tercüman]]
| align="center" | 105
|
|-
| 252
| [[Kullanıcı:Küçükdere iöo gülay|Küçükdere iöo gülay]]
| align="center" | 104
|
|-
| 253
| [[Kullanıcı:Gülşah|Gülşah]]
| align="center" | 102
|
|-
| 254
| [[Kullanıcı:Balamax|Balamax]]
| align="center" | 102
|
|-
| 255
| [[Kullanıcı:Ormanbotanigi|Ormanbotanigi]]
| align="center" | 101
|
|-
| 256
| [[Kullanıcı:Stml 176|Stml 176]]
| align="center" | 101
|
|-
| 257
| [[Kullanıcı:Stml 165|Stml 165]]
| align="center" | 101
|
|-
| 258
| [[Kullanıcı:Kmlblg399|Kmlblg399]]
| align="center" | 100
|
|-
| 259
| [[Kullanıcı:Nerval|Nerval]]
| align="center" | 99
|
|-
| 260
| [[Kullanıcı:Stml 72|Stml 72]]
| align="center" | 98
|
|-
| 261
| [[Kullanıcı:Stryn|Stryn]]
| align="center" | 97
|
|-
| 262
| [[Kullanıcı:Küçükdere iöo sinan|Küçükdere iöo sinan]]
| align="center" | 96
|
|-
| 263
| [[Kullanıcı:Dünya vatandaşı|Dünya vatandaşı]]
| align="center" | 95
|
|-
| 264
| [[Kullanıcı:Muhammet yazıcı|Muhammet yazıcı]]
| align="center" | 95
|
|-
| 265
| [[Kullanıcı:SİHL 729|SİHL 729]]
| align="center" | 95
|
|-
| 266
| [[Kullanıcı:Xorasan|Xorasan]]
| align="center" | 95
|
|-
| 267
| [[Kullanıcı:Fiilçeker|Fiilçeker]]
| align="center" | 94
|
|-
| 268
| [[Kullanıcı:RadiX|RadiX]]
| align="center" | 94
|
|-
| 269
| [[Kullanıcı:Muallim Fatih|Muallim Fatih]]
| align="center" | 94
|
|-
| 270
| [[Kullanıcı:Marrovi|Marrovi]]
| align="center" | 94
|
|-
| 271
| [[Kullanıcı:Spacebirdy|Spacebirdy]]
| align="center" | 93
|
|-
| 272
| [[Kullanıcı:MhmtÖ|MhmtÖ]]
| align="center" | 93
|
|-
| 273
| [[Kullanıcı:Berr.in|Berr.in]]
| align="center" | 92
|
|-
| 274
| [[Kullanıcı:SL-busenur.523|SL-busenur.523]]
| align="center" | 91
|
|-
| 275
| [[Kullanıcı:SLİpek|SLİpek]]
| align="center" | 90
|
|-
| 276
| [[Kullanıcı:Kmlblgözlem|Kmlblgözlem]]
| align="center" | 89
|
|-
| 277
| [[Kullanıcı:Cavlak-FS|Cavlak-FS]]
| align="center" | 88
|
|-
| 278
| [[Kullanıcı:Azminwan|Azminwan]]
| align="center" | 87
|
|-
| 279
| [[Kullanıcı:Moonpulsar|Moonpulsar]]
| align="center" | 87
|
|-
| 280
| [[Kullanıcı:Stml 171|Stml 171]]
| align="center" | 86
|
|-
| 281
| [[Kullanıcı:Turkmen|Turkmen]]
| align="center" | 85
|
|-
| 282
| [[Kullanıcı:Xyz...|Xyz...]]
| align="center" | 84
|
|-
| 283
| [[Kullanıcı:Stml 117|Stml 117]]
| align="center" | 83
|
|-
| 284
| [[Kullanıcı:Juliancolton|Juliancolton]]
| align="center" | 81
|
|-
| 285
| [[Kullanıcı:Ludwig20|Ludwig20]]
| align="center" | 81
|
|-
| 286
| [[Kullanıcı:2004ateş|2004ateş]]
| align="center" | 81
|
|-
| 287
| [[Kullanıcı:Süleymanbedir|Süleymanbedir]]
| align="center" | 80
|
|-
| 288
| [[Kullanıcı:Kıdemli|Kıdemli]]
| align="center" | 80
|
|-
| 289
| [[Kullanıcı:Karabat Kuşu|Karabat Kuşu]]
| align="center" | 79
|
|-
| 290
| [[Kullanıcı:SAML VE EML CEMİL BAYRAM|SAML VE EML CEMİL BAYRAM]]
| align="center" | 79
|
|-
| 291
| [[Kullanıcı:Semlaml111|Semlaml111]]
| align="center" | 79
|
|-
| 292
| [[Kullanıcı:SEML Turan ÇAKIR|SEML Turan ÇAKIR]]
| align="center" | 78
|
|-
| 293
| [[Kullanıcı:Küçükdere iöo celilakyürek|Küçükdere iöo celilakyürek]]
| align="center" | 78
|
|-
| 294
| [[Kullanıcı:SLemine613|SLemine613]]
| align="center" | 78
|
|-
| 295
| [[Kullanıcı:Dokuz sekiz|Dokuz sekiz]]
| align="center" | 78
|
|-
| 296
| [[Kullanıcı:Sayonzei|Sayonzei]]
| align="center" | 78
|
|-
| 297
| [[Kullanıcı:İlhanbiskin|İlhanbiskin]]
| align="center" | 77
|
|-
| 298
| [[Kullanıcı:İsmailhoca|İsmailhoca]]
| align="center" | 77
|
|-
| 299
| [[Kullanıcı:Kmlblg80|Kmlblg80]]
| align="center" | 77
|
|-
| 300
| [[Kullanıcı:Masseman|Masseman]]
| align="center" | 76
|
|-
| 301
| [[Kullanıcı:Ridvanu|Ridvanu]]
| align="center" | 75
|
|-
| 302
| [[Kullanıcı:Ege yalcin~trwiktionary|Ege yalcin~trwiktionary]]
| align="center" | 75
|
|-
| 303
| [[Kullanıcı:İbrhm.kuru|İbrhm.kuru]]
| align="center" | 73
|
|-
| 304
| [[Kullanıcı:Mustafa Eşitgen|Mustafa Eşitgen]]
| align="center" | 73
|
|-
| 305
| [[Kullanıcı:Fagus|Fagus]]
| align="center" | 73
|
|-
| 306
| [[Kullanıcı:SL Pelin|SL Pelin]]
| align="center" | 72
|
|-
| 307
| [[Kullanıcı:Seninbey~trwiktionary|Seninbey~trwiktionary]]
| align="center" | 72
|
|-
| 308
| [[Kullanıcı:EmrahÖ|EmrahÖ]]
| align="center" | 72
|
|-
| 309
| [[Kullanıcı:BitikciKebbenek|BitikciKebbenek]]
| align="center" | 72
|
|-
| 310
| [[Kullanıcı:Emree4|Emree4]]
| align="center" | 71
|
|-
| 311
| [[Kullanıcı:Kontrolcu|Kontrolcu]]
| align="center" | 71
|
|-
| 312
| [[Kullanıcı:Nebra|Nebra]]
| align="center" | 71
|
|-
| 313
| [[Kullanıcı:Pembe karadeniz|Pembe karadeniz]]
| align="center" | 71
|
|-
| 314
| [[Kullanıcı:Stml 147|Stml 147]]
| align="center" | 70
|
|-
| 315
| [[Kullanıcı:SEMLNazımKURŞUN|SEMLNazımKURŞUN]]
| align="center" | 70
|
|-
| 316
| [[Kullanıcı:Giresunlu1993|Giresunlu1993]]
| align="center" | 70
|
|-
| 317
| [[Kullanıcı:Plenumchamber~trwiktionary|Plenumchamber~trwiktionary]]
| align="center" | 69
|
|-
| 318
| [[Kullanıcı:~2026-19042-98|~2026-19042-98]]
| align="center" | 69
|
|-
| 319
| [[Kullanıcı:Rock on She|Rock on She]]
| align="center" | 67
|
|-
| 320
| [[Kullanıcı:SLSerkan|SLSerkan]]
| align="center" | 65
|
|-
| 321
| [[Kullanıcı:Karduelis|Karduelis]]
| align="center" | 65
|
|-
| 322
| [[Kullanıcı:Zontollektuel|Zontollektuel]]
| align="center" | 65
|
|-
| 323
| [[Kullanıcı:Evrenkoruyan|Evrenkoruyan]]
| align="center" | 64
|
|-
| 324
| [[Kullanıcı:Adanalınacak|Adanalınacak]]
| align="center" | 64
|
|-
| 325
| [[Kullanıcı:Hapşu|Hapşu]]
| align="center" | 63
|
|-
| 326
| [[Kullanıcı:Dixtosa|Dixtosa]]
| align="center" | 63
|
|-
| 327
| [[Kullanıcı:El.wikt.user|El.wikt.user]]
| align="center" | 62
|
|-
| 328
| [[Kullanıcı:SL özge üstün|SL özge üstün]]
| align="center" | 62
|
|-
| 329
| [[Kullanıcı:Sal hasanilhan|Sal hasanilhan]]
| align="center" | 61
|
|-
| 330
| [[Kullanıcı:Prenses|Prenses]]
| align="center" | 60
|
|-
| 331
| [[Kullanıcı:Tofeiku|Tofeiku]]
| align="center" | 60
|
|-
| 332
| [[Kullanıcı:Sihl Murat|Sihl Murat]]
| align="center" | 59
|
|-
| 333
| [[Kullanıcı:Renamed user 1e23409a06e0b7922c2dfc98dde51974|Renamed user 1e23409a06e0b7922c2dfc98dde51974]]
| align="center" | 59
|
|-
| 334
| [[Kullanıcı:Arnavutsezgin|Arnavutsezgin]]
| align="center" | 58
|
|-
| 335
| [[Kullanıcı:Erkan Yilmaz|Erkan Yilmaz]]
| align="center" | 58
|
|-
| 336
| [[Kullanıcı:Galibibülent|Galibibülent]]
| align="center" | 57
|
|-
| 337
| [[Kullanıcı:S.orsev.dilek soydan|S.orsev.dilek soydan]]
| align="center" | 57
|
|-
| 338
| [[Kullanıcı:Stml 183|Stml 183]]
| align="center" | 57
|
|-
| 339
| [[Kullanıcı:Buzulkuşu|Buzulkuşu]]
| align="center" | 57
|
|-
| 340
| [[Kullanıcı:Spanier|Spanier]]
| align="center" | 57
|
|-
| 341
| [[Kullanıcı:LejyonerZaza|LejyonerZaza]]
| align="center" | 57
|
|-
| 342
| [[Kullanıcı:SAL 329|SAL 329]]
| align="center" | 56
|
|-
| 343
| [[Kullanıcı:S htk 428|S htk 428]]
| align="center" | 56
|
|-
| 344
| [[Kullanıcı:Stml 161|Stml 161]]
| align="center" | 56
|
|-
| 345
| [[Kullanıcı:HASAN ÖZGÜREN EML|HASAN ÖZGÜREN EML]]
| align="center" | 56
|
|-
| 346
| [[Kullanıcı:SLesra|SLesra]]
| align="center" | 56
|
|-
| 347
| [[Kullanıcı:SLhalit540|SLhalit540]]
| align="center" | 56
|
|-
| 348
| [[Kullanıcı:Stultiwikia|Stultiwikia]]
| align="center" | 56
|
|-
| 349
| [[Kullanıcı:LisafBia|LisafBia]]
| align="center" | 56
|
|-
| 350
| [[Kullanıcı:Quinlan83|Quinlan83]]
| align="center" | 56
|
|-
| 351
| [[Kullanıcı:Şegirt|Şegirt]]
| align="center" | 56
|
|-
| 352
| [[Kullanıcı:TAKASUGI Shinji|TAKASUGI Shinji]]
| align="center" | 55
|
|-
| 353
| [[Kullanıcı:Sarvaturi~trwiktionary|Sarvaturi~trwiktionary]]
| align="center" | 55
|
|-
| 354
| [[Kullanıcı:Ks.Tahd|Ks.Tahd]]
| align="center" | 55
|
|-
| 355
| [[Kullanıcı:Darkhorn~trwiktionary|Darkhorn~trwiktionary]]
| align="center" | 54
|
|-
| 356
| [[Kullanıcı:Anlztrk|Anlztrk]]
| align="center" | 54
|
|-
| 357
| [[Kullanıcı:Samleml özgür aydın|Samleml özgür aydın]]
| align="center" | 53
|
|-
| 358
| [[Kullanıcı:Elvire|Elvire]]
| align="center" | 53
|
|-
| 359
| [[Kullanıcı:İmmortalance|İmmortalance]]
| align="center" | 53
|
|-
| 360
| [[Kullanıcı:Tehonk|Tehonk]]
| align="center" | 53
|
|-
| 361
| [[Kullanıcı:NotBot|NotBot]]
| align="center" | 52
|
|-
| 362
| [[Kullanıcı:SHTK.ARZU|SHTK.ARZU]]
| align="center" | 52
|
|-
| 363
| [[Kullanıcı:SLOĞUZHAN|SLOĞUZHAN]]
| align="center" | 52
|
|-
| 364
| [[Kullanıcı:WikimediaNotifier|WikimediaNotifier]]
| align="center" | 52
|
|-
| 365
| [[Kullanıcı:Gufosowa|Gufosowa]]
| align="center" | 52
|
|-
| 366
| [[Kullanıcı:たまほめ|たまほめ]]
| align="center" | 51
|
|-
| 367
| [[Kullanıcı:Tembelejderha|Tembelejderha]]
| align="center" | 50
|
|-
| 368
| [[Kullanıcı:Kml Nurcan|Kml Nurcan]]
| align="center" | 50
|
|-
| 369
| [[Kullanıcı:Elton|Elton]]
| align="center" | 50
|
|-
| 370
| [[Kullanıcı:Ozgurhatic|Ozgurhatic]]
| align="center" | 50
|
|-
| 371
| [[Kullanıcı:Tıbbiye|Tıbbiye]]
| align="center" | 50
|
|-
| 372
| [[Kullanıcı:Sepkenki|Sepkenki]]
| align="center" | 50
|
|-
| 373
| [[Kullanıcı:Świętokrzyskie3|Świętokrzyskie3]]
| align="center" | 50
|
|-
| 374
| [[Kullanıcı:Stml 188|Stml 188]]
| align="center" | 48
|
|-
| 375
| [[Kullanıcı:Maderibeyza|Maderibeyza]]
| align="center" | 48
|
|-
| 376
| [[Kullanıcı:Müftülükarif|Müftülükarif]]
| align="center" | 48
|
|-
| 377
| [[Kullanıcı:Spl908455|Spl908455]]
| align="center" | 48
|
|-
| 378
| [[Kullanıcı:Gökçe Yörük|Gökçe Yörük]]
| align="center" | 48
|
|-
| 379
| [[Kullanıcı:Greywolf28|Greywolf28]]
| align="center" | 48
|
|-
| 380
| [[Kullanıcı:Sal ruveyda|Sal ruveyda]]
| align="center" | 47
|
|-
| 381
| [[Kullanıcı:S htk 145|S htk 145]]
| align="center" | 47
|
|-
| 382
| [[Kullanıcı:Cmrncntn|Cmrncntn]]
| align="center" | 47
|
|-
| 383
| [[Kullanıcı:Anon0004|Anon0004]]
| align="center" | 47
|
|-
| 384
| [[Kullanıcı:Doruk Salancı~trwiktionary|Doruk Salancı~trwiktionary]]
| align="center" | 46
|
|-
| 385
| [[Kullanıcı:Seml abdullah ege|Seml abdullah ege]]
| align="center" | 46
|
|-
| 386
| [[Kullanıcı:SACAK ALİFATİH|SACAK ALİFATİH]]
| align="center" | 46
|
|-
| 387
| [[Kullanıcı:Adml polat çelik 156|Adml polat çelik 156]]
| align="center" | 46
|
|-
| 388
| [[Kullanıcı:Luckas Blade|Luckas Blade]]
| align="center" | 46
|
|-
| 389
| [[Kullanıcı:BurstPower|BurstPower]]
| align="center" | 46
|
|-
| 390
| [[Kullanıcı:Lord Leatherface~trwiktionary|Lord Leatherface~trwiktionary]]
| align="center" | 45
|
|-
| 391
| [[Kullanıcı:Mambet|Mambet]]
| align="center" | 45
|
|-
| 392
| [[Kullanıcı:Öz Türkçe|Öz Türkçe]]
| align="center" | 45
|
|-
| 393
| [[Kullanıcı:JAn Dudík|JAn Dudík]]
| align="center" | 45
|
|-
| 394
| [[Kullanıcı:Billinghurst|Billinghurst]]
| align="center" | 45
|
|-
| 395
| [[Kullanıcı:Praxidicae|Praxidicae]]
| align="center" | 45
|
|-
| 396
| [[Kullanıcı:MathXplore|MathXplore]]
| align="center" | 45
|
|-
| 397
| [[Kullanıcı:Etimoloji|Etimoloji]]
| align="center" | 45
|
|-
| 398
| [[Kullanıcı:Zbutie3.14|Zbutie3.14]]
| align="center" | 45
|
|-
| 399
| [[Kullanıcı:Brightt11|Brightt11]]
| align="center" | 45
|
|-
| 400
| [[Kullanıcı:Sunshine23|Sunshine23]]
| align="center" | 44
|
|-
| 401
| [[Kullanıcı:BD2412|BD2412]]
| align="center" | 44
|
|-
| 402
| [[Kullanıcı:Wisar12|Wisar12]]
| align="center" | 44
|
|-
| 403
| [[Kullanıcı:Turkittihadcemiyeti|Turkittihadcemiyeti]]
| align="center" | 44
|
|-
| 404
| [[Kullanıcı:Hoo man|Hoo man]]
| align="center" | 43
|
|-
| 405
| [[Kullanıcı:Ghybu|Ghybu]]
| align="center" | 43
|
|-
| 406
| [[Kullanıcı:Zolgoyo|Zolgoyo]]
| align="center" | 43
|
|-
| 407
| [[Kullanıcı:YBK4|YBK4]]
| align="center" | 43
|
|-
| 408
| [[Kullanıcı:Aktas5561|Aktas5561]]
| align="center" | 42
|
|-
| 409
| [[Kullanıcı:Wiki13|Wiki13]]
| align="center" | 42
|
|-
| 410
| [[Kullanıcı:Alexander Mikhalenko|Alexander Mikhalenko]]
| align="center" | 42
|
|-
| 411
| [[Kullanıcı:Orsev.i.o.ayşe kibaroğlu|Orsev.i.o.ayşe kibaroğlu]]
| align="center" | 41
|
|-
| 412
| [[Kullanıcı:Stml 168|Stml 168]]
| align="center" | 41
|
|-
| 413
| [[Kullanıcı:Sihl selim|Sihl selim]]
| align="center" | 41
|
|-
| 414
| [[Kullanıcı:V. Doğan Günay~trwiktionary|V. Doğan Günay~trwiktionary]]
| align="center" | 41
|
|-
| 415
| [[Kullanıcı:Filanca|Filanca]]
| align="center" | 41
|
|-
| 416
| [[Kullanıcı:Nurettincerek|Nurettincerek]]
| align="center" | 41
|
|-
| 417
| [[Kullanıcı:Umitduranist|Umitduranist]]
| align="center" | 41
|
|-
| 418
| [[Kullanıcı:Wentayan|Wentayan]]
| align="center" | 41
|
|-
| 419
| [[Kullanıcı:Abant12|Abant12]]
| align="center" | 41
|
|-
| 420
| [[Kullanıcı:Ahmetcatalkaya|Ahmetcatalkaya]]
| align="center" | 40
|
|-
| 421
| [[Kullanıcı:Ahulusi|Ahulusi]]
| align="center" | 40
|
|-
| 422
| [[Kullanıcı:Savh|Savh]]
| align="center" | 40
|
|-
| 423
| [[Kullanıcı:FocalPoint|FocalPoint]]
| align="center" | 40
|
|-
| 424
| [[Kullanıcı:Bulgu|Bulgu]]
| align="center" | 40
|
|-
| 425
| [[Kullanıcı:Dijan|Dijan]]
| align="center" | 39
|
|-
| 426
| [[Kullanıcı:Saçak fatmamandirali|Saçak fatmamandirali]]
| align="center" | 39
|
|-
| 427
| [[Kullanıcı:Yabancı|Yabancı]]
| align="center" | 39
|
|-
| 428
| [[Kullanıcı:Cekli829|Cekli829]]
| align="center" | 39
|
|-
| 429
| [[Kullanıcı:Vanished user Xorisdtbdfgonugyfs|Vanished user Xorisdtbdfgonugyfs]]
| align="center" | 39
|
|-
| 430
| [[Kullanıcı:Mathonius|Mathonius]]
| align="center" | 39
|
|-
| 431
| [[Kullanıcı:Vincent Vega|Vincent Vega]]
| align="center" | 39
|
|-
| 432
| [[Kullanıcı:Samuele2002|Samuele2002]]
| align="center" | 39
|
|-
| 433
| [[Kullanıcı:CanerDemirci281993|CanerDemirci281993]]
| align="center" | 39
|
|-
| 434
| [[Kullanıcı:Eusbarbosa|Eusbarbosa]]
| align="center" | 38
|
|-
| 435
| [[Kullanıcı:Açak osman|Açak osman]]
| align="center" | 38
|
|-
| 436
| [[Kullanıcı:Araklı|Araklı]]
| align="center" | 38
|
|-
| 437
| [[Kullanıcı:Oylm mat|Oylm mat]]
| align="center" | 37
|
|-
| 438
| [[Kullanıcı:Yalhi|Yalhi]]
| align="center" | 37
|
|-
| 439
| [[Kullanıcı:Admlmuhammet alican|Admlmuhammet alican]]
| align="center" | 37
|
|-
| 440
| [[Kullanıcı:SLHakan|SLHakan]]
| align="center" | 37
|
|-
| 441
| [[Kullanıcı:JackPotte|JackPotte]]
| align="center" | 37
|
|-
| 442
| [[Kullanıcı:Onkaimeon|Onkaimeon]]
| align="center" | 37
|
|-
| 443
| [[Kullanıcı:Vanished user 127237|Vanished user 127237]]
| align="center" | 37
|
|-
| 444
| [[Kullanıcı:Aabdullayev851|Aabdullayev851]]
| align="center" | 37
|
|-
| 445
| [[Kullanıcı:Wooze|Wooze]]
| align="center" | 37
|
|-
| 446
| [[Kullanıcı:Intolerable situation|Intolerable situation]]
| align="center" | 37
|
|-
| 447
| [[Kullanıcı:Tapio Toola|Tapio Toola]]
| align="center" | 37
|
|-
| 448
| [[Kullanıcı:Ykio emelaydın|Ykio emelaydın]]
| align="center" | 36
|
|-
| 449
| [[Kullanıcı:S htk 612|S htk 612]]
| align="center" | 36
|
|-
| 450
| [[Kullanıcı:Flyax|Flyax]]
| align="center" | 36
|
|-
| 451
| [[Kullanıcı:Manco Capac|Manco Capac]]
| align="center" | 36
|
|-
| 452
| [[Kullanıcı:Joseph|Joseph]]
| align="center" | 36
|
|-
| 453
| [[Kullanıcı:Zuzumaykut|Zuzumaykut]]
| align="center" | 36
|
|-
| 454
| [[Kullanıcı:Goktr001|Goktr001]]
| align="center" | 36
|
|-
| 455
| [[Kullanıcı:Rxy|Rxy]]
| align="center" | 36
|
|-
| 456
| [[Kullanıcı:MPF|MPF]]
| align="center" | 36
|
|-
| 457
| [[Kullanıcı:Alperen|Alperen]]
| align="center" | 35
|
|-
| 458
| [[Kullanıcı:S htk 186|S htk 186]]
| align="center" | 35
|
|-
| 459
| [[Kullanıcı:Yonca573|Yonca573]]
| align="center" | 35
|
|-
| 460
| [[Kullanıcı:Pinar|Pinar]]
| align="center" | 34
|
|-
| 461
| [[Kullanıcı:Mardetanha|Mardetanha]]
| align="center" | 34
|
|-
| 462
| [[Kullanıcı:Nanahuatl|Nanahuatl]]
| align="center" | 34
|
|-
| 463
| [[Kullanıcı:Yoldasso|Yoldasso]]
| align="center" | 34
|
|-
| 464
| [[Kullanıcı:Java2002|Java2002]]
| align="center" | 34
|
|-
| 465
| [[Kullanıcı:Hasley|Hasley]]
| align="center" | 34
|
|-
| 466
| [[Kullanıcı:Recep Arslanbaş|Recep Arslanbaş]]
| align="center" | 34
|
|-
| 467
| [[Kullanıcı:TugbekOlek|TugbekOlek]]
| align="center" | 33
|
|-
| 468
| [[Kullanıcı:Vese~trwiktionary|Vese~trwiktionary]]
| align="center" | 33
|
|-
| 469
| [[Kullanıcı:Ozyurekli|Ozyurekli]]
| align="center" | 33
|
|-
| 470
| [[Kullanıcı:EsenBoga|EsenBoga]]
| align="center" | 33
|
|-
| 471
| [[Kullanıcı:Golgelerim|Golgelerim]]
| align="center" | 33
|
|-
| 472
| [[Kullanıcı:Sanya3|Sanya3]]
| align="center" | 33
|
|-
| 473
| [[Kullanıcı:Kırmızı renk|Kırmızı renk]]
| align="center" | 33
|
|-
| 474
| [[Kullanıcı:Stml 153|Stml 153]]
| align="center" | 32
|
|-
| 475
| [[Kullanıcı:SLsoner|SLsoner]]
| align="center" | 32
|
|-
| 476
| [[Kullanıcı:KoreanQuoter|KoreanQuoter]]
| align="center" | 32
|
|-
| 477
| [[Kullanıcı:Sungurer|Sungurer]]
| align="center" | 32
|
|-
| 478
| [[Kullanıcı:HestNewDes|HestNewDes]]
| align="center" | 32
|
|-
| 479
| [[Kullanıcı:Sml-hanife1|Sml-hanife1]]
| align="center" | 31
|
|-
| 480
| [[Kullanıcı:SAL 417|SAL 417]]
| align="center" | 31
|
|-
| 481
| [[Kullanıcı:MercanX|MercanX]]
| align="center" | 31
|
|-
| 482
| [[Kullanıcı:Kokoloc|Kokoloc]]
| align="center" | 31
|
|-
| 483
| [[Kullanıcı:Köksarı|Köksarı]]
| align="center" | 30
|
|-
| 484
| [[Kullanıcı:Akkiz|Akkiz]]
| align="center" | 30
|
|-
| 485
| [[Kullanıcı:Mclovinx|Mclovinx]]
| align="center" | 30
|
|-
| 486
| [[Kullanıcı:Talha1481|Talha1481]]
| align="center" | 30
|
|-
| 487
| [[Kullanıcı:Kumanof|Kumanof]]
| align="center" | 30
|
|-
| 488
| [[Kullanıcı:Tenperver|Tenperver]]
| align="center" | 30
|
|-
| 489
| [[Kullanıcı:~2026-22795-35|~2026-22795-35]]
| align="center" | 30
|
|-
| 490
| [[Kullanıcı:Sı 643|Sı 643]]
| align="center" | 29
|
|-
| 491
| [[Kullanıcı:SAMLEML MURAT ÇEBİ|SAMLEML MURAT ÇEBİ]]
| align="center" | 29
|
|-
| 492
| [[Kullanıcı:Kmlblg318|Kmlblg318]]
| align="center" | 29
|
|-
| 493
| [[Kullanıcı:Merube 89|Merube 89]]
| align="center" | 29
|
|-
| 494
| [[Kullanıcı:Hgav|Hgav]]
| align="center" | 29
|
|-
| 495
| [[Kullanıcı:Alsace38|Alsace38]]
| align="center" | 29
|
|-
| 496
| [[Kullanıcı:Dilşo Kardî|Dilşo Kardî]]
| align="center" | 29
|
|-
| 497
| [[Kullanıcı:Ulukan|Ulukan]]
| align="center" | 29
|
|-
| 498
| [[Kullanıcı:Kaan2121|Kaan2121]]
| align="center" | 29
|
|-
| 499
| [[Kullanıcı:Selahattin ilhan|Selahattin ilhan]]
| align="center" | 29
|
|-
| 500
| [[Kullanıcı:Zartatue|Zartatue]]
| align="center" | 29
|
|}
</center>
kqvc5kc410469oopjleymsynwqmr4hi
бездетность
0
539966
5669928
4024567
2026-06-23T08:07:32Z
MustafaCavlak
59368
/* Rusça */
5669928
wikitext
text/x-wiki
==Rusça==
===Köken===
{{ek|dil=ru|безде́тный|ость}}
===Ad===
{{başlık başı|ru|ad|baş=безде́тность|c=d}}
# [[çocuksuzluk]]
====Çekimleme====
{{ru-ad-3da|безде́тность}}
7eygytyjoyhntcgztjweblm0bl8nyst
Մուրադյան
0
641566
5669820
4591614
2026-06-22T17:15:36Z
MustafaCavlak
59368
/* Ermenice */
5669820
wikitext
text/x-wiki
==Ermenice==
{{vikipedi|dil=hy}}
===Köken===
{{ek|dil=hy|Մուրադ|յան}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
sj1c1wmcjuyvo0k0i602q2b5xsifkwx
magnaient
0
1131745
5669905
5131504
2026-06-22T21:33:08Z
Turgut46
26665
/* Fransızca */ Test.
5669905
wikitext
text/x-wiki
== Fransızca ==
===Eylem===
{{başlık başı|fr|çekimli eylem}}
# {{fr-er-çekimi}}
okfo03kga04w7hbz6qppldygp4ax7ta
ατσαλάκωτος
0
1302656
5669783
5355268
2026-06-22T13:34:31Z
MustafaCavlak
59368
/* Ön ad */
5669783
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# buruşmamış, buruşturulmamış
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ατσαλάκωτ}}
45b6ciby9poay0czhhbg59b3lz8yonc
ατσούγκριστος
0
1302665
5669782
5355277
2026-06-22T13:32:35Z
MustafaCavlak
59368
/* Ön ad */
5669782
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# tokuşturulmamış
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ατσούγκριστ}}
4oux8f2nqg16w57nvyxyu9so97ii6hy
διάδοχος
0
1303292
5669771
5355905
2026-06-22T13:15:38Z
MustafaCavlak
59368
/* Ön ad */
5669771
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[ardıl]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=διάδοχ}}
a9jbh4kj1klqf3vji0x9wbq3kqprbk4
ελευθερόβουλος
0
1303675
5669785
5356289
2026-06-22T13:38:22Z
MustafaCavlak
59368
/* Ön ad */
5669785
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[özgür]] [[iradeli]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ελευθερόβουλ}}
s44uwjk68v2vq58v3e6e077fchkz241
κατάφυτος
0
1304612
5669775
5357226
2026-06-22T13:23:39Z
MustafaCavlak
59368
/* Ön ad */
5669775
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[yeşillik]]le dolu, [[bitki]]yle dolu (yer); [[ağaçlık]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=κατάφυτ}}
mmalxkdlcngc5xgx477ovfv14mricgt
ολόλαμπρος
0
1305534
5669767
5358148
2026-06-22T13:13:27Z
MustafaCavlak
59368
/* Ön ad */
5669767
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[apaydın]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ολόλαμπρ}}
rcu3hjuy8wnuis2gdniln6ukc1p6mh0
ολοφάνερος
0
1305556
5669769
5358170
2026-06-22T13:14:17Z
MustafaCavlak
59368
/* Ön ad */
5669769
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[apaçık]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ολοφάνερ}}
4cbtmj1yvckmxrbhvnti4642zrolfky
ολόφωτος
0
1305557
5669768
5358171
2026-06-22T13:13:49Z
MustafaCavlak
59368
5669768
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[apaydın]]
====Çekimleme====
{{el-çekim-öa|çekim=ος-η-ο|kök=ολόφωτ}}
23ggq0zhc5w79c3a50cyabmrb49lbzl
garantiulo
0
1589089
5669758
2026-06-22T13:00:50Z
MustafaCavlak
59368
Yeni sayfa : ==Esperanto dili== ===Ad=== {{eo-ad}} # [[rehine]] ===Ek okumalar=== * {{R:eo:PIV 2020}}
5669758
wikitext
text/x-wiki
==Esperanto dili==
===Ad===
{{eo-ad}}
# [[rehine]]
===Ek okumalar===
* {{R:eo:PIV 2020}}
j343reg9vfy7varukdew1rmie1fzj4w
όμηρος
0
1589090
5669759
2026-06-22T13:03:50Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|e|c2=d|όμηροι}} # [[rehine]]
5669759
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|e|c2=d|όμηροι}}
# [[rehine]]
4n67px70d7fvy6evnpdql2026dttgrv
τήξη
0
1589091
5669760
2026-06-22T13:05:11Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[erime]]
5669760
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[erime]]
aft0iihsjuab6d63rnuxwcorzjjtqr7
τετράδα
0
1589092
5669761
2026-06-22T13:07:38Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[dörtlü]]
5669761
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[dörtlü]]
2acbds2z6dogyuynz9my5e601vhggzx
τζελατίνα
0
1589093
5669762
2026-06-22T13:09:18Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # {{farklı|dil=el|ζελατίνα}}
5669762
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# {{farklı|dil=el|ζελατίνα}}
5ny8f207z1eiq8hxju4spzz8zk19h6h
ζελατίνα
0
1589094
5669763
2026-06-22T13:09:36Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[jelatin]]
5669763
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[jelatin]]
8504r5n5a0vtm49p9k04fkkurjblg8h
αντίσωμα
0
1589095
5669764
2026-06-22T13:11:21Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[antikor]]
5669764
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[antikor]]
j214rc7mowzno7byaqm16rjkb6kuvd1
ερευνητής
0
1589096
5669765
2026-06-22T13:12:10Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|e}} # [[araştırmacı]]
5669765
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|e}}
# [[araştırmacı]]
eo2ar2rwqheimwo70zokom01eqkisrn
ερευνήτρια
0
1589097
5669766
2026-06-22T13:12:57Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # {{dişil karşılığı|el|ερευνητής}}
5669766
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# {{dişil karşılığı|el|ερευνητής}}
43h2h15elfbn9ra5kmi3gbzy8yb1fc2
βλακεία
0
1589098
5669770
2026-06-22T13:14:42Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[aptallık]]
5669770
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[aptallık]]
ez9m8d0pe609uula0rpjy1yh3yyz4y9
κυψέλη
0
1589099
5669772
2026-06-22T13:16:26Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[arı kovanı]]
5669772
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[arı kovanı]]
98uoe3vfkb55h68m4admhw7dvhb14pn
αλισίβα
0
1589100
5669773
2026-06-22T13:17:59Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[Arap sabunu]]
5669773
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[Arap sabunu]]
bic77naab5r18i3i6ao4yj8ponfz4p8
5669774
5669773
2026-06-22T13:20:30Z
MustafaCavlak
59368
/* Ad */
5669774
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[kül suyu]]
re7hscpukzbxgbyqqlss4l9mryswyul
ψιθύρισμα
0
1589101
5669776
2026-06-22T13:25:02Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[fısıldama]], [[fısıltı]]
5669776
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[fısıldama]], [[fısıltı]]
8wkhx2ii7aser5kb5h3ftzwrxsnsoip
φωτόμετρο
0
1589102
5669777
2026-06-22T13:26:04Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[ışıkölçer]]
5669777
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[ışıkölçer]]
7zhh03u8rebgy7qkzmwdcyb3gx4612c
χάπι
0
1589103
5669778
2026-06-22T13:26:37Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[hap]]
5669778
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[hap]]
gioyqvu3a9tsa76tni0yhaerexjba3d
θύσανος
0
1589104
5669779
2026-06-22T13:27:21Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|e}} # [[püskül]]
5669779
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|e}}
# [[püskül]]
7mz5dn1hfofk2e1tlowv3nhaq51jl7d
παραβλέπω
0
1589105
5669780
2026-06-22T13:28:28Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # [[görmezlikten gelmek]]
5669780
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# [[görmezlikten gelmek]]
fnm07i1ysv82niy0utlxkhuj4wveejo
τσουγκρίζω
0
1589106
5669781
2026-06-22T13:31:53Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # [[tokuşturmak]]
5669781
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# [[tokuşturmak]]
9oc3b56m68zarnf6511h8yapmjixy5o
τσαλακωμένος
0
1589107
5669784
2026-06-22T13:35:58Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ortaç=== {{el-ortaç}} # {{el-ortaç çekimi|τσαλακώνω}}
5669784
wikitext
text/x-wiki
==Yunanca==
===Ortaç===
{{el-ortaç}}
# {{el-ortaç çekimi|τσαλακώνω}}
bsaeluu8h95uu0av97ngn0i2v8zatkl
ιωδισμός
0
1589108
5669786
2026-06-22T13:41:42Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|e}} # [[iyot]] [[zehirlenme]]si
5669786
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|e}}
# [[iyot]] [[zehirlenme]]si
8drjo3a388ydzm84g5x5ruxe3o0jvm4
ιώδιο
0
1589109
5669787
2026-06-22T13:43:26Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[iyot]]
5669787
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[iyot]]
ea6m214ep2bomdoojuolzkddzi9gtcd
ιωδιούχος
0
1589110
5669788
2026-06-22T13:44:53Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ön ad=== {{el-ön ad}} # [[iyotlu]]
5669788
wikitext
text/x-wiki
==Yunanca==
===Ön ad===
{{el-ön ad}}
# [[iyotlu]]
e3j1slq3ppmqljgszmuzcdp6fsx1jz8
ατομικός αριθμός
0
1589111
5669789
2026-06-22T13:46:43Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|e}} # [[atom numarası]]
5669789
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|e}}
# [[atom numarası]]
scw2ka5tmyhl61hzxykw5hrxes7dobu
περπάτημα
0
1589112
5669790
2026-06-22T13:57:25Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[yürüme]], [[yürüyüş]]
5669790
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[yürüme]], [[yürüyüş]]
28hzcql7vdebh0dfz27r4cu2x5pm7n4
βάδισμα
0
1589113
5669791
2026-06-22T13:58:46Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[yürüyüş]]
5669791
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[yürüyüş]]
3ya1pk99lkyvdmnwrkpmch5eh0znah4
σπουδή
0
1589114
5669792
2026-06-22T14:00:05Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[çalışma]]
5669792
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[çalışma]]
65wocg34xln9027vhosmufz4f3iypul
σπουδαστήριο
0
1589115
5669793
2026-06-22T14:00:53Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|n}} # [[çalışma odası]]
5669793
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[çalışma odası]]
nitalu3htv8w4m5hejevlda5bkuewcw
5669794
5669793
2026-06-22T14:01:20Z
MustafaCavlak
59368
/* Ad */
5669794
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|n}}
# [[çalışma odası]], [[etüt odası]]
khsp846ogxrwe1jbq6i816q5u04avf9
εγκλείω
0
1589116
5669795
2026-06-22T14:02:54Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # [[kapatmak]], [[hapsetmek]]
5669795
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# [[kapatmak]], [[hapsetmek]]
geea60pouc0t9fwpjfl70pclvm9k3ga
περιορίζω
0
1589117
5669796
2026-06-22T14:04:06Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # [[kısıtlamak]], [[sınırlamak]]
5669796
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# [[kısıtlamak]], [[sınırlamak]]
osjlibn9l5b7qyjsrq7l1g5p9v4kmkl
έκθεση
0
1589118
5669797
2026-06-22T14:07:17Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Ad=== {{el-ad|d}} # [[sergi]], [[sergileme]], [[fuar]] # [[rapor]]
5669797
wikitext
text/x-wiki
==Yunanca==
===Ad===
{{el-ad|d}}
# [[sergi]], [[sergileme]], [[fuar]]
# [[rapor]]
mts8qpsd0sm5d7o0lmx5cmzg4333a1r
ποικίλλω
0
1589119
5669798
2026-06-22T14:12:27Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # {{t|dil=el|geçişli}} [[süslemek]], [[bezemek]] # {{t|dil=el|geçişli}} [[çeşitlendirmek]] # {{t|dil=el|geçişsiz}} [[çeşitlilik]] [[göstermek]], [[çeşitlenmek]] #: {{ux|el|Οι χαρακτήρες των ανθρώπων '''ποικίλλουν'''. Δεν είμαστε όλοι ίδιοι.|İnsanların karakterleri çeşitlilik gösterir. Hepimiz aynı değiliz.}}
5669798
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# {{t|dil=el|geçişli}} [[süslemek]], [[bezemek]]
# {{t|dil=el|geçişli}} [[çeşitlendirmek]]
# {{t|dil=el|geçişsiz}} [[çeşitlilik]] [[göstermek]], [[çeşitlenmek]]
#: {{ux|el|Οι χαρακτήρες των ανθρώπων '''ποικίλλουν'''. Δεν είμαστε όλοι ίδιοι.|İnsanların karakterleri çeşitlilik gösterir. Hepimiz aynı değiliz.}}
b6nx14bgdhromoucw5y2erfsco8hbix
konstipasyonizm
0
1589120
5669799
2026-06-22T14:16:59Z
~2026-36094-38
68142
Yeni sayfa : ==Türkçe== ===Ad=== {{tr-ad}} # bir süreci kasten geciktirme, engelleme veya kötüye kullanma uygulaması ([[Latince]] constipatio - engelleme, tıkama) ====Köken==== # {{k|dil=tr|fr}} {{ç|fr|constipationnisme}} ({{ek|dil=tr|konstipasyon|izm}})
5669799
wikitext
text/x-wiki
==Türkçe==
===Ad===
{{tr-ad}}
# bir süreci kasten geciktirme, engelleme veya kötüye kullanma uygulaması ([[Latince]] constipatio - engelleme, tıkama)
====Köken====
# {{k|dil=tr|fr}} {{ç|fr|constipationnisme}} ({{ek|dil=tr|konstipasyon|izm}})
gxwqxxeieh3ykd1rmbejnpp6d58ut9q
konflagrasyonizm
0
1589121
5669800
2026-06-22T14:24:01Z
~2026-36094-38
68142
Yeni sayfa : ==Türkçe== ===Ad=== {{tr-ad}} # kötü amaçlarla herhangi bir malı, binayı veya araziyi ateşe verme, kundakçılık ([[Latince]] conflagratio - büyük yangın) ====Köken==== # {{k|dil=tr|fr}} {{ç|fr|conflagrationnisme}} ({{ek|dil=tr|konflagrasyon|izm}})
5669800
wikitext
text/x-wiki
==Türkçe==
===Ad===
{{tr-ad}}
# kötü amaçlarla herhangi bir malı, binayı veya araziyi ateşe verme, kundakçılık ([[Latince]] conflagratio - büyük yangın)
====Köken====
# {{k|dil=tr|fr}} {{ç|fr|conflagrationnisme}} ({{ek|dil=tr|konflagrasyon|izm}})
47xdq2ikl3op00ifo9fcgdnexnxopmb
konsorsiyonizm
0
1589122
5669801
2026-06-22T14:35:06Z
~2026-36094-38
68142
Yeni sayfa : ==Türkçe== ===Ad=== {{tr-ad}} # düşmanla işbirliği ([[Latince]] consortio - gizli anlaşma) ====Köken==== # {{k|dil=tr|fr}} {{ç|fr|consortionnisme}} ({{ek|dil=tr|konsorsiyon|izm}})
5669801
wikitext
text/x-wiki
==Türkçe==
===Ad===
{{tr-ad}}
# düşmanla işbirliği ([[Latince]] consortio - gizli anlaşma)
====Köken====
# {{k|dil=tr|fr}} {{ç|fr|consortionnisme}} ({{ek|dil=tr|konsorsiyon|izm}})
eaqr7lnf2g3gzzvwa9twwy6uwcyrvyc
υποβάλλω
0
1589123
5669802
2026-06-22T15:32:35Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{el-eylem}} # [[sunmak]], [[arz etmek]] # [[maruz bırakmak]] # [[aklına sokmak]], [[zorlamak]], [[empoze etmek]]
5669802
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{el-eylem}}
# [[sunmak]], [[arz etmek]]
# [[maruz bırakmak]]
# [[aklına sokmak]], [[zorlamak]], [[empoze etmek]]
422yjpdddbd43tqzeh0ovlsgjzm8mv0
υποβάλει
0
1589124
5669803
2026-06-22T15:33:15Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{başlık başı|el|çekimli eylem}} # {{çekim|dil=el|υποβάλλω||act|bitimsiz|;|3t|bağımlı}}
5669803
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{başlık başı|el|çekimli eylem}}
# {{çekim|dil=el|υποβάλλω||act|bitimsiz|;|3t|bağımlı}}
0gop4591t4b3dxkmf0sgdghp69i9o6o
υποδουλώσει
0
1589125
5669804
2026-06-22T15:34:12Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{başlık başı|el|çekimli eylem}} # {{çekim|dil=el|υποδουλώνω||act|bitimsiz|;|3t|bağımlı}}
5669804
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{başlık başı|el|çekimli eylem}}
# {{çekim|dil=el|υποδουλώνω||act|bitimsiz|;|3t|bağımlı}}
31495fij0y2l6blyf82ry21pv1k5tp0
νυστάξει
0
1589126
5669805
2026-06-22T15:35:36Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{başlık başı|el|çekimli eylem}} # {{çekim|dil=el|νυστάζω||act|bitimsiz|;|3t|bağımlı}}
5669805
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{başlık başı|el|çekimli eylem}}
# {{çekim|dil=el|νυστάζω||act|bitimsiz|;|3t|bağımlı}}
2trvwkgf4ojfkdp8kp78ldwb1dcz4lb
προχωρήσει
0
1589127
5669806
2026-06-22T16:02:18Z
MustafaCavlak
59368
Yeni sayfa : ==Yunanca== ===Eylem=== {{başlık başı|el|çekimli eylem}} # {{çekim|dil=el|[[προχωράω]] / [[προχωρώ]]||act|bitimsiz|;|3t|bağımlı}}
5669806
wikitext
text/x-wiki
==Yunanca==
===Eylem===
{{başlık başı|el|çekimli eylem}}
# {{çekim|dil=el|[[προχωράω]] / [[προχωρώ]]||act|bitimsiz|;|3t|bağımlı}}
5utnfuzp6n4lkva5h705jlntl4f1q9w
կուղբ
0
1589128
5669807
2026-06-22T16:22:47Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[kunduz]]
5669807
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[kunduz]]
dnibffrqsiyohdiwvkddpgkm1p4asfn
դրամ
0
1589129
5669808
2026-06-22T16:27:45Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # {{t|dil=hy|para}} [[dram]]
5669808
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# {{t|dil=hy|para}} [[dram]]
eq70bbkkbmo70qdl8h84o7a4hxtng5a
Kategori:Ermenicede para
14
1589130
5669809
2026-06-22T16:29:16Z
MustafaCavlak
59368
Yeni sayfa : {{kategori konu|dil=hy|konu=para}}
5669809
wikitext
text/x-wiki
{{kategori konu|dil=hy|konu=para}}
8296hffw05c9czqvejqs9h5f1iuplq3
դրամա
0
1589131
5669810
2026-06-22T16:30:33Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[drama]]
5669810
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[drama]]
4hjnqvgtchzesogsiqhwgbleh7acc5y
երևանյան
0
1589132
5669811
2026-06-22T16:42:12Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ön ad=== {{başlık başı|hy|ön ad}} # [[Erivan]] ile ilgili
5669811
wikitext
text/x-wiki
==Ermenice==
===Ön ad===
{{başlık başı|hy|ön ad}}
# [[Erivan]] ile ilgili
mxcqxk8vvgeebpxknil2sbf4st5zxt4
ահ
0
1589133
5669812
2026-06-22T16:44:18Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[korku]], [[dehşet]]
5669812
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[korku]], [[dehşet]]
d9r42jfla2zeltju2w7bp7mkwa5xom7
սարսափ
0
1589134
5669813
2026-06-22T16:45:43Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[korku]], [[dehşet]]
5669813
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[korku]], [[dehşet]]
d9r42jfla2zeltju2w7bp7mkwa5xom7
երկյուղ
0
1589135
5669814
2026-06-22T16:49:03Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[korku]], [[dehşet]]
5669814
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[korku]], [[dehşet]]
d9r42jfla2zeltju2w7bp7mkwa5xom7
երկյուղել
0
1589136
5669815
2026-06-22T16:52:08Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Eylem=== {{hy-eylem}} # [[korkmak]], [[ürkmek]]
5669815
wikitext
text/x-wiki
==Ermenice==
===Eylem===
{{hy-eylem}}
# [[korkmak]], [[ürkmek]]
qrdsxwxakf1az3pwld7cxeboqoop4ru
Վիքիբառարան
0
1589137
5669816
2026-06-22T16:56:29Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Özel ad=== {{hy-özel ad}} # [[Vikisözlük]]
5669816
wikitext
text/x-wiki
==Ermenice==
===Özel ad===
{{hy-özel ad}}
# [[Vikisözlük]]
nyndv61u1ub4es5p61n14na45tsa9x0
բառարանագիր
0
1589138
5669817
2026-06-22T16:59:01Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[sözlükçü]], [[leksikograf]]
5669817
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[sözlükçü]], [[leksikograf]]
6yj575akljsh0nlks1ummfju2zwcqt3
բառարանագրություն
0
1589139
5669818
2026-06-22T17:00:08Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[leksikografi]], [[sözlükçülük]]
5669818
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[leksikografi]], [[sözlükçülük]]
s69yy12xiwmdpx6la681d6fhxt38qx5
ճպուռ
0
1589140
5669819
2026-06-22T17:12:24Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # {{t|dil=hy|böcekler}} [[yusufçuk]]
5669819
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# {{t|dil=hy|böcekler}} [[yusufçuk]]
8k7zsfe98j73wkrgi99h9h2lliz1un9
Խաչատուրյան
0
1589141
5669821
2026-06-22T17:18:27Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|soyadı}}
5669821
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
3ed14gmtlu670v5z9gfb2bdfh552qbc
Ղալումյան
0
1589142
5669822
2026-06-22T17:21:41Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|soyadı}}
5669822
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
3ed14gmtlu670v5z9gfb2bdfh552qbc
Դավիդյան
0
1589143
5669823
2026-06-22T17:28:36Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|soyadı}}
5669823
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
3ed14gmtlu670v5z9gfb2bdfh552qbc
Դաշյան
0
1589144
5669824
2026-06-22T17:29:47Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|soyadı}}
5669824
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
3ed14gmtlu670v5z9gfb2bdfh552qbc
Շաղոյան
0
1589145
5669825
2026-06-22T17:30:42Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|soyadı}}
5669825
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|soyadı}}
3ed14gmtlu670v5z9gfb2bdfh552qbc
Նիկոլ
0
1589146
5669826
2026-06-22T17:38:47Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669826
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Գառնիկի
0
1589147
5669827
2026-06-22T17:39:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669827
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Արմեն
0
1589148
5669828
2026-06-22T17:40:29Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669828
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Ռոբերտ
0
1589149
5669829
2026-06-22T17:41:19Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669829
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Սերժ
0
1589150
5669830
2026-06-22T17:41:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669830
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Ազատի
0
1589151
5669831
2026-06-22T17:42:20Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669831
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Վովայի
0
1589152
5669832
2026-06-22T17:42:59Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669832
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Վլադիմիր
0
1589153
5669833
2026-06-22T17:43:55Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669833
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Աբազյան
0
1589154
5669834
2026-06-22T17:51:20Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669834
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Աբդիու
0
1589155
5669835
2026-06-22T17:57:04Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669835
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Աբդլմսեհ
0
1589156
5669836
2026-06-22T17:57:29Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669836
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Ահարոն
0
1589157
5669837
2026-06-22T17:58:18Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669837
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Անձրևես
0
1589158
5669838
2026-06-22T17:59:08Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669838
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
Ասլիպարոն
0
1589159
5669839
2026-06-22T17:59:56Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenisteniyor}} ===Özel ad=== {{hy-özel ad}} # {{özel ad|dil=hy|erkek adı}}
5669839
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenisteniyor}}
===Özel ad===
{{hy-özel ad}}
# {{özel ad|dil=hy|erkek adı}}
hur9nd2gh9hk6v3aewh25v3svdcfwod
խանութ
0
1589160
5669840
2026-06-22T18:22:27Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[mağaza]], [[dükkân]]
5669840
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[mağaza]], [[dükkân]]
s7z65cnx8icrbktnqvuqzf7jp4oz7kq
5669841
5669840
2026-06-22T18:24:01Z
MustafaCavlak
59368
/* Ad */
5669841
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[mağaza]], [[dükkân]]
====Çekimleme====
{{hy-ad-ի-ներ}}
8h06wyw3if04ftsxr8qkz0mqelzvrnf
խանութներ
0
1589161
5669842
2026-06-22T18:25:10Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||nom|ç}}
5669842
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||nom|ç}}
jxqwawaybh9gm7xv3av43td0l88bqzk
խանութը
0
1589162
5669843
2026-06-22T18:25:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|nom|t}}
5669843
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|nom|t}}
lsl927emptcicxj3lhmdjj8sw0w623c
խանութն
0
1589163
5669844
2026-06-22T18:26:12Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|nom|t}}
5669844
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|nom|t}}
lsl927emptcicxj3lhmdjj8sw0w623c
խանութները
0
1589164
5669845
2026-06-22T18:26:41Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|nom|ç}}
5669845
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|nom|ç}}
1axhjk0jnlwe4mbvqitjmhup9n6zik1
խանութներն
0
1589165
5669846
2026-06-22T18:26:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|nom|ç}}
5669846
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|nom|ç}}
1axhjk0jnlwe4mbvqitjmhup9n6zik1
խանութի
0
1589166
5669847
2026-06-22T18:27:12Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||dat|t}}
5669847
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||dat|t}}
thjjsxjxvdqeagi6sde83ux3axo30we
խանութների
0
1589167
5669848
2026-06-22T18:27:28Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||dat|ç}}
5669848
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||dat|ç}}
4nbhp8vddhy5ck0ghdzghldqwgvgaqi
խանութին
0
1589168
5669849
2026-06-22T18:27:41Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|dat|t}}
5669849
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|dat|t}}
j8mp7qltwzn747ue6hf9nqcgm27sea1
խանութներին
0
1589169
5669850
2026-06-22T18:27:55Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||def|dat|ç}}
5669850
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||def|dat|ç}}
rj8yfcbbbski54l9i7mm07uco36ey9k
խանութից
0
1589170
5669851
2026-06-22T18:28:09Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||abl|t}}
5669851
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||abl|t}}
g9wz2nhkpr8x5eu411we16w1vy7i6ho
խանութներից
0
1589171
5669852
2026-06-22T18:28:35Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||abl|ç}}
5669852
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||abl|ç}}
6d5628r6e3gsa5m5ngxe61tlebncic3
խանութով
0
1589172
5669853
2026-06-22T18:28:50Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||araç|t}}
5669853
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||araç|t}}
9boahh7mdyihbvvsua5j4l8tw81qlo1
խանութներով
0
1589173
5669854
2026-06-22T18:29:07Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||araç|ç}}
5669854
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||araç|ç}}
g3isj9di0j5yesizp1g1kemsogowzo8
խանութում
0
1589174
5669855
2026-06-22T18:29:21Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||loc|t}}
5669855
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||loc|t}}
m3rrkyfgex0mgnaaj84us9wxus0h1xf
խանութներում
0
1589175
5669856
2026-06-22T18:29:38Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||loc|ç}}
5669856
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||loc|ç}}
5wr24l4va9piyf3r5ygt6w06fmszkt1
խանութս
0
1589176
5669857
2026-06-22T18:29:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|nom|t}}
5669857
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|nom|t}}
0l8ooccra86gmqrw7mmq62ftqddxjfo
խանութներս
0
1589177
5669858
2026-06-22T18:30:09Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|nom|ç}}
5669858
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|nom|ç}}
fujhb6fnehguk2twpnyw1tu5u73pglf
խանութիս
0
1589178
5669859
2026-06-22T18:30:22Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|dat|t}}
5669859
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|dat|t}}
sx0gp10ixxm8stpvofaep24hw55zovx
խանութներիս
0
1589179
5669860
2026-06-22T18:30:38Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|dat|ç}}
5669860
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|dat|ç}}
kyz2mhbnmtm7kzmnypaj1jan7epata2
խանութիցս
0
1589180
5669861
2026-06-22T18:30:52Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|abl|t}}
5669861
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|abl|t}}
0urimbosg8qbfkhu39cn9o9emg8i4gq
խանութներիցս
0
1589181
5669862
2026-06-22T18:31:05Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|abl|ç}}
5669862
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|abl|ç}}
mj17v69atodmlvmz4rgxkr4mlsxpuxy
խանութովս
0
1589182
5669863
2026-06-22T18:31:22Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|araç|t}}
5669863
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|araç|t}}
gpwagm6l0ub3f0hdo8unpryjmlglce0
խանութներովս
0
1589183
5669864
2026-06-22T18:31:36Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|araç|ç}}
5669864
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|araç|ç}}
65exkbn81jr15fw2ljbyf8a7g8xvolw
խանութումս
0
1589184
5669865
2026-06-22T18:31:49Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|loc|t}}
5669865
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|loc|t}}
hc7e6pj096btu5n30n3wo0cxms7ztpf
խանութներումս
0
1589185
5669866
2026-06-22T18:32:11Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||1t|loc|ç}}
5669866
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||1t|loc|ç}}
2ry87eusl9dpbr9ag2lvnml5rvlamie
խանութդ
0
1589186
5669867
2026-06-22T18:32:30Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|nom|t}}
5669867
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|nom|t}}
767lk1qpjmhqp5sm9rx4s2vc5v3681h
խանութներդ
0
1589187
5669868
2026-06-22T18:32:42Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|nom|ç}}
5669868
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|nom|ç}}
altws2gbfmsfvx1bbfu9iw8hkgz8z66
խանութիդ
0
1589188
5669869
2026-06-22T18:32:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|dat|t}}
5669869
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|dat|t}}
5di352l4pravbzssbmkfyrljqrxv9je
խանութներիդ
0
1589189
5669870
2026-06-22T18:33:10Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|dat|ç}}
5669870
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|dat|ç}}
65hgzg70kscixspmi53oip21i3blhfd
խանութիցդ
0
1589190
5669871
2026-06-22T18:33:25Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|abl|t}}
5669871
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|abl|t}}
c9dny6f0d6spgfct1z8adgrtrltjl5o
խանութներիցդ
0
1589191
5669872
2026-06-22T18:33:40Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|abl|ç}}
5669872
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|abl|ç}}
i9eaaeckgs7zz5a36dzfknx20peelvz
խանութովդ
0
1589192
5669873
2026-06-22T18:33:53Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|araç|t}}
5669873
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|araç|t}}
qacphxoc38uwaqfdhyc0ymuz965jjyi
խանութներովդ
0
1589193
5669874
2026-06-22T18:34:09Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|araç|ç}}
5669874
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|araç|ç}}
oszetni4y2g6z672dx05w3wfsf0eaay
խանութումդ
0
1589194
5669875
2026-06-22T18:34:24Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|loc|t}}
5669875
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|loc|t}}
97tlgjoo664j4r826d2amr8ydkb0kvc
խանութներումդ
0
1589195
5669876
2026-06-22T18:34:40Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{başlık başı|hy|çekimli ad}} # {{çekim|dil=hy|խանութ||2t|loc|ç}}
5669876
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{başlık başı|hy|çekimli ad}}
# {{çekim|dil=hy|խանութ||2t|loc|ç}}
oyezee2wucauuwg0t4f6vy4s7nvht5m
մանրաթել
0
1589196
5669877
2026-06-22T18:40:35Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[lif]]
5669877
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[lif]]
kdvpptd14ft5ngmicld2152xqpisrqw
փոքր
0
1589197
5669878
2026-06-22T18:43:13Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ön ad=== {{hy-ön ad}} # [[küçük]]
5669878
wikitext
text/x-wiki
==Ermenice==
===Ön ad===
{{hy-ön ad}}
# [[küçük]]
f8rppenu2t322m5a1egdsvy07m0uumt
պորտալար
0
1589198
5669879
2026-06-22T18:56:18Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[göbek bağı]], [[kordon]]
5669879
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[göbek bağı]], [[kordon]]
30w1d6nc4e5ucnhgouf5xntyr0g7scv
հարսնապար
0
1589199
5669880
2026-06-22T18:58:37Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[gelin]]in kendi düğününde yaptığı [[dans]]; gelin dansı
5669880
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[gelin]]in kendi düğününde yaptığı [[dans]]; gelin dansı
kyiy188ki2gm7wsl69vmez55cdoss0a
պարասրահ
0
1589200
5669881
2026-06-22T19:01:17Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[dans salonu]]
5669881
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[dans salonu]]
5d32j8dpw731xal83ia4abj1tcobv4z
սրահ
0
1589201
5669882
2026-06-22T19:02:18Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[salon]]
5669882
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[salon]]
69uie1u73cx3zpwr12h3u7nuztqkmmt
գահ
0
1589202
5669883
2026-06-22T19:03:21Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[taht]]
5669883
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[taht]]
ey67y6oyxr8letywa0ff6zggxc6br7y
ընթերցել
0
1589203
5669884
2026-06-22T19:05:46Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Eylem=== {{hy-eylem}} # {{t|dil=hy|resmî dil}} [[okumak]] #: {{eş anlamlılar|hy|կարդալ}}
5669884
wikitext
text/x-wiki
==Ermenice==
===Eylem===
{{hy-eylem}}
# {{t|dil=hy|resmî dil}} [[okumak]]
#: {{eş anlamlılar|hy|կարդալ}}
dl4mudvgwx9jyo9u2zm1xwjtocdmogt
Kategori:Ermenice resmî dil
14
1589204
5669885
2026-06-22T19:06:17Z
MustafaCavlak
59368
Yeni sayfa : {{kategori söz|dil=hy|söz=resmî dil}}
5669885
wikitext
text/x-wiki
{{kategori söz|dil=hy|söz=resmî dil}}
8ugv68xbaxwzcl6pufjpv1x0fsut83r
վերընթերցել
0
1589205
5669886
2026-06-22T19:07:40Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Eylem=== {{hy-eylem}} # [[tekrar]] [[okumak]], [[yeniden]] okumak
5669886
wikitext
text/x-wiki
==Ermenice==
===Eylem===
{{hy-eylem}}
# [[tekrar]] [[okumak]], [[yeniden]] okumak
ffaft0v3uh8u9fd6b9ynsf30s5algrp
միջազգային
0
1589206
5669887
2026-06-22T19:09:31Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ön ad=== {{hy-ön ad}} # [[uluslararası]]
5669887
wikitext
text/x-wiki
==Ermenice==
===Ön ad===
{{hy-ön ad}}
# [[uluslararası]]
6jqo8rdmarlls4gs6nj0rxvc1xtpg5v
գալակտիկական
0
1589207
5669888
2026-06-22T19:11:47Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ön ad=== {{başlık başı|hy|ön ad}} # [[galaksi]] ile ilgili
5669888
wikitext
text/x-wiki
==Ermenice==
===Ön ad===
{{başlık başı|hy|ön ad}}
# [[galaksi]] ile ilgili
dok432ihcn7gufwynujy4kg4hzve35t
գալակտիկա
0
1589208
5669889
2026-06-22T19:12:17Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[galaksi]]
5669889
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[galaksi]]
6nqpor2zwyl7ledimvb7sgd52oe45d9
արևելագետ
0
1589209
5669890
2026-06-22T19:15:08Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[oryantalist]]
5669890
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[oryantalist]]
nntrwk1p76oru97odlo8f9pqun454mg
արևելք
0
1589210
5669891
2026-06-22T19:15:43Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[doğu]]
5669891
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[doğu]]
6vnlyv45glkvviva2au956y7727d7co
հյուսիս-արևելք
0
1589211
5669892
2026-06-22T19:16:21Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[kuzeydoğu]]
5669892
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[kuzeydoğu]]
q0ljx2x3d9kc99dc3c5mrqigy4egsg0
հյուսիս
0
1589212
5669893
2026-06-22T19:17:02Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[kuzey]]
5669893
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[kuzey]]
rs01gxoqfy1ami6e0e8fsmmttu0gozt
հյուսիս-արևմուտք
0
1589213
5669894
2026-06-22T19:17:34Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[kuzeybatı]]
5669894
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[kuzeybatı]]
q3sg2m2474bjnfxkpnei23vqohyt7x5
արևմուտք
0
1589214
5669895
2026-06-22T19:18:05Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[batı]]
5669895
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[batı]]
iq8ue4g8r7payau99odj2jcwotb10q1
հարավ-արևմուտք
0
1589215
5669896
2026-06-22T19:18:31Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[güneybatı]]
5669896
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[güneybatı]]
ei4w03vxsh0905ahsosuxpwvfpbbze9
հարավ
0
1589216
5669897
2026-06-22T19:18:58Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[güney]]
5669897
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[güney]]
n8v59lzr7lnywudhosnqwq0irhjkeo9
հարավ-արևելք
0
1589217
5669898
2026-06-22T19:19:25Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[güneydoğu]]
5669898
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[güneydoğu]]
m92wxl0wo7z051mwo71k1j71r74balo
Կորեա
0
1589218
5669899
2026-06-22T19:20:21Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Özel ad=== {{hy-özel ad}} # [[Kore]]
5669899
wikitext
text/x-wiki
==Ermenice==
===Özel ad===
{{hy-özel ad}}
# [[Kore]]
2198efggmjib8khibme0g0fcaxykmlb
Հյուսիսային Կորեա
0
1589219
5669900
2026-06-22T19:23:01Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== {{vikipedi|dil=hy}} ===Özel ad=== {{hy-özel ad|baş=[[հյուսիսային|Հյուսիսային]] [[Կորեա]]}} # {{t|dil=hy|ülkeler}} [[Kuzey Kore]] ====Çekimleme==== {{hy-ad-ի-ներ|çoğulsuz=evet}}
5669900
wikitext
text/x-wiki
==Ermenice==
{{vikipedi|dil=hy}}
===Özel ad===
{{hy-özel ad|baş=[[հյուսիսային|Հյուսիսային]] [[Կորեա]]}}
# {{t|dil=hy|ülkeler}} [[Kuzey Kore]]
====Çekimleme====
{{hy-ad-ի-ներ|çoğulsuz=evet}}
8b1chi7jbx2y93706xsx92xhsc88wwr
հյուսիսային
0
1589220
5669901
2026-06-22T19:24:50Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{kökenli|hy|xcl|հիւսիսային}}. ===Söyleniş=== * {{ses|dil=hy|LL-Q8785 (hye)-Veravi95-հյուսիսային.wav|(Batı Ermenicesi}} ===Ön ad=== {{hy-ön ad|GYB=հիւսիսային}} # [[kuzey]] ====Çekimleme==== {{hy-ad-ի-ներ|n=on}}
5669901
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenli|hy|xcl|հիւսիսային}}.
===Söyleniş===
* {{ses|dil=hy|LL-Q8785 (hye)-Veravi95-հյուսիսային.wav|(Batı Ermenicesi}}
===Ön ad===
{{hy-ön ad|GYB=հիւսիսային}}
# [[kuzey]]
====Çekimleme====
{{hy-ad-ի-ներ|n=on}}
9ozlbe0dnt2ppo64xtrgb43z08r5e8t
5669902
5669901
2026-06-22T19:26:32Z
MustafaCavlak
59368
/* Söyleniş */
5669902
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{kökenli|hy|xcl|հիւսիսային}}.
===Söyleniş===
* {{hy-d}} {{IPA|dil=hy|[hjusisɑjín]}}
* {{hy-b}} {{IPA|dil=hy|[hʏsisɑjín]}}
* {{ses|dil=hy|LL-Q8785 (hye)-Veravi95-հյուսիսային.wav|(Batı Ermenicesi)}}
===Ön ad===
{{hy-ön ad|GYB=հիւսիսային}}
# [[kuzey]]
====Çekimleme====
{{hy-ad-ի-ներ|n=on}}
gw6xzssi02g6jhsg20uj0ikyi2yz7ex
српскохрватски
0
1589221
5669907
2026-06-23T06:38:28Z
MustafaCavlak
59368
Yeni sayfa : ==Sırp-Hırvatça== ===Ön ad=== {{başlık başı|sh|ön ad|baş=српскохр̀ва̄тскӣ|Latin harfleriyle|srpskohr̀vātskī}} # [[Sırp-Hırvatça]] ===Ad=== {{başlık başı|sh|ad|baş=српскохр̀ва̄тскӣ|c=e|Latin harfleriyle|srpskohr̀vātskī}} # {{t|dil=sh|diller}} [[Sırp-Hırvatça]]
5669907
wikitext
text/x-wiki
==Sırp-Hırvatça==
===Ön ad===
{{başlık başı|sh|ön ad|baş=српскохр̀ва̄тскӣ|Latin harfleriyle|srpskohr̀vātskī}}
# [[Sırp-Hırvatça]]
===Ad===
{{başlık başı|sh|ad|baş=српскохр̀ва̄тскӣ|c=e|Latin harfleriyle|srpskohr̀vātskī}}
# {{t|dil=sh|diller}} [[Sırp-Hırvatça]]
rbdp62ppzff0j4vjca3ng111fi26lu8
ուսանող
0
1589222
5669908
2026-06-23T06:39:33Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[üniversite]] [[öğrenci]]si
5669908
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[üniversite]] [[öğrenci]]si
1je0iw0om1g5i1uu9gnbo7flbnk63qh
сковорода
0
1589223
5669909
2026-06-23T06:42:56Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== {{vikipedi|dil=ru}} ===Ad=== {{başlık başı|ru|ad|c=d|baş=сковорода́}} # [[tava]]
5669909
wikitext
text/x-wiki
==Rusça==
{{vikipedi|dil=ru}}
===Ad===
{{başlık başı|ru|ad|c=d|baş=сковорода́}}
# [[tava]]
odvqi7t2jbgtav0hjtpmrgx7imdb3xd
այստեղ
0
1589224
5669910
2026-06-23T06:55:57Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Adıl=== {{başlık başı|hy|adıl}} # [[burada]], [[bura]]ya, [[burası]]
5669910
wikitext
text/x-wiki
==Ermenice==
===Adıl===
{{başlık başı|hy|adıl}}
# [[burada]], [[bura]]ya, [[burası]]
3ioc6xqkv09qzv8kf11v2ahfrdy7rzj
думаю
0
1589225
5669911
2026-06-23T07:00:11Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́маю}} # {{çekim|dil=ru|ду́мать||1t|pres|ind|tamamlanmamış}}
5669911
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́маю}}
# {{çekim|dil=ru|ду́мать||1t|pres|ind|tamamlanmamış}}
fvgqj3bz78zid38gdmzjt8boh5ojyw8
думающий
0
1589226
5669912
2026-06-23T07:01:53Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Ortaç=== {{başlık başı|ru|ortaç|baş=ду́мающий}} # {{çekim|dil=ru|ду́мать||pres|act|tamamlanmamış|part}}
5669912
wikitext
text/x-wiki
==Rusça==
===Ortaç===
{{başlık başı|ru|ortaç|baş=ду́мающий}}
# {{çekim|dil=ru|ду́мать||pres|act|tamamlanmamış|part}}
kizkkf8bojly7f4gofhu2kbeinb3zyr
думавший
0
1589227
5669913
2026-06-23T07:03:06Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Ortaç=== {{başlık başı|ru|ortaç|baş=ду́мавший}} # {{çekim|dil=ru|ду́мать||past|act|tamamlanmamış|part}}
5669913
wikitext
text/x-wiki
==Rusça==
===Ortaç===
{{başlık başı|ru|ortaç|baş=ду́мавший}}
# {{çekim|dil=ru|ду́мать||past|act|tamamlanmamış|part}}
bi1j6lfmh8ce7jbioj9jkkth563nyk1
думаешь
0
1589228
5669914
2026-06-23T07:04:33Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́маешь}} # {{çekim|dil=ru|ду́мать||2t|pres|ind|tamamlanmamış}}
5669914
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́маешь}}
# {{çekim|dil=ru|ду́мать||2t|pres|ind|tamamlanmamış}}
002u48krl867jg4fclnpuqfg4u0ndip
думает
0
1589229
5669915
2026-06-23T07:05:36Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́мает}} # {{çekim|dil=ru|ду́мать||3t|pres|ind|tamamlanmamış}}
5669915
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́мает}}
# {{çekim|dil=ru|ду́мать||3t|pres|ind|tamamlanmamış}}
60pvtv1hb9pfu8yuyawavnyp2pvfffx
думаем
0
1589230
5669916
2026-06-23T07:06:46Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́маем}} # {{çekim|dil=ru|ду́мать||1ç|pres|ind|tamamlanmamış}}
5669916
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́маем}}
# {{çekim|dil=ru|ду́мать||1ç|pres|ind|tamamlanmamış}}
1flcughilmnlon9fznjtga29pb7i54x
думаете
0
1589231
5669917
2026-06-23T07:07:37Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́маете}} # {{çekim|dil=ru|ду́мать||2ç|pres|ind|tamamlanmamış}}
5669917
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́маете}}
# {{çekim|dil=ru|ду́мать||2ç|pres|ind|tamamlanmamış}}
16li9baxqj58p0saa1y8lzkr6cmn249
думают
0
1589232
5669918
2026-06-23T07:08:34Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́мают}} # {{çekim|dil=ru|ду́мать||3ç|pres|ind|tamamlanmamış}}
5669918
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́мают}}
# {{çekim|dil=ru|ду́мать||3ç|pres|ind|tamamlanmamış}}
46ean69r38p5yy40ki0ow2f915ztfi4
думай
0
1589233
5669919
2026-06-23T07:09:43Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́май}} # {{çekim|dil=ru|ду́мать||2t|imp|tamamlanmamış}}
5669919
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́май}}
# {{çekim|dil=ru|ду́мать||2t|imp|tamamlanmamış}}
gq610ywpe200uhofomdb66a7hpg616c
думайте
0
1589234
5669920
2026-06-23T07:10:37Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Eylem=== {{başlık başı|ru|çekimli eylem|baş=ду́майте}} # {{çekim|dil=ru|ду́мать||2ç|imp|tamamlanmamış}}
5669920
wikitext
text/x-wiki
==Rusça==
===Eylem===
{{başlık başı|ru|çekimli eylem|baş=ду́майте}}
# {{çekim|dil=ru|ду́мать||2ç|imp|tamamlanmamış}}
qtsd23ed0ztogewlm9dhnj3zv2fbahr
Modül:ru-ad
828
1589235
5669921
2026-06-23T07:17:49Z
MustafaCavlak
59368
imported from en for testing
5669921
Scribunto
text/plain
--[=[
This module contains functions for creating inflection tables for Russian
nouns.
Author: Benwing, rewritten from early version by Wikitiki89
Form of arguments: One of the following:
1. LEMMA|DECL|PLSTEM (all arguments optional)
2. ACCENT|LEMMA|DECL|PLSTEM (all arguments optional)
3. multiple sets of arguments separated by the literal word "or"
Arguments:
ACCENT: Accent pattern (a b c d e f b' d' f' f''). Multiple values can
be specified, separated by commas. If omitted, defaults to a or b
depending on the position of stress on the lemma or
explicitly-specified declension.
LEMMA: Lemma form (i.e. nom sg or nom pl), with appropriately-placed
stress; or the stem, if an explicit declension is specified
(in this case, the declension usually looks like an ending, and
the stem is the portion of the lemma minus the ending). In the
first argument set (i.e. first set of arguments separated
by "or"), defaults to page name; in later sets, defaults to lemma
of previous set. A plural form can be given, and causes argument
n= to default to n=p (plural only). Normally, an accent is
required if multisyllabic, and unaccented monosyllables with
automatically be stressed; prefix with * to override both behaviors.
DECL: Declension field. Normally omitted to autodetect based on the
lemma form; see below.
PLSTEM: special plural stem (defaults to stem of lemma)
Additional named arguments:
a: animacy (a/an/anim = animate, i/in/inan = inanimate, b/bi/both/ai = both
(listing animate first in the headword), ia = both (listing inanimate
first in the headword), otherwise inanimate)
n: number restriction (p = plural only, s = singular only, b = both;
defaults to both unless the lemma is plural, in which case it
defaults to plural only)
CASE_NUM or acc_NUM_ANIM or par/loc/voc: override (or multiple
values separated by commas) for case/number combination;
forms auto-linked; can have raw links in it, can have an
ending "note" (*, +, 1, 2, 3, etc.)
pltail: Specify something (usually a * or similar) to attach to
the end of the last plural form when there's more than one. Used in
conjunction with notes= to indicate that alternative plural forms
are obsolete, poetic, etc.
pltailall: Similar pltail= but attaches to all plural forms.
Typically used in conjunction with notes= to make a comment about
the plural as a whole (e.g. it's mostly hypothetical, rare and
awkward, etc.).
sgtail, sgtailall: Same as pltail=, pltailall= but for the singular.
obltail, obltailall: Same as pltail=, pltailall= but for oblique cases
(not the nominative or accusative).
CASE_NUM_tail: Attach the argument to the end of the last form
(whether there's one or more than one) for the particular
case/number combination. Note that this doesn't work quite like
pltail= or sgtail= in that it doesn't skip adding the argument
when there's only one form.
CASE_NUM_tailall: Attach the argument to the end of all forms
specified for the particular case/number combination. Similar to
pltailall= or sgtailall=.
suffix: Add a suffix such as ся to all forms.
prefix: Add a prefix to all forms.
plhyp, plhypall, CASE_NUM_hyp, etc.: Same as pltail, pltailall,
CASE_NUM_tal, etc. but specify that the marked forms are
mostly hypothetical or rare/awkard. Generally you will want
plhypall=y to mark the plural as hypothetical.
Per word named arguments:
All of the above named arguments have per-word variants, e.g.
a1, a2, ...; n1, n2, ...; CASE_NUM1, CASE_NUM2, ...;
pltail1, pltail2, ...; etc. These apply to the individual words of a
form.
Case abbreviations:
nom: nominative
gen: genitive
dat: dative
acc: accusative
ins: instrumental
pre: prepositional
par: partitive
loc: locative
voc: vocative
Number abbreviations:
sg: singular
pl: plural
Animacy abbreviations:
an: animate
in: inanimate
Declension field:
One of the following for regular nouns:
(blank)
GENDER
-VARIANT
GENDER-VARIANT
$
DECLTYPE
DECLTYPE/DECLTYPE
(also, can append various special-case markers to any of the above)
Or one of the following for adjectival nouns:
+
+ь
$
+short, +mixed or +proper
+DECLTYPE
GENDER if present is m, f, n or 3f; for regular nouns, required if the
lemma ends in -ь or is plural, ignored otherwise. 3f is the same
as f but in the case of a plural lemma in -и, detects a
third-declension feminine with singular in -ь rather than a
first-declension feminine with singular in -а or -я.
VARIANT is one way of requesting variant declensions (see also special
case (1) for variant nom pls, and special case (2) for variant
gen pls). The currently allowed values are -ья (will select a
mixed declension that has some normal declension in its singular
and the -ья plural declension); -ин (for animate masculine nouns
in -ин with plural in -е -- note that this is autodetected in the
majority of cases where the ending in -янин or -анин); -ишко
(used for inanimate neuter-form diminutive masculine nouns in
-ишко [also сараю́шко] with nom pl -и and colloquial feminine-ending
alternants in some singular cases); -ище (similar to -ишко but
used for *animate* augmentative masculine neuter-form nouns in
-ище). Variants -ишко and -ище must be given with with special
case (1).
$ is indeclinable words. It is principally useful in multiword
expressions where some of the words are indeclinable.
DECLTYPE is an explicit declension type. Normally you shouldn't use
this, and should instead let the declension type be autodetected
based on the ending, supplying the appropriate hint if needed
(gender for regular nouns, +ь for adjectives). If provided, the
declension type is usually the same as the ending, and if present,
the lemma field should be just the stem, without the ending.
Possibilities for regular nouns are (blank) or # for hard-consonant
declension, а, я, о, е or ё, е́, й, ья, ье or ьё, ь-m, ь-f,
ин, ёнок or онок or енок, ёночек or оночек or еночек, мя,
-а or #-а, ь-я, й-я, о-и or о-ы, -ья or #-ья, $ (indeclinable).
Old-style (pre-reform) declensions use ъ instead of (blank), ъ-а
instead of -а, ъ-ья instead of -ья, and инъ, ёнокъ/онокъ/енокъ,
ёночекъ/оночекъ/еночекъ instead of the same without terminating ъ.
The declensions can also be written with an accent on them; this
chooses the same declension (except for е vs. е́), but causes
ACCENT to default to pattern b instead of a.
For adjectival nouns, you should normally supply just + and let
the ending determine the declension; supply +ь in the case of a
possessive adjectival noun in -ий, which have an extra -ь- in
most endings compared with normal adjectival nouns in -ий, but
which can't be distinguished based on the nominative singular.
You can also supply +short, +mixed or +proper, which constrains
the declension appropriately but still autodetects the
gender-specific and stress-specific variant. If you do supply
a specific declension type, as with regular nouns you need to
omit the ending from the lemma field and supply just the stem.
Possibilities are +ый, +ое, +ая, +ій, +ее, +яя, +ой, +о́е, +а́я,
+ьій, +ье, +ья, +-short or +#-short (masc), +о-short,
+о-stressed-short or +о́-short, +а-short, +а-stressed-short or
+а́-short, and similar for -mixed and -proper (except there aren't
any stressed mixed declensions).
DECLTYPE/DECLTYPE is used for nouns with one declension in the
singular and a different one in the plural, for cases that
PLVARIANT and special case (1) below don't cover.
Special-case markers:
(1) for Zaliznyak-style alternative nominative plural ending:
-а or -я for masculine, -и or -ы for neuter
(2) for Zaliznyak-style alternative genitive plural ending:
-ъ/none for masculine, -ей for feminine, -ов(ъ) for neuter,
-ей for plural variant -ья
* for reducibles (nom sg or gen pl has an extra vowel before the
final consonant as compared with the stem found in other cases)
;ё for Zaliznyak-style alternation between last е in stem and ё
TODO:
1. Multi-word issues:
-- FIXME: Make sure internal_notes handled correctly; we may run into issues
with multiple internal notes from different words, if we're not careful
to use different footnote symbols for each type of footnote (which we
don't do currently). [NOT DONE, MAY NOT DO]
-- Handling default lemma: With multiple words, we should probably split
the page name on spaces and default each word in turn [NOT DONE, MAY
NOT DO]
2a. FIXME: For -ишко diminutives and -ище augmentatives, should add an
appropriate category of some sort (currently marked by colloqfem= in
category).
2b. FIXME: Adding a note to dat_sg also adds it to loc_sg when it exists;
seems wrong. See луг.
2c. FIXME: When you have both d' and f in feminines and you use sgtail=*,
you get two *'s. See User:Benwing2/test-ru-noun-debug.
3. ADJECTIVE FIXMES:
3a. FIXME: Change calls to ru-adj11 to use the new proper name support in
ru-adjective.
3b. FIXME: Test that omitting a manual form in ru-adjective leaves the form as
a big dash.
3c. FIXME: какой-либо and какой-то display genitives with translit -go
instead of -vo. To fix this properly requires implementing real
manual translit for adjectives.
3d. FIXME: Implement real manual translit for adjectives.
5. [FIXME: Consider adding an indicator in the header line when the ё/e
alternation occurs. This is a bit tricky to calculate: If special case
;ё is given, but also if ё occurs in the stem and the accent pattern is
as follows -- for sg-only, b' d' f' f'', also b d f if the noun is masc
or 3rd-decl fem (i.e. nom-sg ending is non-syllabic); for pl-only,
e f f' f'', also b b' c if the gen pl is non-syllabic; for sg/pl,
any but a or b, also b if either nom sg or gen pl is non-syllabic. But
it gets more complicated due to overrides. An alternative is to check
all forms to see if ё is present in some but not all; but this is
tricky also because e.g. reducibles frequently have ё/null alternation,
which doesn't count, and some endings have е or ё in them, which also
doesn't count. If we were to do it this way, we'd have to (a) count
the number of е's in the form(s) with ё and verify that there's at least
one form without ё and with one more е than in the form(s) with ё
(and it gets trickier if different forms with ё have different numbers of
е in them, although that is probably rare); and (b) ignore the appropriate
endings (the best way to do this would probably be to look at the actual
suffixes that were generated in args.suffixes and chop off any matching
ending in the actual form(s), but also chop off final -е/ё, as well as
-ев(ъ)/-ёв(ъ)/-ей/-ёй in the gen pl, which is frequently overridden,
unless perhaps the stem ends in the same way). It'd probably not
possible to do this in a 100% foolproof way but can be "good enough" for
nearly all circumstances.] [MIGHT BE TOO MUCH WORK]
6. HEADWORD FIXMES:
6a. FIXME: In ru-headword, create a category for words whose gender doesn't
match the form. (This is easy to do for ru-noun+ but harder for ru-noun.
We would need to do limited autodetection of the ending: for singulars,
-а/я should be feminine, -е/о/ё should be neuter, -ь should be masculine
or feminine, anything else should be masculine; for plurals, -и/ы should
be masculine or feminine, -а/я should be neuter except that -ія can be
feminine or neuter due to old-style adjectival pluralia tantum nouns,
anything else can be any gender.)
6b. FIXME: Recognize indeclinable nouns and indicate as indeclinable. Probably
should work by checking the case forms to see if they're the same.
9. FIXME: Change stress-pattern detection and overriding to happen inside of
looping over the two parts of a slash decl. Requires that the loop over
the two parts happen outside of the loop over stress patterns. Requires
that the category code get split into two parts, one to handle combined
singular/plural categories that goes outside the two loops, and one to
handle everything else that goes inside the two loops.
10. FIXME: override_matches_suffix() had a free variable reference to ARGS
in it, which should have triggered an error whenever there was a nom_sg or
nom_pl override but didn't. Is there an error causing this never to be
called? Check.
11a. FIXME: In a multiword lemma, using loc2=+ causes only the second word
to get linked instead of the whole expression. Same for par2=+, voc2=+.
11b. FIXME: Using loc=+ with a multiword lemma should do the right thing, same
as if locN=+ is specified for each individual word. Instead it generates
the locative as a whole from the dative, which fails e.g. if some of the
words are adjectival. Same for par=+, voc=+.
13. Multi-word issues:
-- Setting n=pl when auto-detecting a plural lemma. How does that interact
with multi-word stuff? (DONE)
-- compute_heading() -- what to do with multiple words? I assume we should
display info on the first noun (non-indeclinable, non-adjectival), and
on the first adjectival word otherwise, and finally on an indeclinable word (DONE)
-- args.genders -- it should presumably come from the same word as is used
in compute_heading(); but we should allow the overall gender to be
overridden, at least in ru-noun+ (DONE)
-- Bug in args.suffix: Gets added to every word in attach_with() and then
again at the end, after pltail and such. Needs to be added to the
last word only, before pltail. Need also suffixN for individual words.
(DONE, NEEDS TESTING)
-- Should have ..N versions of pltail and variants. (DONE, NEEDS TESTING)
-- Need to handle overrides of acc_sg, acc_pl (DONE)
-- Overrides of nom_sg/nom_pl should also override acc_sg/acc_pl if it
was originally empty and the animacy is inanimate; similarly for
gen_sg/gen_pl and animates; this needs to work both for per-word and
overall overrides. (DONE)
-- do_generate_forms(_multi) need to run part of make_table(), enough to
combine all per_word_info into single lists of forms and store back
into args[case]. (DONE, NEEDS TESTING)
-- In generate_forms, should probably check if a=="i" and only return
acc_sg_in as acc_sg=; or if a=="a" and only return acc_sg_an as acc_sg=;
in old/new comparison code, do something similar, also when a=="b"
check if acc_sg_in==acc_sg_an and make it acc_sg; when a=="b" and the
_in and _an variants are different, might need to ignore them or check
that acc_sg_in==nom_sg and acc_sg_an==gen_sg; similarly for _pl
(DONE, NEEDS TESTING)
-- Need to test with multiple words! [DONE]
-- Current handling of <adj> won't work properly with multiple words;
will need to translate word-by-word in that case (should be solved by
manual-translit branch) [DONE]
14. In multiple-words branch, fix ru-decl-noun-multi so it recognizes
things like *, (1), (2) and ; without the need for a separator. Consider
using semicolon as a separator, since we already use it to separate ё
from a previous declension. Maybe use $ or ~ for an indeclinable word; don't
use semicolon. [IMPLEMENTED. NEED TO TEST.]
16. [Consider having ru-noun+ treat par= as a second genitive in
the headword, as is done with край] [WON'T DO]
17. [FIXME: Consider removing slash patterns and instead handling them by
allowing additional declension flags 'sg' and 'pl'. This simplifies the
various special cases caused by slash declensions. It would also be
possible to remove the special plural stem, which would get rid of more
special cases. On the other hand, it makes it more complicated to support
plural variant -ья with all singular types, and the category code that
displays things like "Russian nouns with singular -X and plural -Y"
also gets more complicated, and there's something convenient and intuitive
about plural stems, and slash declensions are also convenient and at least
somewhat intuitive. One possibility is to externally allow slash
declensions and special plural stems and rewrite them internally to
separate stems with 'sg' and 'pl' declension flags; but there are still
the two coding issues mentioned above.]
18. [FIXME: Consider redoing slash patterns so they operate at the outer
level, i.e. things like special cases apply separately in the singular
and plural part of the slash pattern.]
19. In ru-noun, don't recognize -а with m as plural unless (1)
or n=pl is also given, because there are masculine words with the
feminine ending. Check using the test code whether this changes
anything. Also check if there are other similar cases (neuter with
-и isn't parallel because -и is always plural). [IMPLEMENTED. NEED TO TEST.]
19a. Internal notes weren't propagated properly from adjectives.
[IMPLEMENTED. NEED TO TEST.]
19b. Add support for -ишко and -ище variants (p. 74 of Z),
which conversationally and/or colloquially have feminine endings in
certain cases. [IMPLEMENTED. NEED TO TEST. MAKE SURE THE INTERNAL NOTES
APPEAR.]
19d. For masculine animate neuter-form nouns, the accusative singular
ends in -а (-я soft) instead of -о. [IMPLEMENTED. NEED TO TEST.
NOTE: Currently this variant only can be selected using new-style
arguments where the gender can be given. Perhaps we should consider
allowing gender to be specified with old-style explicit declensions.]
21. Put back gender hints for pl adjectival nouns; used by ru-noun+.
[IMPLEMENTED. NEED TO TEST.]
23. Mixed and proper-noun adjectives have built-in notes. We need to
handle those notes with an "internal_notes" section similar to what is used
in the adjective module. [IMPLEMENTED. NEED TO TEST.]
24. Adjective detection code here needs to work the same as for the
adjective module, in particular in the handling of short, stressed-short,
mixed, proper, stressed-proper. [IMPLEMENTED. NEED TO TEST.]
25. Consider simplifying plural-variant code to only allow -ья as a
plural variant [and maybe even change that to be something like (1')].
[IMPLEMENTED REDUCTION OF PLURAL VARIANTS TO -ья; PLURAL-VARIANT CODE
STILL COMPLEX, THOUGH. NEED TO TEST.]
26. Automatically superscript *, numbers and similar things at the
beginning of a note. Also do this in adjective module. [IMPLEMENTED.
NEED TO TEST.]
28. Make the check for multiple stress patterns (categorizing/tracking)
smarter, to keep a list of them and check at the end, so we handle
multiple stress patterns specified through different arg sets.
[IMPLEMENTED; NEED TO TEST.]
29. More sophisticated handling of user-requested plural variant vs. special
case (1) vs. plural-detected variant. [IMPLEMENTED. NEED TO TEST FURTHER.]
30. Solution to ambiguous plural involving gender spec "3f". [IMPLEMENTED;
NEED TO TEST. Use запчасти, новости.]
33. With pluralia tantum adjectival nouns, we don't know the gender.
By default we assume masculine (or feminine for old-style -ія nouns) and
currently this goes into the category, but shouldn't. [IMPLEMENTED.]
39. [Eventually: Even with decl type explicitly given, the full stem with
ending should be included.] [MAY NEVER IMPLEMENT]
40. [Get error "Unable to dereduce" with strange noun ва́йя, what should
happen?] [WILL NOT FIX; USE AN OVERRIDE]
41. In творог, module generates partitive творогу́ when it should copy the
dative творогу́,тво́рогу. (DONE)
42. [[груз 200]] doesn't work. Interprets 200 as a footnote symbol.
43. When converting е -> ё not after cons and with translit, we should
convert e -> o to avoid double j. (DONE)
44. FIXME: In ро́вня/ровня́, similarly with неровня, marks genitive plural
ровня́ as irregular even though it isn't. (NOT OUR ERROR; THE DECLENSIONS
OF THESE NOUNS MARKED THE ENDING-STRESSED VARIANTS WITH (2).)
]=]--
local m_utilities = require("Module:utilities")
local m_table = require("Module:table")
local m_links = require("Module:links")
local com = require("Module:ru-common")
local nom = require("Module:ru-nominal")
local m_ru_adj = require("Module:ru-adjective")
local m_str_utils = require("Module:string utilities")
local scriptutils = require("Module:script utilities")
local m_table_tools = require("Module:table tools")
local m_debug = require("Module:debug")
local export = {}
local lang = require("Module:languages").getByCode("ru")
local Latn = require("Module:scripts").getByCode("Latn")
local format = m_str_utils.format
local rfind = m_str_utils.find
local rsubn = m_str_utils.gsub
local rmatch = m_str_utils.match
local rsplit = m_str_utils.split
local u = m_str_utils.char
local ulower = m_str_utils.lower
local usub = m_str_utils.sub
local ulen = m_str_utils.len
local unpack = unpack or table.unpack -- Lua 5.2 compatibility
-- If enabled, compare this module with new version of module to make
-- sure all declensions are the same. Eventually consider removing this;
-- but useful as new code is created.
local test_new_ru_noun_module = false
local AC = u(0x0301) -- acute = ́
local GR = u(0x0300) -- grave = ̀
local CFLEX = u(0x0302) -- circumflex = ̂
local DIA = u(0x0308) -- diaeresis = ̈
local PSEUDOCONS = u(0xFFF2) -- pseudoconsonant placeholder, matching ru-common
local IRREGMARKER = "△"
local HYPMARKER = "⟐"
local paucal_marker = "*"
local paucal_internal_note = "* Used with the numbers 1.5, 2, 3, 4 and higher numbers after 20 ending in 2, 3, and 4."
-- text class to check lowercase arg against to see if Latin text embedded in it
local latin_text_class = "[a-zščžěáéíóúýàèìòùỳâêîôûŷạẹịọụỵȧėȯẏ]"
-- Forward functions
local generate_forms_1
local determine_decl
local handle_forms_and_overrides
local handle_overall_forms_and_overrides
local concat_word_forms
local make_table
local detect_adj_type
local detect_stress_pattern
local override_stress_pattern
local determine_stress_variant
local determine_stem_variant
local is_reducible
local is_dereducible
local add_bare_suffix
local attach_stressed
local do_stress_pattern
local canonicalize_override
-- version of rsubn() that discards all but the first return value
local function rsub(term, foo, bar)
local retval = rsubn(term, foo, bar)
return retval
end
-- version of rsubn() that returns a 2nd argument boolean indicating whether
-- a substitution was made.
local function rsubb(term, foo, bar)
local retval, nsubs = rsubn(term, foo, bar)
return retval, nsubs > 0
end
-- version of rfind() that lowercases its string first, for case-insensitive matching
local function rlfind(term, foo)
return rfind(ulower(term), foo)
end
local function track(page)
m_debug.track("ru-noun/" .. page)
return true
end
-- version of m_table.insertIfNot() that makes sure 'false' doesn't get inserted by mistake.
local function insert_if_not(foo, bar)
assert(bar ~= false)
m_table.insertIfNot(foo, bar)
end
-- Fancy version of ine() (if-not-empty). Converts empty string to nil,
-- but also strips leading/trailing space and then single or double quotes,
-- to allow for embedded spaces.
local function ine(arg)
if not arg then return nil end
arg = rsub(arg, "^%s*(.-)%s*$", "%1")
if arg == "" then return nil end
local inside_quotes = rmatch(arg, '^"(.*)"$')
if inside_quotes then
return inside_quotes
end
inside_quotes = rmatch(arg, "^'(.*)'$")
if inside_quotes then
return inside_quotes
end
return arg
end
-- FIXME: Move to utils
-- Iterate over a chain of parameters, FIRST then PREF2, PREF3, ...,
-- inserting into LIST (newly created if omitted). Return LIST.
local function get_arg_chain(args, first, pref, list)
if not list then
list = {}
end
local val = args[first]
local i = 2
while val do
table.insert(list, val)
val = args[pref .. i]
i = i + 1
end
return list
end
-- synthesize a frame so that exported functions meant to be called from
-- templates can be called from the debug console.
local function debug_frame(parargs, args)
return {args = args, getParent = function() return {args = parargs} end}
end
local function rutr_pairs_equal(term1, term2)
local ru1, tr1 = term1[1], term1[2]
local ru2, tr2 = term2[1], term2[2]
local ru1entry, ru1notes = m_table_tools.separate_notes(m_links.remove_links(ru1))
local ru2entry, ru2notes = m_table_tools.separate_notes(m_links.remove_links(ru2))
if ru1entry ~= ru2entry then
return false
end
local tr1entry, tr1notes
local tr2entry, tr2notes
if tr1 then
tr1entry, tr1notes = m_table_tools.separate_notes(tr1)
end
if tr2 then
tr2entry, tr2notes = m_table_tools.separate_notes(tr2)
end
if tr1entry == tr2entry then
return true
elseif type(tr1entry) == type(tr2entry) then
return false
else
tr1entry = tr1entry or com.translit_no_links(ru1entry)
tr2entry = tr2entry or com.translit_no_links(ru2entry)
return tr1entry == tr2entry
end
end
local function contains_rutr_pair(list, pair)
for _, item in ipairs(list) do
if rutr_pairs_equal(item, pair) then
return true
end
end
return false
end
-- Clone parent's args while also assigning nil to empty strings.
local function clone_args(frame)
local args = {}
for pname, param in pairs(frame:getParent().args) do
args[pname] = ine(param)
end
return args
end
-- Old-style declensions.
local declensions_old = {}
-- New-style declensions; computed automatically from the old-style ones,
-- for the most part.
local declensions = {}
-- Internal notes for old-style declensions.
local internal_notes_table_old = {}
-- Same for new-style declensions.
local internal_notes_table = {}
-- Category and type information corresponding to declensions: These may
-- contain the following fields: 'singular', 'plural', 'decl', 'hard', 'g',
-- 'suffix', 'gensg', 'irregpl', 'alt_nom_pl', 'cant_reduce', 'ignore_reduce',
-- 'stem_suffix'.
--
-- 'singular' is used to construct a category of the form
-- "Russian nouns SINGULAR". If omitted, a category is constructed of the
-- form "Russian nouns ending in -ENDING", where ENDING is the actual
-- nom sg ending shorn of its acute accents; or "Russian nouns ending
-- in suffix -ENDING", if 'suffix' is true. The value of SINGULAR can be
-- one of the following: a single string, a list of strings, or a function,
-- which is passed one argument (the value of ENDING that would be used to
-- auto-initialize the category), and should return a single string or list
-- of strings. Such a category is only constructed if 'gensg' is true.
--
-- 'plural' is analogous but used to construct a category of the form
-- "Russian nouns with PLURAL", and if omitted, a category is constructed
-- of the form "Russian nouns with plural -ENDING", based on the actual
-- nom pl ending shorn of its acute accents. Currently no plural category
-- is actually constructed.
--
-- In addition, a category may normally constructed from the combination of
-- 'singular' and 'plural', appropriately defaulted; e.g. if both are present,
-- the combined category will be "Russian nouns SINGULAR with PLURAL" and
-- if both are missing, the combined category will be
-- "Russian nouns ending in -SGENDING with plural -PLENDING" (or
-- "Russian nouns ending in suffix -SGENDING with plural -PLENDING" if
-- 'suffix' is true). Note that if either singular or plural or both
-- specifies a list, looping will occur over all combinations. Such a
-- category is constructed only if 'irregpl' or 'alt_nom_pl' or 'suffix'
-- is true or if the declension class is a slash class.
--
-- 'decl' is "1st", "2nd", "3rd" or "indeclinable"; 'hard' is "hard", "soft"
-- or "none"; 'g' is "m", "f", "n" or "none"; these are all traditional
-- declension categories.
--
-- If 'suffix' is true, the declension type includes a long suffix
-- added to the string that itself undergoes reducibility and such, and so
-- reducibility cannot occur in the stem minus the suffix. Categories will
-- be created for the suffix.
--
-- 'alt_nom_pl' indicates that the declension has an alternative nominative
-- plural (corresponding to Zaliznyak's special case 1; compare special case 2
-- for alternative genitive plural). 'irregpl' indicates that the entire
-- plural is irregular.
--
-- In addition to the above categories, additional more specific categories
-- are constructed based on the final letter of the stem, e.g.
-- "Russian velar-stem 1st-declension hard nouns". See calls to
-- com.get_stem_trailing_letter_type(). 'stem_suffix', if present, is added to
-- the end of the stem when get_stem_trailing_letter_type() is called.
-- This is the only place that 'stem_suffix' is used. This is for use with
-- the '-ья' and '-ье' declension types, so that the trailing letter is
-- 'ь' and not whatever precedes it.
--
-- 'enable_categories' is a special hack for testing, which disables all
-- category insertion if false. Delete this as soon as we've verified the
-- working of the category code and created all the necessary categories.
local enable_categories = true
-- Category/type info corresponding to old-style declensions; see above.
local declensions_old_cat = {}
-- Category/type info corresponding to new-style declensions. Computed
-- automatically from the old-style ones, for the most part. Same format
-- as the old-style ones.
local declensions_cat = {}
-- Table listing aliases of old-style declension classes.
local declensions_old_aliases = {}
-- Table listing aliases of new-style declension classes; computed
-- automatically from the old-style ones.
local declensions_aliases = {}
local stress_patterns = {}
-- Set of patterns with ending-stressed genitive plural.
local ending_stressed_gen_pl_patterns = {}
-- Set of patterns with ending-stressed prepositional singular.
local ending_stressed_pre_sg_patterns = {}
-- Set of patterns with ending-stressed dative singular.
local ending_stressed_dat_sg_patterns = {}
-- Set of patterns with all singular forms ending-stressed.
local ending_stressed_sg_patterns = {}
-- Set of patterns with all plural forms ending-stressed.
local ending_stressed_pl_patterns = {}
local declinable_cases_except_accusative = {
"nom_sg", "gen_sg", "dat_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "ins_pl", "pre_pl",
}
local accusative_cases_unsplit_animacy = {
"acc_sg", "acc_pl",
}
local accusative_cases_split_animacy = {
"acc_sg_an", "acc_sg_in", "acc_pl_an", "acc_pl_in",
}
local lemma_linked_cases = {
"nom_sg_linked", "nom_pl_linked",
}
local overridable_only_cases = {
"par", "loc", "voc", "par_pl", "loc_pl", "voc_pl", "count", "pauc",
}
local overridable_only_cases_set = m_table.listToSet(overridable_only_cases)
-- List of all cases that are declined normally.
local decl_cases = m_table.append(declinable_cases_except_accusative, accusative_cases_unsplit_animacy)
-- List of all cases that can be overridden (includes all cases except the "linked" lemma case variants). Also
-- currently the same as the cases returned by export.generate_forms().
local overridable_cases = m_table.append(decl_cases, accusative_cases_split_animacy, overridable_only_cases)
-- List of all cases that can be displayed (includes all cases except plain accusatives).
local displayable_cases = m_table.append(declinable_cases_except_accusative, lemma_linked_cases,
accusative_cases_split_animacy, overridable_only_cases)
-- List of all cases, including those that are declined normally (nom/gen/dat/acc/ins/pre sg and pl), plus
-- animate/inanimate accusative variants (computed automatically as appropriate from the previous cases), plus
-- additional overridable cases (loc/par/voc and plural), plus the "linked" lemma case variants used in ru-noun+
-- headwords (nom_sg_linked, nom_pl_linked, whose values come from nom_sg and nom_pl but may have additional embedded
-- links if they were given in the lemma).
local all_cases = m_table.append(overridable_cases, lemma_linked_cases)
local function english_case_description(case)
if case == "par" or case == "loc" or case == "voc" then
-- For historical reasons, the singular of these cases doens't include "_sg" in their code.
case = case .. "_sg"
end
local engcase = rsub(case, "^([a-z]*)", {
nom="nominative", gen="genitive", dat="dative",
acc="accusative", ins="instrumental", pre="prepositional",
par="partitive", loc="locative", voc="vocative",
count="count form", pauc="paucal form",
})
engcase = rsub(engcase, "(_[a-z]*)", {
_sg=" singular", _pl=" plural",
_an="", _in="",
--_an=" animate", _in=" inanimate"
})
return engcase
end
--------------------------------------------------------------------------
-- Tracking and categorization --
--------------------------------------------------------------------------
-- FIXME! Move below the main code
-- FIXME!! Consider deleting most of this tracking code once we've enabled
-- all the categories. Note that some of the tracking categories aren't
-- completely redundant; e.g. we have tracking pages that combine decl and
-- stress classes, such as "а/a" or "о-и/d'", which are more or less
-- equivalent to stem/gender/stress categories, but we also have the same
-- prefixed by "reducible-stem/" for reducible stems.
local function tracking_code(stress, orig_decl, decl, args, n, islast)
assert(orig_decl)
assert(decl)
local hint_types = com.get_stem_trailing_letter_type(args.stem)
if orig_decl == decl then
orig_decl = nil
end
if args.notes then
track("notes")
end
local function all_pl_irreg()
track("irreg")
for _, case in ipairs(overridable_cases) do
if rfind(case, "_pl") then
track("irreg/" .. case)
end
end
end
local function track_prefix(prefix)
local function dotrack(suf)
track(prefix .. suf)
end
dotrack(stress)
dotrack(decl)
dotrack(decl .. "/" .. stress)
if orig_decl then
dotrack(orig_decl)
dotrack(orig_decl .. "/" .. stress)
end
for _, hint_type in ipairs(hint_types) do
dotrack(hint_type)
dotrack(decl .. "/" .. hint_type)
if orig_decl then
dotrack(orig_decl .. "/" .. hint_type)
end
end
end
track_prefix("")
if args.reducible then
track("reducible-stem")
track_prefix("reducible-stem/")
end
if rlfind(args.stem, "и́?н$") and (decl == "" or decl == "#") then
track("irregular-in")
end
if args.pltail then
track("pltail")
end
if args.sgtail then
track("sgtail")
end
if args.pltailall then
track("pltailall")
end
if args.sgtailall then
track("sgtailall")
end
if args.pl ~= args.stem then
track("irreg-pl-stem")
track("irreg")
end
if args.alt_gen_pl then
track("alt-gen-pl")
track("irreg")
track("irreg/gen_pl")
end
if args.want_sc1 then
track("want-sc1")
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-и$") or rfind(decl, "-а$") or rfind(decl, "-я$") then
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-ья$") then
track("variant-ья")
all_pl_irreg()
end
if rfind(decl, "%(ишк%)$") then
track("variant-ишко")
end
if rfind(decl, "%(ищ%)$") then
track("variant-ище")
end
if args.jo_special then
track("jo-special")
end
if args.manual then
track("manual")
end
if args.explicit_gender then
track("explicit-gender")
track("explicit-gender/" .. args.explicit_gender)
end
for _, case in ipairs(overridable_cases) do
if args[case .. n] and not args.manual then
track("override")
track("override/" .. case .. n)
track("irreg")
track("irreg/" .. case .. n)
-- questionable use: track_prefix("irreg/" .. case .. "/")
-- questionable use: track_prefix("irreg/" .. case .. n .. "/")
end
if islast and args[case] and not args.manual then
track("override")
track("override/" .. case)
track("irreg")
track("irreg/" .. case)
-- questionable use: track_prefix("irreg/" .. case .. "/")
end
if args[case .. "_tail"] then
track("casenum-tail")
track("casenum-tail/" .. case)
end
if args[case .. "_tailall"] then
track("casenum-tailall")
track("casenum-tailall/" .. case)
end
end
end
local gender_to_full = {m="masculine", f="feminine", n="neuter"}
local gender_to_short = {m="masc", f="fem", n="neut"}
-- Insert the category CAT (a string) into list CATEGORIES. String will
-- have "Russian " prepended and ~ substituted for the plural part of speech.
local function insert_category(categories, cat, pos, atbeg)
if enable_categories then
local fullcat = "Russian " .. rsub(cat, "~", pos .. "s")
if atbeg then
table.insert(categories, 1, fullcat)
else
table.insert(categories, fullcat)
end
end
end
-- Insert categories into ARGS.CATEGORIES corresponding to the specified
-- stress and declension classes and to the form of the stem (e.g. velar,
-- sibilant, etc.). Also initialize values used to compute the declension
-- heading that describes similar information. N is the number of the
-- word being processed; ISLAST is true if this is the last word.
local function categorize_and_init_heading(stress, decl, args, n, islast)
local function cat_to_list(cat)
if not cat then
return {}
elseif type(cat) == "string" then
return {cat}
else
assert(type(cat) == "table")
return cat
end
end
-- Insert category CAT into the list of categories in ARGS.
-- CAT may be nil, a single string or a list of strings. We call
-- insert_category() on each string. The strings will have "Russian "
-- prepended and "~" replaced with the plural part of speech.
local function insert_cat(cat)
for _, c in ipairs(cat_to_list(cat)) do
insert_category(args.categories, c, args.pos)
end
end
-- "Resolve" the category spec CATSPEC into the sort of category spec
-- accepted by insert_cat(), i.e. nil, a single string or a list of
-- strings. CATSPEC may be any of these or a function, which takes one
-- argument (SUFFIX) and returns another CATSPEC.
local function resolve_cat(catspec, suffix)
if type(catspec) == "function" then
return resolve_cat(catspec(suffix), suffix)
else
return catspec
end
end
-- Check whether an override for nom_sg or nom_pl still contains the
-- normal suffix (which should already have accents removed) in at least
-- one of its entries. If no override then of course we return true.
local function override_matches_suffix(args, case, n, suffix)
assert(suffix == com.remove_accents(suffix))
-- NOTE: It might not be completely correct to pass in args.forms
-- here when n == ""; this is different behavior from
-- handle_overall_forms_and_overrides(). But args.forms is only used
-- to retrieve the dat_sg for handling loc/par, and we're never
-- called with loc or par as the value of CASE, so it doesn't matter.
-- We add an assert to make sure of this.
assert(case ~= "loc" and case ~= "par")
local override = canonicalize_override(args, case, args.forms, n)
if not override then
return true
end
for _, x in ipairs(override) do
local ru, tr = x[1], x[2]
local entry, notes = m_table_tools.separate_notes(ru)
entry = com.remove_accents(m_links.remove_links(entry))
if rlfind(entry, suffix .. "$") then
return true
end
end
return false
end
if args.manual then
return
end
local h = args.heading_info
assert(decl)
local decl_cats = args.old and declensions_old_cat or declensions_cat
local sgdecl, pldecl
local is_slash_decl = rfind(decl, "/")
if is_slash_decl then
local indiv_decls = rsplit(decl, "/")
sgdecl, pldecl = indiv_decls[1], indiv_decls[2]
else
sgdecl, pldecl = decl, decl
end
local sgdc = decl_cats[sgdecl]
local pldc = decl_cats[pldecl]
assert(sgdc)
assert(pldc)
local sghint_types = com.get_stem_trailing_letter_type(
args.stem .. (sgdc.stem_suffix or ""))
-- insert English version of Zaliznyak stem type
if sgdc.decl == "indeclinable" then
insert_cat("indeclinable ~")
insert_if_not(h.stemetc, "indecl")
else
local stem_type =
sgdc.decl == "3rd" and "3rd-declension" or
m_table.contains(sghint_types, "velar") and "velar-stem" or
m_table.contains(sghint_types, "sibilant") and "sibilant-stem" or
m_table.contains(sghint_types, "c") and "ц-stem" or
m_table.contains(sghint_types, "i") and "i-stem" or
m_table.contains(sghint_types, "vowel") and "vowel-stem" or
m_table.contains(sghint_types, "soft-cons") and "vowel-stem" or
m_table.contains(sghint_types, "palatal") and "vowel-stem" or
sgdc.hard == "soft" and "soft-stem" or
"hard-stem"
local short_stem_type = stem_type == "3rd-declension" and "3rd-decl" or stem_type
if sgdc.adj then
-- Don't include gender for pluralia tantum because it's mostly
-- indeterminate (certainly when specified using a plural lemma,
-- which will be usually; technically it's partly or completely
-- determinate in certain old-style adjectives that distinguish
-- masculine from feminine/neuter, but this is too rare a case
-- to worry about)
local gendertext = args.thisn == "p" and "plural-only" or gender_to_full[sgdc.g]
insert_if_not(h.adjectival, "yes")
if args.thisn ~= "p" then
insert_if_not(h.gender, gender_to_short[sgdc.g])
end
if sgdc.possadj then
insert_cat(sgdc.decl .. " possessive " .. gendertext .. " accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " adjectival ~")
insert_if_not(h.stemetc, sgdc.decl .. " poss")
insert_if_not(h.stress, stress)
elseif stem_type == "soft-stem" or stem_type == "vowel-stem" then
insert_cat(stem_type .. " " .. gendertext .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
else
insert_cat(stem_type .. " " .. gendertext .. " accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
else
-- NOTE: There are 8 Zaliznyak-style stem types and 3 genders, but
-- we don't create a category for masculine-form 3rd-declension
-- nouns (there is such a noun, путь, but it mostly behaves
-- like a feminine noun), so there are 23.
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form ~")
-- NOTE: Here we are creating categories for the combination of
-- stem, gender and accent. There are 10 accent patterns and 23
-- combinations of stem and gender, which potentially makes for
-- 10*23 = 230 such categories, which is a lot. Not all such
-- categories should actually exist; there were maybe 75 former
-- declension templates, each of which was essentially categorized
-- by the same three variables, but some of which dealt with
-- ancillary issues like irregular plurals; this amounts to 67
-- actual stem/gender/accent categories, although there are more
-- of them in Zaliznyak (FIXME, how many? See generate_cats.py).
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " ~")
insert_if_not(h.adjectival, "no")
insert_if_not(h.gender, gender_to_short[sgdc.g])
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
insert_cat("~ with accent pattern " .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")))
end
local sgsuffix = args.suffixes.nom_sg
if sgsuffix then
assert(#sgsuffix == 1) -- If this ever fails, then implement a loop
sgsuffix = com.remove_accents(sgsuffix[1])
-- If we are plural only or if nom_sg is overridden and has
-- an unusual suffix, then don't create category for sg suffix
if args.thisn == "p" or not override_matches_suffix(args, "nom_sg", n, sgsuffix) or islast and not override_matches_suffix(args, "nom_sg", "", sgsuffix) then
sgsuffix = nil
end
end
local plsuffix = args.suffixes.nom_pl
if plsuffix then
assert(#plsuffix == 1) -- If this ever fails, then implement a loop
plsuffix = com.remove_accents(plsuffix[1])
-- If we are a singulare tantum or if nom_pl is overridden and has
-- an unusual suffix, then don't create category for pl suffix
if args.thisn == "s" or not override_matches_suffix(args, "nom_pl", n, plsuffix) or islast and not override_matches_suffix(args, "nom_pl", "", plsuffix) then
plsuffix = nil
end
end
sgsuffix = sgsuffix and rsub(sgsuffix, "ъ$", "")
plsuffix = plsuffix and rsub(plsuffix, "ъ$", "")
local sgcat = sgsuffix and (resolve_cat(sgdc.singular, sgsuffix) or "ending in " .. (sgsuffix == "" and "a consonant" or (sgdc.suffix and "suffix " or "") .. "-" .. sgsuffix))
local plcat = plsuffix and (resolve_cat(pldc.plural, plsuffix) or "plural -" .. plsuffix)
if sgcat and sgdc.gensg then
for _, cat in ipairs(cat_to_list(sgcat)) do
insert_cat("~ " .. cat)
end
end
if sgcat and plcat and (sgdc.suffix or sgdc.alt_nom_pl or sgdc.irregpl or
is_slash_decl and plsuffix == "-ья") then
for _, scat in ipairs(cat_to_list(sgcat)) do
for _, pcat in ipairs(cat_to_list(plcat)) do
insert_cat("~ " .. scat .. " with " .. pcat)
end
end
end
if args.pl ~= args.stem then
insert_cat("~ with irregular plural stem")
end
if args.reducible and not sgdc.ignore_reduce then
insert_cat("~ with reducible stem")
if args.soft_n then
insert_cat("~ with soft final н in reduced stem")
end
insert_if_not(h.reducible, "yes")
else
insert_if_not(h.reducible, "no")
end
if args.alt_gen_pl then
insert_cat("~ with alternative genitive plural")
end
if sgdc.adj then
insert_cat("adjectival ~")
end
end
local function compute_heading(args)
local headings = {}
local h = args.heading_info
table.insert(headings, args.a == "a" and "anim" or args.a == "i" and
"inan" or "bian")
table.insert(headings, args.nonumber and "uncountable" or
args.n == "s" and "sg-only" or args.n == "p" and "pl-only" or nil)
if #h.gender > 0 then
table.insert(headings, table.concat(h.gender, "/") .. "-form")
end
if #h.stemetc > 0 then
table.insert(headings, table.concat(h.stemetc, "/"))
end
if #h.stress > 0 then
local stresses = {}
for _, stress in ipairs(h.stress) do
table.insert(stresses, (stress:gsub("''", "ʺ"):gsub("'", "ʹ")))
end
table.insert(headings, "accent-" .. table.concat(stresses, "/"))
end
local function handle_bool(boolvals, text, into)
into = into or headings
if m_table.contains(boolvals, "yes") and m_table.contains(boolvals, "no") then
table.insert(into, "[" .. text .. "]")
elseif m_table.contains(boolvals, "yes") then
table.insert(into, text)
end
end
handle_bool(h.adjectival, "adj")
handle_bool(h.reducible, "reduc")
return headings
end
local function compute_overall_heading_categories_and_genders(args)
local hinfo = args.per_word_heading_info
local index = 0
-- First try for non-adjectival, non-indeclinable
for i=1,#hinfo do
if not m_table.contains(hinfo[i].stemetc, "indecl") and not m_table.contains(hinfo[i].adjectival, "yes") then
index = i
break
end
end
if index == 0 then
-- Then just non-indeclinable
for i=1,#hinfo do
if not m_table.contains(hinfo[i].stemetc, "indecl") then
index = i
break
end
end
end
-- Finally, do anything
if index == 0 then
index = 1
end
-- Compute final heading
local headings = args.per_word_headings[index]
local categories = args.per_word_categories[index]
if args.any_irreg then
table.insert(headings, "irreg")
insert_category(categories, "irregular ~", args.pos)
end
for _, case in ipairs(overridable_cases) do
local is_pl = rfind(case, "_pl")
if args.n == "s" and is_pl or args.n == "p" and not is_pl then
-- Don't create singular categories when plural-only or vice-versa
elseif overridable_only_cases_set[case] then
if args.any_overridden[case] then
insert_category(categories, "~ with " .. english_case_description(case), args.pos)
end
elseif args.any_irreg_case[case] then
insert_category(categories, "~ with irregular " .. english_case_description(case), args.pos)
end
end
local heading = args.manual and "" or "(<span style=\"font-size: smaller;\">[[Appendix:Russian nouns#Declension tables|" .. table.concat(headings, " ") .. "]]</span>)"
args.heading = heading
args.categories = categories
args.genders = args.per_word_genders[index]
end
--------------------------------------------------------------------------
-- Main code --
--------------------------------------------------------------------------
-- Used by do_generate_forms().
local function arg1_is_stress(arg1)
if not arg1 then return false end
for _, arg in ipairs(rsplit(arg1, ",")) do
if not rfind(arg, "^[a-f]'?'?$") then
return false
end
end
return true
end
-- Used by do_generate_forms(), handling a word joiner argument
-- of the form 'join:JOINER'.
local function extract_word_joiner(spec)
word_joiner = rmatch(spec, "^join:(.*)$")
assert(word_joiner)
return com.split_russian_tr(word_joiner, "dopair")
end
local function determine_headword_gender(args, sgdc, gender)
-- If gender unspecified, use normal gender of declension, except when
-- adjectival nouns that are pluralia tantum, where the gender is
-- mostly indeterminate (FIXME, not completely with old-style declensions
-- but we don't handle that currently).
if not gender then
if sgdc.adj and args.thisn == "p" then
gender = nil
else
gender = sgdc.g
end
end
-- Determine headword genders
gender = gender and gender ~= "none" and gender .. "-" or ""
local plsuffix = args.n == "p" and "-p" or ""
local hgens
if args.a == "a" then
hgens = {gender .. "an" .. plsuffix}
elseif args.a == "i" then
hgens = {gender .. "in" .. plsuffix}
elseif args.a == "ai" then
hgens = {gender .. "an" .. plsuffix, gender .. "in" .. plsuffix}
else
hgens = {gender .. "in" .. plsuffix, gender .. "an" .. plsuffix}
end
-- Insert into list of genders
for _, hgen in ipairs(hgens) do
insert_if_not(args.genders, hgen)
end
end
function export.do_generate_forms(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SETS and JOINER, where ARG_SETS
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Gather arguments into a list of ARG_SET objects, containing (potentially)
-- elements 1, 2, 3, 4, corresponding to accent pattern, stem, declension
-- type, pl stem and coming from consecutive numbered parameters. Sets of
-- declension parameters are separated by the word "or".
local arg_sets = {}
-- Find maximum-numbered arg, allowing for holes
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Now gather the arguments.
local offset = 0
local arg_set = {}
for i=1,(max_arg + 1) do
local end_arg_set = false
local end_word = false -- FIXME, is this correct?
local word_joiner
if i == max_arg + 1 then
end_arg_set = true
end_word = true
word_joiner = {""}
elseif args[i] == "_" then
end_arg_set = true
end_word = true
word_joiner = {" "}
elseif args[i] == "-" then
end_arg_set = true
end_word = true
word_joiner = {"-"}
elseif args[i] and rfind(args[i], "^join:") then
end_arg_set = true
end_word = true
word_joiner = extract_word_joiner(args[i])
elseif args[i] == "or" then
end_arg_set = true
end
if end_arg_set then
table.insert(arg_sets, arg_set)
arg_set = {}
offset = i
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
else
-- If the first argument isn't stress, that means all arguments
-- have been shifted to the left one. We want to shift them
-- back to the right one, so we change the offset so that we
-- get the same effect of skipping a slot in the arg set.
if i - offset == 1 and not arg1_is_stress(args[i]) then
offset = offset - 1
end
if i - offset > 4 then
error("Too many arguments for argument set: arg " .. i .. " = " .. (args[i] or "(blank)"))
end
arg_set[i - offset] = args[i]
end
end
return generate_forms_1(args, per_word_info)
end
function export.do_generate_forms_multi(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SET and JOINER, where ARG_SET
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Find maximum-numbered arg, allowing for holes (FIXME: Is this needed
-- here? Will there be holes?)
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Gather arguments into a list of ARG_SET objects, containing
-- (potentially) elements 1, 2, 3, corresponding to accent pattern,
-- lemma+declension spec, pl stem, exactly as with do_generate_forms()
-- and {{ru-noun-table}} except that the values come from a single argument
-- of the form ACCENTPATTERN:LEMMADECL:PL where all but LEMMADECL may (and
-- probably will be) omitted and LEMMADECL may be of the following forms:
-- LEMMA (for a noun with empty decl spec),
-- LEMMADECL (for a noun with non-empty decl spec beginning with a
-- *, left paren or semicolon),
-- LEMMA^DECL (for a noun with non-empty decl spec),
-- LEMMA$ (for an indeclinable word)
-- LEMMA+ (for an adjective with auto-detected decl class)
-- LEMMA+DECL (for an adjective with explicit decl class)
-- Sets of parameters for the same word are separated by the word "or".
local arg_sets = {}
local continue_arg_sets = true
for i=1,(max_arg + 1) do
local end_word = false
local word_joiner
local process_arg = false
if i == max_arg + 1 then
end_word = true
word_joiner = {""}
elseif args[i] == "-" then
end_word = true
word_joiner = {"-"}
continue_arg_sets = true
elseif rfind(args[i], "^join:") then
end_word = true
word_joiner = extract_word_joiner(args[i])
continue_arg_sets = true
elseif args[i] == "or" then
continue_arg_sets = true
else
if continue_arg_sets then
continue_arg_sets = false
else
end_word = true
word_joiner = {" "}
end
process_arg = true
end
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
if process_arg then
local vals = rsplit(args[i], ":")
if #vals > 3 then
error("Can't specify more than 3 colon-separated params of param set: " .. args[i])
end
local arg_set = {}
if arg1_is_stress(vals[1]) then
arg_set[1] = vals[1]
arg_set[2] = vals[2]
arg_set[4] = vals[3]
else
arg_set[2] = vals[1]
arg_set[4] = vals[2]
end
-- recognize indeclinable
local indecl_stem = rmatch(arg_set[2], "^(.-)%$$")
if indecl_stem then
arg_set[2] = indecl_stem
arg_set[3] = "$"
else
-- recognize adjective
local adj_stem, adj_type = rmatch(arg_set[2], "^(.*)(%+.*)$")
if adj_stem then
arg_set[2] = adj_stem
arg_set[3] = adj_type
else
-- recognize noun with ^
local noun_stem, noun_type = rmatch(arg_set[2], "^(.*)%^(.*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- recognize noun without ^ but with decl spec
noun_stem, noun_type = rmatch(arg_set[2], "^(..-)([;*(].*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- noun without ^ or decl spec; nothing to do
end
end
end
end
table.insert(arg_sets, arg_set)
end
end
return generate_forms_1(args, per_word_info)
end
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence.
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence,
-- as well as show_z() for template {{ru-decl-noun-z}}, which has a
-- subset of the functionality of the other two.
generate_forms_1 = function(args, per_word_info)
local orig_args
if test_new_ru_noun_module then
orig_args = mw.clone(args)
end
local pagename = args.pagename or mw.loadData("Module:headword/data").pagename
local old = args.old
local function verify_animacy_value(val)
if not val then return nil end
if val == "a" or val == "an" or val == "anim" then
return "a"
elseif val == "i" or val == "in" or val == "inan" then
return "i"
elseif val == "b" or val == "bi" or val == "both" or val == "ai" then
return "ai"
elseif val == "ia" then
return "ia"
end
error("Animacy value " .. val .. " should be empty or a/an/anim (animate), i/in/inan (inanimate), b/bi/both/ai (bianimate, listing animate first), or ia (bianimate, listing inanimate first)")
return nil
end
local function verify_number_value(val, allow_none)
if not val then return nil end
local short = usub(val, 1, 1)
if short == "s" or short == "p" or short == "b" or (allow_none and short == "n" or false) then
return short
end
if allow_none then
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), 'b' (both) or 'n' (none)")
else
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), or 'b' (both)")
end
return nil
end
-- Verify and canonicalize animacy, number, prefix, suffix
assert(#per_word_info >= 1)
for i=1,#per_word_info do
args["a" .. i] = verify_animacy_value(args["a" .. i])
args["n" .. i] = verify_number_value(args["n" .. i])
args["prefix" .. i] = com.split_russian_tr(args["prefix" .. i] or "", "dopair")
args["suffix" .. i] = com.split_russian_tr(args["suffix" .. i] or "", "dopair")
end
args.a = verify_animacy_value(args.a) or "i"
-- args.ndef, if set, is the default value for args.n; if unset, it defaults
-- to "both". It is set to "singular" in ru-proper noun+. We store the value
-- of args.n in args.orign before defaulting to args.ndef because we may
-- change it to plural-only later on if it was unspecified (this happens if
-- an individual word's lemma is plural), and to determine whether it was
-- unspecified, we need the original value before defaulting.
args.n = verify_number_value(args.n, "allow none")
-- treat n=none like n=sg but set a flag so "singular" isn't displayed
if args.n == "n" then
args.n = "s"
args.nonumber = true
end
args.ndef = verify_number_value(args.ndef)
args.orign = args.n
args.n = args.n or args.ndef
args.prefix = com.split_russian_tr(args.prefix or "", "dopair")
args.suffix = com.split_russian_tr(args.suffix or "", "dopair")
-- Attach overall prefix to first per-word prefix, similarly for suffix
args.prefix1 = com.concat_paired_russian_tr(args.prefix, args.prefix1)
args["suffix" .. #per_word_info] = com.concat_paired_russian_tr(
args["suffix" .. #per_word_info], args.suffix)
-- Initialize non-word-specific arguments.
--
-- The following is a list of WORD_INFO items, one per word, each of
-- which is a two element list of WORD_FORMS (a table listing the forms for
-- each case) and JOINER (a string, indicating how to join the word with
-- the next one).
args.per_word_info = {}
-- List of HEADING_INFO items, one per word, containing the raw material
-- used to generate the header. Initialized from 'args.heading_info',
-- which is initialized in categorize_and_init_heading().
args.per_word_heading_info = {}
-- List of CATEGORIES items, one per word, containing the categories
-- used to initialize the page categories. Initialized from
-- 'args.categories', which is initialized in categorize_and_init_heading().
args.per_word_categories = {}
-- List of HEADINGS items, one per word, containing the actual words that
-- go into the header if we were to use that word to construct the header.
-- Comes from compute_heading(). We use this to generate the actual header
-- string, which goes into 'args.heading' at the end (done in
-- compute_overall_heading_categories_and_genders()).
args.per_word_headings = {}
-- List of GENDERS items, one per word, containing the headword genders
-- for each word (where "headword gender" is in the format used in
-- headwords and actually includes animacy and number as well, e.g.
-- 'm-in' or 'f-an-p'). We end up selecting one such item and putting
-- it into 'args.genders' at the end
-- (in compute_overall_heading_categories_and_genders()).
args.per_word_genders = {}
args.any_overridden = {}
args.any_non_nil = {}
args.any_irreg = false
args.any_irreg_case = {}
local function insert_cat(cat)
insert_category(args.categories, cat, args.pos)
end
args.internal_notes = {}
local decl_sufs = old and declensions_old or declensions
local decl_cats = old and declensions_old_cat or declensions_cat
local intable = old and internal_notes_table_old or internal_notes_table
-- Default lemma defaults to previous lemma, first one to page name.
-- FIXME: With multiple words, we should probably split the page name
-- on spaces and default each word in turn
local default_lemma
local all_stresses_seen
-- Made into a function to avoid having to indent a lot of code.
-- Process a single arg set of a single word. This inserts the forms
-- for the word into args.forms and sets categories and tracking pages.
local function do_arg_set(arg_set, n, islast)
local stress_arg = arg_set[1]
local decl = arg_set[3] or ""
local pl, pltr
if arg_set[4] then
pl, pltr = com.split_russian_tr(arg_set[4])
end
-- Extract special markers from declension class.
if decl == "manual" then
decl = "$"
args.manual = true
if #per_word_info > 1 or #per_word_info[1][1] > 1 then
error("Can't specify multiple words or argument sets when manual")
end
if pl then
error("Can't specify optional stem parameters when manual")
end
end
decl, args.jo_special = rsubb(decl, "([^/%a])ё$", "%1")
if not args.jo_special then
decl, args.jo_special = rsubb(decl, "([^/%a])ё([^/%a])", "%1%2")
end
decl, args.want_sc1 = rsubb(decl, "%(1%)", "")
decl, args.alt_gen_pl = rsubb(decl, "%(2%)", "")
decl, args.reducible = rsubb(decl, "%*", "")
decl, args.soft_n = rsubb(decl, "%(нь%)", "")
decl = rsub(decl, ";", "")
-- Get the lemma.
local lemma = args.manual and "-" or arg_set[2] or default_lemma
if not lemma then
error("Lemma in first argument set must be specified")
end
default_lemma = lemma
local lemmatr
lemma, lemmatr = com.split_russian_tr(lemma)
-- If we're conjugating a suffix, insert a pseudoconsonant at the beginning
-- of all forms, so they get conjugated as if ending in a consonant.
-- We remove the pseudoconsonant later.
local is_suffix = lemma ~= "-" and rfind(lemma, "^%-")
args.any_suffix = args.any_suffix or is_suffix
local asif_prefix = args["asif_prefix" .. n] or args.asif_prefix or is_suffix and PSEUDOCONS
if asif_prefix then
lemma = rsub(lemma, "^%-", "-" .. asif_prefix)
if lemmatr then
lemmatr = rsub(lemmatr, "^%-", "-" .. com.translit(asif_prefix))
end
end
args.thisa = args["a" .. n] or args.a
args.thisn = args["n" .. n] or args.n
-- Check for explicit allow-unaccented indication.
local allow_unaccented
lemma, allow_unaccented = rsubb(lemma, "^%*", "")
args.allow_unaccented = args.allow_unaccented or allow_unaccented
if args.allow_unaccented then
track("allow-unaccented")
end
args.orig_lemma = lemma
lemma = m_links.remove_links(lemma)
args.lemma_no_links = lemma
args.lemmatr = lemmatr
if args.lemma then
-- Explicit lemma given.
args.explicit_lemma, args.explicit_lemmatr = com.split_russian_tr(args.lemma)
end
-- Treat suffixes without an accent, and suffixes with an accent on the
-- initial hyphen, as if they were preceded with a *, which overrides
-- all the logic that normally (a) normalizes the accent, and (b)
-- complains about multisyllabic words without an accent. Don't do this
-- if lemma is just -, which is used specially in manual declension
-- tables (e.g. сто, три).
if lemma ~= "-" and (rfind(lemma, "^%-" .. AC) or (com.is_unstressed(lemma) and rfind(lemma, "^%-"))) then
args.allow_unaccented = true
end
-- Convert lemma and decl arg into stem and canonicalized decl.
-- This will autodetect the declension from the lemma if an explicit
-- decl isn't given.
local stem, tr, gender, was_accented, was_plural, was_autodetected
if rfind(decl, "^%+") then
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
detect_adj_type(lemma, lemmatr, decl, old)
else
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
determine_decl(lemma, lemmatr, decl, args)
end
if was_plural then
args.n = args.orign or "p"
args.thisn = args["n" .. n] or args.n
elseif decl ~= "$" then
args.thisn = args.thisn or "b"
end
args.explicit_gender = gender
-- If allow-unaccented not given, maybe check for missing accents.
if not args.allow_unaccented and not stress_arg and was_autodetected and com.needs_accents(lemma) then
-- If user gave the full word and expects us to determine the
-- declension and stress, the word should have an accent on the
-- stem or ending. We have a separate check farther below for
-- an accent on a multisyllabic stem, after stripping off any
-- ending; but this way we get an error if the user e.g. writes
-- "гора" without an accent rather than assuming it's stem
-- stressed, as would otherwise happen.
error("Lemma must have an accent in it: " .. lemma)
end
-- If stress not given, auto-determine; else validate/canonicalize
-- stress arg, override in certain cases and convert to list.
if not stress_arg then
stress_arg = {detect_stress_pattern(stem, decl, decl_cats, args.reducible, was_plural, was_accented)}
else
stress_arg = rsplit(stress_arg, ",")
for i=1,#stress_arg do
local stress = stress_arg[i]
stress = override_stress_pattern(decl, stress)
if not stress_patterns[stress] then
error("Unrecognized accent pattern " .. stress)
end
stress_arg[i] = stress
end
end
-- parse slash decl to list
local sub_decls
if rfind(decl, "/") then
track("mixed-decl")
insert_cat("~ with mixed declension")
local indiv_decls = rsplit(decl, "/")
-- Should have been caught in canonicalize_decl()
assert(#indiv_decls == 2)
sub_decls = {{indiv_decls[1], "sg"}, {indiv_decls[2], "pl"}}
else
sub_decls = {{decl}}
end
-- Get singular declension and corresponding category.
local sgdecl = sub_decls[1][1]
local sgdc = decl_cats[sgdecl]
assert(sgdc)
-- Compute headword gender(s). We base it off the singular declension
-- if we have a slash declension -- it's the best we can do.
determine_headword_gender(args, sgdc, gender)
local original_stem, original_tr = stem, tr
local original_pl, original_pltr = pl, pltr
-- Loop over accent patterns in case more than one given.
for _, stress in ipairs(stress_arg) do
args.suffixes = {}
stem, tr = original_stem, original_tr
local bare, baretr
local stem_for_bare, tr_for_bare
pl, pltr = original_pl, original_pltr
insert_if_not(all_stresses_seen, stress)
local stem_was_unstressed = com.is_unstressed(stem)
-- If special case ;ё was given and stem is unstressed,
-- add ё to the stem now; but don't let this interfere with
-- restressing, to handle cases like железа́ with gen pl желёз
-- but nom pl же́лезы.
if stem_was_unstressed and args.jo_special then
-- Beware, Cyrillic еЕ in first rsub, Latin eE in second
local new_stem = rsub(stem, "([еЕ])([^еЕ]*)$",
function(e, rest)
return (e == "Е" and "Ё" or "ё") .. rest
end
)
if stem == new_stem then
error("No е in stem to replace with ё")
end
stem = new_stem
if tr then
local subbed
-- e after j -> o, e not after j -> jo; don't just convert e -> jo
-- and then map jjo -> jo because we want to preserve double j
tr, subbed = rsubb(tr, "([jJ])([eE])([^eE]*)$",
function(j, e, rest)
return j .. (e == "E" and "O" or "o") .. AC .. rest
end
)
if not subbed then
tr = rsub(tr, "([eE])([^eE]*)$",
function(e, rest)
return (e == "E" and "Jo" or "jo") .. AC .. rest
end
)
end
tr = com.j_correction(tr)
end
-- This is used to handle железа́ with gen pl желёз and nom pl
-- же́лезы. We have two stressed stems, one for the gen pl and
-- one for the remaining pl cases, and the variable 'stem' can
-- handle only one, so we put the second (gen pl) stem in
-- stem_for_bare, which goes into the value of 'bare' (used
-- only for gen pl here).
stem_for_bare, tr_for_bare = stem, tr
end
-- Maybe add stress to the stem, depending on whether the
-- stem was unstressed and the stress pattern. Stem pattern f
-- and variants call for initial stress (голова́ -> го́ловы);
-- stem pattern d and variants call for stem-final stress
-- (сапожо́к -> сапо́жки). Stem patterns b and b' apparently
-- call for stem-final stress as well but it's unlikely to
-- make much of a difference (pattern b' only occurs in 3rd-decl
-- feminines, which should already have stress in the stem,
-- and the only place pattern b gets stem stress is in bare
-- forms, i.e. nom sg and/or gen pl depending on the decl type,
-- and nom sg stress should already be in the lemma while
-- gen pl stress is handled by a different ending-stressing
-- mechanism in attach_unstressed(); however, the user is
-- free to leave a masc or 3rd-decl fem lemma completely
-- unstressed with pattern b, and then the stem-final stress
-- *will* make a difference).
local function restress_stem(stem, tr, stress, stem_unstressed)
-- If the user has indicated they purposely are leaving the
-- word unstressed by putting a * at the beginning of the main
-- stem, leave it unstressed. This might indicate lack of
-- knowledge of the stress or a truly unaccented word
-- (e.g. an unaccented suffix).
if args.allow_unaccented then
return stem, tr
end
if tr and com.is_unstressed(stem) ~= com.is_unstressed(tr) then
error("Stem " .. stem .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(stem) then
stem, tr = com.make_ending_stressed(stem, tr)
-- For those patterns that are ending-stressed in the singular
-- nominative (and hence are likely to be expressed without an
-- accent on the stem) it's safe to put a particular accent on
-- the stem depending on the stress type. Otherwise, give an
-- error if no accent.
elseif stem_unstressed then
if rfind(stress, "^f") then
stem, tr = com.make_beginning_stressed(stem, tr)
elseif (rfind(stress, "^[bd]") or
args.thisn == "p" and ending_stressed_pl_patterns[stress]) then
stem, tr = com.make_ending_stressed(stem, tr)
elseif com.needs_accents(stem) then
error("Stem " .. stem .. " requires an accent")
end
end
return stem, tr
end
stem, tr = restress_stem(stem, tr, stress, stem_was_unstressed)
-- Leave pl unaccented if user wants this; see restress_stem().
if pl and not args.allow_unaccented then
if pltr and com.is_unstressed(pl) ~= com.is_unstressed(pltr) then
error("Plural stem " .. pl .. " and translit " .. pltr .. " must have same accent pattern")
end
if com.is_monosyllabic(pl) then
pl, pltr = com.make_ending_stressed(pl, pltr)
end
-- I think this is safe.
if com.needs_accents(pl) then
if ending_stressed_pl_patterns[stress] then
pl, pltr = com.make_ending_stressed(pl, pltr)
elseif not args.allow_unaccented then
error("Plural stem " .. pl .. " requires an accent")
end
end
end
local resolved_bare, resolved_baretr
-- Handle (de)reducibles
-- FIXME! We are dereducing based on the singular declension.
-- In a slash declension things can get weird and we don't
-- handle that. We are also computing the bare value from the
-- singular stem, and again things can get weird with a plural
-- stem. Note that we don't compute a bare value unless we have
-- to (either (de)reducible or stress pattern f/f'/f'' combined
-- with ё special case); the remaining times we generate the bare
-- value directly from the plural stem.
if args.reducible and not sgdc.ignore_reduce then
-- Zaliznyak treats all nouns in -ье and -ья as being
-- reducible. We handle this automatically and don't require
-- the user to specify this, but ignore it if so for
-- compatibility.
if is_reducible(sgdc) then
-- If we derived the stem from a nom pl form, then
-- it's already reduced, and we need to dereduce it to
-- get a bare form; otherwise the stem comes from the
-- nom sg and we need to reduce it to get the real stem.
if was_plural then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
resolved_bare, resolved_baretr = stem, tr
stem, tr = export.reduce_nom_sg_stem(stem, tr,
sgdecl, args.soft_n, "error")
-- Stem will be unstressed if stress was on elided
-- vowel; restress stem the way we did above. (This is
-- needed in at least one word, сапожо́к 3*d(2), with
-- plural stem probably сапо́жк- and gen pl probably
-- сапо́жек.)
stem, tr = restress_stem(stem, tr, stress,
com.is_unstressed(stem))
if stress ~= "a" and stress ~= "b" and args.alt_gen_pl and not pl then
-- Nouns like рожо́к, глазо́к of type 3*d(2) have
-- gen pl's ро́жек, гла́зок; to handle this,
-- dereduce the reduced stem and store in a
-- special place.
args.gen_pl_bare, args.gen_pl_baretr =
export.dereduce_nom_sg_stem(stem, tr,
sgdc, stress, old, "error")
end
end
elseif is_dereducible(sgdc) then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
error("Declension class " .. sgdecl .. " not (de)reducible")
end
elseif stem_for_bare and stem ~= stem_for_bare then
resolved_bare, resolved_baretr =
add_bare_suffix(stem_for_bare, tr_for_bare, old, sgdc, false)
end
-- Leave unaccented if user wants this; see restress_stem().
-- FIXME, we no longer allow the user to specify the bare value
-- so it's unclear if this is needed any more.
if resolved_bare and not args.allow_unaccented then
if resolved_baretr and com.is_unstressed(resolved_bare) ~= com.is_unstressed(resolved_baretr) then
error("Resolved bare stem " .. resolved_bare .. " and translit " .. resolved_baretr .. " must have same accent pattern")
end
if com.is_monosyllabic(resolved_bare) then
resolved_bare, resolved_baretr =
com.make_ending_stressed(resolved_bare, resolved_baretr)
else
if com.needs_accents(resolved_bare) then
error("Resolved bare stem " .. resolved_bare .. " requires an accent")
end
end
end
args.stem, args.stemtr = stem, tr
args.bare, args.baretr = resolved_bare, resolved_baretr
args.ustem, args.ustemtr = com.make_unstressed_once(stem, tr)
if pl then
args.pl, args.pltr = pl, pltr
else
args.pl, args.pltr = stem, tr
end
args.upl, args.upltr = com.make_unstressed_once(args.pl, args.pltr)
-- Special hack for любо́вь and other reducible 3rd-fem nouns,
-- which have the full stem in the ins sg
args.ins_sg_stem = sgdecl == "ь-f" and args.reducible and resolved_bare
args.ins_sg_tr = sgdecl == "ь-f" and args.reducible and resolved_baretr
-- Loop over declension classes (we may have two of them, one for
-- singular and one for plural, in the case of a mixed declension
-- class of the form SGDECL/PLDECL).
for _,decl_spec in ipairs(sub_decls) do
local orig_decl = decl_spec[1]
local number = decl_spec[2]
local real_decl =
determine_stress_variant(orig_decl, stress)
real_decl = determine_stem_variant(real_decl,
number == "pl" and args.pl or args.stem)
-- sanity checking; errors should have been caught in
-- canonicalize_decl()
assert(decl_cats[real_decl], "real_decl " .. real_decl .. " nonexistent")
assert(decl_sufs[real_decl], "real_decl " .. real_decl .. " nonexistent")
tracking_code(stress, orig_decl, real_decl, args, n, islast)
do_stress_pattern(stress, args, real_decl, number, n, islast)
-- handle internal notes
local internal_note = intable[real_decl]
if internal_note then
insert_if_not(args.internal_notes, internal_note)
end
end
categorize_and_init_heading(stress, decl, args, n, islast)
end
end
local n = 0
for _, word_info in ipairs(per_word_info) do
n = n + 1
local islast = n == #per_word_info
local arg_sets, joiner = word_info[1], word_info[2]
args.forms = {}
args.heading_info = {animacy={}, number={}, gender={}, stress={},
stemetc={}, adjectival={}, reducible={}}
args.categories = {}
args.genders = {}
args.this_any_non_nil = {}
args.any_suffix = false
if #arg_sets > 1 then
track("multiple-arg-sets")
insert_cat("~ with multiple argument sets")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
default_lemma = pagename
all_stresses_seen = {}
-- Loop over all arg sets.
for _, arg_set in ipairs(arg_sets) do
do_arg_set(arg_set, n, islast)
end
if #all_stresses_seen > 1 then
track("multiple-accent-patterns")
insert_cat("~ with multiple accent patterns")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
table.insert(args.per_word_heading_info, args.heading_info)
table.insert(args.per_word_categories, args.categories)
local headings = compute_heading(args)
table.insert(args.per_word_headings, headings)
table.insert(args.per_word_genders, args.genders)
handle_forms_and_overrides(args, n, islast)
if args.any_suffix then
-- If we're conjugating a suffix, remove the pseudoconsonant or asif_prefix
-- that we previously inserted at the beginning.
local asif_prefix = args["asif_prefix" .. n] or args.asif_prefix or PSEUDOCONS
local asif_prefix_tr = com.translit(asif_prefix)
for _, case in ipairs(all_cases) do
if args.forms[case] then
local newforms = {}
for _, form in ipairs(args.forms[case]) do
local formru = form[1]
local formtr = form[2]
formru = rsub(formru, "^%-" .. asif_prefix, "-")
if formtr then
formtr = rsub(formtr, "^%-" .. asif_prefix_tr, "-")
end
if formru == "-" then
-- if no ending, insert "(no suffix)".
table.insert(newforms, {"(no suffix)"})
else
table.insert(newforms, {formru, formtr})
end
end
args.forms[case] = newforms
end
end
end
table.insert(args.per_word_info, {args.forms, joiner})
end
handle_overall_forms_and_overrides(args)
compute_overall_heading_categories_and_genders(args)
for _, case in ipairs(all_cases) do
if args[case] then
for _, form in ipairs(args[case]) do
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
ruentry = m_links.remove_links(ruentry)
if rfind(ulower(ruentry), latin_text_class) then
--error("Found Latin text " .. ruentry .. " in case " .. case)
track("latin-text")
track("latin-text/" .. case)
end
end
end
end
-- Test code to compare existing module to new one.
if test_new_ru_noun_module then
local m_new_ru_noun = require("Module:User:Benwing2/ru-noun")
local newargs = m_new_ru_noun.do_generate_forms(orig_args, old)
local difdecl = false
for _, case in ipairs(all_cases) do
local arg = args[case]
local newarg = newargs[case]
local is_pl = rfind(case, "_pl")
if args.thisn == "s" and is_pl or args.thisn == "p" and not is_pl then
-- Don't need to check cases that won't be displayed.
elseif not m_table.deepEquals(arg, newarg) then
local monosyl_accent_diff = false
-- Differences only in monosyllabic accents. Enable if we
-- change the algorithm for these.
--if arg and newarg and #arg == 1 and #newarg == 1 then
-- local ru1, tr1 = arg[1][1], arg[1][2]
-- local ru2, tr2 = newarg[1][1], newarg[1][2]
-- if com.is_monosyllabic(ru1) and com.is_monosyllabic(ru2) then
-- ru1, tr1 = com.remove_accents(ru1, tr1)
-- ru2, tr2 = com.remove_accents(ru2, tr2)
-- if ru1 == ru2 and tr1 == tr2 then
-- monosyl_accent_diff = true
-- end
-- end
--end
if monosyl_accent_diff then
track("monosyl-accent-diff")
difdecl = true
else
-- Uncomment this to display the particular case and
-- differing forms.
--error(case .. " " .. (arg and com.concat_forms(arg) or "nil") .. " || " .. (newarg and com.concat_forms(newarg) or "nil"))
track("different-decl")
difdecl = true
end
break
end
end
if not difdecl then track("same-decl") end
end
return args
end
-- Implementation of main entry point
local function do_show(frame, old)
local args = clone_args(frame)
local args = export.do_generate_forms(args, old)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The main entry point for modern declension tables.
function export.show(frame)
return do_show(frame, false)
end
-- The main entry point for old declension tables.
function export.show_old(frame)
return do_show(frame, true)
end
-- Implementation of new entry point, esp. for multiple words
local function do_show_multi(frame)
local args = clone_args(frame)
local args = export.do_generate_forms_multi(args)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The new entry point, esp. for multiple words (but works fine for
-- single words).
function export.show_multi(frame)
return do_show_multi(frame)
end
local function get_form(forms, preserve_links, raw)
local canon_forms = {}
for _, form in ipairs(forms) do
if raw then
local ru, tr = form[1], form[2]
ru = rsub(ru, "|", "<!>")
if tr then
tr = rsub(tr, "|", "<!>")
end
insert_if_not(canon_forms, {ru, tr})
else
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
-- Skip hypothetical forms (but include in the raw versions)
if not rfind(runotes, HYPMARKER) then
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.separate_notes(tr)
end
if not preserve_links then
ruentry = m_links.remove_links(ruentry)
end
ruentry = rsub(ruentry, "|", "<!>")
if trentry then
trentry = rsub(trentry, "|", "<!>")
end
insert_if_not(canon_forms, {ruentry, trentry})
end
end
end
return com.concat_forms(canon_forms)
end
local function case_will_be_displayed(args, case)
local ispl = rfind(case, "_pl")
local caseok = true
if args.n == "p" then
caseok = ispl
elseif args.n == "s" then
caseok = not ispl
end
for _, override_case in ipairs(overridable_only_cases) do
if case == override_case and not args.any_overridden[override_case] then
caseok = false
break
end
end
if args.a == "a" or args.a == "i" then
if rfind(case, "_[ai]n") then
caseok = false
end
else -- bianimate
-- don't include inanimate/animate variants if combined variant exists
-- (typically because inanimate/animate variants are the same);
-- FIXME: This could conceivably be different from how the display
-- code works, which just checks that the inanimate/animate variants
-- are the same when deciding whether to display them, in particular
-- if there is an override. Here we are following the algorithm of
-- handle_overall_forms_and_overrides().
if (case == "acc_sg_in" or case == "acc_sg_an") and args.acc_sg or
(case == "acc_pl_in" or case == "acc_pl_an") and args.acc_pl then
caseok = false
end
end
if not args[case] then
caseok = false
end
return caseok
end
local function concat_case_args(args, do_all, raw)
local ins_text = {}
for _, case in ipairs(do_all and all_cases or overridable_cases) do
if case_will_be_displayed(args, case) then
local forms = get_form(args[case], rfind(case, "_linked"), raw)
if forms ~= "" then
table.insert(ins_text, case .. (raw and "_raw" or "") .. "=" ..
forms)
end
end
end
return table.concat(ins_text, "|")
end
-- The entry point for 'ru-noun-forms' to generate all noun forms.
-- This returns a single string, with | separating arguments and named
-- arguments of the form NAME=VALUE.
function export.generate_forms(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
return concat_case_args(args)
end
-- The entry point to generate multiple sets of noun forms. This is a hack
-- to speed up calling from a bot, where we often want to compare old and new
-- argument results to make sure they're the same. Each set of arguments is
-- jammed together into a single argument with individual values separated by
-- <!>; named arguments are of the form NAME<->VALUE. The return value for
-- each set of arguments is as in export.generate_forms(), and the return
-- values are concatenated with <!> separating them. NOTE: This will fail if
-- the exact sequences <!> or <-> happen to occur in values (which is unlikely,
-- esp. as we don't even use the characters <, ! or > for anything) and aren't
-- HTML-escaped.
function export.generate_multi_forms(frame)
local retvals = {}
for _, argset in ipairs(frame.args) do
local args = {}
local i = 0
local argvals = rsplit(argset, "<!>")
for _, argval in ipairs(argvals) do
local split_arg = rsplit(argval, "<%->")
if #split_arg == 1 then
i = i + 1
args[i] = ine(split_arg[1])
else
assert(#split_arg == 2)
args[split_arg[1]] = ine(split_arg[2])
end
end
args = export.do_generate_forms(args, false)
table.insert(retvals, concat_case_args(args))
end
return table.concat(retvals, "<!>")
end
-- The entry point for 'ru-noun-form' to generate a particular noun form.
function export.generate_form(frame)
local args = clone_args(frame)
if not args.form then
error("Must specify desired form using form=")
end
local form = args.form
if not m_table.contains(all_cases, form) then
error("Unrecognized form " .. form)
end
local args = export.do_generate_forms(args, false)
if not args[form] then
return ""
else
return get_form(args[form])
end
end
-- The entry point for generating arguments of various sorts, including
-- the case forms, gender, number and animacy.
function export.generate_args(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
local retargs = {}
table.insert(retargs, concat_case_args(args, "doall"))
table.insert(retargs, concat_case_args(args, "doall", "raw"))
table.insert(retargs, "g=" .. table.concat(args.genders, ","))
-- The following is correct even with ndef because if ndef is
-- set we will set it in args.n.
table.insert(retargs, "n=" .. (args.n or "b"))
table.insert(retargs, "a=" .. (args.a or "i"))
return table.concat(retargs, "|")
end
-- The entry point for compatibility with {{ru-decl-noun-z}}.
function export.show_z(frame)
local args = clone_args(frame)
local stem = args[1]
local stress = args[2]
local specific = args[4] or ""
-- Parse gender/animacy spec
local gender, anim = rmatch(args[3], "^([mfn])-([a-z]+)")
if not gender then
error("Unrecognized gender/anim spec " .. args[3])
end
if anim ~= "an" and anim ~= "in" then anim = "both" end
args.a = anim
-- Handle specific
specific = rsub(specific, "ё", ";ё")
-- Compute decl; special case for семьянин (perhaps not necessary)
local decl = com.make_unstressed_once(stem) == "семьянин" and "#" .. specific or gender .. specific
-- Handle overrides
args.pre_sg = args.prp_sg
args.pre_pl = args.prp_pl
args.notes = args.note
if args.par then
args.par = "+"
end
if args.loc then
if args.loc == "в" then
args.loc = "в +"
elseif args.loc == "на" then
args.loc = "на +"
else
args.loc = "в +,на +"
end
end
local arg_set = {}
table.insert(arg_set, stress)
table.insert(arg_set, stem)
table.insert(arg_set, decl)
local per_word_info = {{{arg_set}, ""}}
return generate_forms_1(args, per_word_info)
end
--------------------------------------------------------------------------
-- Autodetection and lemma munging --
--------------------------------------------------------------------------
-- Attempt to detect the type of the lemma based on its ending, separating
-- off the stem and the ending. GENDER must be present with -ь and plural
-- stems, and is otherwise ignored. Return up to three values: The stem
-- (lemma minus ending), the singular lemma ending, and if the lemma was
-- plural, the plural lemma ending. If the lemma was singular, the singular
-- lemma ending will contain any user-given accents; likewise, if the
-- lemma was plural, the plural ending will contain such accents.
-- VARIANT comes from the declension spec and controls certain declension
-- variants.
local function detect_lemma_type(lemma, tr, gender, args, variant)
local base, ending = rmatch(lemma, "^(.*)([еЕ]" .. AC .. ")$") -- accented
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([еЕ])$") -- unaccented
if base then
if variant == "-ище" and not rfind(lemma, "[щЩ][еЕ]$") then
error("With declension variant -ище, lemma should end in -ще: " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), variant == "-ище" and "(ищ)е-и" or "о"
end
if variant == "-ишко" then
base, ending = rmatch(lemma, "^(.*[шШ][кК])([оО])$") -- unaccented
if not base then
error("With declension variant -ишко, lemma should end in -шко: " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), "(ишк)о-и"
end
if variant == "-ин" then
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?[нН][ъЪ]?)$") -- maybe accented
if not base then
error("With declension variant -ин, lemma should end in -ин(ъ): " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
-- Now autodetect -ин; only animate and in -анин/-янин
base, ending = rmatch(lemma, "^(.*[аАяЯ][" .. AC .. GR .. "]?[нН])([иИ][" .. AC .. GR .. "]?[нН][ъЪ]?)$")
-- Need to check the animacy to avoid nouns like маиганин, цианин,
-- меланин, соланин, etc.
if base and args.thisa == "a" then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]" .. AC .. "?[нН][оО][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]" .. AC .. "[нН][оО][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]" .. AC .. "?[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]" .. AC .. "[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([мМ][яЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
--recognize plural endings
if gender == "n" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ][" .. AC .. GR .. "]?)$")
if base then
-- Don't do this; о/-ья is too rare
-- error("Ambiguous plural lemma " .. lemma .. " in -ья, singular could be -о or -ье/-ьё; specify the singular")
return base, com.strip_tr_ending(tr, ending), "ье", ending
end
base, ending = rmatch(lemma, "^(.*)([аяАЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), rfind(ending, "[аА]") and "о" or "е", ending
end
base, ending = rmatch(lemma, "^(.*)([ыиЫИ][" .. AC .. GR .. "]?)$")
if base then
if rfind(ending, "[ыЫ]") or rfind(base, "[" .. com.sib .. com.velar .. "]$") then
return base, com.strip_tr_ending(tr, ending), "о-и", ending
else
-- FIXME, should we return a slash declension?
error("No neuter declension е-и available; use a slash declension")
end
end
end
if gender == "f" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), "ья", ending
end
end
-- Recognize masculines with irregular plurals, but only if the user
-- either explicitly specified that this noun is plural (n=p) or
-- specifically requested the irregular plural. This is necessary
-- because some masculine nouns have feminine endings, which look
-- like irregular plurals.
if gender == "m" then
if args.thisn == "p" or variant == "-ья" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), (args.old and "ъ-ья" or "-ья"), ending
end
end
if args.thisn == "p" or args.want_sc1 then
base, ending = rmatch(lemma, "^(.*)([аА][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), (args.old and "ъ-а" or "-а"), ending
end
base, ending = rmatch(lemma, "^(.*)([яЯ][" .. AC .. GR .. "]?)$")
if base then
if rfind(base, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
return base, com.strip_tr_ending(tr, ending), "й-я", ending
else
return base, com.strip_tr_ending(tr, ending), "ь-я", ending
end
end
end
end
if gender == "m" or gender == "f" then
base, ending = rmatch(lemma, "^(.*[" .. com.sib .. com.velar .. "])([иИ][" .. AC .. GR .. "]?)$")
if not base then
base, ending = rmatch(lemma, "^(.*)([ыЫ][" .. AC .. GR .. "]?)$")
end
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and (args.old and "ъ" or "") or "а", ending
end
base, ending = rmatch(lemma, "^(.*[" .. com.vowel .. "й][" .. AC .. GR .. "]?)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and "й" or "я", ending
end
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and "ь-m" or "я", ending
end
end
if gender == "3f" then
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), "ь-f", ending
end
end
-- end of recognize-plurals code
base, ending = rmatch(lemma, "^(.*)([ьЬ][яеёЯЕЁ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([йаяеоёъЙАЯЕОЁЪ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ьЬ])$")
if base then
if gender == "m" or gender == "f" then
return base, com.strip_tr_ending(tr, ending), "ь-" .. gender
elseif gender == "3f" then
return base, com.strip_tr_ending(tr, ending), "ь-f"
else
error("Need to specify gender m or f with lemma in -ь: ".. lemma)
end
end
if rfind(lemma, "[ыиЫИ][" .. AC .. GR .. "]?$") then
error("If this is a plural lemma, gender must be specified: " .. lemma)
elseif rfind(lemma, "[иИіІуУыЫѣѢэЭюЮѵѴ]" .. DIA .. "?[" .. AC .. GR .. "]?$") then
error("Don't know how to decline lemma ending in this type of vowel: " .. lemma)
end
return lemma, tr, ""
end
local plural_variant_detection_map = {
[""] = {["-ья"]="-ья"},
["ъ"] = {["-ья"]="ъ-ья"},
}
local special_case_1_to_plural_variant = {
[""] = "-а",
["ъ"] = "ъ-а",
["й"] = "й-я",
["ь-m"] = "ь-я",
["о"] = "о-и",
-- these last two are here to avoid getting errors in that checks for
-- compatibility with special case 1; the basic variants of these decls
-- don't actually exist
["(ишк)о"] = "(ишк)о-и",
["(ищ)е"] = "(ищ)е-и",
}
local function map_decl(decl, fun)
if rfind(decl, "/") then
local split_decl = rsplit(decl, "/")
if #split_decl ~= 2 then
error("Mixed declensional class " .. decl
.. "needs exactly two classes, singular and plural")
end
return fun(split_decl[1]) .. "/" .. fun(split_decl[2])
else
return fun(decl)
end
end
-- Canonicalize decl class into non-accented and alias-resolved form;
-- but note that some canonical decl class names with an accent in them
-- (e.g. е́, not the same as е, whose accented version is ё; and various
-- adjective declensions).
local function canonicalize_decl(decl, old)
local function do_canon(decl)
-- remove accents, but not from е́ (for adj decls, accent matters
-- as well but we handle that by mapping the accent to a stress pattern
-- and then to the accented version in determine_stress_variant())
if decl ~= "е́" then
decl = com.remove_accents(decl)
end
local decl_aliases = old and declensions_old_aliases or declensions_aliases
local decl_cats = old and declensions_old_cat or declensions_cat
if decl_aliases[decl] then
-- If we find an alias, map it.
decl = decl_aliases[decl]
elseif not decl_cats[decl] then
error("Unrecognized declension class " .. decl)
end
return decl
end
return map_decl(decl, do_canon)
end
-- Attempt to determine the actual declension (including plural variants)
-- based on a combination of the declension the user specified, what can be
-- detected from the lemma, and special case (1), if given in the declension.
-- DECL is the value the user passed for the declension field, after
-- extraneous annotations (special cases (1) and (2), * for reducible,
-- ё for ё/ё alternation and a ; that may precede ё) have been stripped off.
-- What's left is one of the following:
--
-- 1. Blank, meaning to autodetect the declension from the lemma
-- 2. A hyphen followed by a declension variant (-ья, -ин, -ишко, -ище; see
-- long comment at top of file)
-- 3. A gender (m, f, n, 3f)
-- 4. A gender plus declension variant (e.g. f-ья)
-- 5. An actual declension, possibly including a plural variant (e.g. о-и) or
-- a slash declension (e.g. я/-ья, used for the noun дядя).
--
-- Return seven args: stem (lemma minus ending), translit, canonicalized
-- declension, explicitly specified gender if any (m, f, n or nil), whether
-- the specified declension or detected ending was accented, whether the
-- detected ending was pl, and whether the declension was autodetected
-- (corresponds to cases where a full word with ending attached is required
-- in the lemma field). "Canonicalized" means after autodetection, with
-- accents removed, with any aliases mapped to their canonical versions
-- and with any requested declension variants applied. The result is either a
-- declension that will have a categorization entry (in declensions_cat[] or
-- declensions_old_cat[]) or a slash declension where each part similarly has
-- a categorization entry.
--
-- Note that gender is never required when an explicit declension is given,
-- and in connection with stem autodetection is required only when the lemma
-- either ends in -ь or is plural.
determine_decl = function(lemma, tr, decl, args)
-- Assume we're passed a value for DECL of types 1-4 above, and
-- fetch gender and requested declension variant.
local stem
local want_ya_plural, orig_pl_ending, variant
local was_autodetected
local gender = rmatch(decl, "^(3?[mfn]?)$")
if not gender then
gender, variant = rmatch(decl, "^(3?[mfn]?)(%-[^%-]+)$")
-- But be careful with explicit declensions like -а that look like
-- variants without gender (FIXME, eventually we should maybe do
-- something about the potential ambiguity).
if gender == "" and not m_table.contains({"-ья", "-ин", "-ишко", "-ище"}, variant) then
gender, variant = nil, nil
end
end
-- If DECL is of type 1-4, handle declension variants and detect
-- the actual declension from the lemma.
if gender then
-- Check for declension variants
if variant then
if variant == "-ья" then
want_ya_plural = "-ья"
else
-- Sanity-check remaining declension variants, which need
-- specific values of animacy, gender and special-case (1)
local sc1_needed
local animate_needed
if variant == "-ишко" then
animate_needed = false
sc1_needed = true
elseif variant == "-ище" then
animate_needed = true
sc1_needed = true
elseif variant == "-ин" then
animate_needed = true
sc1_needed = false
else
-- WARNING: If adding another variant, you need to also
-- add to the list farther above.
error("Unrecognized declension variant " .. variant .. ", should be -ья, -ин, -ишко or -ище")
end
if sc1_needed and not args.want_sc1 then
error("Declension variant " .. variant .. " must be used with special case (1)")
elseif sc1_needed == false and args.want_sc1 then
error("Declension variant " .. variant .. " must not be used with special case (1)")
end
if animate_needed and args.thisa ~= "a" then
error("Declension variant " .. variant .. " must be specified as animate")
elseif animate_needed == false and args.thisa == "a" then
error("Declension variant " .. variant .. " must not be specified as animate")
end
if gender ~= "" and gender ~= "m" then
error("Declension variant " .. variant .. " should be used with the masculine gender")
end
end
end
stem, tr, decl, orig_pl_ending = detect_lemma_type(lemma, tr, gender,
args, variant)
was_autodetected = true
else
stem, tr = lemma, tr
end
-- Now canonicalize gender
if gender == "3f" then
gender = "f"
elseif gender == "" then
gender = nil
end
-- The ending should be treated as accented if either the original singular
-- or plural ending was accented, or if the stem is non-syllabic.
local was_accented = com.is_stressed(decl) or
orig_pl_ending and com.is_stressed(orig_pl_ending) or
com.is_nonsyllabic(stem)
local was_plural = not not orig_pl_ending
decl = canonicalize_decl(decl, args.old)
-- The rest of this code concerns plural variants. It's somewhat
-- complicated because there are potentially four sources of plural
-- variants (not to mention plural variants constructed using slash
-- notation):
--
-- 1. A user-requested plural variant in declension types 2 or 4 above
-- (currently only -ья)
-- 2. An explicit plural variant encoded in an explicit declension of
-- type 5 above
-- 3. An autodetected plural variant (which will happen in some cases
-- when autodetection is performed on a nominative plural)
-- 4. A plural variant derived using special case (1).
--
-- Up to three actual plural variants might exist (e.g. if the user
-- specifies a DECL value or 'm-ья(1)' and a STEM ending in -а,
-- although not all three can ever be compatible because -ья and (1)
-- are never compatible). We can't have all four because if there's
-- an explicit plural variant, there won't be a user-requested or
-- autodetected plural variant.
--
-- The goal below is to do two things: Check that all available plural
-- variants are the same, and generate the actual declension.
-- If we have a type-2 or type-3 variant, we already have the actual
-- declension; else we need to use a table to map the basic declension
-- to the one with the plural variant encoded in it.
--
-- NOTE: The code below was written with a more general plural-variant
-- system. It probably can be simplified a lot now.
-- 1: Handle explicit decl with slash variant
if rfind(decl, "/") then
if want_ya_plural then
-- Don't think this can happen
error("Plural variant " .. want_ya_plural .. " not compatible with slash declension " .. decl)
end
if args.want_sc1 then
error("Special case (1) not compatible with slash declension" .. decl)
end
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- 2: Retrieve explicitly specified or autodetected decl and pl. variant
local basic_decl, detected_or_explicit_plural = rmatch(decl, "^(.*)(%-[^mf]+)$")
if basic_decl == "ь" then
basic_decl = "ь-m"
end
basic_decl = basic_decl or decl
-- 3: Any user-requested plural variant must agree with explicit or
-- autodetected variant.
if want_ya_plural and detected_or_explicit_plural and want_ya_plural ~= detected_or_explicit_plural then
error("Plural variant " .. want_ya_plural .. " requested but plural variant " .. detected_or_explicit_plural .. " detected from plural stem")
end
-- 4: Handle special case (1). Derive the full declension, make sure its
-- plural variant matches any other available plural variants, and
-- return the declension.
if args.want_sc1 then
local sc1_decl = special_case_1_to_plural_variant[basic_decl] or
error("Special case (1) not compatible with declension " .. basic_decl)
local sc1_plural = rsub(sc1_decl, "^.*%-", "-")
local other_plural = want_ya_plural or detected_or_explicit_plural
if other_plural and sc1_plural ~= other_plural then
error("Plural variant " .. other_plural .. " specified or detected, but special case (1) calls for plural variant " .. sc1_plural)
end
return stem, tr, sc1_decl, gender, was_accented, was_plural, was_autodetected
end
-- 5: Handle user-requested plural variant without explicit or detected
-- one. (If an explicit or detected one exists, we've already checked
-- that it agrees with the user-requested one, and so we already have
-- our full declension.)
if want_ya_plural and not detected_or_explicit_plural then
local variant_decl
if plural_variant_detection_map[decl] then
variant_decl = plural_variant_detection_map[decl][want_ya_plural]
end
if variant_decl then
return stem, tr, variant_decl, gender, was_accented, was_plural, was_autodetected
else
return stem, tr, decl .. (args.old and "/ъ-ья" or "/-ья"), gender, was_accented, was_plural, was_autodetected
end
end
-- 6: Just return the full declension, which will include any available
-- plural variant in it.
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- Convert soft adjectival declensions into hard ones following certain
-- stem-final consonants. FIXME: We call this in two places, once
-- to handle auto-detection and once to handle explicit declensions; but
-- in the former case we end up calling it twice.
local function determine_adj_stem_variant(decl, stem)
local iend = rmatch(decl, "^%+[іи]([йея]?)$")
-- Convert ій/ий to ый after velar or sibilant. This is important for
-- velars; doesn't really matter one way or the other for sibilants as
-- the sibilant rules will convert both sets of endings to the same
-- thing (whereas there will be a difference with о vs. е for velars).
if iend and rfind(stem, "[" .. com.velar .. com.sib .. "]$") then
decl = "+ы" .. iend
-- The following is necessary for -ц, unclear if makes sense for
-- sibilants. (Would be necessary -- I think -- if we were
-- inferring short adjective forms, but we're not.)
elseif decl == "+ее" and rfind(stem, "[" .. com.sib_c .. "]$") then
decl = "+ое"
end
return decl
end
-- Attempt to determine the actual adjective declension based on a
-- combination of the declension the user specified and what can be detected
-- from the stem. DECL is the value the user passed for the declension field,
-- after extraneous annotations have been removed (although none are probably
-- relevant here). What's left is one of the following, which always begins
-- with +:
--
-- 1. +, meaning to autodetect the declension from the stem
-- 2. +ь, same as + but selects +ьий instead of +ий if lemma ends in -ий
-- 3. +short, +mixed or +proper, with the declension partly specified but
-- the particular gender/number-specific short/mixed variant to be
-- autodetected
-- 4. A gender (+m, +f or +n), used only for detecting the singular of
-- plural-form lemmas (this is primarily used in conjunction with template
-- ru-noun+, to explicitly specify the gender; for the actual declension,
-- it doesn't much matter what singular gender we pick since we're a
-- plural only)
-- 5. A gender plus short/mixed/proper/ь (e.g. +f-mixed), again with the gender
-- used only for detecting the singular of plural-form short/mixed lemmas
-- 6. An actual declension, possibly including a slash declension
-- (WARNING: Unclear if slash declensions will work, especially those
-- that are adjective/noun combinations)
--
-- Returns the same seven args as for determine_decl(). The returned
-- declension will always begin with +.
detect_adj_type = function(lemma, tr, decl, old)
local was_autodetected
local base, ending
local basedecl, g = rmatch(decl, "^(%+)([mfn])$")
if not basedecl then
g, basedecl = rmatch(decl, "^%+([mfn])%-([a-zь]+)$")
if basedecl then
basedecl = "+" .. basedecl
end
end
decl = basedecl or decl
if decl == "+" or decl == "+ь" then
local loc = rfind(lemma, "[аеиіоыя][" .. AC .. GR .. "]?[йея]$") or rfind(lemma, "ь[йея]$")
if loc then
base, ending = usub(lemma, 1, loc - 1), usub(lemma, loc)
end
if ending == "ий" and decl == "+ь" then
decl = "+ьий"
elseif ending == "ій" and decl == "+ь" then
decl = "+ьій"
elseif ending then
decl = "+" .. ending
else
local loc, shortmixed = rfind(lemma, "[аоы][" .. AC .. GR .. "]?$") or rfind(lemma, "ъ?$")
if loc then
base, ending = usub(lemma, 1, loc - 1), usub(lemma, loc)
shortmixed = rfind(base, "^[" .. com.uppercase .. "].*[иы]" .. AC .. "н$") and "stressed-proper" or -- accented
rfind(base, "^[" .. com.uppercase .. "].*[иы]н$") and "proper" or --not accented
rlfind(base, "[ёео][" .. AC .. GR .. "]?в$") and "short" or
rlfind(base, "[ыи]" .. AC .. "н$") and "stressed-short" or -- accented
rlfind(base, "[ыи]н$") and "mixed" --not accented
end
if not shortmixed then
error("Cannot determine stem type of adjective: " .. lemma)
end
decl = "+" .. ending .. "-" .. shortmixed
end
was_autodetected = true
elseif m_table.contains({"+short", "+mixed", "+proper"}, decl) then
base, ending = rmatch(lemma, "^(.-)([оаыъ]?[" .. AC .. GR .. "]?)$")
assert(base)
local shortmixed = usub(decl, 2)
if rlfind(base, "[ыи]" .. AC .. "н$") then -- accented
if shortmixed == "short" then shortmixed = "stressed-short"
elseif shortmixed == "proper" then shortmixed = "stressed-proper"
end
end
decl = "+" .. ending .. "-" .. shortmixed
was_autodetected = true
else
base = lemma
end
if ending and ending ~= "" then
tr = com.strip_tr_ending(tr, ending)
end
-- Remove any accents from the declension, but not their presence.
-- We will convert was_accented into stress pattern b, and convert that
-- back to an accented version in determine_stress_variant(). This way
-- we end up with the stressed version whether the user placed an accent
-- in the ending or decl or specified stress pattern b.
-- FIXME, might not work in the presence of slash declensions
local was_accented = com.is_stressed(decl)
decl = com.remove_accents(decl)
decl = map_decl(decl, function(decl)
return determine_adj_stem_variant(decl, base)
end)
local singdecl
if decl == "+ые" then
singdecl = (g == "m" or not g) and (was_accented and "+ой" or "+ый") or not old and g == "f" and "+ая" or not old and g == "n" and "+ое"
elseif decl == "+ыя" and old then
singdecl = (g == "f" or not g) and "+ая" or g == "n" and "+ое"
elseif decl == "+ие" and not old then
singdecl = (g == "m" or not g) and "+ий" or g == "f" and "+яя" or g == "n" and "+ее"
elseif decl == "+іе" and old and (g == "m" or not g) then
singdecl = "+ій"
elseif decl == "+ія" and old then
singdecl = (g == "f" or not g) and "+яя" or g == "n" and "+ее"
elseif decl == "+ьи" then
singdecl = (g == "m" or not g) and (old and "+ьій" or "+ьий") or g == "f" and "+ья" or g == "n" and "+ье"
elseif rfind(decl, "^%+ы%-") then -- decl +ы-mixed or similar
local beg = (g == "m" or not g) and (old and "ъ" or "") or g == "f" and "а" or g == "n" and "о"
singdecl = beg and "+" .. beg .. usub(decl, 3)
end
if singdecl then
was_plural = true
decl = singdecl
end
return base, tr, canonicalize_decl(decl, old), g, was_accented, was_plural, was_autodetected
end
-- If stress pattern omitted, detect it based on whether ending is stressed
-- or the decl class or stem accent calls for inherent stress, defaulting to
-- pattern a. This is run after alias resolution and accent removal of DECL;
-- WAS_ACCENTED indicates whether the ending was originally stressed.
-- FIXME: This is run before splitting slash patterns but should be run after.
detect_stress_pattern = function(stem, decl, decl_cats, reducible,
was_plural, was_accented)
-- ёнок and ёночек always bear stress
if rfind(decl, "ёнокъ?") or rfind(decl, "ёночекъ?") then
return "b"
-- stressed suffix и́н; missing in plural and true endings don't bear stress
-- (except for exceptional господи́н)
elseif rfind(decl, "инъ?") and was_accented then
return "d"
-- Adjectival -ой always bears the stress
elseif rfind(decl, "%+ой") then
return "b"
-- Adjectival stressed-short, stressed-proper bears the stress
elseif rfind(decl, "^%+.*%-stressed") then
return "b"
-- Pattern b if ending was accented by user
elseif was_accented then
return "b"
-- Nonsyllabic stem means pattern b
elseif com.is_nonsyllabic(stem) then
return "b"
-- Accent on reducible vowel in masc nom sg (not plural) means pattern b.
-- Think about whether we want to enable this.
-- elseif reducible and not was_plural then
-- -- FIXME hack. Eliminate plural part of slash declension.
-- decl = rsub(decl, "/.*", "")
-- if decl_cats[decl] and decl_cats[decl].g == "m" then
-- if com.is_ending_stressed(stem) or com.is_monosyllabic(stem) then
-- return "b"
-- end
-- end
end
return "a"
end
-- In certain special cases, depending on the declension, we override the
-- user-specified stress pattern and convert it to something else.
-- NOTE: This function is run after alias resolution and accent removal.
-- FIXME: It's also run before splitting slash patterns but should be run after.
override_stress_pattern = function(decl, stress)
-- ёнок and ёночек always bear stress; if user specified a,
-- convert to b. Don't do this with slash patterns (see FIXME above).
if stress == "a" and (rfind(decl, "^ёнокъ?$") or rfind(decl, "^ёночекъ?$")) then
return "b"
end
return stress
end
-- Canonicalize an adjectival declension to either the stressed or unstressed
-- variant depending on the stress. Ultimately this is what ensures that
-- the user's stress mark on an adjectival ending is respected.
determine_stress_variant = function(decl, stress)
if stress == "b" then
if decl == "+ая" then
return "+а́я"
elseif decl == "+ое" then
return "+о́е"
else
-- Convert +...-short to +...-stressed-short and same for -proper
local b, e = rmatch(decl, "^%+(.*)%-(short)$")
if not b then
b, e = rmatch(decl, "^%+(.*)%-(proper)$")
end
if b and not rfind(b, "%-stressed") then
return "+" .. b .. "-stressed-" .. e
end
end
end
return decl
end
-- Canonicalize a declension based on the final stem consonant, in
-- particular converting soft declensions to hard ones after velars and/or
-- sibilants. FIXME: We also do this canonicalization earlier on during
-- auto-detection (determine_adj_stem_variant() is called by
-- detect_adj_type(), and code in detect_lemma_type() does the equivalent
-- of the first clause below). Doing it here ensures that explicitly
-- specified declensions get handled as well, but it would be nice to not
-- do the same thing twice in the auto-detection case.
determine_stem_variant = function(decl, stem)
if decl == "е" and rfind(stem, "[" .. com.sib_c .. "]$") then
return "о"
end
return determine_adj_stem_variant(decl, stem)
end
is_reducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "3rd" and decl_cat.g == "f" or decl_cat.g == "m" then
return true
else
return false
end
end
-- Reduce nom sg to stem by eliminating the "epenthetic" vowel. Applies to
-- masculine 2nd-declension hard and soft, and 3rd-declension feminine in
-- -ь. STEM and DECL are after determine_decl(), before converting
-- outward-facing declensions to inward ones.
function export.reduce_nom_sg_stem(stem, tr, decl, soft_n, can_err)
local full_stem = stem .. (decl == "й" and decl or "")
local full_tr = tr and tr .. (decl == "й" and "j" or "")
local ret, rettr = com.reduce_stem(full_stem, full_tr, soft_n)
if not ret and can_err then
error("Unable to reduce stem " .. stem)
end
return ret, rettr
end
is_dereducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "1st" or decl_cat.decl == "2nd" and decl_cat.g == "n" then
return true
else
return false
end
end
-- Add a possible suffix to the bare stem, according to the declension and
-- value of OLD. This may be -ь, -ъ, -й or nothing. We need to do this here
-- because we don't actually attach such a suffix in attach_unstressed() due
-- to situations where we don't want the suffix added, e.g. dereducible nouns
-- in -ня.
add_bare_suffix = function(bare, baretr, old, sgdc, dereduced)
if old and sgdc.hard == "hard" then
-- Final -ъ isn't transliterated
return bare .. "ъ", baretr
elseif sgdc.hard == "soft" or sgdc.hard == "palatal" then
-- This next clause corresponds to a special case in Vitalik's module.
-- It says that nouns in -ня (accent class a) have gen pl without
-- trailing -ь. It appears to apply to most nouns in -ня (possibly
-- all in -льня), but ку́хня (gen pl ку́хонь) and дерéвня (gen pl
-- дереве́нь) is an exception. (Vitalik's module has an extra
-- condition here 'stress == "a"' that would exclude дере́вня but I
-- don't think this condition is in Zaliznyak, as he indicates
-- дере́вня as having an exceptional genitive plural.)
if dereduced and rfind(bare, "[нН]$") and sgdc.decl == "1st" then
-- FIXME: What happens in this case old-style? I assume that
-- -ъ is added, but this is a guess.
-- Final -ъ isn't transliterated
return bare .. (old and "ъ" or ""), baretr
elseif rfind(bare, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
return bare .. "й", baretr and (baretr .. "j")
else
return bare .. "ь", baretr and (baretr .. "ʹ")
end
else
return bare, baretr
end
end
-- Dereduce stem to the form found in the gen pl (and maybe nom sg) by
-- inserting an epenthetic vowel. Applies to 1st declension and 2nd
-- declension neuter, and to 2nd declension masculine when the stem was
-- specified as a plural form (in which case we're deriving the nom sg,
-- and also the gen pl in the alt-gen-pl scenario). STEM and DECL are
-- after determine_decl(), before converting outward-facing declensions
-- to inward ones. STRESS is the stess pattern.
function export.dereduce_nom_sg_stem(stem, tr, sgdc, stress, old, can_err)
local epenthetic_stress = ending_stressed_gen_pl_patterns[stress]
local ret, rettr = com.dereduce_stem(stem, tr, epenthetic_stress)
if not ret then
if can_err then
error("Unable to dereduce stem " .. stem)
else
return nil, nil
end
end
return add_bare_suffix(ret, rettr, old, sgdc, true)
end
--------------------------------------------------------------------------
-- Second-declension masculine --
--------------------------------------------------------------------------
----------------- Masculine hard -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style).
declensions_old["ъ"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and "е́й" or "о́въ"
end,
["alt_gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["ъ"] = { decl="2nd", hard="hard", g="m" }
-- Normal mapping of old ъ would be "" (blank), but we set up "#" as an alias
-- so we have a way of referring to it without defaulting if need be and
-- distinct from auto-detection (e.g. in the second stem of a word, or to
-- override autodetection of -ёнок or -ин -- the latter is necessary in the
-- case of семьянин).
declensions_aliases["#"] = ""
----------------- Masculine hard, irregular plural -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg nom pl -а.
declensions_old["ъ-а"] = mw.clone(declensions_old["ъ"])
declensions_old["ъ-а"]["nom_pl"] = "а́"
declensions_old_cat["ъ-а"] = { decl="2nd", hard="hard", g="m", alt_nom_pl=true }
declensions_aliases["#-a"] = "-a"
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg soft pl -ья.
-- Differs from the normal declension throughout the plural.
declensions_old["ъ-ья"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = "ьёвъ",
["alt_gen_pl"] = "е́й",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ъ-ья"] = { decl="2nd", hard="hard", g="m", irregpl=true }
declensions_aliases["#-ья"] = "-ья"
----------------- Masculine hard, suffixed, irregular plural -------------------
declensions_old["инъ"] = {
["nom_sg"] = "и́нъ",
["gen_sg"] = "и́на",
["dat_sg"] = "и́ну",
["acc_sg"] = nil,
["ins_sg"] = "и́номъ",
["pre_sg"] = "и́нѣ",
["nom_pl"] = "е́",
["gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["инъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old["ёнокъ"] = {
["nom_sg"] = "ёнокъ",
["gen_sg"] = "ёнка",
["dat_sg"] = "ёнку",
["acc_sg"] = nil,
["ins_sg"] = "ёнкомъ",
["pre_sg"] = "ёнкѣ",
["nom_pl"] = "я́та",
["gen_pl"] = "я́тъ",
["dat_pl"] = "я́тамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тами",
["pre_pl"] = "я́тахъ",
}
declensions_old_cat["ёнокъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["онокъ"] = "ёнокъ"
declensions_old_aliases["енокъ"] = "ёнокъ"
declensions_old["ёночекъ"] = {
["nom_sg"] = "ёночекъ",
["gen_sg"] = "ёночка",
["dat_sg"] = "ёночку",
["acc_sg"] = nil,
["ins_sg"] = "ёночкомъ",
["pre_sg"] = "ёночкѣ",
["nom_pl"] = "я́тки",
["gen_pl"] = "я́токъ",
["dat_pl"] = "я́ткамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тками",
["pre_pl"] = "я́ткахъ",
}
declensions_old_cat["ёночекъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["оночекъ"] = "ёночекъ"
declensions_old_aliases["еночекъ"] = "ёночекъ"
----------------- Masculine soft -------------------
-- Normal soft-masculine declension in -ь
declensions_old["ь-m"] = {
["nom_sg"] = "ь",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["alt_gen_pl"] = "ь",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-m"] = { decl="2nd", hard="soft", g="m" }
-- Soft-masculine declension in -ь with irreg nom pl -я
declensions_old["ь-я"] = mw.clone(declensions_old["ь-m"])
declensions_old["ь-я"]["nom_pl"] = "я́"
declensions_old_cat["ь-я"] = { decl="2nd", hard="soft", g="m", alt_nom_pl=true }
----------------- Masculine palatal -------------------
-- Masculine declension in palatal -й
declensions_old["й"] = {
["nom_sg"] = "й",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = "ёвъ",
["alt_gen_pl"] = "й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["й"] = { decl="2nd", hard="palatal", g="m" }
declensions_old["й-я"] = mw.clone(declensions_old["й"])
declensions_old["й-я"]["nom_pl"] = "я́"
declensions_old_cat["й-я"] = { decl="2nd", hard="palatal", g="m", alt_nom_pl=true }
--------------------------------------------------------------------------
-- First-declension feminine --
--------------------------------------------------------------------------
----------------- Feminine hard -------------------
-- Hard-feminine declension in -а
declensions_old["а"] = {
["nom_sg"] = "а́",
["gen_sg"] = "ы́",
["dat_sg"] = "ѣ́",
["acc_sg"] = "у́",
["ins_sg"] = {"о́й<insa>", "о́ю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["а"] = { decl="1st", hard="hard", g="f" }
----------------- Feminine soft -------------------
-- Soft-feminine declension in -я
declensions_old["я"] = {
["nom_sg"] = "я́",
["gen_sg"] = "и́",
["dat_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_dat_sg_patterns[stress] and "и" or "ѣ́"
end,
["acc_sg"] = "ю́",
["ins_sg"] = {"ёй<insa>", "ёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "]́?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["я"] = { decl="1st", hard="soft", g="f" }
-- Soft-feminine declension in -ья.
-- Almost like ь + -я endings except for genitive plural.
declensions_old["ья"] = {
["nom_sg"] = "ья́",
["gen_sg"] = "ьи́",
["dat_sg"] = "ьѣ́",
["acc_sg"] = "ью́",
["ins_sg"] = {"ьёй<insa>", "ьёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ьи́",
["gen_pl"] = function(stem, stress)
-- circumflex accent is a signal that forces stress, particularly
-- in accent pattern d/d'.
return (ending_stressed_gen_pl_patterns[stress] or stress == "d" or stress == "d'") and "е̂й" or "ий"
end,
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ья"] = {
decl="1st", hard="soft", g="f",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
--------------------------------------------------------------------------
-- Second-declension neuter --
--------------------------------------------------------------------------
----------------- Neuter hard -------------------
-- Normal hard-neuter declension in -о
declensions_old["о"] = {
["nom_sg"] = "о́",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "о́" or nil
end,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "а́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "о́въ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["о"] = { decl="2nd", hard="hard", g="n" }
-- Hard-neuter declension in -о with irreg nom pl -и
declensions_old["о-и"] = mw.clone(declensions_old["о"])
declensions_old["о-и"]["nom_pl"] = "ы́"
declensions_old_cat["о-и"] = { decl="2nd", hard="hard", g="n", alt_nom_pl=true }
declensions_old_aliases["о-ы"] = "о-и"
-- Masculine-gender neuter-form declension in -(ишк)о with irreg nom pl -и,
-- with colloquial feminine endings in some of the singular cases
-- (§5 p. 74 of Zaliznyak)
declensions_old["(ишк)о-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ишк)о-и"]["gen_sg"] = {"а́", "ы́1"}
declensions_old["(ишк)о-и"]["dat_sg"] = {"у́", "ѣ́1"}
declensions_old["(ишк)о-и"]["ins_sg"] = {"о́мъ", "о́й1"}
declensions_old_cat["(ишк)о-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ишк)о-и"] = "<sup>1</sup> Colloquial."
-- Masculine-gender animate neuter-form declension in -(ищ)е with irreg
-- nom pl -и, with colloquial feminine endings in some of the singular cases
-- (§4 p. 74 of Zaliznyak)
declensions_old["(ищ)е-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ищ)е-и"]["acc_sg"] = {"а́", "у́1"}
declensions_old["(ищ)е-и"]["gen_sg"] = {"а́", "ы́2"}
declensions_old["(ищ)е-и"]["dat_sg"] = {"у́", "ѣ́2"}
declensions_old["(ищ)е-и"]["ins_sg"] = {"о́мъ", "о́й2"}
declensions_old_cat["(ищ)е-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ищ)е-и"] = "<sup>1</sup> Colloquial.<br /><sup>2</sup> Less common, more colloquial."
----------------- Neuter soft -------------------
-- Soft-neuter declension in -е (stressed -ё)
declensions_old["е"] = {
["nom_sg"] = "ё",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ё" or nil
end,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е"] = {
singular = function(suffix)
if suffix == "ё" then
return "ending in -ё"
else
return {}
end
end,
decl="2nd", hard="soft", g="n", gensg=true
}
-- User-facing declension type "ё" = "е"
declensions_old_aliases["ё"] = "е"
-- Rare soft-neuter declension in stressed -е́ (e.g. муде́, бытие́)
declensions_old["е́"] = {
["nom_sg"] = "е́",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "е́" or nil
end,
["ins_sg"] = "е́мъ",
["pre_sg"] = function(stem, stress)
-- FIXME!!! Are we sure about this condition? This is what was
-- found in the old template, but the related -е declension has
-- -ие prep sg ending -(и)и only when *not* stressed.
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and "и́" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return rlfind(stem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") and "й" or "е́й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е́"] = {
singular = "ending in stressed -е",
decl="2nd", hard="soft", g="n", gensg=true
}
-- Soft-neuter declension in unstressed -ье (stressed -ьё).
declensions_old["ье"] = {
["nom_sg"] = "ьё",
["gen_sg"] = "ья́",
["dat_sg"] = "ью́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ьё" or nil
end,
["ins_sg"] = "ьёмъ",
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and "е́й" or "ий"
end,
["alt_gen_pl"] = "ьёвъ",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ье"] = {
decl="2nd", hard="soft", g="n",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
declensions_old_aliases["ьё"] = "ье"
--------------------------------------------------------------------------
-- Third declension --
--------------------------------------------------------------------------
declensions_old["ь-f"] = {
["nom_sg"] = "ь",
["gen_sg"] = "и́",
["dat_sg"] = "и́",
["acc_sg"] = "ь",
["ins_sg"] = "ью́",
["pre_sg"] = "и́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-f"] = { decl="3rd", hard="soft", g="f" }
declensions_old["мя"] = {
["nom_sg"] = "мя",
["gen_sg"] = "мени",
["dat_sg"] = "мени",
["acc_sg"] = nil,
["ins_sg"] = "менемъ",
["pre_sg"] = "мени",
["nom_pl"] = "мена́",
["gen_pl"] = "мёнъ",
["dat_pl"] = "мена́мъ",
["acc_pl"] = nil,
["ins_pl"] = "мена́ми",
["pre_pl"] = "мена́хъ",
}
declensions_old_cat["мя"] = { decl="3rd", hard="soft", g="n", cant_reduce=true }
--------------------------------------------------------------------------
-- Indeclinable --
--------------------------------------------------------------------------
-- Indeclinable declension; no endings.
declensions_old["$"] = {
["nom_sg"] = "",
["gen_sg"] = "",
["dat_sg"] = "",
["acc_sg"] = nil,
["ins_sg"] = "",
["pre_sg"] = "",
["nom_pl"] = "",
["gen_pl"] = "",
["dat_pl"] = "",
["acc_pl"] = nil,
["ins_pl"] = "",
["pre_pl"] = "",
}
declensions_old_cat["$"] = { decl="indeclinable", hard="none", g="none" }
--------------------------------------------------------------------------
-- Adjectival --
--------------------------------------------------------------------------
-- This needs to be up here because it is called just below.
local function old_to_new(v)
v = rsub(v, "ъ$", "")
v = rsub(v, "^ъ", "")
v = rsub(v, "(%A)ъ", "%1")
v = rsub(v, "ъ(%A)", "%1")
v = rsub(v, "і", "и")
v = rsub(v, "ѣ", "е")
return v
end
-- Meaning of entry is:
-- 1. The declension name in module ru-adjective
-- 2. The masculine declension name in this module
-- 3. The neuter declension name in this module
-- 4. The feminine declension name in this module
-- 5. The value of hard= for the declensions_cat entry
-- 6. The value of decl= for the declensions_cat entry
-- 7. The value of possadj= for the declensions_cat entry (true if possessive
-- or similar type of adjective)
local adj_decl_map = {
{"ый", "ый", "ое", "ая", "hard", "long", false},
{"ій", "ій", "ее", "яя", "soft", "long", false},
{"ой", "ой", "о́е", "а́я", "hard", "long", false},
{"ьій", "ьій", "ье", "ья", "palatal", "long", true},
{"short", "ъ-short", "о-short", "а-short", "hard", "short", true},
{"mixed", "ъ-mixed", "о-mixed", "а-mixed", "hard", "mixed", true},
{"proper", "ъ-proper", "о-proper", "а-proper", "hard", "proper", true},
{"stressed-short", "ъ-stressed-short", "о-stressed-short", "а-stressed-short", "hard", "short", true},
{"stressed-proper", "ъ-stressed-proper", "о-stressed-proper", "а-stressed-proper", "hard", "proper", true},
}
local function get_adjectival_decl(adjtype, gender, old)
local decl, intnotes = m_ru_adj.get_nominal_decl(adjtype, gender, old)
-- hack fem ins_sg to insert <insa>, <insb>; see concat_word_forms_1()
if gender == "f" and type(decl["ins_sg"]) == "table" and #decl["ins_sg"] == 2 then
decl["ins_sg"][1] = decl["ins_sg"][1] .. "<insa>"
decl["ins_sg"][2] = decl["ins_sg"][2] .. "<insb>"
end
return decl, intnotes
end
for _, declspec in ipairs(adj_decl_map) do
local oadjdecl = declspec[1]
local nadjdecl = old_to_new(oadjdecl)
local odecl_by_gender =
{m="+" .. declspec[2], n="+" .. declspec[3], f="+" .. declspec[4]}
local hard = declspec[5]
local decltype = declspec[6]
local possadj = declspec[7]
for _, g in ipairs({"m", "n", "f"}) do
local odecl = odecl_by_gender[g]
local ndecl = old_to_new(odecl)
declensions_old[odecl], internal_notes_table_old[odecl] =
get_adjectival_decl(oadjdecl, g, true)
declensions[ndecl], internal_notes_table[ndecl] =
get_adjectival_decl(nadjdecl, g, false)
declensions_old_cat[odecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
declensions_cat[ndecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
end
end
-- Set up some aliases.
declensions_old_aliases["+о́-short"] = "+о-stressed-short"
declensions_old_aliases["+а́-short"] = "+а-stressed-short"
declensions_old_aliases["+о́-proper"] = "+о-stressed-proper"
declensions_old_aliases["+а́-proper"] = "+а-stressed-proper"
declensions_aliases["+#-short"] = "+-short"
declensions_aliases["+#-mixed"] = "+-mixed"
declensions_aliases["+#-proper"] = "+-proper"
declensions_aliases["+#-stressed-short"] = "+-stressed-short"
declensions_aliases["+#-stressed-proper"] = "+-stressed-proper"
--------------------------------------------------------------------------
-- Populate new from old --
--------------------------------------------------------------------------
-- Function to convert an entry in an old declensions table to new.
local function old_decl_entry_to_new(v)
if not v then
return nil
elseif type(v) == "table" then
local new_entry = {}
for _, i in ipairs(v) do
table.insert(new_entry, old_decl_entry_to_new(i))
end
return new_entry
elseif type(v) == "function" then
return function(stem, suffix, args)
return old_decl_entry_to_new(v(stem, suffix, args))
end
else
return old_to_new(v)
end
end
-- Function to convert an old declensions table to new.
local function old_decl_to_new(odecl)
local ndecl = {}
for k, v in pairs(odecl) do
ndecl[k] = old_decl_entry_to_new(v)
end
return ndecl
end
-- Function to convert an entry in an old declensions_cat table to new.
local function old_decl_cat_entry_to_new(odecl_cat_entry)
if not odecl_cat_entry then
return nil
elseif type(odecl_cat_entry) == "function" then
return function(suffix)
return old_decl_cat_entry_to_new(odecl_cat_entry(suffix))
end
elseif type(odecl_cat_entry) == "table" then
local ndecl_cat_entry = {}
for k, v in pairs(odecl_cat_entry) do
ndecl_cat_entry[k] = old_decl_cat_entry_to_new(v)
end
return ndecl_cat_entry
elseif type(odecl_cat_entry) == "boolean" then
return odecl_cat_entry
else
assert(type(odecl_cat_entry) == "string")
return old_to_new(odecl_cat_entry)
end
end
-- Function to convert an old declensions_cat table to new.
local function old_decl_cat_to_new(odeclcat)
local ndeclcat = {}
for k, v in pairs(odeclcat) do
ndeclcat[k] = old_decl_cat_entry_to_new(v)
end
return ndeclcat
end
-- populate declensions[] from declensions_old[]
for odecltype, odecl in pairs(declensions_old) do
local ndecltype = old_to_new(odecltype)
if not declensions[ndecltype] then
declensions[ndecltype] = old_decl_to_new(odecl)
end
end
-- populate declensions_cat[] from declensions_old_cat[]
for odecltype, odeclcat in pairs(declensions_old_cat) do
local ndecltype = old_to_new(odecltype)
if not declensions_cat[ndecltype] then
declensions_cat[ndecltype] = old_decl_cat_to_new(odeclcat)
end
end
-- populate declensions_aliases[] from declensions_old_aliases[]
for ofrom, oto in pairs(declensions_old_aliases) do
local from = old_to_new(ofrom)
if not declensions_aliases[from] then
declensions_aliases[from] = old_to_new(oto)
end
end
-- populate internal_notes_table[] from internal_notes_table_old[]
for odecl, note in pairs(internal_notes_table_old) do
local ndecl = old_to_new(odecl)
if not internal_notes_table[ndecl] then
-- FIXME, should we be calling old_to_new() here?
internal_notes_table[ndecl] = note
end
end
--------------------------------------------------------------------------
-- Inflection functions --
--------------------------------------------------------------------------
-- Attach the stressed stem (or plural stem, or barestem) out of ARGS
-- to the unstressed suffix SUF, modifying the suffix as necessary for the
-- last letter of the stem (e.g. if it is velar, sibilant or ц). CASE is
-- the case form being created and is used to select the plural stem if
-- needed. Returns two values, the combined form and the modified suffix.
local function attach_unstressed(args, case, suf, was_stressed)
if suf == nil then
return nil, nil
elseif rfind(suf, CFLEX) then -- if suf has circumflex accent, it forces stressed
return attach_stressed(args, case, suf)
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.pl, args.pltr
end
if not stem and case == "ins_sg" then
stem, tr = args.ins_sg_stem, args.ins_sg_tr
end
if not stem then
stem, tr = args.stem, args.stemtr
end
if nom.nonsyllabic_suffixes[suf] then
-- If gen_pl, use special args.gen_pl_bare if given, else regular
-- args.bare if there isn't a plural stem. If nom_sg, always use
-- regular args.bare.
local barearg, bareargtr
if case == "gen_pl" then
barearg, bareargtr = args.gen_pl_bare, args.gen_pl_baretr
if not barearg and args.pl == args.stem then
barearg, bareargtr = args.bare, args.baretr
end
else
barearg, bareargtr = args.bare, args.baretr
end
local barestem = barearg or stem
local barestem, baretr
if barearg then
barestem, baretr = barearg, bareargtr
else
barestem, baretr = stem, tr
end
if was_stressed and case == "gen_pl" then
if not barearg then
local gen_pl_stem, gen_pl_tr = com.make_ending_stressed(stem, tr)
barestem, baretr = gen_pl_stem, gen_pl_tr
end
end
if rlfind(barestem, "[йьъ]$") then
suf = ""
else
if suf == "ъ" then
-- OK
elseif suf == "й" or suf == "ь" then
if barearg and case == "gen_pl" then
-- explicit bare or reducible, don't add -ь
suf = ""
elseif rfind(barestem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
-- not reducible, do add -ь and correct to -й if necessary
suf = "й"
else
suf = "ь"
end
end
end
return com.concat_russian_tr(barestem, baretr, suf, nil, "dopair"), suf
end
suf = com.make_unstressed(suf)
local rules = nom.unstressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Analogous to attach_unstressed() but for the unstressed stem and a
-- stressed suffix.
attach_stressed = function(args, case, suf)
if suf == nil then
return nil, nil
end
-- circumflex forces stress even when the accent pattern calls for no stress
suf = rsub(suf, "̂", AC)
if com.is_unstressed(suf) then
return attach_unstressed(args, case, suf, "was stressed")
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.upl, args.upltr
end
if not stem then
stem, tr = args.ustem, args.ustemtr
end
local rules = nom.stressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Attach the appropriate stressed or unstressed stem (or plural stem as
-- determined by CASE, or barestem) out of ARGS to the suffix SUF, which may
-- be a list of alternative suffixes (e.g. in the inst sg of feminine nouns).
-- Calls FUN (either attach_stressed() or attach_unstressed()) to do the work
-- for an individual suffix. Returns two values, a list of combined forms
-- and a list of the real suffixes used (which may be modified from the
-- passed-in suffixes, e.g. by removing stress marks or modifying vowels in
-- various ways after a stem-final velar, sibilant or ц). Each combined form
-- is a two-element list {stem, tr} (or a one-element list if tr is nil).
-- IRREG is true if this is an irregular form. We are handling the Nth word;
-- ISLAST is true if this is the last one.
local function attach_with(args, case, suf, fun, irreg, n, islast)
if type(suf) == "table" then
local all_combineds = {}
local all_realsufs = {}
for _, x in ipairs(suf) do
local combineds, realsufs =
attach_with(args, case, x, fun, irreg, n, islast)
for _, combined in ipairs(combineds) do
table.insert(all_combineds, combined)
end
for _, realsuf in ipairs(realsufs) do
table.insert(all_realsufs, realsuf)
end
end
return all_combineds, all_realsufs
else
local combined, realsuf = fun(args, case, suf)
local irregsuf = irreg and {IRREGMARKER} or {""}
return {combined and com.concat_paired_russian_tr(
com.concat_paired_russian_tr(args["prefix" .. n], combined),
com.concat_paired_russian_tr(args["suffix" .. n], irregsuf)) or nil},
{realsuf and realsuf .. args["suffix" .. n][1] or nil}
end
end
-- Generate the form(s) and suffix(es) for CASE according to the declension
-- table DECL, using the attachment function FUN (one of attach_stressed()
-- or attach_unstressed()). IS_SLASH is true if this is a slash declension
-- (different declensions for singular and plural). We are handling the Nth
-- word; ISLAST is true if this is the last one.
local function gen_form(args, decl, case, stress, fun, is_slash, n, islast)
local irreg = false
if not args.suffixes[case] then
args.suffixes[case] = {}
end
local decl_sufs = args.old and declensions_old or declensions
decl_sufs = decl_sufs[decl]
local suf = decl_sufs[case]
local decl_cats = args.old and declensions_old_cat or declensions_cat
local ispl = rfind(case, "_pl")
if ispl and (decl_cats[decl].irregpl or args.pl and args.pl ~= args.stem or is_slash) then
irreg = true
end
if case == "nom_pl" and decl_cats[decl].alt_nom_pl then
irreg = true
end
if type(suf) == "function" then
suf = suf(ispl and args.pl or args.stem, stress, args)
end
if case == "gen_pl" and args.alt_gen_pl then
suf = decl_sufs.alt_gen_pl
irreg = true
if not suf then
error("No alternative genitive plural available for this declension class")
end
end
local combineds, realsufs = attach_with(args, case, suf, fun, irreg, n, islast)
for _, realsuf in ipairs(realsufs) do
args.any_non_nil[case] = true
args.this_any_non_nil[case] = true
insert_if_not(args.suffixes[case], realsuf)
end
return combineds
end
local attachers = {
["+"] = attach_stressed,
["-"] = attach_unstressed,
}
do_stress_pattern = function(stress, args, decl, number, n, islast)
local f = {}
for _, case in ipairs(decl_cases) do
if not number or (number == "sg" and rfind(case, "_sg")) or
(number == "pl" and rfind(case, "_pl")) then
f[case] = gen_form(args, decl, case, stress,
attachers[stress_patterns[stress][case]], not not number,
n, islast)
-- Turn empty form lists into nil to facilitate computation of
-- animate/inanimate accusatives below
if f[case] and #f[case] == 0 then
f[case] = nil
end
-- Compute linked versions of potential lemma cases, for use
-- in the ru-noun+ headword. We substitute the original lemma
-- (before removing links) for forms that are the same as the
-- lemma, if the original lemma has links.
if f[case] and (case == "nom_sg" or case == "nom_pl") then
local linked_forms = {}
for _, form in ipairs(f[case]) do
-- Return true if FORM is "close enough" to LEMMA that we can substitute the
-- linked form of the lemma. Currently this means exactly the same except that
-- we ignore acute and grave accent differences in monosyllables, and ignore
-- notes that may have been appended (e.g. the triangle marking irregularity).
local entry, notes = m_table_tools.separate_notes(form[1])
local lemma = args.lemma_no_links
local close_enough_to_lemma =
entry == lemma or (com.is_monosyllabic(entry) and
com.is_monosyllabic(lemma) and
com.remove_accents(entry) == com.remove_accents(lemma))
if close_enough_to_lemma and
rfind(args.orig_lemma, "%[%[") then
table.insert(linked_forms, {args.orig_lemma .. notes, args.lemmatr and args.lemmatr .. notes})
else
table.insert(linked_forms, form)
end
end
f[case .. "_linked"] = linked_forms
end
end
end
-- Set acc an/in variants now as appropriate. We used to do this in
-- handle_forms_and_overrides(), which simplified the handling of
-- nom/gen/acc overrides but caused problems for words like мазло and
-- трепло that had a mixture of nil and non-nil accusative forms.
local an = args.thisa
if not number or number == "sg" then
f.acc_sg_an = f.acc_sg_an or f.acc_sg or an == "i" and f.nom_sg or f.gen_sg
f.acc_sg_in = f.acc_sg_in or f.acc_sg or an == "a" and f.gen_sg or f.nom_sg
end
if not number or number == "pl" then
f.acc_pl_an = f.acc_pl_an or f.acc_pl or an == "i" and f.nom_pl or f.gen_pl
f.acc_pl_in = f.acc_pl_in or f.acc_pl or an == "a" and f.gen_pl or f.nom_pl
end
for case, forms in pairs(f) do
if not args.forms[case] then
args.forms[case] = {}
end
for _, form in ipairs(forms) do
insert_if_not(args.forms[case], form)
end
end
end
stress_patterns["a"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["b"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["b'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["c"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["d"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["d'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["e"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f''"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
ending_stressed_gen_pl_patterns = m_table.listToSet({"b", "b'", "c", "e", "f", "f'", "f''"})
ending_stressed_pre_sg_patterns = m_table.listToSet({"b", "b'", "d", "d'", "f", "f'", "f''"})
ending_stressed_dat_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_pl_patterns = m_table.listToSet({"b", "b'", "c"})
local numbers = {
["s"] = "singular",
["p"] = "plural",
}
local old_title_temp = [=[Pre-reform declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local title_temp = [=[Declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local extra_case_template, extra_case_template_with_plural
local internal_notes_template
local notes_template
local templates = {}
-- Convert a raw override into a canonicalized list of individual overrides.
-- If input is nil, so is output. Certain junk (e.g. <br/>) is removed,
-- and ~ and ~~ are substituted appropriately; ARGS and CASE are required for
-- this purpose. FORMS is the list of existing case forms and is currently
-- used only to retrieve the dat_sg for handling + in loc/par (this means
-- that any overrides of the dat_sg have to be handled before handling
-- overrides of loc/par). N is the suffix used to retrieve the override --
-- either a number for word-specific overrides, or an empty string for
-- overall overrides.
--
-- It will still be necessary to call m_table_tools.separate_notes() to separate
-- off any trailing "notes" (asterisks, superscript numbers, etc.), and
-- m_links.remove_links() to remove any links to get the raw override form.
canonicalize_override = function(args, case, forms, n)
local val = args[case .. n]
if not val then
return nil
end
-- clean <br /> that's in many multi-form entries and messes up linking
val = rsub(val, "<br%s*/>", "")
-- substitute ~ and ~~ and split by commas
local stem, manualtr
if rfind(case, "_pl") then
stem, manualtr = args.pl, args.pltr
end
if not stem then
stem, manualtr = args.stem, args.stemtr
end
local ustem, umanualtr = com.make_unstressed_once(stem, manualtr)
local stemtr, ustemtr = manualtr, umanualtr
local vals = rsplit(val, "%s*,%s*")
local retvals = {}
for _, val in ipairs(vals) do
local valru, valtr = com.split_russian_tr(val)
valru = rsub(valru, "~~", ustem)
valru = rsub(valru, "~", com.is_stressed(val) and ustem or stem)
if rfind(valru, "^%*") then
valru = rsub(valru, "^%*", "") .. HYPMARKER
end
if valtr then
stemtr = stemtr or com.translit_no_links(stem)
ustemtr = ustemtr or com.translit_no_links(ustem)
valtr = rsub(valtr, "~~", ustemtr)
valtr = rsub(valtr, "~", com.is_stressed(val) and ustemtr or stemtr)
elseif val:find("~") and manualtr then
valtr = com.translit_no_links(val)
valtr = rsub(valtr, "~~", umanualtr)
valtr = rsub(valtr, "~", com.is_stressed(val) and umanualtr or manualtr)
end
if valtr then
if valtr:sub(1, 1) == "*" then
valtr = valtr:sub(2) .. HYPMARKER
end
valtr = com.j_correction(valtr)
end
table.insert(retvals, {valru, valtr})
end
vals = retvals
-- handle + in loc/par meaning "the expected form"; NOTE: This requires
-- that dat_sg has already been processed!
if case == "loc" or case == "par" then
local new_vals = {}
for _, rutr in ipairs(vals) do
-- don't just handle + by itself in case the arg has в or на
-- or whatever attached to it
if rfind(rutr[1], "^%+") or rfind(rutr[1], "[%s%[|]%+") then
for _, dat in ipairs(forms["dat_sg"]) do
local ru, tr = rutr[1], rutr[2]
local datru, dattr = dat[1], dat[2]
local valru, valtr
-- separate off any footnote symbols (which may have been
-- introduced by a *tail or *tailall argument, which we need
-- to process before this stage so overrides don't get
-- automatically marked with footnote symbols); if par,
-- try to preserve them, but with loc this can cause
-- problems in case there are multiple dative forms
-- (stress variants) and the last one is marked with a
-- footnote symbol (occurs in грудь)
local datru_entry, datru_notes = m_table_tools.separate_notes(datru)
local dattr_entry, dattr_notes
if dattr then
dattr_entry, dattr_notes = m_table_tools.separate_notes(dattr)
end
if case == "par" then
valru, valtr = datru_entry, dattr_entry
else
valru, valtr = com.make_ending_stressed(datru_entry, dattr_entry)
datru_notes = ""
dattr_notes = ""
end
-- wrap the word in brackets so it's linked; but not if it
-- appears to already be linked
ru = rsub(ru, "^%+", "[[" .. valru .. "]]")
ru = rsub(ru, "([%[|])%+", "%1" .. valru)
ru = rsub(ru, "(%s)%+", "%1[[" .. valru .. "]]")
ru = ru .. datru_notes
-- do the translit; but it shouldn't have brackets in it
if tr or valtr then
tr = tr or com.translit_no_links(rutr[1])
valtr = valtr or com.translit_no_links(valru)
tr = rsub(tr, "^%+", valtr)
tr = rsub(tr, "(%s)%+", "%1" .. valtr)
tr = tr .. dattr_notes
end
table.insert(new_vals, {ru, tr})
end
else
table.insert(new_vals, rutr)
end
end
vals = new_vals
end
-- auto-accent/check for necessary accents
local newvals = {}
for _, v in ipairs(vals) do
local ru, tr = v[1], v[2]
if not args.allow_unaccented then
if tr and com.is_unstressed(ru) ~= com.is_unstressed(tr) then
error("Override " .. ru .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(ru) then
ru, tr = com.make_ending_stressed(ru, tr)
elseif com.needs_accents(ru) then
error("Override " .. ru .. " for case " .. case .. n .. " requires an accent")
end
end
table.insert(newvals, {ru, tr})
end
vals = newvals
return vals
end
local function process_overrides(args, f, n)
local function process_override(case)
if args[case .. n] then
local overrides = canonicalize_override(args, case, f, n)
if not f[case] then
f[case] = {}
end
local new_overrides = {}
for _, form in ipairs(overrides) do
-- Don't consider overrides of loc/par/voc irregular since
-- they're only specified through overrides; FIXME: Theoretically
-- we could consider loc/par irregular if they don't follow the
-- expected forms; but we'd have to figure out how to eliminate
-- the preposition that may be specified
if not overridable_only_cases_set[case] and
not args.manual and
not contains_rutr_pair(f[case], form) then
local formru, formnotes = m_table_tools.separate_notes(form[1])
if formru ~= "-" then
-- don't mark an override of - as irregular, even if
-- it has an attached footnote symbol
form = com.concat_paired_russian_tr(form, {IRREGMARKER})
end
end
if case == "pauc" then
-- Internal note indicating that the form is for numbers 2, 3 and 4.
form = com.concat_paired_russian_tr(form, {paucal_marker})
end
table.insert(new_overrides, form)
end
f[case] = new_overrides
args.any_overridden[case] = true
end
end
-- do dative singular first because it will be used by loc/par
process_override("dat_sg")
-- now do the rest
for _, case in ipairs(overridable_cases) do
if case ~= "dat_sg" then
process_override(case)
end
end
-- if the nominative is overridden, use it to set the linked version unless
-- that is also specifically overridden
if args["nom_sg" .. n] and not args["nom_sg_linked" .. n] then
f.nom_sg_linked = f.nom_sg
end
if args["nom_pl" .. n] and not args["nom_pl_linked" .. n] then
f.nom_pl_linked = f.nom_pl
end
-- convert empty lists to nil to facilitate computation of accusative
-- case variants below.
for _, case in ipairs(all_cases) do
if f[case] then
if type(f[case]) ~= "table" then
error("Logic error, args[case] should be nil or table")
end
if #f[case] == 0 then
f[case] = nil
end
end
end
end
local function process_tail_args(args, f, n)
local function append_note_all(case, value)
value = com.split_russian_tr(value, "dopair")
local function append1(case)
if f[case] then
for i=1,#f[case] do
f[case][i] = com.concat_paired_russian_tr(f[case][i], value)
end
end
end
append1(case)
if case == "acc_sg" then
append1("acc_sg_in")
append1("acc_sg_an")
elseif case == "acc_pl" then
append1("acc_pl_in")
append1("acc_pl_an")
end
end
local function append_note_last(case, value, gt_one)
value = com.split_russian_tr(value, "dopair")
local function append1(case)
if f[case] then
local lastarg = #f[case]
if lastarg > (gt_one and 1 or 0) then
f[case][lastarg] = com.concat_paired_russian_tr(f[case][lastarg], value)
end
end
end
append1(case)
if case == "acc_sg" then
append1("acc_sg_in")
append1("acc_sg_an")
elseif case == "acc_pl" then
append1("acc_pl_in")
append1("acc_pl_an")
end
end
local function handle_tail_hyp(suf)
local arg, val
for _, case in ipairs(all_cases) do
arg = args[case .. "_" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args[case .. "_" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
if not rfind(case, "_pl") then
arg = args["sg" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["sg" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
else
arg = args["pl" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["pl" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
end
if not rfind(case, "nom_") and not rfind(case, "acc_") then
arg = args["obl" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["obl" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
end
end
end
handle_tail_hyp("tail")
handle_tail_hyp("hyp")
end
handle_forms_and_overrides = function(args, n, islast)
local f = args.forms
process_tail_args(args, f, n)
process_overrides(args, f, n)
local an = args.thisa
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). This logic duplicates logic in handle_overall_forms_and_overrides();
-- see long comment there. We need to duplicate the whole logic here to
-- handle words like мать-одиночка, which has an override of acc_sg1.
if not args["acc_sg_an" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_an = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_an = f.acc_sg or an == "i" and f.nom_sg or f.gen_sg or f.acc_sg_an
end
end
if not args["acc_sg_in" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_in = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_in = f.acc_sg or an == "a" and f.gen_sg or f.nom_sg or f.acc_sg_in
end
end
if not args["acc_pl_an" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_an = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_an = f.acc_pl or an== "i" and f.nom_pl or f.gen_pl or f.acc_pl_an
end
end
if not args["acc_pl_in" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_in = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_in = f.acc_pl or an == "a" and f.gen_pl or f.nom_pl or f.acc_pl_in
end
end
f.loc = f.loc or f.pre_sg
f.par = f.par or f.gen_sg
f.voc = f.voc or f.nom_sg
-- Set these in case we have plural only, in which case the
-- singular will also get set to these same values in case we are
-- a plural-only word in a singular-only expression.
f.loc_pl = f.loc_pl or f.pre_pl
f.par_pl = f.par_pl or f.gen_pl
f.voc_pl = f.voc_pl or f.nom_pl
f.count = f.count or f.gen_pl
f.pauc = f.pauc or f.gen_sg
local nu = args.thisn
-- If we have a singular-only, set the plural forms to the singular forms,
-- and vice-versa. This is important so that things work in multi-word
-- expressions that combine different number restrictions (e.g.
-- singular-only with singular/plural or singular-only with plural-only,
-- compare "St. Vincent and the Grenadines" [Сент-Винсент и Гренадины]).
if nu == "s" then
f.nom_pl_linked = f.nom_sg_linked
f.nom_pl = f.nom_sg
f.gen_pl = f.gen_sg
f.dat_pl = f.dat_sg
f.acc_pl = f.acc_sg
f.acc_pl_an = f.acc_sg_an
f.acc_pl_in = f.acc_sg_in
f.ins_pl = f.ins_sg
f.pre_pl = f.pre_sg
f.nom_pl = f.nom_sg
f.loc_pl = f.loc
f.par_pl = f.par
f.voc_pl = f.voc
elseif nu == "p" then
f.nom_sg_linked = f.nom_pl_linked
f.nom_sg = f.nom_pl
f.gen_sg = f.gen_pl
f.dat_sg = f.dat_pl
f.acc_sg = f.acc_pl
f.acc_sg_an = f.acc_pl_an
f.acc_sg_in = f.acc_pl_in
f.ins_sg = f.ins_pl
f.pre_sg = f.pre_pl
f.nom_sg = f.nom_pl
f.loc = f.loc_pl
f.par = f.par_pl
f.voc = f.voc_pl
end
end
handle_overall_forms_and_overrides = function(args)
local overall_forms = {}
for _, case in ipairs(displayable_cases) do
overall_forms[case] = concat_word_forms(args.per_word_info, case)
end
local acc_sg_overridden = args.acc_sg
local acc_pl_overridden = args.acc_pl
process_tail_args(args, overall_forms, "")
process_overrides(args, overall_forms, "")
if case_will_be_displayed(args, "pauc") then
insert_if_not(args.internal_notes, paucal_internal_note)
end
-- if IRREGMARKER is anywhere in text, remove all instances and put
-- at the end before any notes.
local function clean_irreg_marker(case, text)
if rfind(text, IRREGMARKER) then
text = rsub(text, IRREGMARKER, "")
local entry, notes = m_table_tools.separate_notes(text)
if case_will_be_displayed(args, case) then
insert_if_not(args.internal_notes, IRREGMARKER .. " Irregular.")
args.any_irreg = true
args.any_irreg_case[case] = true
end
return entry .. IRREGMARKER .. notes
else
return text
end
end
-- set final args[case] and clean up IRREGMARKER.
for _, case in ipairs(all_cases) do
args[case] = overall_forms[case]
if args[case] then
local cleaned_forms = {}
for _, form in ipairs(args[case]) do
local ru, tr = form[1], form[2]
ru = clean_irreg_marker(case, ru)
if tr then
tr = clean_irreg_marker(case, tr)
end
table.insert(cleaned_forms, {ru, tr})
end
args[case] = cleaned_forms
end
end
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). We need to do this somewhat complicated procedure to get overrides
-- to work correctly, e.g. acc_sg overrides of feminine and neuter nouns
-- (such as мать and полслова) and gen_sg/gen_pl overrides of masculine
-- animate nouns. We also ran into an issue with words like мазло that are
-- neuter-form but can be both masculine animate (in which case the
-- acc sg inherits from the genitive) and neuter animate (in which case the
-- acc sg has the fixed ending -о, same as nominative, despite the animacy).
-- Remember also that the animate/inanimate accusative variants are the ones
-- displayed, not the plain acc_sg/acc_pl ones.
if not args.any_overridden.acc_sg_an then
if acc_sg_overridden then
args.acc_sg_an = args.acc_sg
elseif not args.any_non_nil.acc_sg then
args.acc_sg_an = args.acc_sg or args.a == "i" and args.nom_sg or args.gen_sg or args.acc_sg_an
end
end
if not args.any_overridden.acc_sg_in then
if acc_sg_overridden then
args.acc_sg_in = args.acc_sg
elseif not args.any_non_nil.acc_sg then
args.acc_sg_in = args.acc_sg or args.a == "a" and args.gen_sg or args.nom_sg or args.acc_sg_in
end
end
if not args.any_overridden.acc_pl_an then
if acc_pl_overridden then
args.acc_pl_an = args.acc_pl
elseif not args.any_non_nil.acc_pl then
args.acc_pl_an = args.acc_pl or args.a == "i" and args.nom_pl or args.gen_pl or args.acc_pl_an
end
end
if not args.any_overridden.acc_pl_in then
if acc_pl_overridden then
args.acc_pl_in = args.acc_pl
elseif not args.any_non_nil.acc_pl then
args.acc_pl_in = args.acc_pl or args.a == "a" and args.gen_pl or args.nom_pl or args.acc_pl_in
end
end
-- Try to set the values of acc_sg and acc_pl. The only time we can't is
-- when the noun is bianimate and the anim/inan values are different.
-- This is used primarily for generate_forms() and generate_multi_forms(),
-- since we don't actually display these forms.
if args.a == "a" then
args.acc_sg = args.acc_sg or args.acc_sg_an
args.acc_pl = args.acc_pl or args.acc_pl_an
elseif args.a == "i" then
args.acc_sg = args.acc_sg or args.acc_sg_in
args.acc_pl = args.acc_pl or args.acc_pl_in
else -- bianimate
args.acc_sg = args.acc_sg or m_table.deepEquals(args.acc_sg_in, args.acc_sg_an) and args.acc_sg_in or nil
args.acc_pl = args.acc_pl or m_table.deepEquals(args.acc_pl_in, args.acc_pl_an) and args.acc_pl_in or nil
end
end
-- Subfunction of concat_word_forms(), used to implement recursively
-- generating all combinations of elements from WORD_FORMS (a list, one
-- element per word, of a list of the forms for a word, each of which is a
-- two-element list of {RUSSIAN, TR}) and TRAILING_FORMS (a list of forms, the
-- accumulated suffixes for trailing words so far in the recursion process,
-- again where each form is a two-element list {RUSSIAN, TR}). Each time we
-- recur we take the last FORMS item off of WORD_FORMS and to each form in
-- FORMS we add all elements in TRAILING_FORMS, passing the newly generated
-- list of items down the next recursion level with the shorter WORD_FORMS.
-- We end up returning a list of concatenated forms, where each list item
-- is a two-element list {RUSSIAN, TR}.
local function concat_word_forms_1(word_forms, trailing_forms)
if #word_forms == 0 then
local retforms = {}
for _, form in ipairs(trailing_forms) do
local ru, tr = form[1], form[2]
-- Remove <insa> and <insb> markers; they've served their purpose.
ru = rsub(ru, "<ins[ab]>", "")
tr = tr and rsub(tr, "<ins[ab]>", "")
table.insert(retforms, {ru, tr})
end
return retforms
else
local last_form_info = table.remove(word_forms)
local last_forms, joiner = last_form_info[1], last_form_info[2]
local new_trailing_forms = {}
for _, form in ipairs(last_forms) do
for _, trailing_form in ipairs(trailing_forms) do
-- If form to prepend is empty, don't add the joiner; this
-- is principally used in overall overrides, where we stuff
-- the entire override into the last word
local full_form = form[1] == "" and trailing_form or
com.concat_paired_russian_tr(form,
com.concat_paired_russian_tr(joiner, trailing_form),
"movenotes")
if rfind(full_form[1], "<insa>") and rfind(full_form[1], "<insb>") then
-- REJECT! So we don't get mixtures of the two feminine
-- instrumental singular endings.
else
table.insert(new_trailing_forms, full_form)
end
end
end
return concat_word_forms_1(word_forms, new_trailing_forms)
end
end
-- Generate a list of overall forms by concatenating the per-word forms.
-- PER_WORD_INFO comes from args.per_word_info and is a list of
-- WORD_INFO items, one per word, each of which a two element list of
-- WORD_FORMS (a table listing the forms for each case) and JOINER (a string).
-- We loop over all possible combinations of elements from each word's list
-- of forms for the given case; this requires recursion.
concat_word_forms = function(per_word_info, case)
local word_forms = {}
-- Gather the appropriate word forms. We have to recreate this anew
-- because it will be destructively modified by concat_word_forms_1().
for _, word_info in ipairs(per_word_info) do
table.insert(word_forms, {word_info[1][case], word_info[2]})
end
-- We need to start the recursion with the second parameter containing
-- one blank element rather than no elements, otherwise no elements
-- will be propagated to the next recursion level.
return concat_word_forms_1(word_forms, {{""}})
end
local accel_forms = {
nom_sg = "nom|s",
nom_sg_linked = "nom|s",
nom_pl = "nom|p",
nom_pl_linked = "nom|p",
gen_sg = "gen|s",
gen_pl = "gen|p",
dat_sg = "dat|s",
dat_pl = "dat|p",
acc_sg_an = "an|acc|s",
acc_pl_an = "an|acc|p",
acc_sg_in = "in|acc|s",
acc_pl_in = "in|acc|p",
ins_sg = "ins|s",
ins_pl = "ins|p",
pre_sg = "pre|s",
pre_pl = "pre|p",
loc = "loc|s",
loc_pl = "loc|p",
voc = "voc|s",
voc_pl = "voc|p",
par = "par|s",
par_pl = "par|p",
count = "count form",
pauc = "pau",
}
-- Make the table
make_table = function(args)
local data = {}
data.after_title = " " .. args.heading
data.number = args.nonumber and "" or numbers[args.n]
local lemma_forms = args[args.n == "p" and "nom_pl" or "nom_sg"]
data.lemma = nom.show_form(args.explicit_lemma and {{args.explicit_lemma, args.explicit_lemmatr}} or
args[args.n == "p" and "nom_pl_linked" or "nom_sg_linked"], "lemma", nil, nil)
data.title = args.title or format(args.old and old_title_temp or title_temp, data)
local sg_an_in_equal = m_table.deepEquals(args.acc_sg_an, args.acc_sg_in)
local pl_an_in_equal = m_table.deepEquals(args.acc_pl_an, args.acc_pl_in)
for _, case in ipairs(displayable_cases) do
local accel_form = accel_forms[case]
if not accel_form then
error("Something wrong, can't find accelerator form for " .. case)
end
if (sg_an_in_equal and (case == "acc_sg_an" or case == "acc_sg_in") or
pl_an_in_equal and (case == "acc_pl_an" or case == "acc_pl_in")) then
accel_form = rsub(accel_form, "^[ai]n|", "")
end
if args.n == "p" then
accel_form = rsub(accel_form, "|p$", "")
end
data[case] = nom.show_form(args[case], false, accel_form, lemma_forms, "remove monosyllabic accents only lemma")
end
local temp = nil
if args.n == "s" then
data.nom_x = data.nom_sg
data.gen_x = data.gen_sg
data.dat_x = data.dat_sg
data.acc_x_an = data.acc_sg_an
data.acc_x_in = data.acc_sg_in
data.ins_x = data.ins_sg
data.pre_x = data.pre_sg
if sg_an_in_equal then
temp = "one_number"
else
temp = "one_number_split_animacy"
end
elseif args.n == "p" then
data.nom_x = data.nom_pl
data.gen_x = data.gen_pl
data.dat_x = data.dat_pl
data.acc_x_an = data.acc_pl_an
data.acc_x_in = data.acc_pl_in
data.ins_x = data.ins_pl
data.pre_x = data.pre_pl
data.par = data.par_pl
data.loc = data.loc_pl
data.voc = data.voc_pl
if pl_an_in_equal then
temp = "one_number"
else
temp = "one_number_split_animacy"
end
else
if pl_an_in_equal then
temp = "both_numbers"
elseif sg_an_in_equal then
temp = "both_numbers_split_animacy_plural_only"
else
temp = "both_numbers_split_animacy"
end
end
for _, extra_case in ipairs({
{"par", "partitive"},
{"loc", "locative"},
{"voc", "vocative"},
}) do
local case, engcase, template = unpack(extra_case)
local a = args.a
local colspan = (a == "b" or a == "bi" or a == "both" or a == "ai" or a == "ia") and 2 or 1
if args.n ~= "s" and args.n ~= "p" and args.any_overridden[case .. "_pl"] then
if not args.any_overridden[case] then
data[case] = ""
end
template = extra_case_template_with_plural
elseif args.n ~= "p" and args.any_overridden[case] or args.n == "p" and args.any_overridden[case .. "_pl"] then
template = extra_case_template
end
if template then
template = format(template(colspan), {case=case, engcase=engcase})
data[case .. "_clause"] = format(template, data)
else
data[case .. "_clause"] = ""
end
end
for _, extra_case in ipairs({
{"count", "count form"},
{"pauc", "paucal"},
}) do
local case, engcase, template = unpack(extra_case)
local a = args.a
local colspan = (a == "b" or a == "bi" or a == "both" or a == "ai" or a == "ia") and 2 or 1
if args.n ~= "p" and args.any_overridden[case] then
template = extra_case_template
end
if template then
template = format(template(colspan), {case=case, engcase=engcase})
data[case .. "_clause"] = format(template, data)
else
data[case .. "_clause"] = ""
end
end
local notes = get_arg_chain(args, "notes", "notes")
local all_notes = {}
for _, note in ipairs(args.internal_notes) do
-- Superscript footnote marker at beginning of note, similarly to what's
-- done at end of forms.
local symbol, entry = m_table_tools.get_initial_notes(note)
table.insert(all_notes, symbol .. entry)
end
for _, note in ipairs(notes) do
-- Here too.
local symbol, entry = m_table_tools.get_initial_notes(note)
table.insert(all_notes, symbol .. entry)
end
data.notes = table.concat(all_notes, "<br />")
data.notes_clause = data.notes ~= "" and format(notes_template, data) or ""
return format(templates[temp], data)
end
function extra_case_template(colspan)
colspan = colspan or 1
return ([===[
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=%s | {engcase}
| {\op}{case}{\cl}
|
|-]===]):format(colspan)
end
function extra_case_template_with_plural(colspan)
colspan = colspan or 1
return ([===[
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=%s | {engcase}
| {\op}{case}{\cl}
| {\op}{case}_pl{\cl}
|-]===]):format(colspan)
end
notes_template = [===[
<div style="width:100%;text-align:left;background:var(--wikt-palette-lightblue, #d9ebff);">
<div style="display:inline-block;text-align:left;padding-left:1em;padding-right:1em">
{notes}
</div></div>
]===]
local function template_prelude(min_width)
min_width = min_width or "70"
return rsub([===[
<div>
<div class="NavFrame" data-toggle-category="declension" style="max-width:MINWIDTHem">
<div class="NavHead" style="background:var(--wikt-palette-lighterblue, #ebf4ff);">{title}<span style="font-weight:normal;">{after_title}</span> </div>
<div class="NavContent">
{\op}| style="table-layout:fixed; text-align:center; max-width:MINWIDTHem; width:100%;" class="inflection inflection-table"
|- class="rowgroup"
]===], "MINWIDTH", min_width)
end
local function template_postlude()
return [===[|-{par_clause}{loc_clause}{voc_clause}{count_clause}{pauc_clause}
|{\cl}{notes_clause}</div></div></div>]===]
end
templates["both_numbers"] = template_prelude("45") .. [===[
! style="width:7em;background:var(--wikt-palette-lightblue, #d9ebff);" |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | accusative
| data-accel-col=1 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["both_numbers_split_animacy"] = template_prelude("50") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| data-accel-col=1 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| data-accel-col=1 | {acc_sg_in}
| data-accel-col=2 | {acc_pl_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["both_numbers_split_animacy_plural_only"] = template_prelude("50") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| data-accel-col=1 rowspan=2 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| data-accel-col=2 | {acc_pl_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["one_number"] = template_prelude("30") .. [===[
! style="width:7em;background:var(--wikt-palette-lightblue, #d9ebff);" |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | {number}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | nominative
| {nom_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | genitive
| {gen_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | dative
| {dat_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | accusative
| {acc_x_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | instrumental
| {ins_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | prepositional
| {pre_x}
]===] .. template_postlude()
templates["one_number_split_animacy"] = template_prelude("35") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | {number}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| {nom_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| {gen_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| {dat_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| {acc_x_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| {acc_x_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| {ins_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| {pre_x}
]===] .. template_postlude()
return export
4xe9f3vlbxgqj57x0p9pa3akucutgdj
5669926
5669921
2026-06-23T07:33:24Z
MustafaCavlak
59368
5669926
Scribunto
text/plain
--[=[
This module contains functions for creating inflection tables for Russian
nouns.
Author: Benwing, rewritten from early version by Wikitiki89
Form of arguments: One of the following:
1. LEMMA|DECL|PLSTEM (all arguments optional)
2. ACCENT|LEMMA|DECL|PLSTEM (all arguments optional)
3. multiple sets of arguments separated by the literal word "or"
Arguments:
ACCENT: Accent pattern (a b c d e f b' d' f' f''). Multiple values can
be specified, separated by commas. If omitted, defaults to a or b
depending on the position of stress on the lemma or
explicitly-specified declension.
LEMMA: Lemma form (i.e. nom sg or nom pl), with appropriately-placed
stress; or the stem, if an explicit declension is specified
(in this case, the declension usually looks like an ending, and
the stem is the portion of the lemma minus the ending). In the
first argument set (i.e. first set of arguments separated
by "or"), defaults to page name; in later sets, defaults to lemma
of previous set. A plural form can be given, and causes argument
n= to default to n=p (plural only). Normally, an accent is
required if multisyllabic, and unaccented monosyllables with
automatically be stressed; prefix with * to override both behaviors.
DECL: Declension field. Normally omitted to autodetect based on the
lemma form; see below.
PLSTEM: special plural stem (defaults to stem of lemma)
Additional named arguments:
a: animacy (a/an/anim = animate, i/in/inan = inanimate, b/bi/both/ai = both
(listing animate first in the headword), ia = both (listing inanimate
first in the headword), otherwise inanimate)
n: number restriction (p = plural only, s = singular only, b = both;
defaults to both unless the lemma is plural, in which case it
defaults to plural only)
CASE_NUM or acc_NUM_ANIM or par/loc/voc: override (or multiple
values separated by commas) for case/number combination;
forms auto-linked; can have raw links in it, can have an
ending "note" (*, +, 1, 2, 3, etc.)
pltail: Specify something (usually a * or similar) to attach to
the end of the last plural form when there's more than one. Used in
conjunction with notes= to indicate that alternative plural forms
are obsolete, poetic, etc.
pltailall: Similar pltail= but attaches to all plural forms.
Typically used in conjunction with notes= to make a comment about
the plural as a whole (e.g. it's mostly hypothetical, rare and
awkward, etc.).
sgtail, sgtailall: Same as pltail=, pltailall= but for the singular.
obltail, obltailall: Same as pltail=, pltailall= but for oblique cases
(not the nominative or accusative).
CASE_NUM_tail: Attach the argument to the end of the last form
(whether there's one or more than one) for the particular
case/number combination. Note that this doesn't work quite like
pltail= or sgtail= in that it doesn't skip adding the argument
when there's only one form.
CASE_NUM_tailall: Attach the argument to the end of all forms
specified for the particular case/number combination. Similar to
pltailall= or sgtailall=.
suffix: Add a suffix such as ся to all forms.
prefix: Add a prefix to all forms.
plhyp, plhypall, CASE_NUM_hyp, etc.: Same as pltail, pltailall,
CASE_NUM_tal, etc. but specify that the marked forms are
mostly hypothetical or rare/awkard. Generally you will want
plhypall=y to mark the plural as hypothetical.
Per word named arguments:
All of the above named arguments have per-word variants, e.g.
a1, a2, ...; n1, n2, ...; CASE_NUM1, CASE_NUM2, ...;
pltail1, pltail2, ...; etc. These apply to the individual words of a
form.
Case abbreviations:
nom: nominative
gen: genitive
dat: dative
acc: accusative
ins: instrumental
pre: prepositional
par: partitive
loc: locative
voc: vocative
Number abbreviations:
sg: singular
pl: plural
Animacy abbreviations:
an: animate
in: inanimate
Declension field:
One of the following for regular nouns:
(blank)
GENDER
-VARIANT
GENDER-VARIANT
$
DECLTYPE
DECLTYPE/DECLTYPE
(also, can append various special-case markers to any of the above)
Or one of the following for adjectival nouns:
+
+ь
$
+short, +mixed or +proper
+DECLTYPE
GENDER if present is m, f, n or 3f; for regular nouns, required if the
lemma ends in -ь or is plural, ignored otherwise. 3f is the same
as f but in the case of a plural lemma in -и, detects a
third-declension feminine with singular in -ь rather than a
first-declension feminine with singular in -а or -я.
VARIANT is one way of requesting variant declensions (see also special
case (1) for variant nom pls, and special case (2) for variant
gen pls). The currently allowed values are -ья (will select a
mixed declension that has some normal declension in its singular
and the -ья plural declension); -ин (for animate masculine nouns
in -ин with plural in -е -- note that this is autodetected in the
majority of cases where the ending in -янин or -анин); -ишко
(used for inanimate neuter-form diminutive masculine nouns in
-ишко [also сараю́шко] with nom pl -и and colloquial feminine-ending
alternants in some singular cases); -ище (similar to -ишко but
used for *animate* augmentative masculine neuter-form nouns in
-ище). Variants -ишко and -ище must be given with with special
case (1).
$ is indeclinable words. It is principally useful in multiword
expressions where some of the words are indeclinable.
DECLTYPE is an explicit declension type. Normally you shouldn't use
this, and should instead let the declension type be autodetected
based on the ending, supplying the appropriate hint if needed
(gender for regular nouns, +ь for adjectives). If provided, the
declension type is usually the same as the ending, and if present,
the lemma field should be just the stem, without the ending.
Possibilities for regular nouns are (blank) or # for hard-consonant
declension, а, я, о, е or ё, е́, й, ья, ье or ьё, ь-m, ь-f,
ин, ёнок or онок or енок, ёночек or оночек or еночек, мя,
-а or #-а, ь-я, й-я, о-и or о-ы, -ья or #-ья, $ (indeclinable).
Old-style (pre-reform) declensions use ъ instead of (blank), ъ-а
instead of -а, ъ-ья instead of -ья, and инъ, ёнокъ/онокъ/енокъ,
ёночекъ/оночекъ/еночекъ instead of the same without terminating ъ.
The declensions can also be written with an accent on them; this
chooses the same declension (except for е vs. е́), but causes
ACCENT to default to pattern b instead of a.
For adjectival nouns, you should normally supply just + and let
the ending determine the declension; supply +ь in the case of a
possessive adjectival noun in -ий, which have an extra -ь- in
most endings compared with normal adjectival nouns in -ий, but
which can't be distinguished based on the nominative singular.
You can also supply +short, +mixed or +proper, which constrains
the declension appropriately but still autodetects the
gender-specific and stress-specific variant. If you do supply
a specific declension type, as with regular nouns you need to
omit the ending from the lemma field and supply just the stem.
Possibilities are +ый, +ое, +ая, +ій, +ее, +яя, +ой, +о́е, +а́я,
+ьій, +ье, +ья, +-short or +#-short (masc), +о-short,
+о-stressed-short or +о́-short, +а-short, +а-stressed-short or
+а́-short, and similar for -mixed and -proper (except there aren't
any stressed mixed declensions).
DECLTYPE/DECLTYPE is used for nouns with one declension in the
singular and a different one in the plural, for cases that
PLVARIANT and special case (1) below don't cover.
Special-case markers:
(1) for Zaliznyak-style alternative nominative plural ending:
-а or -я for masculine, -и or -ы for neuter
(2) for Zaliznyak-style alternative genitive plural ending:
-ъ/none for masculine, -ей for feminine, -ов(ъ) for neuter,
-ей for plural variant -ья
* for reducibles (nom sg or gen pl has an extra vowel before the
final consonant as compared with the stem found in other cases)
;ё for Zaliznyak-style alternation between last е in stem and ё
TODO:
1. Multi-word issues:
-- FIXME: Make sure internal_notes handled correctly; we may run into issues
with multiple internal notes from different words, if we're not careful
to use different footnote symbols for each type of footnote (which we
don't do currently). [NOT DONE, MAY NOT DO]
-- Handling default lemma: With multiple words, we should probably split
the page name on spaces and default each word in turn [NOT DONE, MAY
NOT DO]
2a. FIXME: For -ишко diminutives and -ище augmentatives, should add an
appropriate category of some sort (currently marked by colloqfem= in
category).
2b. FIXME: Adding a note to dat_sg also adds it to loc_sg when it exists;
seems wrong. See луг.
2c. FIXME: When you have both d' and f in feminines and you use sgtail=*,
you get two *'s. See User:Benwing2/test-ru-noun-debug.
3. ADJECTIVE FIXMES:
3a. FIXME: Change calls to ru-adj11 to use the new proper name support in
ru-adjective.
3b. FIXME: Test that omitting a manual form in ru-adjective leaves the form as
a big dash.
3c. FIXME: какой-либо and какой-то display genitives with translit -go
instead of -vo. To fix this properly requires implementing real
manual translit for adjectives.
3d. FIXME: Implement real manual translit for adjectives.
5. [FIXME: Consider adding an indicator in the header line when the ё/e
alternation occurs. This is a bit tricky to calculate: If special case
;ё is given, but also if ё occurs in the stem and the accent pattern is
as follows -- for sg-only, b' d' f' f'', also b d f if the noun is masc
or 3rd-decl fem (i.e. nom-sg ending is non-syllabic); for pl-only,
e f f' f'', also b b' c if the gen pl is non-syllabic; for sg/pl,
any but a or b, also b if either nom sg or gen pl is non-syllabic. But
it gets more complicated due to overrides. An alternative is to check
all forms to see if ё is present in some but not all; but this is
tricky also because e.g. reducibles frequently have ё/null alternation,
which doesn't count, and some endings have е or ё in them, which also
doesn't count. If we were to do it this way, we'd have to (a) count
the number of е's in the form(s) with ё and verify that there's at least
one form without ё and with one more е than in the form(s) with ё
(and it gets trickier if different forms with ё have different numbers of
е in them, although that is probably rare); and (b) ignore the appropriate
endings (the best way to do this would probably be to look at the actual
suffixes that were generated in args.suffixes and chop off any matching
ending in the actual form(s), but also chop off final -е/ё, as well as
-ев(ъ)/-ёв(ъ)/-ей/-ёй in the gen pl, which is frequently overridden,
unless perhaps the stem ends in the same way). It'd probably not
possible to do this in a 100% foolproof way but can be "good enough" for
nearly all circumstances.] [MIGHT BE TOO MUCH WORK]
6. HEADWORD FIXMES:
6a. FIXME: In ru-headword, create a category for words whose gender doesn't
match the form. (This is easy to do for ru-noun+ but harder for ru-noun.
We would need to do limited autodetection of the ending: for singulars,
-а/я should be feminine, -е/о/ё should be neuter, -ь should be masculine
or feminine, anything else should be masculine; for plurals, -и/ы should
be masculine or feminine, -а/я should be neuter except that -ія can be
feminine or neuter due to old-style adjectival pluralia tantum nouns,
anything else can be any gender.)
6b. FIXME: Recognize indeclinable nouns and indicate as indeclinable. Probably
should work by checking the case forms to see if they're the same.
9. FIXME: Change stress-pattern detection and overriding to happen inside of
looping over the two parts of a slash decl. Requires that the loop over
the two parts happen outside of the loop over stress patterns. Requires
that the category code get split into two parts, one to handle combined
singular/plural categories that goes outside the two loops, and one to
handle everything else that goes inside the two loops.
10. FIXME: override_matches_suffix() had a free variable reference to ARGS
in it, which should have triggered an error whenever there was a nom_sg or
nom_pl override but didn't. Is there an error causing this never to be
called? Check.
11a. FIXME: In a multiword lemma, using loc2=+ causes only the second word
to get linked instead of the whole expression. Same for par2=+, voc2=+.
11b. FIXME: Using loc=+ with a multiword lemma should do the right thing, same
as if locN=+ is specified for each individual word. Instead it generates
the locative as a whole from the dative, which fails e.g. if some of the
words are adjectival. Same for par=+, voc=+.
13. Multi-word issues:
-- Setting n=pl when auto-detecting a plural lemma. How does that interact
with multi-word stuff? (DONE)
-- compute_heading() -- what to do with multiple words? I assume we should
display info on the first noun (non-indeclinable, non-adjectival), and
on the first adjectival word otherwise, and finally on an indeclinable word (DONE)
-- args.genders -- it should presumably come from the same word as is used
in compute_heading(); but we should allow the overall gender to be
overridden, at least in ru-noun+ (DONE)
-- Bug in args.suffix: Gets added to every word in attach_with() and then
again at the end, after pltail and such. Needs to be added to the
last word only, before pltail. Need also suffixN for individual words.
(DONE, NEEDS TESTING)
-- Should have ..N versions of pltail and variants. (DONE, NEEDS TESTING)
-- Need to handle overrides of acc_sg, acc_pl (DONE)
-- Overrides of nom_sg/nom_pl should also override acc_sg/acc_pl if it
was originally empty and the animacy is inanimate; similarly for
gen_sg/gen_pl and animates; this needs to work both for per-word and
overall overrides. (DONE)
-- do_generate_forms(_multi) need to run part of make_table(), enough to
combine all per_word_info into single lists of forms and store back
into args[case]. (DONE, NEEDS TESTING)
-- In generate_forms, should probably check if a=="i" and only return
acc_sg_in as acc_sg=; or if a=="a" and only return acc_sg_an as acc_sg=;
in old/new comparison code, do something similar, also when a=="b"
check if acc_sg_in==acc_sg_an and make it acc_sg; when a=="b" and the
_in and _an variants are different, might need to ignore them or check
that acc_sg_in==nom_sg and acc_sg_an==gen_sg; similarly for _pl
(DONE, NEEDS TESTING)
-- Need to test with multiple words! [DONE]
-- Current handling of <adj> won't work properly with multiple words;
will need to translate word-by-word in that case (should be solved by
manual-translit branch) [DONE]
14. In multiple-words branch, fix ru-decl-noun-multi so it recognizes
things like *, (1), (2) and ; without the need for a separator. Consider
using semicolon as a separator, since we already use it to separate ё
from a previous declension. Maybe use $ or ~ for an indeclinable word; don't
use semicolon. [IMPLEMENTED. NEED TO TEST.]
16. [Consider having ru-noun+ treat par= as a second genitive in
the headword, as is done with край] [WON'T DO]
17. [FIXME: Consider removing slash patterns and instead handling them by
allowing additional declension flags 'sg' and 'pl'. This simplifies the
various special cases caused by slash declensions. It would also be
possible to remove the special plural stem, which would get rid of more
special cases. On the other hand, it makes it more complicated to support
plural variant -ья with all singular types, and the category code that
displays things like "Russian nouns with singular -X and plural -Y"
also gets more complicated, and there's something convenient and intuitive
about plural stems, and slash declensions are also convenient and at least
somewhat intuitive. One possibility is to externally allow slash
declensions and special plural stems and rewrite them internally to
separate stems with 'sg' and 'pl' declension flags; but there are still
the two coding issues mentioned above.]
18. [FIXME: Consider redoing slash patterns so they operate at the outer
level, i.e. things like special cases apply separately in the singular
and plural part of the slash pattern.]
19. In ru-noun, don't recognize -а with m as plural unless (1)
or n=pl is also given, because there are masculine words with the
feminine ending. Check using the test code whether this changes
anything. Also check if there are other similar cases (neuter with
-и isn't parallel because -и is always plural). [IMPLEMENTED. NEED TO TEST.]
19a. Internal notes weren't propagated properly from adjectives.
[IMPLEMENTED. NEED TO TEST.]
19b. Add support for -ишко and -ище variants (p. 74 of Z),
which conversationally and/or colloquially have feminine endings in
certain cases. [IMPLEMENTED. NEED TO TEST. MAKE SURE THE INTERNAL NOTES
APPEAR.]
19d. For masculine animate neuter-form nouns, the accusative singular
ends in -а (-я soft) instead of -о. [IMPLEMENTED. NEED TO TEST.
NOTE: Currently this variant only can be selected using new-style
arguments where the gender can be given. Perhaps we should consider
allowing gender to be specified with old-style explicit declensions.]
21. Put back gender hints for pl adjectival nouns; used by ru-noun+.
[IMPLEMENTED. NEED TO TEST.]
23. Mixed and proper-noun adjectives have built-in notes. We need to
handle those notes with an "internal_notes" section similar to what is used
in the adjective module. [IMPLEMENTED. NEED TO TEST.]
24. Adjective detection code here needs to work the same as for the
adjective module, in particular in the handling of short, stressed-short,
mixed, proper, stressed-proper. [IMPLEMENTED. NEED TO TEST.]
25. Consider simplifying plural-variant code to only allow -ья as a
plural variant [and maybe even change that to be something like (1')].
[IMPLEMENTED REDUCTION OF PLURAL VARIANTS TO -ья; PLURAL-VARIANT CODE
STILL COMPLEX, THOUGH. NEED TO TEST.]
26. Automatically superscript *, numbers and similar things at the
beginning of a note. Also do this in adjective module. [IMPLEMENTED.
NEED TO TEST.]
28. Make the check for multiple stress patterns (categorizing/tracking)
smarter, to keep a list of them and check at the end, so we handle
multiple stress patterns specified through different arg sets.
[IMPLEMENTED; NEED TO TEST.]
29. More sophisticated handling of user-requested plural variant vs. special
case (1) vs. plural-detected variant. [IMPLEMENTED. NEED TO TEST FURTHER.]
30. Solution to ambiguous plural involving gender spec "3f". [IMPLEMENTED;
NEED TO TEST. Use запчасти, новости.]
33. With pluralia tantum adjectival nouns, we don't know the gender.
By default we assume masculine (or feminine for old-style -ія nouns) and
currently this goes into the category, but shouldn't. [IMPLEMENTED.]
39. [Eventually: Even with decl type explicitly given, the full stem with
ending should be included.] [MAY NEVER IMPLEMENT]
40. [Get error "Unable to dereduce" with strange noun ва́йя, what should
happen?] [WILL NOT FIX; USE AN OVERRIDE]
41. In творог, module generates partitive творогу́ when it should copy the
dative творогу́,тво́рогу. (DONE)
42. [[груз 200]] doesn't work. Interprets 200 as a footnote symbol.
43. When converting е -> ё not after cons and with translit, we should
convert e -> o to avoid double j. (DONE)
44. FIXME: In ро́вня/ровня́, similarly with неровня, marks genitive plural
ровня́ as irregular even though it isn't. (NOT OUR ERROR; THE DECLENSIONS
OF THESE NOUNS MARKED THE ENDING-STRESSED VARIANTS WITH (2).)
]=]--
local m_utilities = require("Modül:araçlar")
local m_table = require("Modül:table")
local m_links = require("Modül:bağlantılar")
local com = require("Modül:ru-genel")
local nom = require("Modül:ru-nominal")
local m_ru_adj = require("Modül:ru-ön ad")
local m_str_utils = require("Modül:string araçları")
local scriptutils = require("Modül:alfabe araçları")
local m_table_tools = require("Modül:table tools")
local m_debug = require("Modül:debug")
local export = {}
local lang = require("Modül:diller").getirKodaGore("ru")
local Latn = require("Modül:alfabeler").getirKodaGore("Latn")
local format = m_str_utils.format
local rfind = m_str_utils.find
local rsubn = m_str_utils.gsub
local rmatch = m_str_utils.match
local rsplit = m_str_utils.split
local u = m_str_utils.char
local ulower = m_str_utils.lower
local usub = m_str_utils.sub
local ulen = m_str_utils.len
local unpack = unpack or table.unpack -- Lua 5.2 compatibility
-- If enabled, compare this module with new version of module to make
-- sure all declensions are the same. Eventually consider removing this;
-- but useful as new code is created.
local test_new_ru_noun_module = false
local AC = u(0x0301) -- acute = ́
local GR = u(0x0300) -- grave = ̀
local CFLEX = u(0x0302) -- circumflex = ̂
local DIA = u(0x0308) -- diaeresis = ̈
local PSEUDOCONS = u(0xFFF2) -- pseudoconsonant placeholder, matching ru-common
local IRREGMARKER = "△"
local HYPMARKER = "⟐"
local paucal_marker = "*"
local paucal_internal_note = "* Used with the numbers 1.5, 2, 3, 4 and higher numbers after 20 ending in 2, 3, and 4."
-- text class to check lowercase arg against to see if Latin text embedded in it
local latin_text_class = "[a-zščžěáéíóúýàèìòùỳâêîôûŷạẹịọụỵȧėȯẏ]"
-- Forward functions
local generate_forms_1
local determine_decl
local handle_forms_and_overrides
local handle_overall_forms_and_overrides
local concat_word_forms
local make_table
local detect_adj_type
local detect_stress_pattern
local override_stress_pattern
local determine_stress_variant
local determine_stem_variant
local is_reducible
local is_dereducible
local add_bare_suffix
local attach_stressed
local do_stress_pattern
local canonicalize_override
-- version of rsubn() that discards all but the first return value
local function rsub(term, foo, bar)
local retval = rsubn(term, foo, bar)
return retval
end
-- version of rsubn() that returns a 2nd argument boolean indicating whether
-- a substitution was made.
local function rsubb(term, foo, bar)
local retval, nsubs = rsubn(term, foo, bar)
return retval, nsubs > 0
end
-- version of rfind() that lowercases its string first, for case-insensitive matching
local function rlfind(term, foo)
return rfind(ulower(term), foo)
end
local function track(page)
m_debug.track("ru-noun/" .. page)
return true
end
-- version of m_table.insertIfNot() that makes sure 'false' doesn't get inserted by mistake.
local function insert_if_not(foo, bar)
assert(bar ~= false)
m_table.insertIfNot(foo, bar)
end
-- Fancy version of ine() (if-not-empty). Converts empty string to nil,
-- but also strips leading/trailing space and then single or double quotes,
-- to allow for embedded spaces.
local function ine(arg)
if not arg then return nil end
arg = rsub(arg, "^%s*(.-)%s*$", "%1")
if arg == "" then return nil end
local inside_quotes = rmatch(arg, '^"(.*)"$')
if inside_quotes then
return inside_quotes
end
inside_quotes = rmatch(arg, "^'(.*)'$")
if inside_quotes then
return inside_quotes
end
return arg
end
-- FIXME: Move to utils
-- Iterate over a chain of parameters, FIRST then PREF2, PREF3, ...,
-- inserting into LIST (newly created if omitted). Return LIST.
local function get_arg_chain(args, first, pref, list)
if not list then
list = {}
end
local val = args[first]
local i = 2
while val do
table.insert(list, val)
val = args[pref .. i]
i = i + 1
end
return list
end
-- synthesize a frame so that exported functions meant to be called from
-- templates can be called from the debug console.
local function debug_frame(parargs, args)
return {args = args, getParent = function() return {args = parargs} end}
end
local function rutr_pairs_equal(term1, term2)
local ru1, tr1 = term1[1], term1[2]
local ru2, tr2 = term2[1], term2[2]
local ru1entry, ru1notes = m_table_tools.separate_notes(m_links.remove_links(ru1))
local ru2entry, ru2notes = m_table_tools.separate_notes(m_links.remove_links(ru2))
if ru1entry ~= ru2entry then
return false
end
local tr1entry, tr1notes
local tr2entry, tr2notes
if tr1 then
tr1entry, tr1notes = m_table_tools.separate_notes(tr1)
end
if tr2 then
tr2entry, tr2notes = m_table_tools.separate_notes(tr2)
end
if tr1entry == tr2entry then
return true
elseif type(tr1entry) == type(tr2entry) then
return false
else
tr1entry = tr1entry or com.translit_no_links(ru1entry)
tr2entry = tr2entry or com.translit_no_links(ru2entry)
return tr1entry == tr2entry
end
end
local function contains_rutr_pair(list, pair)
for _, item in ipairs(list) do
if rutr_pairs_equal(item, pair) then
return true
end
end
return false
end
-- Clone parent's args while also assigning nil to empty strings.
local function clone_args(frame)
local args = {}
for pname, param in pairs(frame:getParent().args) do
args[pname] = ine(param)
end
return args
end
-- Old-style declensions.
local declensions_old = {}
-- New-style declensions; computed automatically from the old-style ones,
-- for the most part.
local declensions = {}
-- Internal notes for old-style declensions.
local internal_notes_table_old = {}
-- Same for new-style declensions.
local internal_notes_table = {}
-- Category and type information corresponding to declensions: These may
-- contain the following fields: 'singular', 'plural', 'decl', 'hard', 'g',
-- 'suffix', 'gensg', 'irregpl', 'alt_nom_pl', 'cant_reduce', 'ignore_reduce',
-- 'stem_suffix'.
--
-- 'singular' is used to construct a category of the form
-- "Russian nouns SINGULAR". If omitted, a category is constructed of the
-- form "Russian nouns ending in -ENDING", where ENDING is the actual
-- nom sg ending shorn of its acute accents; or "Russian nouns ending
-- in suffix -ENDING", if 'suffix' is true. The value of SINGULAR can be
-- one of the following: a single string, a list of strings, or a function,
-- which is passed one argument (the value of ENDING that would be used to
-- auto-initialize the category), and should return a single string or list
-- of strings. Such a category is only constructed if 'gensg' is true.
--
-- 'plural' is analogous but used to construct a category of the form
-- "Russian nouns with PLURAL", and if omitted, a category is constructed
-- of the form "Russian nouns with plural -ENDING", based on the actual
-- nom pl ending shorn of its acute accents. Currently no plural category
-- is actually constructed.
--
-- In addition, a category may normally constructed from the combination of
-- 'singular' and 'plural', appropriately defaulted; e.g. if both are present,
-- the combined category will be "Russian nouns SINGULAR with PLURAL" and
-- if both are missing, the combined category will be
-- "Russian nouns ending in -SGENDING with plural -PLENDING" (or
-- "Russian nouns ending in suffix -SGENDING with plural -PLENDING" if
-- 'suffix' is true). Note that if either singular or plural or both
-- specifies a list, looping will occur over all combinations. Such a
-- category is constructed only if 'irregpl' or 'alt_nom_pl' or 'suffix'
-- is true or if the declension class is a slash class.
--
-- 'decl' is "1st", "2nd", "3rd" or "indeclinable"; 'hard' is "hard", "soft"
-- or "none"; 'g' is "m", "f", "n" or "none"; these are all traditional
-- declension categories.
--
-- If 'suffix' is true, the declension type includes a long suffix
-- added to the string that itself undergoes reducibility and such, and so
-- reducibility cannot occur in the stem minus the suffix. Categories will
-- be created for the suffix.
--
-- 'alt_nom_pl' indicates that the declension has an alternative nominative
-- plural (corresponding to Zaliznyak's special case 1; compare special case 2
-- for alternative genitive plural). 'irregpl' indicates that the entire
-- plural is irregular.
--
-- In addition to the above categories, additional more specific categories
-- are constructed based on the final letter of the stem, e.g.
-- "Russian velar-stem 1st-declension hard nouns". See calls to
-- com.get_stem_trailing_letter_type(). 'stem_suffix', if present, is added to
-- the end of the stem when get_stem_trailing_letter_type() is called.
-- This is the only place that 'stem_suffix' is used. This is for use with
-- the '-ья' and '-ье' declension types, so that the trailing letter is
-- 'ь' and not whatever precedes it.
--
-- 'enable_categories' is a special hack for testing, which disables all
-- category insertion if false. Delete this as soon as we've verified the
-- working of the category code and created all the necessary categories.
local enable_categories = true
-- Category/type info corresponding to old-style declensions; see above.
local declensions_old_cat = {}
-- Category/type info corresponding to new-style declensions. Computed
-- automatically from the old-style ones, for the most part. Same format
-- as the old-style ones.
local declensions_cat = {}
-- Table listing aliases of old-style declension classes.
local declensions_old_aliases = {}
-- Table listing aliases of new-style declension classes; computed
-- automatically from the old-style ones.
local declensions_aliases = {}
local stress_patterns = {}
-- Set of patterns with ending-stressed genitive plural.
local ending_stressed_gen_pl_patterns = {}
-- Set of patterns with ending-stressed prepositional singular.
local ending_stressed_pre_sg_patterns = {}
-- Set of patterns with ending-stressed dative singular.
local ending_stressed_dat_sg_patterns = {}
-- Set of patterns with all singular forms ending-stressed.
local ending_stressed_sg_patterns = {}
-- Set of patterns with all plural forms ending-stressed.
local ending_stressed_pl_patterns = {}
local declinable_cases_except_accusative = {
"nom_sg", "gen_sg", "dat_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "ins_pl", "pre_pl",
}
local accusative_cases_unsplit_animacy = {
"acc_sg", "acc_pl",
}
local accusative_cases_split_animacy = {
"acc_sg_an", "acc_sg_in", "acc_pl_an", "acc_pl_in",
}
local lemma_linked_cases = {
"nom_sg_linked", "nom_pl_linked",
}
local overridable_only_cases = {
"par", "loc", "voc", "par_pl", "loc_pl", "voc_pl", "count", "pauc",
}
local overridable_only_cases_set = m_table.listToSet(overridable_only_cases)
-- List of all cases that are declined normally.
local decl_cases = m_table.append(declinable_cases_except_accusative, accusative_cases_unsplit_animacy)
-- List of all cases that can be overridden (includes all cases except the "linked" lemma case variants). Also
-- currently the same as the cases returned by export.generate_forms().
local overridable_cases = m_table.append(decl_cases, accusative_cases_split_animacy, overridable_only_cases)
-- List of all cases that can be displayed (includes all cases except plain accusatives).
local displayable_cases = m_table.append(declinable_cases_except_accusative, lemma_linked_cases,
accusative_cases_split_animacy, overridable_only_cases)
-- List of all cases, including those that are declined normally (nom/gen/dat/acc/ins/pre sg and pl), plus
-- animate/inanimate accusative variants (computed automatically as appropriate from the previous cases), plus
-- additional overridable cases (loc/par/voc and plural), plus the "linked" lemma case variants used in ru-noun+
-- headwords (nom_sg_linked, nom_pl_linked, whose values come from nom_sg and nom_pl but may have additional embedded
-- links if they were given in the lemma).
local all_cases = m_table.append(overridable_cases, lemma_linked_cases)
local function english_case_description(case)
if case == "par" or case == "loc" or case == "voc" then
-- For historical reasons, the singular of these cases doens't include "_sg" in their code.
case = case .. "_sg"
end
local engcase = rsub(case, "^([a-z]*)", {
nom="nominative", gen="genitive", dat="dative",
acc="accusative", ins="instrumental", pre="prepositional",
par="partitive", loc="locative", voc="vocative",
count="count form", pauc="paucal form",
})
engcase = rsub(engcase, "(_[a-z]*)", {
_sg=" singular", _pl=" plural",
_an="", _in="",
--_an=" animate", _in=" inanimate"
})
return engcase
end
--------------------------------------------------------------------------
-- Tracking and categorization --
--------------------------------------------------------------------------
-- FIXME! Move below the main code
-- FIXME!! Consider deleting most of this tracking code once we've enabled
-- all the categories. Note that some of the tracking categories aren't
-- completely redundant; e.g. we have tracking pages that combine decl and
-- stress classes, such as "а/a" or "о-и/d'", which are more or less
-- equivalent to stem/gender/stress categories, but we also have the same
-- prefixed by "reducible-stem/" for reducible stems.
local function tracking_code(stress, orig_decl, decl, args, n, islast)
assert(orig_decl)
assert(decl)
local hint_types = com.get_stem_trailing_letter_type(args.stem)
if orig_decl == decl then
orig_decl = nil
end
if args.notes then
track("notes")
end
local function all_pl_irreg()
track("irreg")
for _, case in ipairs(overridable_cases) do
if rfind(case, "_pl") then
track("irreg/" .. case)
end
end
end
local function track_prefix(prefix)
local function dotrack(suf)
track(prefix .. suf)
end
dotrack(stress)
dotrack(decl)
dotrack(decl .. "/" .. stress)
if orig_decl then
dotrack(orig_decl)
dotrack(orig_decl .. "/" .. stress)
end
for _, hint_type in ipairs(hint_types) do
dotrack(hint_type)
dotrack(decl .. "/" .. hint_type)
if orig_decl then
dotrack(orig_decl .. "/" .. hint_type)
end
end
end
track_prefix("")
if args.reducible then
track("reducible-stem")
track_prefix("reducible-stem/")
end
if rlfind(args.stem, "и́?н$") and (decl == "" or decl == "#") then
track("irregular-in")
end
if args.pltail then
track("pltail")
end
if args.sgtail then
track("sgtail")
end
if args.pltailall then
track("pltailall")
end
if args.sgtailall then
track("sgtailall")
end
if args.pl ~= args.stem then
track("irreg-pl-stem")
track("irreg")
end
if args.alt_gen_pl then
track("alt-gen-pl")
track("irreg")
track("irreg/gen_pl")
end
if args.want_sc1 then
track("want-sc1")
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-и$") or rfind(decl, "-а$") or rfind(decl, "-я$") then
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-ья$") then
track("variant-ья")
all_pl_irreg()
end
if rfind(decl, "%(ишк%)$") then
track("variant-ишко")
end
if rfind(decl, "%(ищ%)$") then
track("variant-ище")
end
if args.jo_special then
track("jo-special")
end
if args.manual then
track("manual")
end
if args.explicit_gender then
track("explicit-gender")
track("explicit-gender/" .. args.explicit_gender)
end
for _, case in ipairs(overridable_cases) do
if args[case .. n] and not args.manual then
track("override")
track("override/" .. case .. n)
track("irreg")
track("irreg/" .. case .. n)
-- questionable use: track_prefix("irreg/" .. case .. "/")
-- questionable use: track_prefix("irreg/" .. case .. n .. "/")
end
if islast and args[case] and not args.manual then
track("override")
track("override/" .. case)
track("irreg")
track("irreg/" .. case)
-- questionable use: track_prefix("irreg/" .. case .. "/")
end
if args[case .. "_tail"] then
track("casenum-tail")
track("casenum-tail/" .. case)
end
if args[case .. "_tailall"] then
track("casenum-tailall")
track("casenum-tailall/" .. case)
end
end
end
local gender_to_full = {m="masculine", f="feminine", n="neuter"}
local gender_to_short = {m="masc", f="fem", n="neut"}
-- Insert the category CAT (a string) into list CATEGORIES. String will
-- have "Russian " prepended and ~ substituted for the plural part of speech.
local function insert_category(categories, cat, pos, atbeg)
if enable_categories then
local fullcat = "Russian " .. rsub(cat, "~", pos .. "s")
if atbeg then
table.insert(categories, 1, fullcat)
else
table.insert(categories, fullcat)
end
end
end
-- Insert categories into ARGS.CATEGORIES corresponding to the specified
-- stress and declension classes and to the form of the stem (e.g. velar,
-- sibilant, etc.). Also initialize values used to compute the declension
-- heading that describes similar information. N is the number of the
-- word being processed; ISLAST is true if this is the last word.
local function categorize_and_init_heading(stress, decl, args, n, islast)
local function cat_to_list(cat)
if not cat then
return {}
elseif type(cat) == "string" then
return {cat}
else
assert(type(cat) == "table")
return cat
end
end
-- Insert category CAT into the list of categories in ARGS.
-- CAT may be nil, a single string or a list of strings. We call
-- insert_category() on each string. The strings will have "Russian "
-- prepended and "~" replaced with the plural part of speech.
local function insert_cat(cat)
for _, c in ipairs(cat_to_list(cat)) do
insert_category(args.categories, c, args.pos)
end
end
-- "Resolve" the category spec CATSPEC into the sort of category spec
-- accepted by insert_cat(), i.e. nil, a single string or a list of
-- strings. CATSPEC may be any of these or a function, which takes one
-- argument (SUFFIX) and returns another CATSPEC.
local function resolve_cat(catspec, suffix)
if type(catspec) == "function" then
return resolve_cat(catspec(suffix), suffix)
else
return catspec
end
end
-- Check whether an override for nom_sg or nom_pl still contains the
-- normal suffix (which should already have accents removed) in at least
-- one of its entries. If no override then of course we return true.
local function override_matches_suffix(args, case, n, suffix)
assert(suffix == com.remove_accents(suffix))
-- NOTE: It might not be completely correct to pass in args.forms
-- here when n == ""; this is different behavior from
-- handle_overall_forms_and_overrides(). But args.forms is only used
-- to retrieve the dat_sg for handling loc/par, and we're never
-- called with loc or par as the value of CASE, so it doesn't matter.
-- We add an assert to make sure of this.
assert(case ~= "loc" and case ~= "par")
local override = canonicalize_override(args, case, args.forms, n)
if not override then
return true
end
for _, x in ipairs(override) do
local ru, tr = x[1], x[2]
local entry, notes = m_table_tools.separate_notes(ru)
entry = com.remove_accents(m_links.remove_links(entry))
if rlfind(entry, suffix .. "$") then
return true
end
end
return false
end
if args.manual then
return
end
local h = args.heading_info
assert(decl)
local decl_cats = args.old and declensions_old_cat or declensions_cat
local sgdecl, pldecl
local is_slash_decl = rfind(decl, "/")
if is_slash_decl then
local indiv_decls = rsplit(decl, "/")
sgdecl, pldecl = indiv_decls[1], indiv_decls[2]
else
sgdecl, pldecl = decl, decl
end
local sgdc = decl_cats[sgdecl]
local pldc = decl_cats[pldecl]
assert(sgdc)
assert(pldc)
local sghint_types = com.get_stem_trailing_letter_type(
args.stem .. (sgdc.stem_suffix or ""))
-- insert English version of Zaliznyak stem type
if sgdc.decl == "indeclinable" then
insert_cat("indeclinable ~")
insert_if_not(h.stemetc, "indecl")
else
local stem_type =
sgdc.decl == "3rd" and "3rd-declension" or
m_table.contains(sghint_types, "velar") and "velar-stem" or
m_table.contains(sghint_types, "sibilant") and "sibilant-stem" or
m_table.contains(sghint_types, "c") and "ц-stem" or
m_table.contains(sghint_types, "i") and "i-stem" or
m_table.contains(sghint_types, "vowel") and "vowel-stem" or
m_table.contains(sghint_types, "soft-cons") and "vowel-stem" or
m_table.contains(sghint_types, "palatal") and "vowel-stem" or
sgdc.hard == "soft" and "soft-stem" or
"hard-stem"
local short_stem_type = stem_type == "3rd-declension" and "3rd-decl" or stem_type
if sgdc.adj then
-- Don't include gender for pluralia tantum because it's mostly
-- indeterminate (certainly when specified using a plural lemma,
-- which will be usually; technically it's partly or completely
-- determinate in certain old-style adjectives that distinguish
-- masculine from feminine/neuter, but this is too rare a case
-- to worry about)
local gendertext = args.thisn == "p" and "plural-only" or gender_to_full[sgdc.g]
insert_if_not(h.adjectival, "yes")
if args.thisn ~= "p" then
insert_if_not(h.gender, gender_to_short[sgdc.g])
end
if sgdc.possadj then
insert_cat(sgdc.decl .. " possessive " .. gendertext .. " accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " adjectival ~")
insert_if_not(h.stemetc, sgdc.decl .. " poss")
insert_if_not(h.stress, stress)
elseif stem_type == "soft-stem" or stem_type == "vowel-stem" then
insert_cat(stem_type .. " " .. gendertext .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
else
insert_cat(stem_type .. " " .. gendertext .. " accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
else
-- NOTE: There are 8 Zaliznyak-style stem types and 3 genders, but
-- we don't create a category for masculine-form 3rd-declension
-- nouns (there is such a noun, путь, but it mostly behaves
-- like a feminine noun), so there are 23.
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form ~")
-- NOTE: Here we are creating categories for the combination of
-- stem, gender and accent. There are 10 accent patterns and 23
-- combinations of stem and gender, which potentially makes for
-- 10*23 = 230 such categories, which is a lot. Not all such
-- categories should actually exist; there were maybe 75 former
-- declension templates, each of which was essentially categorized
-- by the same three variables, but some of which dealt with
-- ancillary issues like irregular plurals; this amounts to 67
-- actual stem/gender/accent categories, although there are more
-- of them in Zaliznyak (FIXME, how many? See generate_cats.py).
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form accent-" .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")) .. " ~")
insert_if_not(h.adjectival, "no")
insert_if_not(h.gender, gender_to_short[sgdc.g])
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
insert_cat("~ with accent pattern " .. (stress:gsub("''", "ʺ"):gsub("'", "ʹ")))
end
local sgsuffix = args.suffixes.nom_sg
if sgsuffix then
assert(#sgsuffix == 1) -- If this ever fails, then implement a loop
sgsuffix = com.remove_accents(sgsuffix[1])
-- If we are plural only or if nom_sg is overridden and has
-- an unusual suffix, then don't create category for sg suffix
if args.thisn == "p" or not override_matches_suffix(args, "nom_sg", n, sgsuffix) or islast and not override_matches_suffix(args, "nom_sg", "", sgsuffix) then
sgsuffix = nil
end
end
local plsuffix = args.suffixes.nom_pl
if plsuffix then
assert(#plsuffix == 1) -- If this ever fails, then implement a loop
plsuffix = com.remove_accents(plsuffix[1])
-- If we are a singulare tantum or if nom_pl is overridden and has
-- an unusual suffix, then don't create category for pl suffix
if args.thisn == "s" or not override_matches_suffix(args, "nom_pl", n, plsuffix) or islast and not override_matches_suffix(args, "nom_pl", "", plsuffix) then
plsuffix = nil
end
end
sgsuffix = sgsuffix and rsub(sgsuffix, "ъ$", "")
plsuffix = plsuffix and rsub(plsuffix, "ъ$", "")
local sgcat = sgsuffix and (resolve_cat(sgdc.singular, sgsuffix) or "ending in " .. (sgsuffix == "" and "a consonant" or (sgdc.suffix and "suffix " or "") .. "-" .. sgsuffix))
local plcat = plsuffix and (resolve_cat(pldc.plural, plsuffix) or "plural -" .. plsuffix)
if sgcat and sgdc.gensg then
for _, cat in ipairs(cat_to_list(sgcat)) do
insert_cat("~ " .. cat)
end
end
if sgcat and plcat and (sgdc.suffix or sgdc.alt_nom_pl or sgdc.irregpl or
is_slash_decl and plsuffix == "-ья") then
for _, scat in ipairs(cat_to_list(sgcat)) do
for _, pcat in ipairs(cat_to_list(plcat)) do
insert_cat("~ " .. scat .. " with " .. pcat)
end
end
end
if args.pl ~= args.stem then
insert_cat("~ with irregular plural stem")
end
if args.reducible and not sgdc.ignore_reduce then
insert_cat("~ with reducible stem")
if args.soft_n then
insert_cat("~ with soft final н in reduced stem")
end
insert_if_not(h.reducible, "yes")
else
insert_if_not(h.reducible, "no")
end
if args.alt_gen_pl then
insert_cat("~ with alternative genitive plural")
end
if sgdc.adj then
insert_cat("adjectival ~")
end
end
local function compute_heading(args)
local headings = {}
local h = args.heading_info
table.insert(headings, args.a == "a" and "anim" or args.a == "i" and
"inan" or "bian")
table.insert(headings, args.nonumber and "uncountable" or
args.n == "s" and "sg-only" or args.n == "p" and "pl-only" or nil)
if #h.gender > 0 then
table.insert(headings, table.concat(h.gender, "/") .. "-form")
end
if #h.stemetc > 0 then
table.insert(headings, table.concat(h.stemetc, "/"))
end
if #h.stress > 0 then
local stresses = {}
for _, stress in ipairs(h.stress) do
table.insert(stresses, (stress:gsub("''", "ʺ"):gsub("'", "ʹ")))
end
table.insert(headings, "accent-" .. table.concat(stresses, "/"))
end
local function handle_bool(boolvals, text, into)
into = into or headings
if m_table.contains(boolvals, "yes") and m_table.contains(boolvals, "no") then
table.insert(into, "[" .. text .. "]")
elseif m_table.contains(boolvals, "yes") then
table.insert(into, text)
end
end
handle_bool(h.adjectival, "adj")
handle_bool(h.reducible, "reduc")
return headings
end
local function compute_overall_heading_categories_and_genders(args)
local hinfo = args.per_word_heading_info
local index = 0
-- First try for non-adjectival, non-indeclinable
for i=1,#hinfo do
if not m_table.contains(hinfo[i].stemetc, "indecl") and not m_table.contains(hinfo[i].adjectival, "yes") then
index = i
break
end
end
if index == 0 then
-- Then just non-indeclinable
for i=1,#hinfo do
if not m_table.contains(hinfo[i].stemetc, "indecl") then
index = i
break
end
end
end
-- Finally, do anything
if index == 0 then
index = 1
end
-- Compute final heading
local headings = args.per_word_headings[index]
local categories = args.per_word_categories[index]
if args.any_irreg then
table.insert(headings, "irreg")
insert_category(categories, "irregular ~", args.pos)
end
for _, case in ipairs(overridable_cases) do
local is_pl = rfind(case, "_pl")
if args.n == "s" and is_pl or args.n == "p" and not is_pl then
-- Don't create singular categories when plural-only or vice-versa
elseif overridable_only_cases_set[case] then
if args.any_overridden[case] then
insert_category(categories, "~ with " .. english_case_description(case), args.pos)
end
elseif args.any_irreg_case[case] then
insert_category(categories, "~ with irregular " .. english_case_description(case), args.pos)
end
end
local heading = args.manual and "" or "(<span style=\"font-size: smaller;\">[[Appendix:Russian nouns#Declension tables|" .. table.concat(headings, " ") .. "]]</span>)"
args.heading = heading
args.categories = categories
args.genders = args.per_word_genders[index]
end
--------------------------------------------------------------------------
-- Main code --
--------------------------------------------------------------------------
-- Used by do_generate_forms().
local function arg1_is_stress(arg1)
if not arg1 then return false end
for _, arg in ipairs(rsplit(arg1, ",")) do
if not rfind(arg, "^[a-f]'?'?$") then
return false
end
end
return true
end
-- Used by do_generate_forms(), handling a word joiner argument
-- of the form 'join:JOINER'.
local function extract_word_joiner(spec)
word_joiner = rmatch(spec, "^join:(.*)$")
assert(word_joiner)
return com.split_russian_tr(word_joiner, "dopair")
end
local function determine_headword_gender(args, sgdc, gender)
-- If gender unspecified, use normal gender of declension, except when
-- adjectival nouns that are pluralia tantum, where the gender is
-- mostly indeterminate (FIXME, not completely with old-style declensions
-- but we don't handle that currently).
if not gender then
if sgdc.adj and args.thisn == "p" then
gender = nil
else
gender = sgdc.g
end
end
-- Determine headword genders
gender = gender and gender ~= "none" and gender .. "-" or ""
local plsuffix = args.n == "p" and "-p" or ""
local hgens
if args.a == "a" then
hgens = {gender .. "an" .. plsuffix}
elseif args.a == "i" then
hgens = {gender .. "in" .. plsuffix}
elseif args.a == "ai" then
hgens = {gender .. "an" .. plsuffix, gender .. "in" .. plsuffix}
else
hgens = {gender .. "in" .. plsuffix, gender .. "an" .. plsuffix}
end
-- Insert into list of genders
for _, hgen in ipairs(hgens) do
insert_if_not(args.genders, hgen)
end
end
function export.do_generate_forms(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SETS and JOINER, where ARG_SETS
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Gather arguments into a list of ARG_SET objects, containing (potentially)
-- elements 1, 2, 3, 4, corresponding to accent pattern, stem, declension
-- type, pl stem and coming from consecutive numbered parameters. Sets of
-- declension parameters are separated by the word "or".
local arg_sets = {}
-- Find maximum-numbered arg, allowing for holes
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Now gather the arguments.
local offset = 0
local arg_set = {}
for i=1,(max_arg + 1) do
local end_arg_set = false
local end_word = false -- FIXME, is this correct?
local word_joiner
if i == max_arg + 1 then
end_arg_set = true
end_word = true
word_joiner = {""}
elseif args[i] == "_" then
end_arg_set = true
end_word = true
word_joiner = {" "}
elseif args[i] == "-" then
end_arg_set = true
end_word = true
word_joiner = {"-"}
elseif args[i] and rfind(args[i], "^join:") then
end_arg_set = true
end_word = true
word_joiner = extract_word_joiner(args[i])
elseif args[i] == "or" then
end_arg_set = true
end
if end_arg_set then
table.insert(arg_sets, arg_set)
arg_set = {}
offset = i
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
else
-- If the first argument isn't stress, that means all arguments
-- have been shifted to the left one. We want to shift them
-- back to the right one, so we change the offset so that we
-- get the same effect of skipping a slot in the arg set.
if i - offset == 1 and not arg1_is_stress(args[i]) then
offset = offset - 1
end
if i - offset > 4 then
error("Too many arguments for argument set: arg " .. i .. " = " .. (args[i] or "(blank)"))
end
arg_set[i - offset] = args[i]
end
end
return generate_forms_1(args, per_word_info)
end
function export.do_generate_forms_multi(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SET and JOINER, where ARG_SET
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Find maximum-numbered arg, allowing for holes (FIXME: Is this needed
-- here? Will there be holes?)
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Gather arguments into a list of ARG_SET objects, containing
-- (potentially) elements 1, 2, 3, corresponding to accent pattern,
-- lemma+declension spec, pl stem, exactly as with do_generate_forms()
-- and {{ru-noun-table}} except that the values come from a single argument
-- of the form ACCENTPATTERN:LEMMADECL:PL where all but LEMMADECL may (and
-- probably will be) omitted and LEMMADECL may be of the following forms:
-- LEMMA (for a noun with empty decl spec),
-- LEMMADECL (for a noun with non-empty decl spec beginning with a
-- *, left paren or semicolon),
-- LEMMA^DECL (for a noun with non-empty decl spec),
-- LEMMA$ (for an indeclinable word)
-- LEMMA+ (for an adjective with auto-detected decl class)
-- LEMMA+DECL (for an adjective with explicit decl class)
-- Sets of parameters for the same word are separated by the word "or".
local arg_sets = {}
local continue_arg_sets = true
for i=1,(max_arg + 1) do
local end_word = false
local word_joiner
local process_arg = false
if i == max_arg + 1 then
end_word = true
word_joiner = {""}
elseif args[i] == "-" then
end_word = true
word_joiner = {"-"}
continue_arg_sets = true
elseif rfind(args[i], "^join:") then
end_word = true
word_joiner = extract_word_joiner(args[i])
continue_arg_sets = true
elseif args[i] == "or" then
continue_arg_sets = true
else
if continue_arg_sets then
continue_arg_sets = false
else
end_word = true
word_joiner = {" "}
end
process_arg = true
end
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
if process_arg then
local vals = rsplit(args[i], ":")
if #vals > 3 then
error("Can't specify more than 3 colon-separated params of param set: " .. args[i])
end
local arg_set = {}
if arg1_is_stress(vals[1]) then
arg_set[1] = vals[1]
arg_set[2] = vals[2]
arg_set[4] = vals[3]
else
arg_set[2] = vals[1]
arg_set[4] = vals[2]
end
-- recognize indeclinable
local indecl_stem = rmatch(arg_set[2], "^(.-)%$$")
if indecl_stem then
arg_set[2] = indecl_stem
arg_set[3] = "$"
else
-- recognize adjective
local adj_stem, adj_type = rmatch(arg_set[2], "^(.*)(%+.*)$")
if adj_stem then
arg_set[2] = adj_stem
arg_set[3] = adj_type
else
-- recognize noun with ^
local noun_stem, noun_type = rmatch(arg_set[2], "^(.*)%^(.*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- recognize noun without ^ but with decl spec
noun_stem, noun_type = rmatch(arg_set[2], "^(..-)([;*(].*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- noun without ^ or decl spec; nothing to do
end
end
end
end
table.insert(arg_sets, arg_set)
end
end
return generate_forms_1(args, per_word_info)
end
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence.
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence,
-- as well as show_z() for template {{ru-decl-noun-z}}, which has a
-- subset of the functionality of the other two.
generate_forms_1 = function(args, per_word_info)
local orig_args
if test_new_ru_noun_module then
orig_args = mw.clone(args)
end
local pagename = args.pagename or mw.loadData("Modül:başlık başı/veri").pagename
local old = args.old
local function verify_animacy_value(val)
if not val then return nil end
if val == "a" or val == "an" or val == "anim" then
return "a"
elseif val == "i" or val == "in" or val == "inan" then
return "i"
elseif val == "b" or val == "bi" or val == "both" or val == "ai" then
return "ai"
elseif val == "ia" then
return "ia"
end
error("Animacy value " .. val .. " should be empty or a/an/anim (animate), i/in/inan (inanimate), b/bi/both/ai (bianimate, listing animate first), or ia (bianimate, listing inanimate first)")
return nil
end
local function verify_number_value(val, allow_none)
if not val then return nil end
local short = usub(val, 1, 1)
if short == "s" or short == "p" or short == "b" or (allow_none and short == "n" or false) then
return short
end
if allow_none then
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), 'b' (both) or 'n' (none)")
else
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), or 'b' (both)")
end
return nil
end
-- Verify and canonicalize animacy, number, prefix, suffix
assert(#per_word_info >= 1)
for i=1,#per_word_info do
args["a" .. i] = verify_animacy_value(args["a" .. i])
args["n" .. i] = verify_number_value(args["n" .. i])
args["prefix" .. i] = com.split_russian_tr(args["prefix" .. i] or "", "dopair")
args["suffix" .. i] = com.split_russian_tr(args["suffix" .. i] or "", "dopair")
end
args.a = verify_animacy_value(args.a) or "i"
-- args.ndef, if set, is the default value for args.n; if unset, it defaults
-- to "both". It is set to "singular" in ru-proper noun+. We store the value
-- of args.n in args.orign before defaulting to args.ndef because we may
-- change it to plural-only later on if it was unspecified (this happens if
-- an individual word's lemma is plural), and to determine whether it was
-- unspecified, we need the original value before defaulting.
args.n = verify_number_value(args.n, "allow none")
-- treat n=none like n=sg but set a flag so "singular" isn't displayed
if args.n == "n" then
args.n = "s"
args.nonumber = true
end
args.ndef = verify_number_value(args.ndef)
args.orign = args.n
args.n = args.n or args.ndef
args.prefix = com.split_russian_tr(args.prefix or "", "dopair")
args.suffix = com.split_russian_tr(args.suffix or "", "dopair")
-- Attach overall prefix to first per-word prefix, similarly for suffix
args.prefix1 = com.concat_paired_russian_tr(args.prefix, args.prefix1)
args["suffix" .. #per_word_info] = com.concat_paired_russian_tr(
args["suffix" .. #per_word_info], args.suffix)
-- Initialize non-word-specific arguments.
--
-- The following is a list of WORD_INFO items, one per word, each of
-- which is a two element list of WORD_FORMS (a table listing the forms for
-- each case) and JOINER (a string, indicating how to join the word with
-- the next one).
args.per_word_info = {}
-- List of HEADING_INFO items, one per word, containing the raw material
-- used to generate the header. Initialized from 'args.heading_info',
-- which is initialized in categorize_and_init_heading().
args.per_word_heading_info = {}
-- List of CATEGORIES items, one per word, containing the categories
-- used to initialize the page categories. Initialized from
-- 'args.categories', which is initialized in categorize_and_init_heading().
args.per_word_categories = {}
-- List of HEADINGS items, one per word, containing the actual words that
-- go into the header if we were to use that word to construct the header.
-- Comes from compute_heading(). We use this to generate the actual header
-- string, which goes into 'args.heading' at the end (done in
-- compute_overall_heading_categories_and_genders()).
args.per_word_headings = {}
-- List of GENDERS items, one per word, containing the headword genders
-- for each word (where "headword gender" is in the format used in
-- headwords and actually includes animacy and number as well, e.g.
-- 'm-in' or 'f-an-p'). We end up selecting one such item and putting
-- it into 'args.genders' at the end
-- (in compute_overall_heading_categories_and_genders()).
args.per_word_genders = {}
args.any_overridden = {}
args.any_non_nil = {}
args.any_irreg = false
args.any_irreg_case = {}
local function insert_cat(cat)
insert_category(args.categories, cat, args.pos)
end
args.internal_notes = {}
local decl_sufs = old and declensions_old or declensions
local decl_cats = old and declensions_old_cat or declensions_cat
local intable = old and internal_notes_table_old or internal_notes_table
-- Default lemma defaults to previous lemma, first one to page name.
-- FIXME: With multiple words, we should probably split the page name
-- on spaces and default each word in turn
local default_lemma
local all_stresses_seen
-- Made into a function to avoid having to indent a lot of code.
-- Process a single arg set of a single word. This inserts the forms
-- for the word into args.forms and sets categories and tracking pages.
local function do_arg_set(arg_set, n, islast)
local stress_arg = arg_set[1]
local decl = arg_set[3] or ""
local pl, pltr
if arg_set[4] then
pl, pltr = com.split_russian_tr(arg_set[4])
end
-- Extract special markers from declension class.
if decl == "manual" then
decl = "$"
args.manual = true
if #per_word_info > 1 or #per_word_info[1][1] > 1 then
error("Can't specify multiple words or argument sets when manual")
end
if pl then
error("Can't specify optional stem parameters when manual")
end
end
decl, args.jo_special = rsubb(decl, "([^/%a])ё$", "%1")
if not args.jo_special then
decl, args.jo_special = rsubb(decl, "([^/%a])ё([^/%a])", "%1%2")
end
decl, args.want_sc1 = rsubb(decl, "%(1%)", "")
decl, args.alt_gen_pl = rsubb(decl, "%(2%)", "")
decl, args.reducible = rsubb(decl, "%*", "")
decl, args.soft_n = rsubb(decl, "%(нь%)", "")
decl = rsub(decl, ";", "")
-- Get the lemma.
local lemma = args.manual and "-" or arg_set[2] or default_lemma
if not lemma then
error("Lemma in first argument set must be specified")
end
default_lemma = lemma
local lemmatr
lemma, lemmatr = com.split_russian_tr(lemma)
-- If we're conjugating a suffix, insert a pseudoconsonant at the beginning
-- of all forms, so they get conjugated as if ending in a consonant.
-- We remove the pseudoconsonant later.
local is_suffix = lemma ~= "-" and rfind(lemma, "^%-")
args.any_suffix = args.any_suffix or is_suffix
local asif_prefix = args["asif_prefix" .. n] or args.asif_prefix or is_suffix and PSEUDOCONS
if asif_prefix then
lemma = rsub(lemma, "^%-", "-" .. asif_prefix)
if lemmatr then
lemmatr = rsub(lemmatr, "^%-", "-" .. com.translit(asif_prefix))
end
end
args.thisa = args["a" .. n] or args.a
args.thisn = args["n" .. n] or args.n
-- Check for explicit allow-unaccented indication.
local allow_unaccented
lemma, allow_unaccented = rsubb(lemma, "^%*", "")
args.allow_unaccented = args.allow_unaccented or allow_unaccented
if args.allow_unaccented then
track("allow-unaccented")
end
args.orig_lemma = lemma
lemma = m_links.remove_links(lemma)
args.lemma_no_links = lemma
args.lemmatr = lemmatr
if args.lemma then
-- Explicit lemma given.
args.explicit_lemma, args.explicit_lemmatr = com.split_russian_tr(args.lemma)
end
-- Treat suffixes without an accent, and suffixes with an accent on the
-- initial hyphen, as if they were preceded with a *, which overrides
-- all the logic that normally (a) normalizes the accent, and (b)
-- complains about multisyllabic words without an accent. Don't do this
-- if lemma is just -, which is used specially in manual declension
-- tables (e.g. сто, три).
if lemma ~= "-" and (rfind(lemma, "^%-" .. AC) or (com.is_unstressed(lemma) and rfind(lemma, "^%-"))) then
args.allow_unaccented = true
end
-- Convert lemma and decl arg into stem and canonicalized decl.
-- This will autodetect the declension from the lemma if an explicit
-- decl isn't given.
local stem, tr, gender, was_accented, was_plural, was_autodetected
if rfind(decl, "^%+") then
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
detect_adj_type(lemma, lemmatr, decl, old)
else
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
determine_decl(lemma, lemmatr, decl, args)
end
if was_plural then
args.n = args.orign or "p"
args.thisn = args["n" .. n] or args.n
elseif decl ~= "$" then
args.thisn = args.thisn or "b"
end
args.explicit_gender = gender
-- If allow-unaccented not given, maybe check for missing accents.
if not args.allow_unaccented and not stress_arg and was_autodetected and com.needs_accents(lemma) then
-- If user gave the full word and expects us to determine the
-- declension and stress, the word should have an accent on the
-- stem or ending. We have a separate check farther below for
-- an accent on a multisyllabic stem, after stripping off any
-- ending; but this way we get an error if the user e.g. writes
-- "гора" without an accent rather than assuming it's stem
-- stressed, as would otherwise happen.
error("Lemma must have an accent in it: " .. lemma)
end
-- If stress not given, auto-determine; else validate/canonicalize
-- stress arg, override in certain cases and convert to list.
if not stress_arg then
stress_arg = {detect_stress_pattern(stem, decl, decl_cats, args.reducible, was_plural, was_accented)}
else
stress_arg = rsplit(stress_arg, ",")
for i=1,#stress_arg do
local stress = stress_arg[i]
stress = override_stress_pattern(decl, stress)
if not stress_patterns[stress] then
error("Unrecognized accent pattern " .. stress)
end
stress_arg[i] = stress
end
end
-- parse slash decl to list
local sub_decls
if rfind(decl, "/") then
track("mixed-decl")
insert_cat("~ with mixed declension")
local indiv_decls = rsplit(decl, "/")
-- Should have been caught in canonicalize_decl()
assert(#indiv_decls == 2)
sub_decls = {{indiv_decls[1], "sg"}, {indiv_decls[2], "pl"}}
else
sub_decls = {{decl}}
end
-- Get singular declension and corresponding category.
local sgdecl = sub_decls[1][1]
local sgdc = decl_cats[sgdecl]
assert(sgdc)
-- Compute headword gender(s). We base it off the singular declension
-- if we have a slash declension -- it's the best we can do.
determine_headword_gender(args, sgdc, gender)
local original_stem, original_tr = stem, tr
local original_pl, original_pltr = pl, pltr
-- Loop over accent patterns in case more than one given.
for _, stress in ipairs(stress_arg) do
args.suffixes = {}
stem, tr = original_stem, original_tr
local bare, baretr
local stem_for_bare, tr_for_bare
pl, pltr = original_pl, original_pltr
insert_if_not(all_stresses_seen, stress)
local stem_was_unstressed = com.is_unstressed(stem)
-- If special case ;ё was given and stem is unstressed,
-- add ё to the stem now; but don't let this interfere with
-- restressing, to handle cases like железа́ with gen pl желёз
-- but nom pl же́лезы.
if stem_was_unstressed and args.jo_special then
-- Beware, Cyrillic еЕ in first rsub, Latin eE in second
local new_stem = rsub(stem, "([еЕ])([^еЕ]*)$",
function(e, rest)
return (e == "Е" and "Ё" or "ё") .. rest
end
)
if stem == new_stem then
error("No е in stem to replace with ё")
end
stem = new_stem
if tr then
local subbed
-- e after j -> o, e not after j -> jo; don't just convert e -> jo
-- and then map jjo -> jo because we want to preserve double j
tr, subbed = rsubb(tr, "([jJ])([eE])([^eE]*)$",
function(j, e, rest)
return j .. (e == "E" and "O" or "o") .. AC .. rest
end
)
if not subbed then
tr = rsub(tr, "([eE])([^eE]*)$",
function(e, rest)
return (e == "E" and "Jo" or "jo") .. AC .. rest
end
)
end
tr = com.j_correction(tr)
end
-- This is used to handle железа́ with gen pl желёз and nom pl
-- же́лезы. We have two stressed stems, one for the gen pl and
-- one for the remaining pl cases, and the variable 'stem' can
-- handle only one, so we put the second (gen pl) stem in
-- stem_for_bare, which goes into the value of 'bare' (used
-- only for gen pl here).
stem_for_bare, tr_for_bare = stem, tr
end
-- Maybe add stress to the stem, depending on whether the
-- stem was unstressed and the stress pattern. Stem pattern f
-- and variants call for initial stress (голова́ -> го́ловы);
-- stem pattern d and variants call for stem-final stress
-- (сапожо́к -> сапо́жки). Stem patterns b and b' apparently
-- call for stem-final stress as well but it's unlikely to
-- make much of a difference (pattern b' only occurs in 3rd-decl
-- feminines, which should already have stress in the stem,
-- and the only place pattern b gets stem stress is in bare
-- forms, i.e. nom sg and/or gen pl depending on the decl type,
-- and nom sg stress should already be in the lemma while
-- gen pl stress is handled by a different ending-stressing
-- mechanism in attach_unstressed(); however, the user is
-- free to leave a masc or 3rd-decl fem lemma completely
-- unstressed with pattern b, and then the stem-final stress
-- *will* make a difference).
local function restress_stem(stem, tr, stress, stem_unstressed)
-- If the user has indicated they purposely are leaving the
-- word unstressed by putting a * at the beginning of the main
-- stem, leave it unstressed. This might indicate lack of
-- knowledge of the stress or a truly unaccented word
-- (e.g. an unaccented suffix).
if args.allow_unaccented then
return stem, tr
end
if tr and com.is_unstressed(stem) ~= com.is_unstressed(tr) then
error("Stem " .. stem .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(stem) then
stem, tr = com.make_ending_stressed(stem, tr)
-- For those patterns that are ending-stressed in the singular
-- nominative (and hence are likely to be expressed without an
-- accent on the stem) it's safe to put a particular accent on
-- the stem depending on the stress type. Otherwise, give an
-- error if no accent.
elseif stem_unstressed then
if rfind(stress, "^f") then
stem, tr = com.make_beginning_stressed(stem, tr)
elseif (rfind(stress, "^[bd]") or
args.thisn == "p" and ending_stressed_pl_patterns[stress]) then
stem, tr = com.make_ending_stressed(stem, tr)
elseif com.needs_accents(stem) then
error("Stem " .. stem .. " requires an accent")
end
end
return stem, tr
end
stem, tr = restress_stem(stem, tr, stress, stem_was_unstressed)
-- Leave pl unaccented if user wants this; see restress_stem().
if pl and not args.allow_unaccented then
if pltr and com.is_unstressed(pl) ~= com.is_unstressed(pltr) then
error("Plural stem " .. pl .. " and translit " .. pltr .. " must have same accent pattern")
end
if com.is_monosyllabic(pl) then
pl, pltr = com.make_ending_stressed(pl, pltr)
end
-- I think this is safe.
if com.needs_accents(pl) then
if ending_stressed_pl_patterns[stress] then
pl, pltr = com.make_ending_stressed(pl, pltr)
elseif not args.allow_unaccented then
error("Plural stem " .. pl .. " requires an accent")
end
end
end
local resolved_bare, resolved_baretr
-- Handle (de)reducibles
-- FIXME! We are dereducing based on the singular declension.
-- In a slash declension things can get weird and we don't
-- handle that. We are also computing the bare value from the
-- singular stem, and again things can get weird with a plural
-- stem. Note that we don't compute a bare value unless we have
-- to (either (de)reducible or stress pattern f/f'/f'' combined
-- with ё special case); the remaining times we generate the bare
-- value directly from the plural stem.
if args.reducible and not sgdc.ignore_reduce then
-- Zaliznyak treats all nouns in -ье and -ья as being
-- reducible. We handle this automatically and don't require
-- the user to specify this, but ignore it if so for
-- compatibility.
if is_reducible(sgdc) then
-- If we derived the stem from a nom pl form, then
-- it's already reduced, and we need to dereduce it to
-- get a bare form; otherwise the stem comes from the
-- nom sg and we need to reduce it to get the real stem.
if was_plural then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
resolved_bare, resolved_baretr = stem, tr
stem, tr = export.reduce_nom_sg_stem(stem, tr,
sgdecl, args.soft_n, "error")
-- Stem will be unstressed if stress was on elided
-- vowel; restress stem the way we did above. (This is
-- needed in at least one word, сапожо́к 3*d(2), with
-- plural stem probably сапо́жк- and gen pl probably
-- сапо́жек.)
stem, tr = restress_stem(stem, tr, stress,
com.is_unstressed(stem))
if stress ~= "a" and stress ~= "b" and args.alt_gen_pl and not pl then
-- Nouns like рожо́к, глазо́к of type 3*d(2) have
-- gen pl's ро́жек, гла́зок; to handle this,
-- dereduce the reduced stem and store in a
-- special place.
args.gen_pl_bare, args.gen_pl_baretr =
export.dereduce_nom_sg_stem(stem, tr,
sgdc, stress, old, "error")
end
end
elseif is_dereducible(sgdc) then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
error("Declension class " .. sgdecl .. " not (de)reducible")
end
elseif stem_for_bare and stem ~= stem_for_bare then
resolved_bare, resolved_baretr =
add_bare_suffix(stem_for_bare, tr_for_bare, old, sgdc, false)
end
-- Leave unaccented if user wants this; see restress_stem().
-- FIXME, we no longer allow the user to specify the bare value
-- so it's unclear if this is needed any more.
if resolved_bare and not args.allow_unaccented then
if resolved_baretr and com.is_unstressed(resolved_bare) ~= com.is_unstressed(resolved_baretr) then
error("Resolved bare stem " .. resolved_bare .. " and translit " .. resolved_baretr .. " must have same accent pattern")
end
if com.is_monosyllabic(resolved_bare) then
resolved_bare, resolved_baretr =
com.make_ending_stressed(resolved_bare, resolved_baretr)
else
if com.needs_accents(resolved_bare) then
error("Resolved bare stem " .. resolved_bare .. " requires an accent")
end
end
end
args.stem, args.stemtr = stem, tr
args.bare, args.baretr = resolved_bare, resolved_baretr
args.ustem, args.ustemtr = com.make_unstressed_once(stem, tr)
if pl then
args.pl, args.pltr = pl, pltr
else
args.pl, args.pltr = stem, tr
end
args.upl, args.upltr = com.make_unstressed_once(args.pl, args.pltr)
-- Special hack for любо́вь and other reducible 3rd-fem nouns,
-- which have the full stem in the ins sg
args.ins_sg_stem = sgdecl == "ь-f" and args.reducible and resolved_bare
args.ins_sg_tr = sgdecl == "ь-f" and args.reducible and resolved_baretr
-- Loop over declension classes (we may have two of them, one for
-- singular and one for plural, in the case of a mixed declension
-- class of the form SGDECL/PLDECL).
for _,decl_spec in ipairs(sub_decls) do
local orig_decl = decl_spec[1]
local number = decl_spec[2]
local real_decl =
determine_stress_variant(orig_decl, stress)
real_decl = determine_stem_variant(real_decl,
number == "pl" and args.pl or args.stem)
-- sanity checking; errors should have been caught in
-- canonicalize_decl()
assert(decl_cats[real_decl], "real_decl " .. real_decl .. " nonexistent")
assert(decl_sufs[real_decl], "real_decl " .. real_decl .. " nonexistent")
tracking_code(stress, orig_decl, real_decl, args, n, islast)
do_stress_pattern(stress, args, real_decl, number, n, islast)
-- handle internal notes
local internal_note = intable[real_decl]
if internal_note then
insert_if_not(args.internal_notes, internal_note)
end
end
categorize_and_init_heading(stress, decl, args, n, islast)
end
end
local n = 0
for _, word_info in ipairs(per_word_info) do
n = n + 1
local islast = n == #per_word_info
local arg_sets, joiner = word_info[1], word_info[2]
args.forms = {}
args.heading_info = {animacy={}, number={}, gender={}, stress={},
stemetc={}, adjectival={}, reducible={}}
args.categories = {}
args.genders = {}
args.this_any_non_nil = {}
args.any_suffix = false
if #arg_sets > 1 then
track("multiple-arg-sets")
insert_cat("~ with multiple argument sets")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
default_lemma = pagename
all_stresses_seen = {}
-- Loop over all arg sets.
for _, arg_set in ipairs(arg_sets) do
do_arg_set(arg_set, n, islast)
end
if #all_stresses_seen > 1 then
track("multiple-accent-patterns")
insert_cat("~ with multiple accent patterns")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
table.insert(args.per_word_heading_info, args.heading_info)
table.insert(args.per_word_categories, args.categories)
local headings = compute_heading(args)
table.insert(args.per_word_headings, headings)
table.insert(args.per_word_genders, args.genders)
handle_forms_and_overrides(args, n, islast)
if args.any_suffix then
-- If we're conjugating a suffix, remove the pseudoconsonant or asif_prefix
-- that we previously inserted at the beginning.
local asif_prefix = args["asif_prefix" .. n] or args.asif_prefix or PSEUDOCONS
local asif_prefix_tr = com.translit(asif_prefix)
for _, case in ipairs(all_cases) do
if args.forms[case] then
local newforms = {}
for _, form in ipairs(args.forms[case]) do
local formru = form[1]
local formtr = form[2]
formru = rsub(formru, "^%-" .. asif_prefix, "-")
if formtr then
formtr = rsub(formtr, "^%-" .. asif_prefix_tr, "-")
end
if formru == "-" then
-- if no ending, insert "(no suffix)".
table.insert(newforms, {"(no suffix)"})
else
table.insert(newforms, {formru, formtr})
end
end
args.forms[case] = newforms
end
end
end
table.insert(args.per_word_info, {args.forms, joiner})
end
handle_overall_forms_and_overrides(args)
compute_overall_heading_categories_and_genders(args)
for _, case in ipairs(all_cases) do
if args[case] then
for _, form in ipairs(args[case]) do
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
ruentry = m_links.remove_links(ruentry)
if rfind(ulower(ruentry), latin_text_class) then
--error("Found Latin text " .. ruentry .. " in case " .. case)
track("latin-text")
track("latin-text/" .. case)
end
end
end
end
-- Test code to compare existing module to new one.
if test_new_ru_noun_module then
local m_new_ru_noun = require("Modül:User:Benwing2/ru-noun")
local newargs = m_new_ru_noun.do_generate_forms(orig_args, old)
local difdecl = false
for _, case in ipairs(all_cases) do
local arg = args[case]
local newarg = newargs[case]
local is_pl = rfind(case, "_pl")
if args.thisn == "s" and is_pl or args.thisn == "p" and not is_pl then
-- Don't need to check cases that won't be displayed.
elseif not m_table.deepEquals(arg, newarg) then
local monosyl_accent_diff = false
-- Differences only in monosyllabic accents. Enable if we
-- change the algorithm for these.
--if arg and newarg and #arg == 1 and #newarg == 1 then
-- local ru1, tr1 = arg[1][1], arg[1][2]
-- local ru2, tr2 = newarg[1][1], newarg[1][2]
-- if com.is_monosyllabic(ru1) and com.is_monosyllabic(ru2) then
-- ru1, tr1 = com.remove_accents(ru1, tr1)
-- ru2, tr2 = com.remove_accents(ru2, tr2)
-- if ru1 == ru2 and tr1 == tr2 then
-- monosyl_accent_diff = true
-- end
-- end
--end
if monosyl_accent_diff then
track("monosyl-accent-diff")
difdecl = true
else
-- Uncomment this to display the particular case and
-- differing forms.
--error(case .. " " .. (arg and com.concat_forms(arg) or "nil") .. " || " .. (newarg and com.concat_forms(newarg) or "nil"))
track("different-decl")
difdecl = true
end
break
end
end
if not difdecl then track("same-decl") end
end
return args
end
-- Implementation of main entry point
local function do_show(frame, old)
local args = clone_args(frame)
local args = export.do_generate_forms(args, old)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The main entry point for modern declension tables.
function export.show(frame)
return do_show(frame, false)
end
-- The main entry point for old declension tables.
function export.show_old(frame)
return do_show(frame, true)
end
-- Implementation of new entry point, esp. for multiple words
local function do_show_multi(frame)
local args = clone_args(frame)
local args = export.do_generate_forms_multi(args)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The new entry point, esp. for multiple words (but works fine for
-- single words).
function export.show_multi(frame)
return do_show_multi(frame)
end
local function get_form(forms, preserve_links, raw)
local canon_forms = {}
for _, form in ipairs(forms) do
if raw then
local ru, tr = form[1], form[2]
ru = rsub(ru, "|", "<!>")
if tr then
tr = rsub(tr, "|", "<!>")
end
insert_if_not(canon_forms, {ru, tr})
else
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
-- Skip hypothetical forms (but include in the raw versions)
if not rfind(runotes, HYPMARKER) then
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.separate_notes(tr)
end
if not preserve_links then
ruentry = m_links.remove_links(ruentry)
end
ruentry = rsub(ruentry, "|", "<!>")
if trentry then
trentry = rsub(trentry, "|", "<!>")
end
insert_if_not(canon_forms, {ruentry, trentry})
end
end
end
return com.concat_forms(canon_forms)
end
local function case_will_be_displayed(args, case)
local ispl = rfind(case, "_pl")
local caseok = true
if args.n == "p" then
caseok = ispl
elseif args.n == "s" then
caseok = not ispl
end
for _, override_case in ipairs(overridable_only_cases) do
if case == override_case and not args.any_overridden[override_case] then
caseok = false
break
end
end
if args.a == "a" or args.a == "i" then
if rfind(case, "_[ai]n") then
caseok = false
end
else -- bianimate
-- don't include inanimate/animate variants if combined variant exists
-- (typically because inanimate/animate variants are the same);
-- FIXME: This could conceivably be different from how the display
-- code works, which just checks that the inanimate/animate variants
-- are the same when deciding whether to display them, in particular
-- if there is an override. Here we are following the algorithm of
-- handle_overall_forms_and_overrides().
if (case == "acc_sg_in" or case == "acc_sg_an") and args.acc_sg or
(case == "acc_pl_in" or case == "acc_pl_an") and args.acc_pl then
caseok = false
end
end
if not args[case] then
caseok = false
end
return caseok
end
local function concat_case_args(args, do_all, raw)
local ins_text = {}
for _, case in ipairs(do_all and all_cases or overridable_cases) do
if case_will_be_displayed(args, case) then
local forms = get_form(args[case], rfind(case, "_linked"), raw)
if forms ~= "" then
table.insert(ins_text, case .. (raw and "_raw" or "") .. "=" ..
forms)
end
end
end
return table.concat(ins_text, "|")
end
-- The entry point for 'ru-noun-forms' to generate all noun forms.
-- This returns a single string, with | separating arguments and named
-- arguments of the form NAME=VALUE.
function export.generate_forms(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
return concat_case_args(args)
end
-- The entry point to generate multiple sets of noun forms. This is a hack
-- to speed up calling from a bot, where we often want to compare old and new
-- argument results to make sure they're the same. Each set of arguments is
-- jammed together into a single argument with individual values separated by
-- <!>; named arguments are of the form NAME<->VALUE. The return value for
-- each set of arguments is as in export.generate_forms(), and the return
-- values are concatenated with <!> separating them. NOTE: This will fail if
-- the exact sequences <!> or <-> happen to occur in values (which is unlikely,
-- esp. as we don't even use the characters <, ! or > for anything) and aren't
-- HTML-escaped.
function export.generate_multi_forms(frame)
local retvals = {}
for _, argset in ipairs(frame.args) do
local args = {}
local i = 0
local argvals = rsplit(argset, "<!>")
for _, argval in ipairs(argvals) do
local split_arg = rsplit(argval, "<%->")
if #split_arg == 1 then
i = i + 1
args[i] = ine(split_arg[1])
else
assert(#split_arg == 2)
args[split_arg[1]] = ine(split_arg[2])
end
end
args = export.do_generate_forms(args, false)
table.insert(retvals, concat_case_args(args))
end
return table.concat(retvals, "<!>")
end
-- The entry point for 'ru-noun-form' to generate a particular noun form.
function export.generate_form(frame)
local args = clone_args(frame)
if not args.form then
error("Must specify desired form using form=")
end
local form = args.form
if not m_table.contains(all_cases, form) then
error("Unrecognized form " .. form)
end
local args = export.do_generate_forms(args, false)
if not args[form] then
return ""
else
return get_form(args[form])
end
end
-- The entry point for generating arguments of various sorts, including
-- the case forms, gender, number and animacy.
function export.generate_args(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
local retargs = {}
table.insert(retargs, concat_case_args(args, "doall"))
table.insert(retargs, concat_case_args(args, "doall", "raw"))
table.insert(retargs, "g=" .. table.concat(args.genders, ","))
-- The following is correct even with ndef because if ndef is
-- set we will set it in args.n.
table.insert(retargs, "n=" .. (args.n or "b"))
table.insert(retargs, "a=" .. (args.a or "i"))
return table.concat(retargs, "|")
end
-- The entry point for compatibility with {{ru-decl-noun-z}}.
function export.show_z(frame)
local args = clone_args(frame)
local stem = args[1]
local stress = args[2]
local specific = args[4] or ""
-- Parse gender/animacy spec
local gender, anim = rmatch(args[3], "^([mfn])-([a-z]+)")
if not gender then
error("Unrecognized gender/anim spec " .. args[3])
end
if anim ~= "an" and anim ~= "in" then anim = "both" end
args.a = anim
-- Handle specific
specific = rsub(specific, "ё", ";ё")
-- Compute decl; special case for семьянин (perhaps not necessary)
local decl = com.make_unstressed_once(stem) == "семьянин" and "#" .. specific or gender .. specific
-- Handle overrides
args.pre_sg = args.prp_sg
args.pre_pl = args.prp_pl
args.notes = args.note
if args.par then
args.par = "+"
end
if args.loc then
if args.loc == "в" then
args.loc = "в +"
elseif args.loc == "на" then
args.loc = "на +"
else
args.loc = "в +,на +"
end
end
local arg_set = {}
table.insert(arg_set, stress)
table.insert(arg_set, stem)
table.insert(arg_set, decl)
local per_word_info = {{{arg_set}, ""}}
return generate_forms_1(args, per_word_info)
end
--------------------------------------------------------------------------
-- Autodetection and lemma munging --
--------------------------------------------------------------------------
-- Attempt to detect the type of the lemma based on its ending, separating
-- off the stem and the ending. GENDER must be present with -ь and plural
-- stems, and is otherwise ignored. Return up to three values: The stem
-- (lemma minus ending), the singular lemma ending, and if the lemma was
-- plural, the plural lemma ending. If the lemma was singular, the singular
-- lemma ending will contain any user-given accents; likewise, if the
-- lemma was plural, the plural ending will contain such accents.
-- VARIANT comes from the declension spec and controls certain declension
-- variants.
local function detect_lemma_type(lemma, tr, gender, args, variant)
local base, ending = rmatch(lemma, "^(.*)([еЕ]" .. AC .. ")$") -- accented
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([еЕ])$") -- unaccented
if base then
if variant == "-ище" and not rfind(lemma, "[щЩ][еЕ]$") then
error("With declension variant -ище, lemma should end in -ще: " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), variant == "-ище" and "(ищ)е-и" or "о"
end
if variant == "-ишко" then
base, ending = rmatch(lemma, "^(.*[шШ][кК])([оО])$") -- unaccented
if not base then
error("With declension variant -ишко, lemma should end in -шко: " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), "(ишк)о-и"
end
if variant == "-ин" then
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?[нН][ъЪ]?)$") -- maybe accented
if not base then
error("With declension variant -ин, lemma should end in -ин(ъ): " .. lemma)
end
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
-- Now autodetect -ин; only animate and in -анин/-янин
base, ending = rmatch(lemma, "^(.*[аАяЯ][" .. AC .. GR .. "]?[нН])([иИ][" .. AC .. GR .. "]?[нН][ъЪ]?)$")
-- Need to check the animacy to avoid nouns like маиганин, цианин,
-- меланин, соланин, etc.
if base and args.thisa == "a" then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]" .. AC .. "?[нН][оО][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]" .. AC .. "[нН][оО][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]" .. AC .. "?[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]" .. AC .. "[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([мМ][яЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
--recognize plural endings
if gender == "n" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ][" .. AC .. GR .. "]?)$")
if base then
-- Don't do this; о/-ья is too rare
-- error("Ambiguous plural lemma " .. lemma .. " in -ья, singular could be -о or -ье/-ьё; specify the singular")
return base, com.strip_tr_ending(tr, ending), "ье", ending
end
base, ending = rmatch(lemma, "^(.*)([аяАЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), rfind(ending, "[аА]") and "о" or "е", ending
end
base, ending = rmatch(lemma, "^(.*)([ыиЫИ][" .. AC .. GR .. "]?)$")
if base then
if rfind(ending, "[ыЫ]") or rfind(base, "[" .. com.sib .. com.velar .. "]$") then
return base, com.strip_tr_ending(tr, ending), "о-и", ending
else
-- FIXME, should we return a slash declension?
error("No neuter declension е-и available; use a slash declension")
end
end
end
if gender == "f" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), "ья", ending
end
end
-- Recognize masculines with irregular plurals, but only if the user
-- either explicitly specified that this noun is plural (n=p) or
-- specifically requested the irregular plural. This is necessary
-- because some masculine nouns have feminine endings, which look
-- like irregular plurals.
if gender == "m" then
if args.thisn == "p" or variant == "-ья" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), (args.old and "ъ-ья" or "-ья"), ending
end
end
if args.thisn == "p" or args.want_sc1 then
base, ending = rmatch(lemma, "^(.*)([аА][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), (args.old and "ъ-а" or "-а"), ending
end
base, ending = rmatch(lemma, "^(.*)([яЯ][" .. AC .. GR .. "]?)$")
if base then
if rfind(base, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
return base, com.strip_tr_ending(tr, ending), "й-я", ending
else
return base, com.strip_tr_ending(tr, ending), "ь-я", ending
end
end
end
end
if gender == "m" or gender == "f" then
base, ending = rmatch(lemma, "^(.*[" .. com.sib .. com.velar .. "])([иИ][" .. AC .. GR .. "]?)$")
if not base then
base, ending = rmatch(lemma, "^(.*)([ыЫ][" .. AC .. GR .. "]?)$")
end
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and (args.old and "ъ" or "") or "а", ending
end
base, ending = rmatch(lemma, "^(.*[" .. com.vowel .. "й][" .. AC .. GR .. "]?)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and "й" or "я", ending
end
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), gender == "m" and "ь-m" or "я", ending
end
end
if gender == "3f" then
base, ending = rmatch(lemma, "^(.*)([иИ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), "ь-f", ending
end
end
-- end of recognize-plurals code
base, ending = rmatch(lemma, "^(.*)([ьЬ][яеёЯЕЁ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([йаяеоёъЙАЯЕОЁЪ][" .. AC .. GR .. "]?)$")
if base then
return base, com.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ьЬ])$")
if base then
if gender == "m" or gender == "f" then
return base, com.strip_tr_ending(tr, ending), "ь-" .. gender
elseif gender == "3f" then
return base, com.strip_tr_ending(tr, ending), "ь-f"
else
error("Need to specify gender m or f with lemma in -ь: ".. lemma)
end
end
if rfind(lemma, "[ыиЫИ][" .. AC .. GR .. "]?$") then
error("If this is a plural lemma, gender must be specified: " .. lemma)
elseif rfind(lemma, "[иИіІуУыЫѣѢэЭюЮѵѴ]" .. DIA .. "?[" .. AC .. GR .. "]?$") then
error("Don't know how to decline lemma ending in this type of vowel: " .. lemma)
end
return lemma, tr, ""
end
local plural_variant_detection_map = {
[""] = {["-ья"]="-ья"},
["ъ"] = {["-ья"]="ъ-ья"},
}
local special_case_1_to_plural_variant = {
[""] = "-а",
["ъ"] = "ъ-а",
["й"] = "й-я",
["ь-m"] = "ь-я",
["о"] = "о-и",
-- these last two are here to avoid getting errors in that checks for
-- compatibility with special case 1; the basic variants of these decls
-- don't actually exist
["(ишк)о"] = "(ишк)о-и",
["(ищ)е"] = "(ищ)е-и",
}
local function map_decl(decl, fun)
if rfind(decl, "/") then
local split_decl = rsplit(decl, "/")
if #split_decl ~= 2 then
error("Mixed declensional class " .. decl
.. "needs exactly two classes, singular and plural")
end
return fun(split_decl[1]) .. "/" .. fun(split_decl[2])
else
return fun(decl)
end
end
-- Canonicalize decl class into non-accented and alias-resolved form;
-- but note that some canonical decl class names with an accent in them
-- (e.g. е́, not the same as е, whose accented version is ё; and various
-- adjective declensions).
local function canonicalize_decl(decl, old)
local function do_canon(decl)
-- remove accents, but not from е́ (for adj decls, accent matters
-- as well but we handle that by mapping the accent to a stress pattern
-- and then to the accented version in determine_stress_variant())
if decl ~= "е́" then
decl = com.remove_accents(decl)
end
local decl_aliases = old and declensions_old_aliases or declensions_aliases
local decl_cats = old and declensions_old_cat or declensions_cat
if decl_aliases[decl] then
-- If we find an alias, map it.
decl = decl_aliases[decl]
elseif not decl_cats[decl] then
error("Unrecognized declension class " .. decl)
end
return decl
end
return map_decl(decl, do_canon)
end
-- Attempt to determine the actual declension (including plural variants)
-- based on a combination of the declension the user specified, what can be
-- detected from the lemma, and special case (1), if given in the declension.
-- DECL is the value the user passed for the declension field, after
-- extraneous annotations (special cases (1) and (2), * for reducible,
-- ё for ё/ё alternation and a ; that may precede ё) have been stripped off.
-- What's left is one of the following:
--
-- 1. Blank, meaning to autodetect the declension from the lemma
-- 2. A hyphen followed by a declension variant (-ья, -ин, -ишко, -ище; see
-- long comment at top of file)
-- 3. A gender (m, f, n, 3f)
-- 4. A gender plus declension variant (e.g. f-ья)
-- 5. An actual declension, possibly including a plural variant (e.g. о-и) or
-- a slash declension (e.g. я/-ья, used for the noun дядя).
--
-- Return seven args: stem (lemma minus ending), translit, canonicalized
-- declension, explicitly specified gender if any (m, f, n or nil), whether
-- the specified declension or detected ending was accented, whether the
-- detected ending was pl, and whether the declension was autodetected
-- (corresponds to cases where a full word with ending attached is required
-- in the lemma field). "Canonicalized" means after autodetection, with
-- accents removed, with any aliases mapped to their canonical versions
-- and with any requested declension variants applied. The result is either a
-- declension that will have a categorization entry (in declensions_cat[] or
-- declensions_old_cat[]) or a slash declension where each part similarly has
-- a categorization entry.
--
-- Note that gender is never required when an explicit declension is given,
-- and in connection with stem autodetection is required only when the lemma
-- either ends in -ь or is plural.
determine_decl = function(lemma, tr, decl, args)
-- Assume we're passed a value for DECL of types 1-4 above, and
-- fetch gender and requested declension variant.
local stem
local want_ya_plural, orig_pl_ending, variant
local was_autodetected
local gender = rmatch(decl, "^(3?[mfn]?)$")
if not gender then
gender, variant = rmatch(decl, "^(3?[mfn]?)(%-[^%-]+)$")
-- But be careful with explicit declensions like -а that look like
-- variants without gender (FIXME, eventually we should maybe do
-- something about the potential ambiguity).
if gender == "" and not m_table.contains({"-ья", "-ин", "-ишко", "-ище"}, variant) then
gender, variant = nil, nil
end
end
-- If DECL is of type 1-4, handle declension variants and detect
-- the actual declension from the lemma.
if gender then
-- Check for declension variants
if variant then
if variant == "-ья" then
want_ya_plural = "-ья"
else
-- Sanity-check remaining declension variants, which need
-- specific values of animacy, gender and special-case (1)
local sc1_needed
local animate_needed
if variant == "-ишко" then
animate_needed = false
sc1_needed = true
elseif variant == "-ище" then
animate_needed = true
sc1_needed = true
elseif variant == "-ин" then
animate_needed = true
sc1_needed = false
else
-- WARNING: If adding another variant, you need to also
-- add to the list farther above.
error("Unrecognized declension variant " .. variant .. ", should be -ья, -ин, -ишко or -ище")
end
if sc1_needed and not args.want_sc1 then
error("Declension variant " .. variant .. " must be used with special case (1)")
elseif sc1_needed == false and args.want_sc1 then
error("Declension variant " .. variant .. " must not be used with special case (1)")
end
if animate_needed and args.thisa ~= "a" then
error("Declension variant " .. variant .. " must be specified as animate")
elseif animate_needed == false and args.thisa == "a" then
error("Declension variant " .. variant .. " must not be specified as animate")
end
if gender ~= "" and gender ~= "m" then
error("Declension variant " .. variant .. " should be used with the masculine gender")
end
end
end
stem, tr, decl, orig_pl_ending = detect_lemma_type(lemma, tr, gender,
args, variant)
was_autodetected = true
else
stem, tr = lemma, tr
end
-- Now canonicalize gender
if gender == "3f" then
gender = "f"
elseif gender == "" then
gender = nil
end
-- The ending should be treated as accented if either the original singular
-- or plural ending was accented, or if the stem is non-syllabic.
local was_accented = com.is_stressed(decl) or
orig_pl_ending and com.is_stressed(orig_pl_ending) or
com.is_nonsyllabic(stem)
local was_plural = not not orig_pl_ending
decl = canonicalize_decl(decl, args.old)
-- The rest of this code concerns plural variants. It's somewhat
-- complicated because there are potentially four sources of plural
-- variants (not to mention plural variants constructed using slash
-- notation):
--
-- 1. A user-requested plural variant in declension types 2 or 4 above
-- (currently only -ья)
-- 2. An explicit plural variant encoded in an explicit declension of
-- type 5 above
-- 3. An autodetected plural variant (which will happen in some cases
-- when autodetection is performed on a nominative plural)
-- 4. A plural variant derived using special case (1).
--
-- Up to three actual plural variants might exist (e.g. if the user
-- specifies a DECL value or 'm-ья(1)' and a STEM ending in -а,
-- although not all three can ever be compatible because -ья and (1)
-- are never compatible). We can't have all four because if there's
-- an explicit plural variant, there won't be a user-requested or
-- autodetected plural variant.
--
-- The goal below is to do two things: Check that all available plural
-- variants are the same, and generate the actual declension.
-- If we have a type-2 or type-3 variant, we already have the actual
-- declension; else we need to use a table to map the basic declension
-- to the one with the plural variant encoded in it.
--
-- NOTE: The code below was written with a more general plural-variant
-- system. It probably can be simplified a lot now.
-- 1: Handle explicit decl with slash variant
if rfind(decl, "/") then
if want_ya_plural then
-- Don't think this can happen
error("Plural variant " .. want_ya_plural .. " not compatible with slash declension " .. decl)
end
if args.want_sc1 then
error("Special case (1) not compatible with slash declension" .. decl)
end
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- 2: Retrieve explicitly specified or autodetected decl and pl. variant
local basic_decl, detected_or_explicit_plural = rmatch(decl, "^(.*)(%-[^mf]+)$")
if basic_decl == "ь" then
basic_decl = "ь-m"
end
basic_decl = basic_decl or decl
-- 3: Any user-requested plural variant must agree with explicit or
-- autodetected variant.
if want_ya_plural and detected_or_explicit_plural and want_ya_plural ~= detected_or_explicit_plural then
error("Plural variant " .. want_ya_plural .. " requested but plural variant " .. detected_or_explicit_plural .. " detected from plural stem")
end
-- 4: Handle special case (1). Derive the full declension, make sure its
-- plural variant matches any other available plural variants, and
-- return the declension.
if args.want_sc1 then
local sc1_decl = special_case_1_to_plural_variant[basic_decl] or
error("Special case (1) not compatible with declension " .. basic_decl)
local sc1_plural = rsub(sc1_decl, "^.*%-", "-")
local other_plural = want_ya_plural or detected_or_explicit_plural
if other_plural and sc1_plural ~= other_plural then
error("Plural variant " .. other_plural .. " specified or detected, but special case (1) calls for plural variant " .. sc1_plural)
end
return stem, tr, sc1_decl, gender, was_accented, was_plural, was_autodetected
end
-- 5: Handle user-requested plural variant without explicit or detected
-- one. (If an explicit or detected one exists, we've already checked
-- that it agrees with the user-requested one, and so we already have
-- our full declension.)
if want_ya_plural and not detected_or_explicit_plural then
local variant_decl
if plural_variant_detection_map[decl] then
variant_decl = plural_variant_detection_map[decl][want_ya_plural]
end
if variant_decl then
return stem, tr, variant_decl, gender, was_accented, was_plural, was_autodetected
else
return stem, tr, decl .. (args.old and "/ъ-ья" or "/-ья"), gender, was_accented, was_plural, was_autodetected
end
end
-- 6: Just return the full declension, which will include any available
-- plural variant in it.
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- Convert soft adjectival declensions into hard ones following certain
-- stem-final consonants. FIXME: We call this in two places, once
-- to handle auto-detection and once to handle explicit declensions; but
-- in the former case we end up calling it twice.
local function determine_adj_stem_variant(decl, stem)
local iend = rmatch(decl, "^%+[іи]([йея]?)$")
-- Convert ій/ий to ый after velar or sibilant. This is important for
-- velars; doesn't really matter one way or the other for sibilants as
-- the sibilant rules will convert both sets of endings to the same
-- thing (whereas there will be a difference with о vs. е for velars).
if iend and rfind(stem, "[" .. com.velar .. com.sib .. "]$") then
decl = "+ы" .. iend
-- The following is necessary for -ц, unclear if makes sense for
-- sibilants. (Would be necessary -- I think -- if we were
-- inferring short adjective forms, but we're not.)
elseif decl == "+ее" and rfind(stem, "[" .. com.sib_c .. "]$") then
decl = "+ое"
end
return decl
end
-- Attempt to determine the actual adjective declension based on a
-- combination of the declension the user specified and what can be detected
-- from the stem. DECL is the value the user passed for the declension field,
-- after extraneous annotations have been removed (although none are probably
-- relevant here). What's left is one of the following, which always begins
-- with +:
--
-- 1. +, meaning to autodetect the declension from the stem
-- 2. +ь, same as + but selects +ьий instead of +ий if lemma ends in -ий
-- 3. +short, +mixed or +proper, with the declension partly specified but
-- the particular gender/number-specific short/mixed variant to be
-- autodetected
-- 4. A gender (+m, +f or +n), used only for detecting the singular of
-- plural-form lemmas (this is primarily used in conjunction with template
-- ru-noun+, to explicitly specify the gender; for the actual declension,
-- it doesn't much matter what singular gender we pick since we're a
-- plural only)
-- 5. A gender plus short/mixed/proper/ь (e.g. +f-mixed), again with the gender
-- used only for detecting the singular of plural-form short/mixed lemmas
-- 6. An actual declension, possibly including a slash declension
-- (WARNING: Unclear if slash declensions will work, especially those
-- that are adjective/noun combinations)
--
-- Returns the same seven args as for determine_decl(). The returned
-- declension will always begin with +.
detect_adj_type = function(lemma, tr, decl, old)
local was_autodetected
local base, ending
local basedecl, g = rmatch(decl, "^(%+)([mfn])$")
if not basedecl then
g, basedecl = rmatch(decl, "^%+([mfn])%-([a-zь]+)$")
if basedecl then
basedecl = "+" .. basedecl
end
end
decl = basedecl or decl
if decl == "+" or decl == "+ь" then
local loc = rfind(lemma, "[аеиіоыя][" .. AC .. GR .. "]?[йея]$") or rfind(lemma, "ь[йея]$")
if loc then
base, ending = usub(lemma, 1, loc - 1), usub(lemma, loc)
end
if ending == "ий" and decl == "+ь" then
decl = "+ьий"
elseif ending == "ій" and decl == "+ь" then
decl = "+ьій"
elseif ending then
decl = "+" .. ending
else
local loc, shortmixed = rfind(lemma, "[аоы][" .. AC .. GR .. "]?$") or rfind(lemma, "ъ?$")
if loc then
base, ending = usub(lemma, 1, loc - 1), usub(lemma, loc)
shortmixed = rfind(base, "^[" .. com.uppercase .. "].*[иы]" .. AC .. "н$") and "stressed-proper" or -- accented
rfind(base, "^[" .. com.uppercase .. "].*[иы]н$") and "proper" or --not accented
rlfind(base, "[ёео][" .. AC .. GR .. "]?в$") and "short" or
rlfind(base, "[ыи]" .. AC .. "н$") and "stressed-short" or -- accented
rlfind(base, "[ыи]н$") and "mixed" --not accented
end
if not shortmixed then
error("Cannot determine stem type of adjective: " .. lemma)
end
decl = "+" .. ending .. "-" .. shortmixed
end
was_autodetected = true
elseif m_table.contains({"+short", "+mixed", "+proper"}, decl) then
base, ending = rmatch(lemma, "^(.-)([оаыъ]?[" .. AC .. GR .. "]?)$")
assert(base)
local shortmixed = usub(decl, 2)
if rlfind(base, "[ыи]" .. AC .. "н$") then -- accented
if shortmixed == "short" then shortmixed = "stressed-short"
elseif shortmixed == "proper" then shortmixed = "stressed-proper"
end
end
decl = "+" .. ending .. "-" .. shortmixed
was_autodetected = true
else
base = lemma
end
if ending and ending ~= "" then
tr = com.strip_tr_ending(tr, ending)
end
-- Remove any accents from the declension, but not their presence.
-- We will convert was_accented into stress pattern b, and convert that
-- back to an accented version in determine_stress_variant(). This way
-- we end up with the stressed version whether the user placed an accent
-- in the ending or decl or specified stress pattern b.
-- FIXME, might not work in the presence of slash declensions
local was_accented = com.is_stressed(decl)
decl = com.remove_accents(decl)
decl = map_decl(decl, function(decl)
return determine_adj_stem_variant(decl, base)
end)
local singdecl
if decl == "+ые" then
singdecl = (g == "m" or not g) and (was_accented and "+ой" or "+ый") or not old and g == "f" and "+ая" or not old and g == "n" and "+ое"
elseif decl == "+ыя" and old then
singdecl = (g == "f" or not g) and "+ая" or g == "n" and "+ое"
elseif decl == "+ие" and not old then
singdecl = (g == "m" or not g) and "+ий" or g == "f" and "+яя" or g == "n" and "+ее"
elseif decl == "+іе" and old and (g == "m" or not g) then
singdecl = "+ій"
elseif decl == "+ія" and old then
singdecl = (g == "f" or not g) and "+яя" or g == "n" and "+ее"
elseif decl == "+ьи" then
singdecl = (g == "m" or not g) and (old and "+ьій" or "+ьий") or g == "f" and "+ья" or g == "n" and "+ье"
elseif rfind(decl, "^%+ы%-") then -- decl +ы-mixed or similar
local beg = (g == "m" or not g) and (old and "ъ" or "") or g == "f" and "а" or g == "n" and "о"
singdecl = beg and "+" .. beg .. usub(decl, 3)
end
if singdecl then
was_plural = true
decl = singdecl
end
return base, tr, canonicalize_decl(decl, old), g, was_accented, was_plural, was_autodetected
end
-- If stress pattern omitted, detect it based on whether ending is stressed
-- or the decl class or stem accent calls for inherent stress, defaulting to
-- pattern a. This is run after alias resolution and accent removal of DECL;
-- WAS_ACCENTED indicates whether the ending was originally stressed.
-- FIXME: This is run before splitting slash patterns but should be run after.
detect_stress_pattern = function(stem, decl, decl_cats, reducible,
was_plural, was_accented)
-- ёнок and ёночек always bear stress
if rfind(decl, "ёнокъ?") or rfind(decl, "ёночекъ?") then
return "b"
-- stressed suffix и́н; missing in plural and true endings don't bear stress
-- (except for exceptional господи́н)
elseif rfind(decl, "инъ?") and was_accented then
return "d"
-- Adjectival -ой always bears the stress
elseif rfind(decl, "%+ой") then
return "b"
-- Adjectival stressed-short, stressed-proper bears the stress
elseif rfind(decl, "^%+.*%-stressed") then
return "b"
-- Pattern b if ending was accented by user
elseif was_accented then
return "b"
-- Nonsyllabic stem means pattern b
elseif com.is_nonsyllabic(stem) then
return "b"
-- Accent on reducible vowel in masc nom sg (not plural) means pattern b.
-- Think about whether we want to enable this.
-- elseif reducible and not was_plural then
-- -- FIXME hack. Eliminate plural part of slash declension.
-- decl = rsub(decl, "/.*", "")
-- if decl_cats[decl] and decl_cats[decl].g == "m" then
-- if com.is_ending_stressed(stem) or com.is_monosyllabic(stem) then
-- return "b"
-- end
-- end
end
return "a"
end
-- In certain special cases, depending on the declension, we override the
-- user-specified stress pattern and convert it to something else.
-- NOTE: This function is run after alias resolution and accent removal.
-- FIXME: It's also run before splitting slash patterns but should be run after.
override_stress_pattern = function(decl, stress)
-- ёнок and ёночек always bear stress; if user specified a,
-- convert to b. Don't do this with slash patterns (see FIXME above).
if stress == "a" and (rfind(decl, "^ёнокъ?$") or rfind(decl, "^ёночекъ?$")) then
return "b"
end
return stress
end
-- Canonicalize an adjectival declension to either the stressed or unstressed
-- variant depending on the stress. Ultimately this is what ensures that
-- the user's stress mark on an adjectival ending is respected.
determine_stress_variant = function(decl, stress)
if stress == "b" then
if decl == "+ая" then
return "+а́я"
elseif decl == "+ое" then
return "+о́е"
else
-- Convert +...-short to +...-stressed-short and same for -proper
local b, e = rmatch(decl, "^%+(.*)%-(short)$")
if not b then
b, e = rmatch(decl, "^%+(.*)%-(proper)$")
end
if b and not rfind(b, "%-stressed") then
return "+" .. b .. "-stressed-" .. e
end
end
end
return decl
end
-- Canonicalize a declension based on the final stem consonant, in
-- particular converting soft declensions to hard ones after velars and/or
-- sibilants. FIXME: We also do this canonicalization earlier on during
-- auto-detection (determine_adj_stem_variant() is called by
-- detect_adj_type(), and code in detect_lemma_type() does the equivalent
-- of the first clause below). Doing it here ensures that explicitly
-- specified declensions get handled as well, but it would be nice to not
-- do the same thing twice in the auto-detection case.
determine_stem_variant = function(decl, stem)
if decl == "е" and rfind(stem, "[" .. com.sib_c .. "]$") then
return "о"
end
return determine_adj_stem_variant(decl, stem)
end
is_reducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "3rd" and decl_cat.g == "f" or decl_cat.g == "m" then
return true
else
return false
end
end
-- Reduce nom sg to stem by eliminating the "epenthetic" vowel. Applies to
-- masculine 2nd-declension hard and soft, and 3rd-declension feminine in
-- -ь. STEM and DECL are after determine_decl(), before converting
-- outward-facing declensions to inward ones.
function export.reduce_nom_sg_stem(stem, tr, decl, soft_n, can_err)
local full_stem = stem .. (decl == "й" and decl or "")
local full_tr = tr and tr .. (decl == "й" and "j" or "")
local ret, rettr = com.reduce_stem(full_stem, full_tr, soft_n)
if not ret and can_err then
error("Unable to reduce stem " .. stem)
end
return ret, rettr
end
is_dereducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "1st" or decl_cat.decl == "2nd" and decl_cat.g == "n" then
return true
else
return false
end
end
-- Add a possible suffix to the bare stem, according to the declension and
-- value of OLD. This may be -ь, -ъ, -й or nothing. We need to do this here
-- because we don't actually attach such a suffix in attach_unstressed() due
-- to situations where we don't want the suffix added, e.g. dereducible nouns
-- in -ня.
add_bare_suffix = function(bare, baretr, old, sgdc, dereduced)
if old and sgdc.hard == "hard" then
-- Final -ъ isn't transliterated
return bare .. "ъ", baretr
elseif sgdc.hard == "soft" or sgdc.hard == "palatal" then
-- This next clause corresponds to a special case in Vitalik's module.
-- It says that nouns in -ня (accent class a) have gen pl without
-- trailing -ь. It appears to apply to most nouns in -ня (possibly
-- all in -льня), but ку́хня (gen pl ку́хонь) and дерéвня (gen pl
-- дереве́нь) is an exception. (Vitalik's module has an extra
-- condition here 'stress == "a"' that would exclude дере́вня but I
-- don't think this condition is in Zaliznyak, as he indicates
-- дере́вня as having an exceptional genitive plural.)
if dereduced and rfind(bare, "[нН]$") and sgdc.decl == "1st" then
-- FIXME: What happens in this case old-style? I assume that
-- -ъ is added, but this is a guess.
-- Final -ъ isn't transliterated
return bare .. (old and "ъ" or ""), baretr
elseif rfind(bare, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
return bare .. "й", baretr and (baretr .. "j")
else
return bare .. "ь", baretr and (baretr .. "ʹ")
end
else
return bare, baretr
end
end
-- Dereduce stem to the form found in the gen pl (and maybe nom sg) by
-- inserting an epenthetic vowel. Applies to 1st declension and 2nd
-- declension neuter, and to 2nd declension masculine when the stem was
-- specified as a plural form (in which case we're deriving the nom sg,
-- and also the gen pl in the alt-gen-pl scenario). STEM and DECL are
-- after determine_decl(), before converting outward-facing declensions
-- to inward ones. STRESS is the stess pattern.
function export.dereduce_nom_sg_stem(stem, tr, sgdc, stress, old, can_err)
local epenthetic_stress = ending_stressed_gen_pl_patterns[stress]
local ret, rettr = com.dereduce_stem(stem, tr, epenthetic_stress)
if not ret then
if can_err then
error("Unable to dereduce stem " .. stem)
else
return nil, nil
end
end
return add_bare_suffix(ret, rettr, old, sgdc, true)
end
--------------------------------------------------------------------------
-- Second-declension masculine --
--------------------------------------------------------------------------
----------------- Masculine hard -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style).
declensions_old["ъ"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and "е́й" or "о́въ"
end,
["alt_gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["ъ"] = { decl="2nd", hard="hard", g="m" }
-- Normal mapping of old ъ would be "" (blank), but we set up "#" as an alias
-- so we have a way of referring to it without defaulting if need be and
-- distinct from auto-detection (e.g. in the second stem of a word, or to
-- override autodetection of -ёнок or -ин -- the latter is necessary in the
-- case of семьянин).
declensions_aliases["#"] = ""
----------------- Masculine hard, irregular plural -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg nom pl -а.
declensions_old["ъ-а"] = mw.clone(declensions_old["ъ"])
declensions_old["ъ-а"]["nom_pl"] = "а́"
declensions_old_cat["ъ-а"] = { decl="2nd", hard="hard", g="m", alt_nom_pl=true }
declensions_aliases["#-a"] = "-a"
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg soft pl -ья.
-- Differs from the normal declension throughout the plural.
declensions_old["ъ-ья"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = "ьёвъ",
["alt_gen_pl"] = "е́й",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ъ-ья"] = { decl="2nd", hard="hard", g="m", irregpl=true }
declensions_aliases["#-ья"] = "-ья"
----------------- Masculine hard, suffixed, irregular plural -------------------
declensions_old["инъ"] = {
["nom_sg"] = "и́нъ",
["gen_sg"] = "и́на",
["dat_sg"] = "и́ну",
["acc_sg"] = nil,
["ins_sg"] = "и́номъ",
["pre_sg"] = "и́нѣ",
["nom_pl"] = "е́",
["gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["инъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old["ёнокъ"] = {
["nom_sg"] = "ёнокъ",
["gen_sg"] = "ёнка",
["dat_sg"] = "ёнку",
["acc_sg"] = nil,
["ins_sg"] = "ёнкомъ",
["pre_sg"] = "ёнкѣ",
["nom_pl"] = "я́та",
["gen_pl"] = "я́тъ",
["dat_pl"] = "я́тамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тами",
["pre_pl"] = "я́тахъ",
}
declensions_old_cat["ёнокъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["онокъ"] = "ёнокъ"
declensions_old_aliases["енокъ"] = "ёнокъ"
declensions_old["ёночекъ"] = {
["nom_sg"] = "ёночекъ",
["gen_sg"] = "ёночка",
["dat_sg"] = "ёночку",
["acc_sg"] = nil,
["ins_sg"] = "ёночкомъ",
["pre_sg"] = "ёночкѣ",
["nom_pl"] = "я́тки",
["gen_pl"] = "я́токъ",
["dat_pl"] = "я́ткамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тками",
["pre_pl"] = "я́ткахъ",
}
declensions_old_cat["ёночекъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["оночекъ"] = "ёночекъ"
declensions_old_aliases["еночекъ"] = "ёночекъ"
----------------- Masculine soft -------------------
-- Normal soft-masculine declension in -ь
declensions_old["ь-m"] = {
["nom_sg"] = "ь",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["alt_gen_pl"] = "ь",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-m"] = { decl="2nd", hard="soft", g="m" }
-- Soft-masculine declension in -ь with irreg nom pl -я
declensions_old["ь-я"] = mw.clone(declensions_old["ь-m"])
declensions_old["ь-я"]["nom_pl"] = "я́"
declensions_old_cat["ь-я"] = { decl="2nd", hard="soft", g="m", alt_nom_pl=true }
----------------- Masculine palatal -------------------
-- Masculine declension in palatal -й
declensions_old["й"] = {
["nom_sg"] = "й",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = "ёвъ",
["alt_gen_pl"] = "й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["й"] = { decl="2nd", hard="palatal", g="m" }
declensions_old["й-я"] = mw.clone(declensions_old["й"])
declensions_old["й-я"]["nom_pl"] = "я́"
declensions_old_cat["й-я"] = { decl="2nd", hard="palatal", g="m", alt_nom_pl=true }
--------------------------------------------------------------------------
-- First-declension feminine --
--------------------------------------------------------------------------
----------------- Feminine hard -------------------
-- Hard-feminine declension in -а
declensions_old["а"] = {
["nom_sg"] = "а́",
["gen_sg"] = "ы́",
["dat_sg"] = "ѣ́",
["acc_sg"] = "у́",
["ins_sg"] = {"о́й<insa>", "о́ю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["а"] = { decl="1st", hard="hard", g="f" }
----------------- Feminine soft -------------------
-- Soft-feminine declension in -я
declensions_old["я"] = {
["nom_sg"] = "я́",
["gen_sg"] = "и́",
["dat_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_dat_sg_patterns[stress] and "и" or "ѣ́"
end,
["acc_sg"] = "ю́",
["ins_sg"] = {"ёй<insa>", "ёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "]́?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["я"] = { decl="1st", hard="soft", g="f" }
-- Soft-feminine declension in -ья.
-- Almost like ь + -я endings except for genitive plural.
declensions_old["ья"] = {
["nom_sg"] = "ья́",
["gen_sg"] = "ьи́",
["dat_sg"] = "ьѣ́",
["acc_sg"] = "ью́",
["ins_sg"] = {"ьёй<insa>", "ьёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ьи́",
["gen_pl"] = function(stem, stress)
-- circumflex accent is a signal that forces stress, particularly
-- in accent pattern d/d'.
return (ending_stressed_gen_pl_patterns[stress] or stress == "d" or stress == "d'") and "е̂й" or "ий"
end,
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ья"] = {
decl="1st", hard="soft", g="f",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
--------------------------------------------------------------------------
-- Second-declension neuter --
--------------------------------------------------------------------------
----------------- Neuter hard -------------------
-- Normal hard-neuter declension in -о
declensions_old["о"] = {
["nom_sg"] = "о́",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "о́" or nil
end,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "а́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "о́въ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["о"] = { decl="2nd", hard="hard", g="n" }
-- Hard-neuter declension in -о with irreg nom pl -и
declensions_old["о-и"] = mw.clone(declensions_old["о"])
declensions_old["о-и"]["nom_pl"] = "ы́"
declensions_old_cat["о-и"] = { decl="2nd", hard="hard", g="n", alt_nom_pl=true }
declensions_old_aliases["о-ы"] = "о-и"
-- Masculine-gender neuter-form declension in -(ишк)о with irreg nom pl -и,
-- with colloquial feminine endings in some of the singular cases
-- (§5 p. 74 of Zaliznyak)
declensions_old["(ишк)о-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ишк)о-и"]["gen_sg"] = {"а́", "ы́1"}
declensions_old["(ишк)о-и"]["dat_sg"] = {"у́", "ѣ́1"}
declensions_old["(ишк)о-и"]["ins_sg"] = {"о́мъ", "о́й1"}
declensions_old_cat["(ишк)о-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ишк)о-и"] = "<sup>1</sup> Colloquial."
-- Masculine-gender animate neuter-form declension in -(ищ)е with irreg
-- nom pl -и, with colloquial feminine endings in some of the singular cases
-- (§4 p. 74 of Zaliznyak)
declensions_old["(ищ)е-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ищ)е-и"]["acc_sg"] = {"а́", "у́1"}
declensions_old["(ищ)е-и"]["gen_sg"] = {"а́", "ы́2"}
declensions_old["(ищ)е-и"]["dat_sg"] = {"у́", "ѣ́2"}
declensions_old["(ищ)е-и"]["ins_sg"] = {"о́мъ", "о́й2"}
declensions_old_cat["(ищ)е-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ищ)е-и"] = "<sup>1</sup> Colloquial.<br /><sup>2</sup> Less common, more colloquial."
----------------- Neuter soft -------------------
-- Soft-neuter declension in -е (stressed -ё)
declensions_old["е"] = {
["nom_sg"] = "ё",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ё" or nil
end,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е"] = {
singular = function(suffix)
if suffix == "ё" then
return "ending in -ё"
else
return {}
end
end,
decl="2nd", hard="soft", g="n", gensg=true
}
-- User-facing declension type "ё" = "е"
declensions_old_aliases["ё"] = "е"
-- Rare soft-neuter declension in stressed -е́ (e.g. муде́, бытие́)
declensions_old["е́"] = {
["nom_sg"] = "е́",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "е́" or nil
end,
["ins_sg"] = "е́мъ",
["pre_sg"] = function(stem, stress)
-- FIXME!!! Are we sure about this condition? This is what was
-- found in the old template, but the related -е declension has
-- -ие prep sg ending -(и)и only when *not* stressed.
return rlfind(stem, "[іи][" .. AC .. GR .. "]?$") and "и́" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return rlfind(stem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") and "й" or "е́й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е́"] = {
singular = "ending in stressed -е",
decl="2nd", hard="soft", g="n", gensg=true
}
-- Soft-neuter declension in unstressed -ье (stressed -ьё).
declensions_old["ье"] = {
["nom_sg"] = "ьё",
["gen_sg"] = "ья́",
["dat_sg"] = "ью́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ьё" or nil
end,
["ins_sg"] = "ьёмъ",
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and "е́й" or "ий"
end,
["alt_gen_pl"] = "ьёвъ",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ье"] = {
decl="2nd", hard="soft", g="n",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
declensions_old_aliases["ьё"] = "ье"
--------------------------------------------------------------------------
-- Third declension --
--------------------------------------------------------------------------
declensions_old["ь-f"] = {
["nom_sg"] = "ь",
["gen_sg"] = "и́",
["dat_sg"] = "и́",
["acc_sg"] = "ь",
["ins_sg"] = "ью́",
["pre_sg"] = "и́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-f"] = { decl="3rd", hard="soft", g="f" }
declensions_old["мя"] = {
["nom_sg"] = "мя",
["gen_sg"] = "мени",
["dat_sg"] = "мени",
["acc_sg"] = nil,
["ins_sg"] = "менемъ",
["pre_sg"] = "мени",
["nom_pl"] = "мена́",
["gen_pl"] = "мёнъ",
["dat_pl"] = "мена́мъ",
["acc_pl"] = nil,
["ins_pl"] = "мена́ми",
["pre_pl"] = "мена́хъ",
}
declensions_old_cat["мя"] = { decl="3rd", hard="soft", g="n", cant_reduce=true }
--------------------------------------------------------------------------
-- Indeclinable --
--------------------------------------------------------------------------
-- Indeclinable declension; no endings.
declensions_old["$"] = {
["nom_sg"] = "",
["gen_sg"] = "",
["dat_sg"] = "",
["acc_sg"] = nil,
["ins_sg"] = "",
["pre_sg"] = "",
["nom_pl"] = "",
["gen_pl"] = "",
["dat_pl"] = "",
["acc_pl"] = nil,
["ins_pl"] = "",
["pre_pl"] = "",
}
declensions_old_cat["$"] = { decl="indeclinable", hard="none", g="none" }
--------------------------------------------------------------------------
-- Adjectival --
--------------------------------------------------------------------------
-- This needs to be up here because it is called just below.
local function old_to_new(v)
v = rsub(v, "ъ$", "")
v = rsub(v, "^ъ", "")
v = rsub(v, "(%A)ъ", "%1")
v = rsub(v, "ъ(%A)", "%1")
v = rsub(v, "і", "и")
v = rsub(v, "ѣ", "е")
return v
end
-- Meaning of entry is:
-- 1. The declension name in module ru-adjective
-- 2. The masculine declension name in this module
-- 3. The neuter declension name in this module
-- 4. The feminine declension name in this module
-- 5. The value of hard= for the declensions_cat entry
-- 6. The value of decl= for the declensions_cat entry
-- 7. The value of possadj= for the declensions_cat entry (true if possessive
-- or similar type of adjective)
local adj_decl_map = {
{"ый", "ый", "ое", "ая", "hard", "long", false},
{"ій", "ій", "ее", "яя", "soft", "long", false},
{"ой", "ой", "о́е", "а́я", "hard", "long", false},
{"ьій", "ьій", "ье", "ья", "palatal", "long", true},
{"short", "ъ-short", "о-short", "а-short", "hard", "short", true},
{"mixed", "ъ-mixed", "о-mixed", "а-mixed", "hard", "mixed", true},
{"proper", "ъ-proper", "о-proper", "а-proper", "hard", "proper", true},
{"stressed-short", "ъ-stressed-short", "о-stressed-short", "а-stressed-short", "hard", "short", true},
{"stressed-proper", "ъ-stressed-proper", "о-stressed-proper", "а-stressed-proper", "hard", "proper", true},
}
local function get_adjectival_decl(adjtype, gender, old)
local decl, intnotes = m_ru_adj.get_nominal_decl(adjtype, gender, old)
-- hack fem ins_sg to insert <insa>, <insb>; see concat_word_forms_1()
if gender == "f" and type(decl["ins_sg"]) == "table" and #decl["ins_sg"] == 2 then
decl["ins_sg"][1] = decl["ins_sg"][1] .. "<insa>"
decl["ins_sg"][2] = decl["ins_sg"][2] .. "<insb>"
end
return decl, intnotes
end
for _, declspec in ipairs(adj_decl_map) do
local oadjdecl = declspec[1]
local nadjdecl = old_to_new(oadjdecl)
local odecl_by_gender =
{m="+" .. declspec[2], n="+" .. declspec[3], f="+" .. declspec[4]}
local hard = declspec[5]
local decltype = declspec[6]
local possadj = declspec[7]
for _, g in ipairs({"m", "n", "f"}) do
local odecl = odecl_by_gender[g]
local ndecl = old_to_new(odecl)
declensions_old[odecl], internal_notes_table_old[odecl] =
get_adjectival_decl(oadjdecl, g, true)
declensions[ndecl], internal_notes_table[ndecl] =
get_adjectival_decl(nadjdecl, g, false)
declensions_old_cat[odecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
declensions_cat[ndecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
end
end
-- Set up some aliases.
declensions_old_aliases["+о́-short"] = "+о-stressed-short"
declensions_old_aliases["+а́-short"] = "+а-stressed-short"
declensions_old_aliases["+о́-proper"] = "+о-stressed-proper"
declensions_old_aliases["+а́-proper"] = "+а-stressed-proper"
declensions_aliases["+#-short"] = "+-short"
declensions_aliases["+#-mixed"] = "+-mixed"
declensions_aliases["+#-proper"] = "+-proper"
declensions_aliases["+#-stressed-short"] = "+-stressed-short"
declensions_aliases["+#-stressed-proper"] = "+-stressed-proper"
--------------------------------------------------------------------------
-- Populate new from old --
--------------------------------------------------------------------------
-- Function to convert an entry in an old declensions table to new.
local function old_decl_entry_to_new(v)
if not v then
return nil
elseif type(v) == "table" then
local new_entry = {}
for _, i in ipairs(v) do
table.insert(new_entry, old_decl_entry_to_new(i))
end
return new_entry
elseif type(v) == "function" then
return function(stem, suffix, args)
return old_decl_entry_to_new(v(stem, suffix, args))
end
else
return old_to_new(v)
end
end
-- Function to convert an old declensions table to new.
local function old_decl_to_new(odecl)
local ndecl = {}
for k, v in pairs(odecl) do
ndecl[k] = old_decl_entry_to_new(v)
end
return ndecl
end
-- Function to convert an entry in an old declensions_cat table to new.
local function old_decl_cat_entry_to_new(odecl_cat_entry)
if not odecl_cat_entry then
return nil
elseif type(odecl_cat_entry) == "function" then
return function(suffix)
return old_decl_cat_entry_to_new(odecl_cat_entry(suffix))
end
elseif type(odecl_cat_entry) == "table" then
local ndecl_cat_entry = {}
for k, v in pairs(odecl_cat_entry) do
ndecl_cat_entry[k] = old_decl_cat_entry_to_new(v)
end
return ndecl_cat_entry
elseif type(odecl_cat_entry) == "boolean" then
return odecl_cat_entry
else
assert(type(odecl_cat_entry) == "string")
return old_to_new(odecl_cat_entry)
end
end
-- Function to convert an old declensions_cat table to new.
local function old_decl_cat_to_new(odeclcat)
local ndeclcat = {}
for k, v in pairs(odeclcat) do
ndeclcat[k] = old_decl_cat_entry_to_new(v)
end
return ndeclcat
end
-- populate declensions[] from declensions_old[]
for odecltype, odecl in pairs(declensions_old) do
local ndecltype = old_to_new(odecltype)
if not declensions[ndecltype] then
declensions[ndecltype] = old_decl_to_new(odecl)
end
end
-- populate declensions_cat[] from declensions_old_cat[]
for odecltype, odeclcat in pairs(declensions_old_cat) do
local ndecltype = old_to_new(odecltype)
if not declensions_cat[ndecltype] then
declensions_cat[ndecltype] = old_decl_cat_to_new(odeclcat)
end
end
-- populate declensions_aliases[] from declensions_old_aliases[]
for ofrom, oto in pairs(declensions_old_aliases) do
local from = old_to_new(ofrom)
if not declensions_aliases[from] then
declensions_aliases[from] = old_to_new(oto)
end
end
-- populate internal_notes_table[] from internal_notes_table_old[]
for odecl, note in pairs(internal_notes_table_old) do
local ndecl = old_to_new(odecl)
if not internal_notes_table[ndecl] then
-- FIXME, should we be calling old_to_new() here?
internal_notes_table[ndecl] = note
end
end
--------------------------------------------------------------------------
-- Inflection functions --
--------------------------------------------------------------------------
-- Attach the stressed stem (or plural stem, or barestem) out of ARGS
-- to the unstressed suffix SUF, modifying the suffix as necessary for the
-- last letter of the stem (e.g. if it is velar, sibilant or ц). CASE is
-- the case form being created and is used to select the plural stem if
-- needed. Returns two values, the combined form and the modified suffix.
local function attach_unstressed(args, case, suf, was_stressed)
if suf == nil then
return nil, nil
elseif rfind(suf, CFLEX) then -- if suf has circumflex accent, it forces stressed
return attach_stressed(args, case, suf)
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.pl, args.pltr
end
if not stem and case == "ins_sg" then
stem, tr = args.ins_sg_stem, args.ins_sg_tr
end
if not stem then
stem, tr = args.stem, args.stemtr
end
if nom.nonsyllabic_suffixes[suf] then
-- If gen_pl, use special args.gen_pl_bare if given, else regular
-- args.bare if there isn't a plural stem. If nom_sg, always use
-- regular args.bare.
local barearg, bareargtr
if case == "gen_pl" then
barearg, bareargtr = args.gen_pl_bare, args.gen_pl_baretr
if not barearg and args.pl == args.stem then
barearg, bareargtr = args.bare, args.baretr
end
else
barearg, bareargtr = args.bare, args.baretr
end
local barestem = barearg or stem
local barestem, baretr
if barearg then
barestem, baretr = barearg, bareargtr
else
barestem, baretr = stem, tr
end
if was_stressed and case == "gen_pl" then
if not barearg then
local gen_pl_stem, gen_pl_tr = com.make_ending_stressed(stem, tr)
barestem, baretr = gen_pl_stem, gen_pl_tr
end
end
if rlfind(barestem, "[йьъ]$") then
suf = ""
else
if suf == "ъ" then
-- OK
elseif suf == "й" or suf == "ь" then
if barearg and case == "gen_pl" then
-- explicit bare or reducible, don't add -ь
suf = ""
elseif rfind(barestem, "[" .. com.vowel .. "][" .. AC .. GR .. "]?$") then
-- not reducible, do add -ь and correct to -й if necessary
suf = "й"
else
suf = "ь"
end
end
end
return com.concat_russian_tr(barestem, baretr, suf, nil, "dopair"), suf
end
suf = com.make_unstressed(suf)
local rules = nom.unstressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Analogous to attach_unstressed() but for the unstressed stem and a
-- stressed suffix.
attach_stressed = function(args, case, suf)
if suf == nil then
return nil, nil
end
-- circumflex forces stress even when the accent pattern calls for no stress
suf = rsub(suf, "̂", AC)
if com.is_unstressed(suf) then
return attach_unstressed(args, case, suf, "was stressed")
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.upl, args.upltr
end
if not stem then
stem, tr = args.ustem, args.ustemtr
end
local rules = nom.stressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Attach the appropriate stressed or unstressed stem (or plural stem as
-- determined by CASE, or barestem) out of ARGS to the suffix SUF, which may
-- be a list of alternative suffixes (e.g. in the inst sg of feminine nouns).
-- Calls FUN (either attach_stressed() or attach_unstressed()) to do the work
-- for an individual suffix. Returns two values, a list of combined forms
-- and a list of the real suffixes used (which may be modified from the
-- passed-in suffixes, e.g. by removing stress marks or modifying vowels in
-- various ways after a stem-final velar, sibilant or ц). Each combined form
-- is a two-element list {stem, tr} (or a one-element list if tr is nil).
-- IRREG is true if this is an irregular form. We are handling the Nth word;
-- ISLAST is true if this is the last one.
local function attach_with(args, case, suf, fun, irreg, n, islast)
if type(suf) == "table" then
local all_combineds = {}
local all_realsufs = {}
for _, x in ipairs(suf) do
local combineds, realsufs =
attach_with(args, case, x, fun, irreg, n, islast)
for _, combined in ipairs(combineds) do
table.insert(all_combineds, combined)
end
for _, realsuf in ipairs(realsufs) do
table.insert(all_realsufs, realsuf)
end
end
return all_combineds, all_realsufs
else
local combined, realsuf = fun(args, case, suf)
local irregsuf = irreg and {IRREGMARKER} or {""}
return {combined and com.concat_paired_russian_tr(
com.concat_paired_russian_tr(args["prefix" .. n], combined),
com.concat_paired_russian_tr(args["suffix" .. n], irregsuf)) or nil},
{realsuf and realsuf .. args["suffix" .. n][1] or nil}
end
end
-- Generate the form(s) and suffix(es) for CASE according to the declension
-- table DECL, using the attachment function FUN (one of attach_stressed()
-- or attach_unstressed()). IS_SLASH is true if this is a slash declension
-- (different declensions for singular and plural). We are handling the Nth
-- word; ISLAST is true if this is the last one.
local function gen_form(args, decl, case, stress, fun, is_slash, n, islast)
local irreg = false
if not args.suffixes[case] then
args.suffixes[case] = {}
end
local decl_sufs = args.old and declensions_old or declensions
decl_sufs = decl_sufs[decl]
local suf = decl_sufs[case]
local decl_cats = args.old and declensions_old_cat or declensions_cat
local ispl = rfind(case, "_pl")
if ispl and (decl_cats[decl].irregpl or args.pl and args.pl ~= args.stem or is_slash) then
irreg = true
end
if case == "nom_pl" and decl_cats[decl].alt_nom_pl then
irreg = true
end
if type(suf) == "function" then
suf = suf(ispl and args.pl or args.stem, stress, args)
end
if case == "gen_pl" and args.alt_gen_pl then
suf = decl_sufs.alt_gen_pl
irreg = true
if not suf then
error("No alternative genitive plural available for this declension class")
end
end
local combineds, realsufs = attach_with(args, case, suf, fun, irreg, n, islast)
for _, realsuf in ipairs(realsufs) do
args.any_non_nil[case] = true
args.this_any_non_nil[case] = true
insert_if_not(args.suffixes[case], realsuf)
end
return combineds
end
local attachers = {
["+"] = attach_stressed,
["-"] = attach_unstressed,
}
do_stress_pattern = function(stress, args, decl, number, n, islast)
local f = {}
for _, case in ipairs(decl_cases) do
if not number or (number == "sg" and rfind(case, "_sg")) or
(number == "pl" and rfind(case, "_pl")) then
f[case] = gen_form(args, decl, case, stress,
attachers[stress_patterns[stress][case]], not not number,
n, islast)
-- Turn empty form lists into nil to facilitate computation of
-- animate/inanimate accusatives below
if f[case] and #f[case] == 0 then
f[case] = nil
end
-- Compute linked versions of potential lemma cases, for use
-- in the ru-noun+ headword. We substitute the original lemma
-- (before removing links) for forms that are the same as the
-- lemma, if the original lemma has links.
if f[case] and (case == "nom_sg" or case == "nom_pl") then
local linked_forms = {}
for _, form in ipairs(f[case]) do
-- Return true if FORM is "close enough" to LEMMA that we can substitute the
-- linked form of the lemma. Currently this means exactly the same except that
-- we ignore acute and grave accent differences in monosyllables, and ignore
-- notes that may have been appended (e.g. the triangle marking irregularity).
local entry, notes = m_table_tools.separate_notes(form[1])
local lemma = args.lemma_no_links
local close_enough_to_lemma =
entry == lemma or (com.is_monosyllabic(entry) and
com.is_monosyllabic(lemma) and
com.remove_accents(entry) == com.remove_accents(lemma))
if close_enough_to_lemma and
rfind(args.orig_lemma, "%[%[") then
table.insert(linked_forms, {args.orig_lemma .. notes, args.lemmatr and args.lemmatr .. notes})
else
table.insert(linked_forms, form)
end
end
f[case .. "_linked"] = linked_forms
end
end
end
-- Set acc an/in variants now as appropriate. We used to do this in
-- handle_forms_and_overrides(), which simplified the handling of
-- nom/gen/acc overrides but caused problems for words like мазло and
-- трепло that had a mixture of nil and non-nil accusative forms.
local an = args.thisa
if not number or number == "sg" then
f.acc_sg_an = f.acc_sg_an or f.acc_sg or an == "i" and f.nom_sg or f.gen_sg
f.acc_sg_in = f.acc_sg_in or f.acc_sg or an == "a" and f.gen_sg or f.nom_sg
end
if not number or number == "pl" then
f.acc_pl_an = f.acc_pl_an or f.acc_pl or an == "i" and f.nom_pl or f.gen_pl
f.acc_pl_in = f.acc_pl_in or f.acc_pl or an == "a" and f.gen_pl or f.nom_pl
end
for case, forms in pairs(f) do
if not args.forms[case] then
args.forms[case] = {}
end
for _, form in ipairs(forms) do
insert_if_not(args.forms[case], form)
end
end
end
stress_patterns["a"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["b"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["b'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["c"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["d"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["d'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["e"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f''"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
ending_stressed_gen_pl_patterns = m_table.listToSet({"b", "b'", "c", "e", "f", "f'", "f''"})
ending_stressed_pre_sg_patterns = m_table.listToSet({"b", "b'", "d", "d'", "f", "f'", "f''"})
ending_stressed_dat_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_pl_patterns = m_table.listToSet({"b", "b'", "c"})
local numbers = {
["s"] = "singular",
["p"] = "plural",
}
local old_title_temp = [=[Pre-reform declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local title_temp = [=[Declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local extra_case_template, extra_case_template_with_plural
local internal_notes_template
local notes_template
local templates = {}
-- Convert a raw override into a canonicalized list of individual overrides.
-- If input is nil, so is output. Certain junk (e.g. <br/>) is removed,
-- and ~ and ~~ are substituted appropriately; ARGS and CASE are required for
-- this purpose. FORMS is the list of existing case forms and is currently
-- used only to retrieve the dat_sg for handling + in loc/par (this means
-- that any overrides of the dat_sg have to be handled before handling
-- overrides of loc/par). N is the suffix used to retrieve the override --
-- either a number for word-specific overrides, or an empty string for
-- overall overrides.
--
-- It will still be necessary to call m_table_tools.separate_notes() to separate
-- off any trailing "notes" (asterisks, superscript numbers, etc.), and
-- m_links.remove_links() to remove any links to get the raw override form.
canonicalize_override = function(args, case, forms, n)
local val = args[case .. n]
if not val then
return nil
end
-- clean <br /> that's in many multi-form entries and messes up linking
val = rsub(val, "<br%s*/>", "")
-- substitute ~ and ~~ and split by commas
local stem, manualtr
if rfind(case, "_pl") then
stem, manualtr = args.pl, args.pltr
end
if not stem then
stem, manualtr = args.stem, args.stemtr
end
local ustem, umanualtr = com.make_unstressed_once(stem, manualtr)
local stemtr, ustemtr = manualtr, umanualtr
local vals = rsplit(val, "%s*,%s*")
local retvals = {}
for _, val in ipairs(vals) do
local valru, valtr = com.split_russian_tr(val)
valru = rsub(valru, "~~", ustem)
valru = rsub(valru, "~", com.is_stressed(val) and ustem or stem)
if rfind(valru, "^%*") then
valru = rsub(valru, "^%*", "") .. HYPMARKER
end
if valtr then
stemtr = stemtr or com.translit_no_links(stem)
ustemtr = ustemtr or com.translit_no_links(ustem)
valtr = rsub(valtr, "~~", ustemtr)
valtr = rsub(valtr, "~", com.is_stressed(val) and ustemtr or stemtr)
elseif val:find("~") and manualtr then
valtr = com.translit_no_links(val)
valtr = rsub(valtr, "~~", umanualtr)
valtr = rsub(valtr, "~", com.is_stressed(val) and umanualtr or manualtr)
end
if valtr then
if valtr:sub(1, 1) == "*" then
valtr = valtr:sub(2) .. HYPMARKER
end
valtr = com.j_correction(valtr)
end
table.insert(retvals, {valru, valtr})
end
vals = retvals
-- handle + in loc/par meaning "the expected form"; NOTE: This requires
-- that dat_sg has already been processed!
if case == "loc" or case == "par" then
local new_vals = {}
for _, rutr in ipairs(vals) do
-- don't just handle + by itself in case the arg has в or на
-- or whatever attached to it
if rfind(rutr[1], "^%+") or rfind(rutr[1], "[%s%[|]%+") then
for _, dat in ipairs(forms["dat_sg"]) do
local ru, tr = rutr[1], rutr[2]
local datru, dattr = dat[1], dat[2]
local valru, valtr
-- separate off any footnote symbols (which may have been
-- introduced by a *tail or *tailall argument, which we need
-- to process before this stage so overrides don't get
-- automatically marked with footnote symbols); if par,
-- try to preserve them, but with loc this can cause
-- problems in case there are multiple dative forms
-- (stress variants) and the last one is marked with a
-- footnote symbol (occurs in грудь)
local datru_entry, datru_notes = m_table_tools.separate_notes(datru)
local dattr_entry, dattr_notes
if dattr then
dattr_entry, dattr_notes = m_table_tools.separate_notes(dattr)
end
if case == "par" then
valru, valtr = datru_entry, dattr_entry
else
valru, valtr = com.make_ending_stressed(datru_entry, dattr_entry)
datru_notes = ""
dattr_notes = ""
end
-- wrap the word in brackets so it's linked; but not if it
-- appears to already be linked
ru = rsub(ru, "^%+", "[[" .. valru .. "]]")
ru = rsub(ru, "([%[|])%+", "%1" .. valru)
ru = rsub(ru, "(%s)%+", "%1[[" .. valru .. "]]")
ru = ru .. datru_notes
-- do the translit; but it shouldn't have brackets in it
if tr or valtr then
tr = tr or com.translit_no_links(rutr[1])
valtr = valtr or com.translit_no_links(valru)
tr = rsub(tr, "^%+", valtr)
tr = rsub(tr, "(%s)%+", "%1" .. valtr)
tr = tr .. dattr_notes
end
table.insert(new_vals, {ru, tr})
end
else
table.insert(new_vals, rutr)
end
end
vals = new_vals
end
-- auto-accent/check for necessary accents
local newvals = {}
for _, v in ipairs(vals) do
local ru, tr = v[1], v[2]
if not args.allow_unaccented then
if tr and com.is_unstressed(ru) ~= com.is_unstressed(tr) then
error("Override " .. ru .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(ru) then
ru, tr = com.make_ending_stressed(ru, tr)
elseif com.needs_accents(ru) then
error("Override " .. ru .. " for case " .. case .. n .. " requires an accent")
end
end
table.insert(newvals, {ru, tr})
end
vals = newvals
return vals
end
local function process_overrides(args, f, n)
local function process_override(case)
if args[case .. n] then
local overrides = canonicalize_override(args, case, f, n)
if not f[case] then
f[case] = {}
end
local new_overrides = {}
for _, form in ipairs(overrides) do
-- Don't consider overrides of loc/par/voc irregular since
-- they're only specified through overrides; FIXME: Theoretically
-- we could consider loc/par irregular if they don't follow the
-- expected forms; but we'd have to figure out how to eliminate
-- the preposition that may be specified
if not overridable_only_cases_set[case] and
not args.manual and
not contains_rutr_pair(f[case], form) then
local formru, formnotes = m_table_tools.separate_notes(form[1])
if formru ~= "-" then
-- don't mark an override of - as irregular, even if
-- it has an attached footnote symbol
form = com.concat_paired_russian_tr(form, {IRREGMARKER})
end
end
if case == "pauc" then
-- Internal note indicating that the form is for numbers 2, 3 and 4.
form = com.concat_paired_russian_tr(form, {paucal_marker})
end
table.insert(new_overrides, form)
end
f[case] = new_overrides
args.any_overridden[case] = true
end
end
-- do dative singular first because it will be used by loc/par
process_override("dat_sg")
-- now do the rest
for _, case in ipairs(overridable_cases) do
if case ~= "dat_sg" then
process_override(case)
end
end
-- if the nominative is overridden, use it to set the linked version unless
-- that is also specifically overridden
if args["nom_sg" .. n] and not args["nom_sg_linked" .. n] then
f.nom_sg_linked = f.nom_sg
end
if args["nom_pl" .. n] and not args["nom_pl_linked" .. n] then
f.nom_pl_linked = f.nom_pl
end
-- convert empty lists to nil to facilitate computation of accusative
-- case variants below.
for _, case in ipairs(all_cases) do
if f[case] then
if type(f[case]) ~= "table" then
error("Logic error, args[case] should be nil or table")
end
if #f[case] == 0 then
f[case] = nil
end
end
end
end
local function process_tail_args(args, f, n)
local function append_note_all(case, value)
value = com.split_russian_tr(value, "dopair")
local function append1(case)
if f[case] then
for i=1,#f[case] do
f[case][i] = com.concat_paired_russian_tr(f[case][i], value)
end
end
end
append1(case)
if case == "acc_sg" then
append1("acc_sg_in")
append1("acc_sg_an")
elseif case == "acc_pl" then
append1("acc_pl_in")
append1("acc_pl_an")
end
end
local function append_note_last(case, value, gt_one)
value = com.split_russian_tr(value, "dopair")
local function append1(case)
if f[case] then
local lastarg = #f[case]
if lastarg > (gt_one and 1 or 0) then
f[case][lastarg] = com.concat_paired_russian_tr(f[case][lastarg], value)
end
end
end
append1(case)
if case == "acc_sg" then
append1("acc_sg_in")
append1("acc_sg_an")
elseif case == "acc_pl" then
append1("acc_pl_in")
append1("acc_pl_an")
end
end
local function handle_tail_hyp(suf)
local arg, val
for _, case in ipairs(all_cases) do
arg = args[case .. "_" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args[case .. "_" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
if not rfind(case, "_pl") then
arg = args["sg" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["sg" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
else
arg = args["pl" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["pl" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
end
if not rfind(case, "nom_") and not rfind(case, "acc_") then
arg = args["obl" .. suf .. "all" .. n]
if arg then
append_note_all(case, suf == "hyp" and HYPMARKER or arg)
end
arg = args["obl" .. suf .. n]
if arg then
append_note_last(case, suf == "hyp" and HYPMARKER or arg, ">1")
end
end
end
end
handle_tail_hyp("tail")
handle_tail_hyp("hyp")
end
handle_forms_and_overrides = function(args, n, islast)
local f = args.forms
process_tail_args(args, f, n)
process_overrides(args, f, n)
local an = args.thisa
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). This logic duplicates logic in handle_overall_forms_and_overrides();
-- see long comment there. We need to duplicate the whole logic here to
-- handle words like мать-одиночка, which has an override of acc_sg1.
if not args["acc_sg_an" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_an = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_an = f.acc_sg or an == "i" and f.nom_sg or f.gen_sg or f.acc_sg_an
end
end
if not args["acc_sg_in" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_in = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_in = f.acc_sg or an == "a" and f.gen_sg or f.nom_sg or f.acc_sg_in
end
end
if not args["acc_pl_an" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_an = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_an = f.acc_pl or an== "i" and f.nom_pl or f.gen_pl or f.acc_pl_an
end
end
if not args["acc_pl_in" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_in = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_in = f.acc_pl or an == "a" and f.gen_pl or f.nom_pl or f.acc_pl_in
end
end
f.loc = f.loc or f.pre_sg
f.par = f.par or f.gen_sg
f.voc = f.voc or f.nom_sg
-- Set these in case we have plural only, in which case the
-- singular will also get set to these same values in case we are
-- a plural-only word in a singular-only expression.
f.loc_pl = f.loc_pl or f.pre_pl
f.par_pl = f.par_pl or f.gen_pl
f.voc_pl = f.voc_pl or f.nom_pl
f.count = f.count or f.gen_pl
f.pauc = f.pauc or f.gen_sg
local nu = args.thisn
-- If we have a singular-only, set the plural forms to the singular forms,
-- and vice-versa. This is important so that things work in multi-word
-- expressions that combine different number restrictions (e.g.
-- singular-only with singular/plural or singular-only with plural-only,
-- compare "St. Vincent and the Grenadines" [Сент-Винсент и Гренадины]).
if nu == "s" then
f.nom_pl_linked = f.nom_sg_linked
f.nom_pl = f.nom_sg
f.gen_pl = f.gen_sg
f.dat_pl = f.dat_sg
f.acc_pl = f.acc_sg
f.acc_pl_an = f.acc_sg_an
f.acc_pl_in = f.acc_sg_in
f.ins_pl = f.ins_sg
f.pre_pl = f.pre_sg
f.nom_pl = f.nom_sg
f.loc_pl = f.loc
f.par_pl = f.par
f.voc_pl = f.voc
elseif nu == "p" then
f.nom_sg_linked = f.nom_pl_linked
f.nom_sg = f.nom_pl
f.gen_sg = f.gen_pl
f.dat_sg = f.dat_pl
f.acc_sg = f.acc_pl
f.acc_sg_an = f.acc_pl_an
f.acc_sg_in = f.acc_pl_in
f.ins_sg = f.ins_pl
f.pre_sg = f.pre_pl
f.nom_sg = f.nom_pl
f.loc = f.loc_pl
f.par = f.par_pl
f.voc = f.voc_pl
end
end
handle_overall_forms_and_overrides = function(args)
local overall_forms = {}
for _, case in ipairs(displayable_cases) do
overall_forms[case] = concat_word_forms(args.per_word_info, case)
end
local acc_sg_overridden = args.acc_sg
local acc_pl_overridden = args.acc_pl
process_tail_args(args, overall_forms, "")
process_overrides(args, overall_forms, "")
if case_will_be_displayed(args, "pauc") then
insert_if_not(args.internal_notes, paucal_internal_note)
end
-- if IRREGMARKER is anywhere in text, remove all instances and put
-- at the end before any notes.
local function clean_irreg_marker(case, text)
if rfind(text, IRREGMARKER) then
text = rsub(text, IRREGMARKER, "")
local entry, notes = m_table_tools.separate_notes(text)
if case_will_be_displayed(args, case) then
insert_if_not(args.internal_notes, IRREGMARKER .. " Irregular.")
args.any_irreg = true
args.any_irreg_case[case] = true
end
return entry .. IRREGMARKER .. notes
else
return text
end
end
-- set final args[case] and clean up IRREGMARKER.
for _, case in ipairs(all_cases) do
args[case] = overall_forms[case]
if args[case] then
local cleaned_forms = {}
for _, form in ipairs(args[case]) do
local ru, tr = form[1], form[2]
ru = clean_irreg_marker(case, ru)
if tr then
tr = clean_irreg_marker(case, tr)
end
table.insert(cleaned_forms, {ru, tr})
end
args[case] = cleaned_forms
end
end
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). We need to do this somewhat complicated procedure to get overrides
-- to work correctly, e.g. acc_sg overrides of feminine and neuter nouns
-- (such as мать and полслова) and gen_sg/gen_pl overrides of masculine
-- animate nouns. We also ran into an issue with words like мазло that are
-- neuter-form but can be both masculine animate (in which case the
-- acc sg inherits from the genitive) and neuter animate (in which case the
-- acc sg has the fixed ending -о, same as nominative, despite the animacy).
-- Remember also that the animate/inanimate accusative variants are the ones
-- displayed, not the plain acc_sg/acc_pl ones.
if not args.any_overridden.acc_sg_an then
if acc_sg_overridden then
args.acc_sg_an = args.acc_sg
elseif not args.any_non_nil.acc_sg then
args.acc_sg_an = args.acc_sg or args.a == "i" and args.nom_sg or args.gen_sg or args.acc_sg_an
end
end
if not args.any_overridden.acc_sg_in then
if acc_sg_overridden then
args.acc_sg_in = args.acc_sg
elseif not args.any_non_nil.acc_sg then
args.acc_sg_in = args.acc_sg or args.a == "a" and args.gen_sg or args.nom_sg or args.acc_sg_in
end
end
if not args.any_overridden.acc_pl_an then
if acc_pl_overridden then
args.acc_pl_an = args.acc_pl
elseif not args.any_non_nil.acc_pl then
args.acc_pl_an = args.acc_pl or args.a == "i" and args.nom_pl or args.gen_pl or args.acc_pl_an
end
end
if not args.any_overridden.acc_pl_in then
if acc_pl_overridden then
args.acc_pl_in = args.acc_pl
elseif not args.any_non_nil.acc_pl then
args.acc_pl_in = args.acc_pl or args.a == "a" and args.gen_pl or args.nom_pl or args.acc_pl_in
end
end
-- Try to set the values of acc_sg and acc_pl. The only time we can't is
-- when the noun is bianimate and the anim/inan values are different.
-- This is used primarily for generate_forms() and generate_multi_forms(),
-- since we don't actually display these forms.
if args.a == "a" then
args.acc_sg = args.acc_sg or args.acc_sg_an
args.acc_pl = args.acc_pl or args.acc_pl_an
elseif args.a == "i" then
args.acc_sg = args.acc_sg or args.acc_sg_in
args.acc_pl = args.acc_pl or args.acc_pl_in
else -- bianimate
args.acc_sg = args.acc_sg or m_table.deepEquals(args.acc_sg_in, args.acc_sg_an) and args.acc_sg_in or nil
args.acc_pl = args.acc_pl or m_table.deepEquals(args.acc_pl_in, args.acc_pl_an) and args.acc_pl_in or nil
end
end
-- Subfunction of concat_word_forms(), used to implement recursively
-- generating all combinations of elements from WORD_FORMS (a list, one
-- element per word, of a list of the forms for a word, each of which is a
-- two-element list of {RUSSIAN, TR}) and TRAILING_FORMS (a list of forms, the
-- accumulated suffixes for trailing words so far in the recursion process,
-- again where each form is a two-element list {RUSSIAN, TR}). Each time we
-- recur we take the last FORMS item off of WORD_FORMS and to each form in
-- FORMS we add all elements in TRAILING_FORMS, passing the newly generated
-- list of items down the next recursion level with the shorter WORD_FORMS.
-- We end up returning a list of concatenated forms, where each list item
-- is a two-element list {RUSSIAN, TR}.
local function concat_word_forms_1(word_forms, trailing_forms)
if #word_forms == 0 then
local retforms = {}
for _, form in ipairs(trailing_forms) do
local ru, tr = form[1], form[2]
-- Remove <insa> and <insb> markers; they've served their purpose.
ru = rsub(ru, "<ins[ab]>", "")
tr = tr and rsub(tr, "<ins[ab]>", "")
table.insert(retforms, {ru, tr})
end
return retforms
else
local last_form_info = table.remove(word_forms)
local last_forms, joiner = last_form_info[1], last_form_info[2]
local new_trailing_forms = {}
for _, form in ipairs(last_forms) do
for _, trailing_form in ipairs(trailing_forms) do
-- If form to prepend is empty, don't add the joiner; this
-- is principally used in overall overrides, where we stuff
-- the entire override into the last word
local full_form = form[1] == "" and trailing_form or
com.concat_paired_russian_tr(form,
com.concat_paired_russian_tr(joiner, trailing_form),
"movenotes")
if rfind(full_form[1], "<insa>") and rfind(full_form[1], "<insb>") then
-- REJECT! So we don't get mixtures of the two feminine
-- instrumental singular endings.
else
table.insert(new_trailing_forms, full_form)
end
end
end
return concat_word_forms_1(word_forms, new_trailing_forms)
end
end
-- Generate a list of overall forms by concatenating the per-word forms.
-- PER_WORD_INFO comes from args.per_word_info and is a list of
-- WORD_INFO items, one per word, each of which a two element list of
-- WORD_FORMS (a table listing the forms for each case) and JOINER (a string).
-- We loop over all possible combinations of elements from each word's list
-- of forms for the given case; this requires recursion.
concat_word_forms = function(per_word_info, case)
local word_forms = {}
-- Gather the appropriate word forms. We have to recreate this anew
-- because it will be destructively modified by concat_word_forms_1().
for _, word_info in ipairs(per_word_info) do
table.insert(word_forms, {word_info[1][case], word_info[2]})
end
-- We need to start the recursion with the second parameter containing
-- one blank element rather than no elements, otherwise no elements
-- will be propagated to the next recursion level.
return concat_word_forms_1(word_forms, {{""}})
end
local accel_forms = {
nom_sg = "nom|s",
nom_sg_linked = "nom|s",
nom_pl = "nom|p",
nom_pl_linked = "nom|p",
gen_sg = "gen|s",
gen_pl = "gen|p",
dat_sg = "dat|s",
dat_pl = "dat|p",
acc_sg_an = "an|acc|s",
acc_pl_an = "an|acc|p",
acc_sg_in = "in|acc|s",
acc_pl_in = "in|acc|p",
ins_sg = "ins|s",
ins_pl = "ins|p",
pre_sg = "pre|s",
pre_pl = "pre|p",
loc = "loc|s",
loc_pl = "loc|p",
voc = "voc|s",
voc_pl = "voc|p",
par = "par|s",
par_pl = "par|p",
count = "count form",
pauc = "pau",
}
-- Make the table
make_table = function(args)
local data = {}
data.after_title = " " .. args.heading
data.number = args.nonumber and "" or numbers[args.n]
local lemma_forms = args[args.n == "p" and "nom_pl" or "nom_sg"]
data.lemma = nom.show_form(args.explicit_lemma and {{args.explicit_lemma, args.explicit_lemmatr}} or
args[args.n == "p" and "nom_pl_linked" or "nom_sg_linked"], "lemma", nil, nil)
data.title = args.title or format(args.old and old_title_temp or title_temp, data)
local sg_an_in_equal = m_table.deepEquals(args.acc_sg_an, args.acc_sg_in)
local pl_an_in_equal = m_table.deepEquals(args.acc_pl_an, args.acc_pl_in)
for _, case in ipairs(displayable_cases) do
local accel_form = accel_forms[case]
if not accel_form then
error("Something wrong, can't find accelerator form for " .. case)
end
if (sg_an_in_equal and (case == "acc_sg_an" or case == "acc_sg_in") or
pl_an_in_equal and (case == "acc_pl_an" or case == "acc_pl_in")) then
accel_form = rsub(accel_form, "^[ai]n|", "")
end
if args.n == "p" then
accel_form = rsub(accel_form, "|p$", "")
end
data[case] = nom.show_form(args[case], false, accel_form, lemma_forms, "remove monosyllabic accents only lemma")
end
local temp = nil
if args.n == "s" then
data.nom_x = data.nom_sg
data.gen_x = data.gen_sg
data.dat_x = data.dat_sg
data.acc_x_an = data.acc_sg_an
data.acc_x_in = data.acc_sg_in
data.ins_x = data.ins_sg
data.pre_x = data.pre_sg
if sg_an_in_equal then
temp = "one_number"
else
temp = "one_number_split_animacy"
end
elseif args.n == "p" then
data.nom_x = data.nom_pl
data.gen_x = data.gen_pl
data.dat_x = data.dat_pl
data.acc_x_an = data.acc_pl_an
data.acc_x_in = data.acc_pl_in
data.ins_x = data.ins_pl
data.pre_x = data.pre_pl
data.par = data.par_pl
data.loc = data.loc_pl
data.voc = data.voc_pl
if pl_an_in_equal then
temp = "one_number"
else
temp = "one_number_split_animacy"
end
else
if pl_an_in_equal then
temp = "both_numbers"
elseif sg_an_in_equal then
temp = "both_numbers_split_animacy_plural_only"
else
temp = "both_numbers_split_animacy"
end
end
for _, extra_case in ipairs({
{"par", "partitive"},
{"loc", "locative"},
{"voc", "vocative"},
}) do
local case, engcase, template = unpack(extra_case)
local a = args.a
local colspan = (a == "b" or a == "bi" or a == "both" or a == "ai" or a == "ia") and 2 or 1
if args.n ~= "s" and args.n ~= "p" and args.any_overridden[case .. "_pl"] then
if not args.any_overridden[case] then
data[case] = ""
end
template = extra_case_template_with_plural
elseif args.n ~= "p" and args.any_overridden[case] or args.n == "p" and args.any_overridden[case .. "_pl"] then
template = extra_case_template
end
if template then
template = format(template(colspan), {case=case, engcase=engcase})
data[case .. "_clause"] = format(template, data)
else
data[case .. "_clause"] = ""
end
end
for _, extra_case in ipairs({
{"count", "count form"},
{"pauc", "paucal"},
}) do
local case, engcase, template = unpack(extra_case)
local a = args.a
local colspan = (a == "b" or a == "bi" or a == "both" or a == "ai" or a == "ia") and 2 or 1
if args.n ~= "p" and args.any_overridden[case] then
template = extra_case_template
end
if template then
template = format(template(colspan), {case=case, engcase=engcase})
data[case .. "_clause"] = format(template, data)
else
data[case .. "_clause"] = ""
end
end
local notes = get_arg_chain(args, "notes", "notes")
local all_notes = {}
for _, note in ipairs(args.internal_notes) do
-- Superscript footnote marker at beginning of note, similarly to what's
-- done at end of forms.
local symbol, entry = m_table_tools.get_initial_notes(note)
table.insert(all_notes, symbol .. entry)
end
for _, note in ipairs(notes) do
-- Here too.
local symbol, entry = m_table_tools.get_initial_notes(note)
table.insert(all_notes, symbol .. entry)
end
data.notes = table.concat(all_notes, "<br />")
data.notes_clause = data.notes ~= "" and format(notes_template, data) or ""
return format(templates[temp], data)
end
function extra_case_template(colspan)
colspan = colspan or 1
return ([===[
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=%s | {engcase}
| {\op}{case}{\cl}
|
|-]===]):format(colspan)
end
function extra_case_template_with_plural(colspan)
colspan = colspan or 1
return ([===[
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=%s | {engcase}
| {\op}{case}{\cl}
| {\op}{case}_pl{\cl}
|-]===]):format(colspan)
end
notes_template = [===[
<div style="width:100%;text-align:left;background:var(--wikt-palette-lightblue, #d9ebff);">
<div style="display:inline-block;text-align:left;padding-left:1em;padding-right:1em">
{notes}
</div></div>
]===]
local function template_prelude(min_width)
min_width = min_width or "70"
return rsub([===[
<div>
<div class="NavFrame" data-toggle-category="declension" style="max-width:MINWIDTHem">
<div class="NavHead" style="background:var(--wikt-palette-lighterblue, #ebf4ff);">{title}<span style="font-weight:normal;">{after_title}</span> </div>
<div class="NavContent">
{\op}| style="table-layout:fixed; text-align:center; max-width:MINWIDTHem; width:100%;" class="inflection inflection-table"
|- class="rowgroup"
]===], "MINWIDTH", min_width)
end
local function template_postlude()
return [===[|-{par_clause}{loc_clause}{voc_clause}{count_clause}{pauc_clause}
|{\cl}{notes_clause}</div></div></div>]===]
end
templates["both_numbers"] = template_prelude("45") .. [===[
! style="width:7em;background:var(--wikt-palette-lightblue, #d9ebff);" |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | accusative
| data-accel-col=1 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["both_numbers_split_animacy"] = template_prelude("50") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| data-accel-col=1 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| data-accel-col=1 | {acc_sg_in}
| data-accel-col=2 | {acc_pl_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["both_numbers_split_animacy_plural_only"] = template_prelude("50") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | singular
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | plural
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| data-accel-col=1 | {nom_sg}
| data-accel-col=2 | {nom_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| data-accel-col=1 | {gen_sg}
| data-accel-col=2 | {gen_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| data-accel-col=1 | {dat_sg}
| data-accel-col=2 | {dat_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| data-accel-col=1 rowspan=2 | {acc_sg_an}
| data-accel-col=2 | {acc_pl_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| data-accel-col=2 | {acc_pl_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| data-accel-col=1 | {ins_sg}
| data-accel-col=2 | {ins_pl}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| data-accel-col=1 | {pre_sg}
| data-accel-col=2 | {pre_pl}
]===] .. template_postlude()
templates["one_number"] = template_prelude("30") .. [===[
! style="width:7em;background:var(--wikt-palette-lightblue, #d9ebff);" |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | {number}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | nominative
| {nom_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | genitive
| {gen_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | dative
| {dat_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | accusative
| {acc_x_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | instrumental
| {ins_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | prepositional
| {pre_x}
]===] .. template_postlude()
templates["one_number_split_animacy"] = template_prelude("35") .. [===[
! style="width:15em;background:var(--wikt-palette-lightblue, #d9ebff);" colspan=2 |
! style="background:var(--wikt-palette-lightblue, #d9ebff);" | {number}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | nominative
| {nom_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | genitive
| {gen_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | dative
| {dat_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | animate
| {acc_x_an}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" | inanimate
| {acc_x_in}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | instrumental
| {ins_x}
|-
! style="background:var(--wikt-palette-lighterblue, #ebf4ff);" colspan=2 | prepositional
| {pre_x}
]===] .. template_postlude()
return export
l4f28v6akyf5m5p3slq8zohc8q0rf6j
Modül:ru-genel
828
1589236
5669922
2026-06-23T07:23:41Z
MustafaCavlak
59368
imported from en
5669922
Scribunto
text/plain
local export = {}
--[==[ intro:
Author: Benwing; some very early work by Rua and Atitarev
This module holds some commonly used functions for the Russian language. It's generally for use from other modules, not
`#invoke`, although some functions can be invoked from a template (`export.iotation()`, `export.reduce_stem()`,
`export.dereduce_stem()`, actually added to support calling from a bot script rather than from a user template; and
`export.make_unstressed()`, called from {{tl|R:ru:Vasmer}}). There may be issues when invoking such functions from
templates when transliteration is present, due to the need for the transliteration to be decomposed, as mentioned below
(all strings from Wiktionary pages are normally in composed form).
NOTE NOTE NOTE: All functions assume that transliteration (but not Russian) has had its acute and grave accents
decomposed using `export.decompose()`. This is the first thing that should be done to all user-specified
transliteration and any transliteration we compute that we expect to work with.
]==]
local m_str_utils = require("Module:string utilities")
local m_table_tools = require("Module:table tools")
-- Prevents an infinite require loop since ru-translit requires a different function in this module.
local m_ru_translit = require("Module:require when needed")("Module:ru-translit")
local m_headword_utilities = require("Module:require when needed")("Module:headword utilities")
local m_table = require("Module:require when needed")("Module:table")
local gsplit = m_str_utils.gsplit
local split = m_str_utils.split
local toNFD = mw.ustring.toNFD
local u = m_str_utils.char
local ufind = mw.ustring.find
local ugsub = mw.ustring.gsub
local ulen = m_str_utils.len
local ulower = m_str_utils.lower
local umatch = mw.ustring.match
local unpack = unpack or table.unpack -- Lua 5.2 compatibility
local usub = m_str_utils.sub
local insert = table.insert
local concat = table.concat
local AC = u(0x0301) -- acute = ́
local GR = u(0x0300) -- grave = ̀
local CFLEX = u(0x0302) -- circumflex = ̂
local BREVE = u(0x0306) -- breve ̆
local DIA = u(0x0308) -- diaeresis = ̈
local CARON = u(0x030C) -- caron ̌
local OGONEK = u(0x0328) -- ogonek ̨
local PSEUDOVOWEL = u(0xFFF1) -- pseudovowel placeholder
local PSEUDOCONS = u(0xFFF2) -- pseudoconsonant placeholder
-- any stress
export.stress = AC .. GR
-- regex for any optional stress(s)
export.opt_stress = "[" .. export.stress .. "]*"
-- any accent
export.accent = export.stress .. DIA .. BREVE .. CARON .. OGONEK
-- regex for any optional accent(s)
export.opt_accent = "[" .. export.accent .. "]*"
-- any composed Cyrillic vowel with grave accent
export.composed_grave_vowel = "ѐЀѝЍ"
-- any Cyrillic vowel except ёЁ
export.vowel_no_jo = "аАеЕиИӥӤіІїЇоОуУыЫѣѢэЭюЮяЯѵѴ" .. PSEUDOVOWEL .. export.composed_grave_vowel
-- any Cyrillic vowel, including ёЁ
export.vowel = export.vowel_no_jo .. "ёЁ"
-- any vowel in transliteration
export.tr_vowel = "aAeEěĚɛƐiIoOǒǑǫǪuUyY" .. PSEUDOVOWEL
-- any consonant in transliteration, omitting "j" and the soft/hard sign
export.tr_cons_no_approx = "bBcCčČdDfFgGkKlLmMnNpPrRsSšŠtTvVxXzZžŽ" .. PSEUDOCONS
-- any consonant in transliteration, including soft/hard sign
export.tr_cons = export.tr_cons_no_approx .. "jJʹʺ"
-- regex for any consonant in transliteration, including soft/hard sign,
-- optionally followed by any accent
export.tr_cons_acc_re = "[" .. export.tr_cons .. "]" .. export.opt_accent
-- any Cyrillic consonant except sibilants and ц
export.cons_except_sib_c = "бБвВгГдДзЗйЙкКлЛмМнНпПрРсСтТфФхХъЪьЬѳѲ" .. PSEUDOCONS
-- Cyrillic sibilant consonants
export.sib = "жЖчЧшШщЩ"
-- Cyrillic sibilant consonants and ц
export.sib_c = export.sib .. "цЦ"
-- any Cyrillic consonant
export.cons = export.cons_except_sib_c .. export.sib_c
-- Cyrillic velar consonants
export.velar = "гГкКхХ"
-- uppercase Cyrillic
export.uppercase = "АБВГДЕЀЁЖЗИЍӤІЇЙКЛМНОПРСТУФХЦЧШЩЪЫЬѢЭЮЯѲѴ"
local function ine(x)
return x ~= "" and x or nil
end
--[==[
Apply Proto-Slavic iotation. This is the change that is affected by a Slavic -j- after a consonant.
`stem` is the Cyrillic stem and `tr` its optional manual translit. If `shch` is given, т iotates to щ instead of ч.
Normally return two values, the iotated Cyrillic stem and corresponding transliteration (which will be {nil} if passed
in as {nil}). If invoked from a template or bot, however, the return value will be a single string; if manual
transliteration is present, the return value will be a string of the form `CYRILLIC//TRANSLIT` with two slahes
separating the Cyrillic from the translit.
]==]
function export.iotation(stem, tr, shch)
local combine_tr = false
-- so this can be called from a template (or usually, a bot)
if type(stem) == "table" then
stem, tr, shch = ine(stem.args[1]), ine(stem.args[2]), ine(stem.args[3])
combine_tr = true
end
stem = ugsub(stem, "[сх]$", "ш")
stem = ugsub(stem, "с[кт]$", "щ")
stem = ugsub(stem, "[кц]$", "ч")
-- normally "т" is iotated as "ч" but there are many verbs that are iotated with "щ"
if shch == "щ" then
stem = stem:gsub("т$", "щ")
else
stem = stem:gsub("т$", "ч")
end
stem = ugsub(stem, "[гдз]$", "ж")
stem = ugsub(stem, "[бвмпфѳ]$", "%0л")
-- ѵ is equivalent to в after certain vowels
stem = ugsub(stem, "[аАеЕѐЀиИѝЍӥӤіІїЇѣѢэЭяЯ]" .. export.opt_stress .. "ѵ$", "%0л")
if tr then
tr = tr:gsub("[sx]$", "š")
:gsub("s[kt]$", "šč")
:gsub("[kc]$", "č")
-- normally "т" is iotated as "ч" but there are many verbs that are iotated with "щ"
if shch == "щ" then
tr = tr:gsub("t$", "šč")
else
tr = tr:gsub("t$", "č")
end
tr = tr:gsub("[dgz]$", "ž")
:gsub("[bfmpv]$", "%0l")
end
if combine_tr then
return export.combine_russian_tr(stem, tr)
end
return stem, tr
end
do
-- A word needs accents if it is unstressed and contains more than one
-- vowel, unless it's a prefix or suffix
local function word_needs_accents(word)
return not (word:sub(1, 1) == "-" or word:sub(-1) == "-") and
export.is_unstressed(word) and not export.is_monosyllabic(word)
end
-- Does a set of Cyrillic words in connected text need accents? We need to
-- split by word and check each one.
function export.needs_accents(text)
for word in gsplit(text, "%s+") do
if word_needs_accents(word) then
return true
end
end
return false
end
end
--[==[
True if either:
# A vowel is marked with explicit primary stress.
# The word has a single jo in it (which doesn't have secondary stress).
This is because a word with multiple jos requires explicit stress to be marked, since we can't infer where primary
stress lies. `Jo` can be `ё`, `ѣ̈`, or `я̈` (e.g. `сѣ̈дла`, plural of `сѣдло́`).
]==]
function export.is_stressed(word)
if word:find(AC) then
return true
end
word = toNFD(word)
return (
select(2, ugsub(word, "[еѣяЕѢЯ]" .. DIA, "")) == 1 and
not umatch(word, "[еѣяЕѢЯ]" .. DIA .. GR)
)
end
--[==[
True if a Cyrillic word requires explicit stress.
]==]
function export.is_unstressed(word)
return not export.is_stressed(word)
end
--[==[
True if Cyrillic word is stressed on the last syllable.
]==]
function export.is_ending_stressed(word)
return ufind(word, "[ёЁ][^" .. export.vowel .. "]*$") or
ufind(word, "[" .. export.vowel .. "][́̈][^" .. export.vowel .. "]*$")
end
do
local function word_is_multi_stressed(word)
-- Allow ё along with another stress, e.g. [[трёхэта́жный]], [[гёрлфре́нд]] or [[Са́мсё]].
return ufind(word, "[" .. export.vowel .. "][́̈].*[" .. export.vowel .. "][́̈]")
end
--[==[
True if a Cyrillic term has two or more stresses (acute or diaeresis) in a given word.
]==]
function export.is_multi_stressed(text)
for word in gsplit(text, "[%s-]+") do
if word_is_multi_stressed(word) then
return true
end
end
return false
end
end
--[==[
True if Cyrillic word is stressed on the first syllable.
]==]
function export.is_beginning_stressed(word)
return ufind(word, "^[^" .. export.vowel .. "]*[ёЁ]") or
ufind(word, "^[^" .. export.vowel .. "]*[" .. export.vowel .. "]́")
end
--[==[
True if Cyrillic word has no vowel. Don't treat suffixes as nonsyllabic even if they have no vowel, as they are
generally added onto words with vowels.
]==]
function export.is_nonsyllabic(word)
return not ufind(word, "^%-") and not ufind(word, "[" .. export.vowel .. "]")
end
--[==[
True if Cyrillic word has no more than one vowel; includes non-syllabic stems such as `льд-`.
]==]
function export.is_monosyllabic(word)
return not ufind(word, "[" .. export.vowel .. "].*[" .. export.vowel .. "]")
end
local recomposer = {
-- Cyrillic letters
["е" .. DIA] = "ё",
["Е" .. DIA] = "Ё",
["и" .. DIA] = "ӥ",
["И" .. DIA] = "Ӥ",
["и" .. BREVE] = "й",
["И" .. BREVE] = "Й",
["і" .. DIA] = "ї",
["І" .. DIA] = "Ї",
-- Latin letters
["c" .. CARON] = "č",
["C" .. CARON] = "Č",
["e" .. CARON] = "ě",
["E" .. CARON] = "Ě",
["o" .. CARON] = "ǒ",
["O" .. CARON] = "Ǒ",
["o" .. OGONEK] = "ǫ",
["O" .. OGONEK] = "Ǫ",
["s" .. CARON] = "š",
["S" .. CARON] = "Š",
["z" .. CARON] = "ž",
["Z" .. CARON] = "Ž",
-- used in ru-pron:
["ж" .. BREVE] = "ӂ", -- used in ru-pron
["Ж" .. BREVE] = "Ӂ",
["j" .. CFLEX] = "ĵ",
["J" .. CFLEX] = "Ĵ",
["j" .. CARON] = "ǰ",
-- no composed uppercase equivalent of J-caron
["ʒ" .. CARON] = "ǯ",
["Ʒ" .. CARON] = "Ǯ",
}
--[==[
Decompose acute, grave, etc. on letters (esp. Latin) into individivual character + combining accent. But recompose
Cyrillic and Latin characters that we want to treat as units and get caught in the crossfire. We mostly want acute and
grave decomposed; perhaps should just explicitly decompose those and no others.
]==]
function export.decompose(text)
return (ugsub(toNFD(text), ".[" .. BREVE .. DIA .. CARON .. OGONEK .. "]", recomposer))
end
function export.assert_decomposed(text)
assert(not ufind(text, "[áéíóúýàèìòùỳäëïöüÿÁÉÍÓÚÝÀÈÌÒÙỲÄËÏÖÜŸ]"))
end
--[==[
Transliterate text and then apply acute/grave decomposition.
]==]
function export.translit(text, no_include_monosyllabic_jo_accent)
local jo_accent = not no_include_monosyllabic_jo_accent and "mono" or nil
return export.decompose(m_ru_translit.tr(text, nil, nil, jo_accent))
end
--[==[
Recompose acutes and graves into preceding vowels. Probably not necessary.
]==]
function export.recompose(text)
return mw.ustring.toNFC(text)
end
local grave_decomposer = {
["ѐ"] = "е" .. GR,
["Ѐ"] = "Е" .. GR,
["ѝ"] = "и" .. GR,
["Ѝ"] = "И" .. GR,
}
--[==[
Decompose precomposed Cyrillic chars w/grave accent; not necessary for acute accent as there aren't precomposed Cyrillic
chars w/acute accent, and undesirable for precomposed `ё` and `Ё`.
]==]
function export.decompose_grave(word)
return (ugsub(word, "[ѐЀѝЍ]", grave_decomposer))
end
local grave_deaccenter = {
[GR] = "", -- grave accent
["ѐ"] = "е", -- composed Cyrillic chars w/grave accent
["Ѐ"] = "Е",
["ѝ"] = "и",
["Ѝ"] = "И",
}
local deaccenter = mw.clone(grave_deaccenter)
deaccenter[AC] = "" -- acute accent
--[==[
Remove acute and grave accents; don't affect composed diaeresis in `ёЁ` or uncomposed diaeresis in `-ѣ̈-` (as in plural
`сѣ̈дла` of `сѣдло́`). NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.remove_accents(word, tr)
local ru_removed = ugsub(word, "[́̀ѐЀѝЍ]", deaccenter)
if not tr then
return ru_removed, nil
end
return ru_removed, (ugsub(tr, "[" .. AC .. GR .. "]", deaccenter))
end
--[==[
Remove grave accents; don't affect acute or composed diaeresis in `ёЁ` or uncomposed diaeresis in `-ѣ̈-` (as in plural
`сѣ̈дла` of `сѣдло́`). NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.remove_grave_accents(word, tr)
local ru_removed = ugsub(word, "[̀ѐЀѝЍ]", grave_deaccenter)
if not tr then
return ru_removed, nil
end
return ru_removed, (ugsub(tr, GR, ""))
end
--[==[
Remove acute and grave accents in monosyllabic words; don't affect diaeresis (composed or uncomposed) because it
indicates a change in vowel quality, which still applies to monosyllabic words. Don't change suffixes, where a
"monosyllabic" stress is still significant (e.g. `-ча́т` short masculine of `-ча́тый`, vs. `-́чат` short masculine of
`-́чатый`). NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.remove_monosyllabic_accents(word, tr)
if export.is_monosyllabic(word) and not ufind(word, "^%-") then
return export.remove_accents(word, tr)
else
return word, tr
end
end
local destresser = mw.clone(deaccenter)
destresser["ё"] = "е"
destresser["Ё"] = "Е"
destresser[DIA] = "" -- diaeresis
-- Subfunction of split_syllables(). On input we get sections of text consisting of CONSONANT - VOWEL - CONSONANT -
-- VOWEL ... - CONSONANT, where CONSONANT consists of zero or more consonants and VOWEL consists of exactly one vowel
-- plus any following accent(s); we combine these into syllables as required by split_syllables().
local function combine_captures(captures)
if captures[1] and not captures[2] then
return captures
end
local combined = {}
for i = 1, (#captures - 1), 2 do
insert(combined, captures[i] .. captures[i+1])
end
combined[#combined] = combined[#combined] .. captures[#captures]
return combined
end
--[==[
Split Russian text and transliteration into syllables. Syllables end with vowel + accent(s), except for the last
syllable, which includes any trailing consonants. NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.split_syllables(ru, tr)
-- Split into alternating consonant/vowel sequences, as described in
-- combine_captures().
local rusyllables = combine_captures(split(ru, "([" .. export.vowel .. "]" .. export.opt_accent .. ")"))
local trsyllables
if tr then
export.assert_decomposed(tr)
trsyllables = combine_captures(split(tr, "([" .. export.tr_vowel .. "]" .. export.opt_accent .. ")"))
if #rusyllables ~= #trsyllables then
error("Russian " .. ru .. " doesn't have same number of syllables as translit " .. tr)
end
end
--error(concat(rusyllables, "/") .. "(" .. #rusyllables .. (trsyllables and (") || " .. concat(trsyllables, "/") .. "(" .. #trsyllables .. ")") or ""))
return rusyllables, trsyllables
end
--[==[
Split Russian word and transliteration into hyphen-separated components. Rejoining with `concat(..., "-")` will
recover the original word. If the original word ends in a hyphen, that hyphen gets included with the preceding component(this is the only case when an individual component has a hyphen in it).
]==]
function export.split_hyphens(ru, tr)
local rucomponents = split(ru, "%-")
if rucomponents[#rucomponents] == "" and #rucomponents > 1 then
rucomponents[#rucomponents - 1] = rucomponents[#rucomponents - 1] .. "-"
table.remove(rucomponents)
end
local trcomponents
if tr then
trcomponents = split(tr, "%-")
if trcomponents[#trcomponents] == "" and #trcomponents > 1 then
trcomponents[#trcomponents - 1] = trcomponents[#trcomponents - 1] .. "-"
table.remove(trcomponents)
end
if #rucomponents ~= #trcomponents then
error("Russian " .. ru .. " doesn't have same number of hyphenated components as translit " .. tr)
end
end
return rucomponents, trcomponents
end
--[==[
Apply `j` correction, converting `je` to `e` after consonants, `jo` to `o` after a sibilant, `ju` to `u` after hard
sibilant. NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.j_correction(tr)
tr = ugsub(tr, "([" .. export.tr_cons_no_approx .. "]" .. export.opt_accent ..")[Jj]([EeĚě])", "%1%2")
tr = ugsub(tr, "([žščŽŠČ])[Jj]([Oo])", "%1%2")
return (ugsub(tr, "([žšŽŠ])[Jj]([Uu])", "%1%2"))
end
local function make_unstressed_ru(ru)
-- The following regexp has grave+acute+diaeresis after the bracket
return (ugsub(ru, "[̀́̈ёЁѐЀѝЍ]", destresser))
end
--[==[
Remove all stress marks (acute, grave, diaeresis). NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.make_unstressed(ru, tr)
-- so this can be called from a template (specifically {{R:ru:Vasmer}})
if type(ru) == "table" then
ru, tr = ine(ru.args[1]), ine(ru.args[2])
end
if not tr then
return make_unstressed_ru(ru), nil
end
-- In the presence of TR, we need to do things the hard way: Splitting
-- into syllables and only converting Latin o to e opposite a ё.
local rusyl, trsyl = export.split_syllables(ru, tr)
for i=1,#rusyl do
if ufind(rusyl[i], "[ёЁ]") then
trsyl[i] = ugsub(trsyl[i], "[Oo]", {["O"] = "E", ["o"] = "e"})
end
rusyl[i] = make_unstressed_ru(rusyl[i])
-- the following should still work as it will affect accents only
trsyl[i] = make_unstressed_ru(trsyl[i])
end
-- Also need to apply j correction as otherwise we'll have je after cons, etc.
return concat(rusyl, ""), export.j_correction(concat(trsyl, ""))
end
local function remove_jo_ru(word)
return (ugsub(word, "[̈ёЁ]", destresser))
end
local function make_unstressed_once_ru(word)
-- leave graves alone
return (ugsub(word, "([́̈ёЁ])([^́̈ёЁ]*)$", function(x, rest) return destresser[x] .. rest; end, 1))
end
--[==[
Remove diaeresis stress marks only. NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.remove_jo(ru, tr)
if not tr then
return remove_jo_ru(ru), nil
end
-- In the presence of TR, we need to do things the hard way: Splitting
-- into syllables and only converting Latin o to e opposite a ё.
local rusyl, trsyl = export.split_syllables(ru, tr)
for i=1,#rusyl do
if ufind(rusyl[i], "[ёЁ]") then
trsyl[i] = ugsub(trsyl[i], "[Oo]", {["O"] = "E", ["o"] = "e"})
end
rusyl[i] = remove_jo_ru(rusyl[i])
-- the following should still work as it will affect accents only
trsyl[i] = make_unstressed_once_ru(trsyl[i])
end
-- Also need to apply j correction as otherwise we'll have je after cons, etc.
return concat(rusyl, ""),
export.j_correction(concat(trsyl, ""))
end
local function map_last_hyphenated_component(fn, ru, tr)
if ufind(ru, "%-") then
-- If there is a hyphen, do it the hard way by splitting into
-- individual components and doing the last one. Otherwise we just do
-- the whole string.
local rucomponents, trcomponents = export.split_hyphens(ru, tr)
local lastru, lasttr = fn(rucomponents[#rucomponents],
trcomponents and trcomponents[#trcomponents] or nil)
rucomponents[#rucomponents] = lastru
ru = concat(rucomponents, "-")
if trcomponents then
trcomponents[#trcomponents] = lasttr
tr = concat(trcomponents, "-")
end
return ru, tr
end
return fn(ru, tr)
end
-- Make last stressed syllable (acute or diaeresis) unstressed; leave unstressed; leave graves alone; if `NOCONCAT`,
-- return individual syllables. NOTE: Translit must already be decomposed! See comment at top.
local function make_unstressed_once_after_hyphen_split(ru, tr, noconcat)
if not tr then
return make_unstressed_once_ru(ru), nil
end
-- In the presence of TR, we need to do things the hard way, as with
-- make_unstressed().
local rusyl, trsyl = export.split_syllables(ru, tr)
for i=#rusyl,1,-1 do
local stressed = export.is_stressed(rusyl[i])
if stressed then
if ufind(rusyl[i], "[ёЁ]") then
trsyl[i] = ugsub(trsyl[i], "[Oo]", {["O"] = "E", ["o"] = "e"})
end
rusyl[i] = make_unstressed_once_ru(rusyl[i])
-- the following should still work as it will affect accents only
trsyl[i] = make_unstressed_once_ru(trsyl[i])
break
end
end
if noconcat then
return rusyl, trsyl
end
-- Also need to apply j correction as otherwise we'll have je after cons
return concat(rusyl, ""),
export.j_correction(concat(trsyl, ""))
end
--[==[
Make last stressed syllable (acute or diaeresis) to the right of any hyphen unstressed (unless the hyphen is
word-final); leave graves alone. We don't destress a syllable to the left of a hyphen unless the hyphen is word-final
(i.e. a prefix). Otherwise e.g. the accents in the first part of words like ко́е-како́й and а́льфа-лу́ч won't remain. NOTE:
Translit must already be decomposed! See comment at top.
]==]
function export.make_unstressed_once(ru, tr)
return map_last_hyphenated_component(make_unstressed_once_after_hyphen_split, ru, tr)
end
local function make_unstressed_once_at_beginning_ru(word)
-- leave graves alone
return (ugsub(word, "^([^́̈ёЁ]*)([́̈ёЁ])", function(rest, x) return rest .. destresser[x]; end, 1))
end
--[==[
Make first stressed syllable (acute or diaeresis) unstressed; leave graves alone; if `NOCONCAT`, return individual
syllables. NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.make_unstressed_once_at_beginning(ru, tr, noconcat)
if not tr then
return make_unstressed_once_at_beginning_ru(ru), nil
end
-- In the presence of TR, we need to do things the hard way, as with
-- make_unstressed().
local rusyl, trsyl = export.split_syllables(ru, tr)
for i=1,#rusyl do
local stressed = export.is_stressed(rusyl[i])
if stressed then
if ufind(rusyl[i], "[ёЁ]") then
trsyl[i] = ugsub(trsyl[i], "[Oo]", {["O"] = "E", ["o"] = "e"})
end
rusyl[i] = make_unstressed_once_at_beginning_ru(rusyl[i])
-- the following should still work as it will affect accents only
trsyl[i] = make_unstressed_once_at_beginning_ru(trsyl[i])
break
end
end
if noconcat then
return rusyl, trsyl
end
-- Also need to apply j correction as otherwise we'll have je after cons
return concat(rusyl, ""),
export.j_correction(concat(trsyl, ""))
end
--[==[
Subfunction of `make_ending_stressed()`, `make_beginning_stressed()`, which add an acute accent to a syllable that may
already have a grave accent; in such a case, remove the grave. NOTE: Translit must already be decomposed! See comment at
top.
]==]
function export.correct_grave_acute_clash(word, tr)
word = ugsub(word, "([̀ѐЀѝЍ])́", function(x) return grave_deaccenter[x] .. AC; end)
word = ugsub(word, AC .. GR, AC)
if not tr then
return word, nil
end
tr = ugsub(tr, GR .. AC, AC)
tr = ugsub(tr, AC .. GR, AC)
return word, tr
end
local function make_ending_stressed_ru(word)
-- If already ending stressed, just return word so we don't mess up ё
if export.is_ending_stressed(word) then
return word
end
-- Destress the last stressed syllable
word = make_unstressed_once_ru(word)
-- Add an acute to the last syllable
word = ugsub(word, "([" .. export.vowel_no_jo .. "])([^" .. export.vowel .. "]*)$",
"%1́%2")
-- If that caused an acute and grave next to each other, remove the grave
return export.correct_grave_acute_clash(word)
end
-- Remove the last primary stress from the word and put it on the final syllable. Leave grave accents alone except in
-- the last syllable. If final syllable already has primary stress, do nothing. NOTE: Translit must already be
-- decomposed! See comment at top.
local function make_ending_stressed_after_hyphen_split(ru, tr)
if not tr then
return make_ending_stressed_ru(ru), nil
end
-- If already ending stressed, just return ru/tr so we don't mess up ё
if export.is_ending_stressed(ru) then
return ru, tr
end
-- Destress the last stressed syllable; pass in "noconcat" so we get
-- the individual syllables back
local rusyl, trsyl = make_unstressed_once_after_hyphen_split(ru, tr, "noconcat")
-- Add an acute to the last syllable of both Russian and translit
rusyl[#rusyl] = ugsub(rusyl[#rusyl], "([" .. export.vowel_no_jo .. "])",
"%1" .. AC)
trsyl[#trsyl] = ugsub(trsyl[#trsyl], "([" .. export.tr_vowel .. "])",
"%1" .. AC)
-- If that caused an acute and grave next to each other, remove the grave
rusyl[#rusyl], trsyl[#trsyl] =
export.correct_grave_acute_clash(rusyl[#rusyl], trsyl[#trsyl])
-- j correction didn't get applied in make_unstressed_once because
-- we short-circuited it and made it return lists of syllables
return concat(rusyl, ""),
export.j_correction(concat(trsyl, ""))
end
--[==[
Remove the last primary stress from the portion of the word to the right of any hyphen (unless the hyphen is word-final)
and put it on the final syllable. Leave grave accents alone except in the last syllable. If final syllable already has
primary stress, do nothing. (See `make_unstressed_once()` for why we don't affect stresses to the left of a hyphen.)
NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.make_ending_stressed(ru, tr)
return map_last_hyphenated_component(make_ending_stressed_after_hyphen_split, ru, tr)
end
local function make_beginning_stressed_ru(word)
-- If already beginning stressed, just return word so we don't mess up ё
if export.is_beginning_stressed(word) then
return word
end
-- Destress the first stressed syllable
word = make_unstressed_once_at_beginning_ru(word)
-- Add an acute to the first syllable
word = ugsub(word, "^([^" .. export.vowel .. "]*)([" .. export.vowel_no_jo .. "])",
"%1%2́")
-- If that caused an acute and grave next to each other, remove the grave
return export.correct_grave_acute_clash(word)
end
--[==[
Remove the first primary stress from the word and put it on the initial syllable. Leave grave accents alone except in
the first syllable. If initial syllable already has primary stress, do nothing. NOTE: Translit must already be
decomposed! See comment at top.
]==]
function export.make_beginning_stressed(ru, tr)
if not tr then
return make_beginning_stressed_ru(ru), nil
end
-- If already beginning stressed, just return ru/tr so we don't mess up ё
if export.is_beginning_stressed(ru) then
return ru, tr
end
-- Destress the first stressed syllable; pass in "noconcat" so we get
-- the individual syllables back
local rusyl, trsyl = export.make_unstressed_once_at_beginning(ru, tr, "noconcat")
-- Add an acute to the first syllable of both Russian and translit
rusyl[1] = ugsub(rusyl[1], "([" .. export.vowel_no_jo .. "])",
"%1" .. AC)
trsyl[1] = ugsub(trsyl[1], "([" .. export.tr_vowel .. "])",
"%1" .. AC)
-- If that caused an acute and grave next to each other, remove the grave
rusyl[1], trsyl[1] = export.correct_grave_acute_clash(rusyl[1], trsyl[1])
-- j correction didn't get applied in make_unstressed_once_at_beginning
-- because we short-circuited it and made it return lists of syllables
return concat(rusyl, ""),
export.j_correction(concat(trsyl, ""))
end
-- used for tracking and categorization
local hard_cons = {"hard-cons", "cons"}
local hard_vowel = {"vowel", "hard-vowel"}
local i_vowel = {"i", "vowel", "soft-vowel"}
local sibilant_cons = {"sibilant", "cons"}
local soft_vowel = {"vowel", "soft-vowel"}
local velar_cons = {"velar", "cons"}
local trailing_letter_type = {
["ш"] = sibilant_cons,
["щ"] = sibilant_cons,
["ч"] = sibilant_cons,
["ж"] = sibilant_cons,
["ц"] = {"c", "cons"},
["к"] = velar_cons,
["г"] = velar_cons,
["х"] = velar_cons,
["ь"] = {"soft-cons", "cons"},
["ъ"] = hard_cons,
["й"] = {"palatal", "cons"},
["а"] = hard_vowel,
["я"] = soft_vowel,
["э"] = hard_vowel,
["е"] = soft_vowel,
["ѐ"] = soft_vowel,
["ѣ"] = soft_vowel,
["и"] = i_vowel,
["ѝ"] = i_vowel,
["ӥ"] = i_vowel,
["і"] = i_vowel,
["ї"] = i_vowel,
["ѵ"] = i_vowel,
["ы"] = hard_vowel,
["о"] = hard_vowel,
["ё"] = soft_vowel,
["у"] = hard_vowel,
["ю"] = soft_vowel,
}
function export.get_stem_trailing_letter_type(stem)
return trailing_letter_type[ulower(usub(export.remove_accents(stem), -1))] or hard_cons
end
--[==[
Reduce stem by eliminating the "epenthetic" vowel. Applies to nominative singular masculine 2nd-declension hard and
soft, and 3rd-declension feminine in -ь (e.g. [[любовь]]). `STEM` should be the result after calling
`detect_stem_type()`, but with final -й if present. Normally returns two arguments (`STEM` and `TR`), but can be called
from a template or bot using `#invoke` and will return one argument (`STEM`, or `STEM//TR` if `TR` is present). Returns
{nil} if unable to reduce. NOTE: Translit must already be decomposed! See comment at top.
]==]
function export.reduce_stem(stem, tr, soft_n)
local pre, letter, post
local pretr, lettertr, posttr
local combine_tr = false
-- test cases with translit:
-- =p.reduce_stem("фе́ез", "fɛ́jez") -> фе́йз, fɛ́jz
-- =p.reduce_stem("фе́йез", "fɛ́jez") -> фе́йз, fɛ́jz
-- =p.reduce_stem("фе́без", "fɛ́bez") -> фе́бз, fɛ́bz
-- =p.reduce_stem("фе́лез", "fɛ́lez") -> фе́льз, fɛ́lʹz
-- =p.reduce_stem("феёз", p.decompose("fɛjóz")) -> фейз, fɛjz
-- =p.reduce_stem("фейёз", p.decompose("fɛjjóz")) -> фейз, fɛjz
-- =p.reduce_stem("фебёз", p.decompose("fɛbjóz")) -> фебз, fɛbz
-- =p.reduce_stem("фелёз", p.decompose("fɛljóz")) -> фельз, fɛlʹz
-- =p.reduce_stem("фе́бей", "fɛ́bej") -> фе́бь, fɛ́bʹ
-- =p.reduce_stem("фебёй", p.decompose("fɛbjój")) -> фебь, fɛbʹ
-- =p.reduce_stem("фе́ей", "fɛ́jej") -> фе́йй, fɛ́jj
-- =p.reduce_stem("феёй", p.decompose("fɛjój")) -> фейй, fɛjj
-- so this can be called from a template (or usually, a bot)
if type(stem) == "table" then
stem, tr = ine(stem.args[1]), ine(stem.args[2])
combine_tr = true
end
pre, letter, post = umatch(stem, "^(.*)([оОеЕёЁ])́?([" .. export.cons .. "]+)$")
if not pre then
return nil, nil
end
if tr then
-- FIXME, may not be necessary to write the posttr portion as a
-- consonant + zero or more consonant/accent combinations -- when will
-- we ever get an accent after a consonant? That would indicate a
-- failure of the decompose mechanism.
pretr, lettertr, posttr = umatch(tr, "^(.*)([oOeE])́?([" .. export.tr_cons .. "][" .. export.tr_cons .. export.accent .. "]*)$")
if not pretr then
return nil, nil -- should not happen unless tr is really messed up
end
-- Remove any trailing j's on the translit stem which don't have a
-- corresponding й on the Cyrillic stem. This avoids excess j's with
-- cases like индонези́ец//indonɛzíjec, but retains multiple j's where
-- necessary.
pretr = ugsub(pretr, "[jJ]+$", function(m)
return usub(m, 1, ulen(umatch(pre, "[йЙ]*$")))
end)
end
if letter == "О" or letter == "о" then
-- FIXME, what about when the accent is on the removed letter?
if post == "й" or post == "Й" then
-- FIXME, is this correct?
return nil, nil
end
letter = ""
else
local is_upper = ufind(post, "[" .. export.uppercase .. "]")
if ufind(pre, "[" .. export.vowel .. "]́?$") then
letter = is_upper and "Й" or "й"
elseif post == "й" or post == "Й" then
letter = is_upper and "Ь" or "ь"
post = ""
if posttr then
posttr = ""
end
elseif ufind(post, "[" .. export.velar .. "]$") and ufind(pre, "[" .. export.cons_except_sib_c .. "]$") or
ufind(post, "[^йЙ" .. export.velar .. "]$") and ufind(pre, "[лЛ]$") or
soft_n and ufind(pre, "[нН]$") then
letter = is_upper and "Ь" or "ь"
else
letter = ""
end
end
stem = pre .. letter .. post
if tr then
tr = pretr .. export.translit(letter) .. posttr
end
if combine_tr then
return export.combine_russian_tr(stem, tr)
else
return stem, tr
end
end
--[==[
Generate the dereduced stem given `STEM` and `EPENTHETIC_STRESS` (which indicates whether the epenthetic vowel should be
stressed); this is without any terminating non-syllabic ending, which is added if needed by the calling function.
Normally returns two arguments (`STEM` and `TR`), but can be called from a template using `#invoke` (or by bot) and will
return one argument (`STEM`, or `STEM//TR` if `TR` is present). Returns {nil} if unable to dereduce. NOTE: Translit
must already be decomposed! See comment at top.
]==]
function export.dereduce_stem(stem, tr, epenthetic_stress)
local combine_tr = false
-- so this can be called from a template (or usually, a bot)
if type(stem) == "table" then
stem, tr, epenthetic_stress = ine(stem.args[1]), ine(stem.args[2]), ine(stem.args[3])
combine_tr = true
end
if epenthetic_stress then
stem, tr = export.make_unstressed_once(stem, tr)
end
local pre, letter, post
local pretr, lettertr, posttr
-- FIXME!!! Deal with this special case
--if not (z.stem_type == "soft" and _.equals(z.stress_type, {"b", "f"}) -- we should ignore asterix for 2*b and 2*f (so to process it just like 2b or 2f)
-- or _.contains(z.specific, "(2)") and _.equals(z.stem_type, {"velar", "letter-ц", "vowel"})) -- and also the same for (2)-specific and 3,5,6 stem-types
--then
-- I think this corresponds to our -ья and -ье types, which we
-- handle separately
--if z.stem_type == "vowel" then -- 1).
-- if _.equals(z.stress_type, {"b", "c", "e", "f", "f'", "b'" }) then -- gen_pl ending stressed -- TODO: special vars for that
-- z.stems["gen_pl"] = _.replace(z.stems["gen_pl"], "ь$", "е́")
-- else
-- z.stems["gen_pl"] = _.replace(z.stems["gen_pl"], "ь$", "и")
-- end
--end
pre, letter, post = umatch(stem, "^(.*)([" .. export.cons .. "])([" .. export.cons .. "])$")
if tr then
pretr, lettertr, posttr = umatch(tr, "^(.*)(" .. export.tr_cons_acc_re .. ")(" .. export.tr_cons_acc_re .. ")$")
if pre and not pretr then
return nil, nil -- should not happen unless tr is really messed up
end
end
if pre then
local is_upper = ufind(post, "[" .. export.uppercase .. "]")
local epvowel
if ufind(letter, "[ьйЬЙ]") then
letter = ""
lettertr = ""
if ufind(post, "[цЦ]$") or not epenthetic_stress then
epvowel = is_upper and "Е" or "е"
else
epvowel = is_upper and "Ё" or "ё"
end
elseif ufind(letter, "[" .. export.cons_except_sib_c .. "]") and ufind(post, "[" .. export.velar .. "]") or
ufind(letter, "[" .. export.velar .. "]") then
epvowel = is_upper and "О" or "о"
elseif post == "ц" or post == "Ц" then
epvowel = is_upper and "Е" or "е"
elseif epenthetic_stress then
if ufind(letter, "[" .. export.sib .. "]") then
epvowel = is_upper and "О́" or "о́"
else
epvowel = is_upper and "Ё" or "ё"
end
else
epvowel = is_upper and "Е" or "е"
end
assert(epvowel)
stem = pre .. letter .. epvowel .. post
if tr then
tr = pretr .. lettertr .. export.translit(epvowel) .. posttr
tr = export.j_correction(tr)
end
if epenthetic_stress then
stem, tr = export.make_ending_stressed(stem, tr)
end
if combine_tr then
return export.combine_russian_tr(stem, tr)
else
return stem, tr
end
end
return nil, nil
end
--[==[
Parse an entry that potentially has final footnote symbols and initial `*` for a hypothetical entry into initial
symbols, text and final symbols.
]==]
function export.split_symbols(entry, do_subscript)
local prefentry, finalnotes = m_table_tools.separate_notes(entry)
local initnotes, text = umatch(prefentry, "(%*?)(.*)$")
return initnotes, text, finalnotes
end
--------------------------------------------------------------------------
-- Used for manual translit --
--------------------------------------------------------------------------
--[==[
Transliterate text after removing any links. The translit is decomposed in the process.
]==]
function export.translit_no_links(text)
return export.translit(require("Module:links").remove_links(text))
end
--[==[
Split a combined Cyrillic/translit string (either in the form `CYRILLIC//TRANSLIT` or just `CYRILLIC`) into its
components, decompose the translit, and either return a two-element list of `CYRILLIC` and `TRANSLIT` (if `dopair` is
given), or two separate return values (if `dopair` is not specified).
]==]
function export.split_russian_tr(term, dopair)
local ru, tr
if not term:find("//") then
ru = term
else
local splitvals = split(term, "//")
if #splitvals ~= 2 then
error("Must have at most one // in a Russian//translit expr: '" .. term .. "'")
end
ru, tr = splitvals[1], export.decompose(splitvals[2])
end
if dopair then
return {ru, tr}
else
return ru, tr
end
end
--[==[
Combine Cyrillic and translit into a single string separated by `//`. If translit is {nil}, the return value will be
the same as the passed-in Cyrillic. `ru` can be a string (which any manual translit in `tr`) or a two-element table of
`{CYRILLIC, TRANSLIT}`.
]==]
function export.combine_russian_tr(ru, tr)
if type(ru) == "table" then
ru, tr = unpack(ru)
end
if tr then
return ru .. "//" .. tr
else
return ru
end
end
local function concat_maybe_moving_notes(x, y, movenotes)
if movenotes then
local xentry, xnotes = m_table_tools.separate_notes(x)
local yentry, ynotes = m_table_tools.separate_notes(y)
return xentry .. yentry .. xnotes .. ynotes
else
return x .. y
end
end
--[==[
Concatenate two Cyrillic strings `RU1` and `RU2` that may have corresponding manual transliteration `TR1` and `TR2`
(which should be {nil} if there is no manual translit, and otherwise should be decomposed). If `DOPAIR`, return a
two-item list of the combined Cyrillic and manual translit (which will be {nil} if both `TR1` and `TR2` are {nil});
else, return two values, the combined Cyrillic and manual translit. If `MOVENOTES`, extract any footnote symbols at the
end of `RU1` and move them to the end of the concatenated string, before any footnote symbols for `RU2`; same thing goes
for `TR1` and `TR2`.
]==]
function export.concat_russian_tr(ru1, tr1, ru2, tr2, dopair, movenotes)
local ru, tr
if not tr1 and not tr2 then
ru = concat_maybe_moving_notes(ru1, ru2, movenotes)
else
if not tr1 then
tr1 = export.translit_no_links(ru1)
end
if not tr2 then
tr2 = export.translit_no_links(ru2)
end
ru, tr = concat_maybe_moving_notes(ru1, ru2, movenotes), export.j_correction(concat_maybe_moving_notes(tr1, tr2, movenotes))
end
if dopair then
return {ru, tr}
else
return ru, tr
end
end
--[==[
Concatenate two Cyrillic/translit combinations (where each combination is a two-element list of `{CYRILLIC, TRANSLIT}`
where `TRANSLIT` may be nil) by individually concatenating the Cyrillic and translit portions, and return a concatenated
combination as a two-element list. If the manual translit portions of both terms on entry are {nil}, the result will
also have {nil} manual translit. If `MOVENOTES`, extract any footnote symbols at the end of `TERM1` and move them after
the concatenated string and before any footnote symbols at the end of `TERM2`.
]==]
function export.concat_paired_russian_tr(term1, term2, movenotes)
assert(type(term1) == "table")
assert(type(term2) == "table")
local ru1, tr1 = term1[1], term1[2]
local ru2, tr2 = term2[1], term2[2]
return export.concat_russian_tr(ru1, tr1, ru2, tr2, "dopair", movenotes)
end
function export.concat_forms(forms)
local joined_rutr = {}
for _, form in ipairs(forms) do
insert(joined_rutr, export.combine_russian_tr(form))
end
return concat(joined_rutr, ",")
end
--[==[
Given a list of forms, where each form is a two-element list of `{CYRILLIC, TRANSLIT}`, strip footnote symbols from the
end of the Cyrillic and translit.
]==]
function export.strip_notes_from_forms(forms)
local newforms = {}
for _, form in ipairs(forms) do
local ru, tr = form[1], form[2]
ru, _ = m_table_tools.separate_notes(ru)
if tr then
tr, _ = m_table_tools.separate_notes(tr)
end
insert(newforms, {ru, tr})
end
return newforms
end
--[==[
Given a list of forms, where each form is a two-element list of `{CYRILLIC, TRANSLIT}`, unzip into parallel lists of
Cyrillic and translit. The latter list may have gaps in it.
]==]
function export.unzip_forms(forms)
local rulist = {}
local trlist = {}
for i, form in ipairs(forms) do
local ru, tr = form[1], form[2]
rulist[i] = ru
trlist[i] = tr
end
return rulist, trlist
end
--[==[
Given parallel lists of Cyrillic and translit (where the latter list may have gaps in it), return a list of forms,
where each form is a two-element list of `{CYRILLIC, TRANSLIT}`.
]==]
function export.zip_forms(rulist, trlist)
local forms = {}
for i, ru in ipairs(rulist) do
insert(forms, {ru, trlist[i]})
end
return forms
end
local function any_forms_have_translit(forms)
for _, form in ipairs(forms) do
if form[2] then
return true
end
end
return false
end
--[==[
Given a list of forms, where each form is a two-element list of `{CYRILLIC, TRANSLIT}`, combine forms with identical
Cyrillic, concatenating the translit with a comma+space in between.
]==]
function export.combine_translit_of_duplicate_forms(forms)
if not forms[1] then
return forms
end
-- Optimization to avoid creating a new list in the majority case when no translit exists.
if not any_forms_have_translit(forms) then
return forms
end
local newforms = {}
insert(newforms, {forms[1][1], forms[1][2]})
for i = 2, #forms do
local found_duplicate = false
for j = 1, #newforms do
-- If the Russian of the next form is the same as that of the last one, combine their translits and modify
-- newforms[] in-place. Otherwise add the next form to newforms[]. Make sure to clone the form rather than
-- just appending it directly since we may modify it in-place; we don't want to side-effect `forms` as
-- passed in.
if forms[i][1] == newforms[j][1] then
local tr1 = newforms[j][2]
local tr2 = forms[i][2]
if not tr1 and not tr2 then
-- this shouldn't normally happen
else
tr1 = tr1 or export.translit_no_links(newforms[j][1])
tr2 = tr2 or export.translit_no_links(forms[i][1])
if tr1 == tr2 then
-- this shouldn't normally happen
else
newforms[j][2] = tr1 .. ", " .. tr2
end
end
found_duplicate = true
break
end
end
if not found_duplicate then
insert(newforms, {forms[i][1], forms[i][2]})
end
end
return newforms
end
function export.any_termobjs_have_translit(termobjs)
for _, termobj in ipairs(termobjs) do
if termobj.tr then
return true
end
end
return false
end
--[==[
Given a list of term objects, where each object is a table of `{term = CYRILLIC, tr = TRANSLIT, ...}`, combine forms
with identical Cyrillic and identical ancillary properties, concatenating the translit with a comma+space in between.
Also recompose all transliteration.
]==]
function export.combine_translit_of_duplicate_termobjs_and_recompose(termobjs)
if not termobjs[1] then
return termobjs
end
-- Optimization to avoid creating a new list in the majority case when no translit exists.
if not export.any_termobjs_have_translit(termobjs) then
return termobjs
end
local newtermobjs = {}
insert(newtermobjs, m_table.shallowCopy(termobjs[1]))
for i = 2, #termobjs do
local found_duplicate = false
for j = 1, #newtermobjs do
-- If the Russian of the next termobj is the same as that of the last one, combine their translits and
-- modify newtermobjs[] in-place. Otherwise add the next termobj to newtermobjs[]. Make sure to clone the
-- termobj rather than just appending it directly since we may modify it in-place; we don't want to
-- side-effect `termobjs` as passed in.
if termobjs[i].term == newtermobjs[j].term and
m_headword_utilities.termobj_ancillary_properties_equal(termobjs[i], newtermobjs[j]) then
local tr1 = newtermobjs[j].tr
local tr2 = termobjs[i].tr
if not tr1 and not tr2 then
-- this shouldn't normally happen
else
tr1 = tr1 or export.translit_no_links(newtermobjs[j].term)
tr2 = tr2 or export.translit_no_links(termobjs[i].term)
if tr1 == tr2 then
-- this shouldn't normally happen
else
newtermobjs[j].tr = tr1 .. ", " .. tr2
end
end
found_duplicate = true
break
end
end
if not found_duplicate then
insert(newtermobjs, m_table.shallowCopy(termobjs[i]))
end
end
for _, newtermobj in ipairs(newtermobjs) do
if newtermobj.tr then
newtermobj.tr = export.recompose(newtermobj.tr)
end
end
return newtermobjs
end
--[==[
Given a list of forms, where each form is a two-element list of `{CYRILLIC, TRANSLIT}`, split cases where two different
transliterations have been packed into a single translit field by creating two adjacent term/translit pairs. This is
the opposite operation of `combine_translit_of_duplicate_forms()`.
]==]
function export.split_translit_of_duplicate_forms(forms)
if not forms[1] then
return forms
end
-- Optimization to avoid creating a new list in the majority case when no translit exists.
if not export.any_forms_have_translit(forms) then
return forms
end
local newforms = {}
for _, form in ipairs(forms) do
local ru, tr = unpack(form)
if not tr or not tr:find(",") or ru:find(",") then
insert(newforms, form)
else
local split_trs = split(tr, ",%s*")
local default_tr = export.translit_no_links(ru)
for _, split_tr in ipairs(split_trs) do
if split_tr == default_tr then
split_tr = nil
end
insert(newforms, {ru, split_tr})
end
end
end
return newforms
end
--[==[
Given a list of forms, where each form is a two-element list of `{CYRILLIC, TRANSLIT}`, split cases where two different
transliterations have been packed into a single translit field by creating two adjacent term/translit pairs. This is
the opposite operation of `combine_translit_of_duplicate_forms()`.
]==]
function export.split_translit_of_duplicate_termobjs_and_decompose(termobjs)
if not termobjs[1] then
return termobjs
end
-- Optimization to avoid creating a new list in the majority case when no translit exists.
if not export.any_termobjs_have_translit(termobjs) then
return termobjs
end
local newobjs = {}
for _, obj in ipairs(termobjs) do
local tr = obj.tr
if tr then
tr = export.decompose(tr)
end
if not tr or not tr:find(",") or obj.term:find(",[^%s]") then
if not tr or tr == obj.tr then
insert(newobjs, obj)
else
obj = m_table.shallowCopy(obj)
obj.tr = tr
insert(newobjs, obj)
end
else
local term = obj.term
local embedded_comma_in_term = not not term:find(",")
-- If there's an embedded comma in the term, we want to split on commas not followed by a space; otherwise,
-- we split on commas whether or not followed by a space. To implement "comma not followed by space", we
-- can use a "frontier pattern"; the one below matches if the character after the comma is in the specified
-- set (which is anything but a comma or space) and the comma itself is not in the specified set (in other
-- words, it is a comma or space, which matches because it's a comma). This doesn't work correctly if there
-- are two commas in a row, but that should not happen as there should not be an empty translit.
local split_trs = split(tr, embedded_comma_in_term and ",%f[^,%s]" or ",%s*")
local default_tr = export.translit_no_links(term)
for _, split_tr in ipairs(split_trs) do
if split_tr == default_tr then
split_tr = nil
end
local newobj = m_table.shallowCopy(obj)
newobj.tr = split_tr
insert(newobjs, newobj)
end
end
end
return newobjs
end
function export.strip_ending(ru, tr, ending)
local strippedru = ugsub(ru, ending .. "$", "")
if strippedru == ru then
error("Argument " .. ru .. " doesn't end with expected ending " .. ending)
end
ru = strippedru
tr = export.strip_tr_ending(tr, ending)
return ru, tr
end
function export.strip_tr_ending(tr, ending)
if not tr then return nil end
local endingtr = ugsub(export.translit_no_links(ending), "^([Jj])", "%1?")
local strippedtr = ugsub(tr, endingtr .. "$", "")
if strippedtr == tr then
error("Translit " .. tr .. " doesn't end with expected ending " .. endingtr)
end
return strippedtr
end
return export
jifqt1url7tz4h3rtigfh31hoqhxgrp
Modül:ru-nominal
828
1589237
5669923
2026-06-23T07:25:54Z
MustafaCavlak
59368
imported from en
5669923
Scribunto
text/plain
--[[
This module holds functions shared between ru-noun and ru-adjective.
]]
local export = {}
local lang = require("Module:languages").getByCode("ru")
local m_links = require("Module:links")
local m_table = require("Module:table")
local com = require("Module:ru-common")
local m_ru_translit = require("Module:ru-translit")
local m_table_tools = require("Module:table tools")
local u = mw.ustring.char
local rfind = mw.ustring.find
local rsubn = mw.ustring.gsub
local unpack = unpack or table.unpack -- Lua 5.2 compatibility
local usub = mw.ustring.sub
local HYPMARKER = "⟐"
-- version of rsubn() that discards all but the first return value
local function rsub(term, foo, bar)
local retval = rsubn(term, foo, bar)
return retval
end
-- Insert an entry into an existing list if not already present, comparing the entry to items in the existing list
-- using a key function. If entry already found, combine it into the existing entry using combine_func, a function of
-- two arguments (the existing and new entries), which should return the combined entry. Return false if entry already
-- found, true if new entry inserted. If combine_func not specified, the existing entry is left alone. If combine_func
-- is specified, the return value will be written over the existing value (i.e. the existing list will be modified
-- in-place).
--
-- FIXME: General enough to consider moving to [[Module:table]].
local function insert_if_not_by_key(list, new_entry, keyfunc, combine_func)
local new_entry_key = keyfunc(new_entry)
for i, item in ipairs(list) do
local item_key = keyfunc(item)
if m_table.deepEquals(item_key, new_entry_key) then
if combine_func then
list[i] = combine_func(item, new_entry)
end
return false
end
end
table.insert(list, new_entry)
return true
end
--------------------------------------------------------------------------
-- Used for manual translit --
--------------------------------------------------------------------------
function export.combine_stem_and_suffix(stem, tr, suf, rules, old)
local first = usub(suf, 1, 1)
if rules then
local conv = rules[first]
if conv then
local ending = usub(suf, 2)
-- The following regexp is not quite the same as com.vowels. For one thing
-- it includes й, which is important. It leaves out ы, which may or may not
-- be important.
if old and conv == "и" and rfind(ending, "^́?[аеёиійоуэюяѣ]") then
conv = "і"
end
suf = conv .. ending
end
end
-- If <adj> is present in the suffix, it means we need to translate it
-- specially; do that now.
local is_adj = rfind(suf, "<adj>")
suf = rsub(suf, "<adj>", "")
local suftr = is_adj and m_ru_translit.tr_adj(suf, "mono")
return com.concat_russian_tr(stem, tr, suf, suftr, "dopair"), suf
end
--------------------------------------------------------------------------
-- Formatting forms for display --
--------------------------------------------------------------------------
-- Generate a string to substitute into a particular form in a Wiki-markup table. `forms` is the list of forms,
-- generated by concat_word_forms(). `is_lemma` is true if we're formatting the entry for use in displaying the lemma
-- in the declension table title. In this case, we don't include the translit, and remove monosyllabic accents from the
-- Cyrillic (but not in multiword expressions). `accel_form` is the form code to speicfy in the accelerator, e.g.
-- 'nom|m|s', or nil for no accelerator. `lemma_forms` is the list of {RU, TR} lemma forms for use in the accelerator,
-- or nil if `accel_form` is nil. `remove_monosyllabic_accents_lemma_only` indicates that monosyllabic accents should
-- be removed only in the lemma; otherwise we remove them from all forms. (FIXME: Rethink why we have this flag; we
-- should be consistent.)
function export.show_form(forms, is_lemma, accel_form, lemma_forms, remove_monosyllabic_accents_lemma_only)
local russianvals = {}
local latinvals = {}
local lemmavals = {}
-- First fetch the lemma forms and translit. If there are adjacent forms that have identical Russian including
-- stress but different translit (e.g. азербайджа́нец with translits 'azerbajdžánec' and 'azɛrbajdžánec'), we
-- combine the translits, comma-separating them. (This is necessary because there is currently only one tr=
-- field per term.) We don't do this when processing the forms below; instead we handle this in a different
-- and more general fashion (see below).
local lemmaru, lemmatr
if accel_form and lemma_forms and lemma_forms[1] ~= "-" then
lemma_forms = com.combine_translit_of_duplicate_forms(com.strip_notes_from_forms(lemma_forms))
for i, form in ipairs(lemma_forms) do
local ru, tr = unpack(lemma_forms[i])
ru, tr = com.remove_monosyllabic_accents(ru, tr)
lemma_forms[i] = {ru, tr}
end
lemmaru, lemmatr = com.unzip_forms(lemma_forms)
end
-- Accumulate separately the Russian and transliteration into RUSSIANVALS and LATINVALS, then concatenate each down
-- below. We need a fair amount of logic here:
-- (1) to separate out footnote symbols;
-- (2) to separate out the hypothetical marker (a footnote symbol but causes display of the Russian and translit
-- in a special font);
-- (3) to maybe remove monosyllabic accents;
-- (4) to deduplicate repeated forms.
-- We used to generate the display (HTML) as we went, but this prevented proper deduplication because the
-- accelerator classes included in the HTML were different for otherwise identical forms. (Specifically, if two
-- forms are the same in the Russian but different in the translit, the Russian will have different display forms
-- because the translit is included in the accelerator classes.) So what we do is accumulate, separately for the
-- Russian and translit, objects containing the entry (Russian or translit), the separated footnote symbols,
-- whether the entry is hypothetical, and (for Russian only) the corresponding translit(s), for accelerator
-- generation. As we accumulate, we duduplicate based only on comparing the entries of two objects. If we need to
-- deduplicate two objects, we also need to combine their footnotes and transliteration (in the latter case, by
-- comma-separating; we do this because there is only one tr= field for each term).
for _, form in ipairs(forms) do
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.separate_notes(tr)
trnotes = rsub(trnotes, HYPMARKER, "")
end
if (is_lemma or not remove_monosyllabic_accents_lemma_only) then
ruentry, trentry = com.remove_monosyllabic_accents(ruentry, trentry)
end
local ishyp = rfind(runotes, HYPMARKER)
if ishyp then
runotes = rsub(runotes, HYPMARKER, "")
end
local ruobj = {entry = ruentry, tr = {trentry or true}, ishyp = ishyp, notes = runotes}
if not trentry then
trentry = com.translit_no_links(ruentry)
end
if not trnotes then
trnotes = com.translit_no_links(runotes)
end
local trobj = {entry = trentry, ishyp = ishyp, notes = trnotes}
local function keyfunc(obj)
return obj.entry
end
local function combine_func_ru(obj1, obj2)
for _, tr in ipairs(obj2.tr) do
m_table.insertIfNot(obj1.tr, tr)
end
obj1.notes = obj1.notes .. obj2.notes
obj1.ishyp = obj1.ishyp or obj2.ishyp
return obj1
end
local function combine_func_tr(obj1, obj2)
obj1.notes = obj1.notes .. obj2.notes
obj1.ishyp = obj1.ishyp or obj2.ishyp
return obj1
end
if is_lemma then
-- m_table.insertIfNot(lemmavals, ruspan .. " (" .. trspan .. ")")
insert_if_not_by_key(lemmavals, ruobj, keyfunc, combine_func_ru)
else
insert_if_not_by_key(russianvals, ruobj, keyfunc, combine_func_ru)
insert_if_not_by_key(latinvals, trobj, keyfunc, combine_func_tr)
end
end
-- Now finally format each object and concatenate them together.
local function concatenate_ru(objs)
local is_missing = false
for i, obj in ipairs(objs) do
local accel = nil
if lemmaru then
local translit = nil
if #obj.tr == 1 and obj.tr[1] == true then
-- no translit
else
for j, tr in ipairs(obj.tr) do
if tr == true then
obj.tr[j] = com.translit_no_links(obj.entry)
end
end
translit = table.concat(obj.tr, ", ")
end
accel = {form = accel_form, translit = translit, lemma = lemmaru, lemma_translit = lemmatr}
end
if obj.entry == "-" and #forms == 1 then
objs[i] = "—"
is_missing = true
end
if obj.ishyp then
-- no accelerator for hypothetical forms
objs[i] = m_links.full_link({lang = lang, term = nil, alt = obj.entry, tr = "-"}, "hypothetical")
else
objs[i] = m_links.full_link({lang = lang, term = obj.entry, tr = "-", accel = accel})
end
objs[i] = objs[i] .. m_table_tools.superscript_notes(obj.notes)
end
return table.concat(objs, ", "), is_missing
end
local function concatenate_tr(objs)
local scriptutils = require("Module:script utilities")
for i, obj in ipairs(objs) do
local trspan = m_links.remove_links(obj.entry) .. m_table_tools.superscript_notes(obj.notes)
if obj.ishyp then
-- FIXME, in the old [[Module:ru-noun]] code, notes were omitted from hypothetical entries. Correct?
objs[i] = scriptutils.tag_text(trspan, lang, require("Module:scripts").getByCode("Latn"),
"hypothetical")
else
objs[i] = scriptutils.tag_translit(trspan, lang, "default", " style=\"color: var(--wikt-palette-grey-8,#888);\"")
end
end
return table.concat(objs, ", ")
end
if is_lemma then
local russian_span, is_missing = concatenate_ru(lemmavals)
return russian_span
else
local russian_span, is_missing = concatenate_ru(russianvals)
if is_missing then
return russian_span
end
local latin_span = concatenate_tr(latinvals)
return russian_span .. "<br />" .. latin_span
end
end
--------------------------------------------------------------------------
-- Sibilant/Velar/ц rules --
--------------------------------------------------------------------------
local stressed_sibilant_rules = {
["я"] = "а",
["ы"] = "и",
["ё"] = "о́",
["ю"] = "у",
}
local stressed_c_rules = {
["я"] = "а",
["ё"] = "о́",
["ю"] = "у",
}
local unstressed_sibilant_rules = {
["я"] = "а",
["ы"] = "и",
["о"] = "е",
["ю"] = "у",
}
local unstressed_c_rules = {
["я"] = "а",
["о"] = "е",
["ю"] = "у",
}
local velar_rules = {
["ы"] = "и",
}
export.stressed_rules = {
["ш"] = stressed_sibilant_rules,
["щ"] = stressed_sibilant_rules,
["ч"] = stressed_sibilant_rules,
["ж"] = stressed_sibilant_rules,
["ц"] = stressed_c_rules,
["к"] = velar_rules,
["г"] = velar_rules,
["х"] = velar_rules,
}
export.unstressed_rules = {
["ш"] = unstressed_sibilant_rules,
["щ"] = unstressed_sibilant_rules,
["ч"] = unstressed_sibilant_rules,
["ж"] = unstressed_sibilant_rules,
["ц"] = unstressed_c_rules,
["к"] = velar_rules,
["г"] = velar_rules,
["х"] = velar_rules,
}
export.nonsyllabic_suffixes = m_table.listToSet({"", "ъ", "ь", "й"})
export.sibilant_suffixes = m_table.listToSet({"ш", "щ", "ч", "ж"})
return export
1uawletg0fhb4bjt39sk8u6mg2vk01k
Modül:ru-ön ad
828
1589238
5669924
2026-06-23T07:27:06Z
MustafaCavlak
59368
imported from en
5669924
Scribunto
text/plain
--[=[
This module contains functions for creating inflection tables for Russian
adjectives.
Author: Benwing, rewritten from early version by Wikitiki89
Arguments:
1: lemma; nom sg, or just the stem if an explicit declension type is
given in arg 2
2: declension type (usually omitted to autodetect based on the lemma),
along with any short accent type and optional irregular short stem;
see below.
CASE_NUMGEN: Override a given form; see abbreviations below
suffix: any suffix to attach unchanged to the end of each form
notes: Notes to add to the end of the table
title: Override the title
shorttail: Footnote (e.g. *, 1, 2, etc.) to add to short forms if
there's more than one; automatically superscripted
shorttailall: Same as shorttail= but applies to all short forms even
if there's only one
CASE_NUMGEN_tail: Like shorttailall but only for a specific form
nofull: Short forms only
Case abbreviations:
nom: nominative
gen: genitive
dat: dative
acc: accusative
ins: instrumental
pre: prepositional
short: short form
Number/gender abbreviations:
m: masculine singular
n: neuter singular
f: feminine singular
p: plural
mp: masculine plural (old-style and special numeral tables only)
fp: masculine plural (special numeral tables only)
Animacy abbreviations:
an: animate
in: inanimate
Declension-type argument (arg 2):
Form is DECLSPEC or DECLSPEC,DECLSPEC,... where DECLSPEC is
one of the following:
DECLTYPE:SHORTACCENT:SHORTSTEM
DECLTYPE:SHORTACCENT
DECLTYPE
SHORTACCENT:SHORTSTEM
SHORTACCENT
(blank)
DECLTYPE should normally be omitted, and the declension autodetected
from the ending; or it should be ь, to indicate that an adjective
in -ий is of the possessive variety with an extra -ь- in most of
the endings. Alternatively, it can be an explicit declension type,
in which case the lemma field needs to be replaced with the bare
stem; the following are the possibilities:
ый ий ой ьий short mixed manual (new-style)
ый ій ьій short stressed-short mixed proper stressed-proper
ъ-short ъ-stressed-short ъ-mixed ъ-proper ъ-stressed-proper
manual (old-style, where ъ-* is a synonym for *, for any *)
SHORTACCENT is one of a a' b b' c c' c'' to auto-generate the
short forms with the specified accent pattern (following
Zaliznyak); if omitted, no short forms will be auto-generated.
SHORTACCENT can contain a * in it for "reducible" adjectives,
where the short masculine singular has an epenthetic vowel
in the final syllable. It can also contain (1) or (2) to
indicate special cases. Both are used in conjunction with
adjectives in -нный/-нний. Special case (1) causes the
short masculine singular to end in -н instead of -нн; special
case (2) causes all short forms to end this way.
SHORTSTEM, if present, is used as the short stem to base the
short forms off of, instead of the normal adjective long stem
(possibly with a final-syllable accent added in the case of
declension type -о́й).
TODO:
1. Figure out what the symbol X-inside-square (⊠) means, which seems to go with
all adjectives in -о́й with multi-syllabic stems. It may mean that the
masculine singular short form is missing. If this indeed a regular thing,
we need to implement it (and if it's regular but means something else,
we need to implement that, too). Also figure out what the other signs
on pages 68-76 etc. mean: -, X, ~, П₂, Р₂, diamond (♢), triangle (△; might
simply mean a misc. irregularity; explained on p. 61).
2. Mark irregular overridden forms with △ (esp. appropriate for short forms).
3. Should non-reducible adjectives in -нный and -нний default to special case
(1)?
4. In the case of a non-dereducible short masc sing of stress type b, we don't
currently move the stress to the last syllable. Should we?
5. FIXME: In decline(), we used to default acc_n to nom_n. Now we do that in
handle_forms_and_overrides(). Verify that this is more correct.
6. FIXME: Allow multiple heads, both to handle cases where two manual translits
exist (or rather, one automatic and one manual), and cases where two
stresses are possible, e.g. мину́вший or ми́нувший.
7. FIXME: Add code to generate the regular comparative. The rules are as
follows: If the stem doesn't end in к г х, add -ee with no stress change
if the short form is type a, else add -е́е (including type a'). If the stem
ends in к г х, turn the last consonant into ч ж ш, add -е, and place the
stress on the syllable preceding the ending. (E.g. дорого́й -> доро́же)
]=]--
local m_utilities = require("Module:utilities")
local m_links = require("Module:links")
local m_table = require("Module:table")
local com = require("Module:ru-common")
local nom = require("Module:ru-nominal")
local strutils = require("Module:string utilities")
local m_table_tools = require("Module:table tools")
local m_debug = require("Module:debug")
local export = {}
local lang = require("Module:languages").getByCode("ru")
----------------------------- Utility functions ------------------------
local rfind = mw.ustring.find
local rsubn = mw.ustring.gsub
local rmatch = mw.ustring.match
local rsplit = mw.text.split
local ulower = mw.ustring.lower
local usub = mw.ustring.sub
-- version of rsubn() that discards all but the first return value
local function rsub(term, foo, bar)
local retval = rsubn(term, foo, bar)
return retval
end
-- version of rsubn() that returns a 2nd argument boolean indicating whether
-- a substitution was made.
local function rsubb(term, foo, bar)
local retval, nsubs = rsubn(term, foo, bar)
return retval, nsubs > 0
end
-- Insert a single form (consisting of {RUSSIAN, TR}) or a list of such
-- forms into an existing list of such forms, adding NOTESYM (a string or nil)
-- to the new forms if not nil. Return whether an insertion was performed.
local function insert_forms_into_existing_forms(existing, newforms, notesym)
if type(newforms) ~= "table" then
newforms = {newforms}
end
local inserted = false
for _, item in ipairs(newforms) do
if not m_table.contains(existing, item) then
if notesym then
item = com.concat_paired_russian_tr(item, {notesym})
end
table.insert(existing, item)
inserted = true
end
end
return inserted
end
local function track(page)
m_debug.track("ru-adjective/" .. page)
return true
end
-- Fancy version of ine() (if-not-empty). Converts empty string to nil,
-- but also strips leading/trailing space and then single or double quotes,
-- to allow for embedded spaces.
local function ine(arg)
if not arg then return nil end
arg = rsub(arg, "^%s*(.-)%s*$", "%1")
if arg == "" then return nil end
local inside_quotes = rmatch(arg, '^"(.*)"$')
if inside_quotes then
return inside_quotes
end
inside_quotes = rmatch(arg, "^'(.*)'$")
if inside_quotes then
return inside_quotes
end
return arg
end
-------------------- Global declension/case/etc. variables -------------------
-- 'enable_categories' is a special hack for testing, which disables all
-- category insertion if false. Delete this as soon as we've verified the
-- working of the category code and created all the necessary categories.
local enable_categories = true
local declensions = {}
local declensions_old = {}
local internal_notes_table = {}
local internal_notes_table_old = {}
local internal_notes_genders = {}
local internal_notes_genders_old = {}
local short_declensions = {}
local short_declensions_old = {}
-- Formerly used for the ой-rare type, which has been removed; but keep
-- the code around in case we need it later.
local short_internal_notes_table = {}
local short_internal_notes_table_old = {}
local short_stress_patterns = {}
local long_cases = {
"nom_m", "nom_n", "nom_f", "nom_p",
"gen_m", "gen_f", "gen_p",
"dat_m", "dat_f", "dat_p",
"acc_m_an", "acc_m_in", "acc_p_an", "acc_p_in", "acc_f", "acc_n",
"ins_m", "ins_f", "ins_p",
"pre_m", "pre_f", "pre_p",
-- extra old long cases
"nom_mp",
"acc_mp_an", "acc_mp_in",
-- extra dva cases
"nom_fp",
"acc_fp_an", "acc_fp_in",
-- extra oba cases
"gen_mp", "gen_fp",
"dat_mp", "dat_fp",
"ins_mp", "ins_fp",
"pre_mp", "pre_fp",
-- extra cases for compounds of два
"acc_mp", "acc_fp",
}
-- Short cases and corresponding numbered arguments
local short_cases = {
"short_m", "short_n", "short_f", "short_p"
}
-- Create master list of all possible cases (actually case/number/gender pairs)
local all_cases = mw.clone(long_cases)
for _, case in ipairs(short_cases) do
m_table.insertIfNot(all_cases, case)
end
-- If enabled, compare this module with new version of module to make
-- sure all declensions are the same.
local test_new_ru_adjective_module = false
-- Forward references to functions
local tracking_code
local categorize
local detect_stem_and_accent_type
local construct_bare_and_short_stem
local decline
local canonicalize_override
local handle_forms_and_overrides
local decline_short
local make_table
--------------------------------------------------------------------------
-- Main code --
--------------------------------------------------------------------------
-- Implementation of main entry point
function export.do_generate_forms(args, old, manual)
local orig_args
if test_new_ru_adjective_module then
orig_args = mw.clone(args)
end
if args[3] or args[4] or args[5] or args[6] then
error("Numbered short forms no longer supported")
end
if args.shorttailall then
track("shorttailall")
end
local SUBPAGENAME = mw.loadData("Module:headword/data").pagename
args.forms = {}
args.categories = {}
old = old or args.old
args.old = old
args.suffix = com.split_russian_tr(args.suffix or "", "dopair")
args.internal_notes = {}
-- Superscript footnote marker at beginning of note, similarly to what's
-- done at end of forms.
if args.notes then
local notes, entry = m_table_tools.get_initial_notes(args.notes)
args.notes = notes .. entry
end
local overall_short_forms_allowed
manual = manual or args[2] == "manual"
args.manual = manual
local decl_types = manual and "$" or args[2] or ""
local lemmas = manual and "-" or args[1] or SUBPAGENAME
local normal_short_classes, rare_short_classes, dated_short_classes
local saw_no_short
for _, lemma_and_tr in ipairs(rsplit(lemmas, ",")) do
-- reset these for each lemma so we get the short classes of the last
-- lemma (doesn't really matter, as they should be the same for all
-- lemmas)
normal_short_classes = {}
rare_short_classes = {}
dated_short_classes = {}
saw_no_short = false
for _, decl_type in ipairs(rsplit(decl_types, ",")) do
local lemma, lemmatr = com.split_russian_tr(lemma_and_tr)
-- if lemma ends with -ся, strip it and act as if suffix=ся given
-- (or rather, prepend ся to suffix)
local active_base = rmatch(lemma, "^(.*)ся$")
if active_base then
lemma = active_base
lemmatr = com.strip_tr_ending(lemmatr, "ся")
args.refl = true
args.real_suffix = com.concat_paired_russian_tr({"ся"}, args.suffix)
else
args.refl = false
args.real_suffix = args.suffix
end
-- Auto-detect actual decl type, and get short accent and overriding
-- short stem, if specified.
local stem, stemtr, short_accent, short_stem, short_stemtr, datedrare
stem, stemtr, decl_type, short_accent, short_stem, datedrare =
detect_stem_and_accent_type(lemma, lemmatr, decl_type, args)
if rfind(decl_type, "^[іи]й$") and rfind(stem, "[" .. com.velar .. com.sib .. "]$") then
decl_type = "ый"
end
if datedrare == "none" then
saw_no_short = true
end
local proper_decl = m_table.contains({"proper", "stressed-proper"}, decl_type)
-- Use args.real_surname to avoid overwriting args.surname, since we may be overwriting
-- `args` multiple times. FIXME: Use a separate data structure for this.
args.real_surname = proper_decl or args.surname
local short_title
if short_accent then
short_title = short_accent
elseif args.real_surname then
short_title = "surname"
elseif m_table.contains({"ьій", "ьий", "ьей", "short", "stressed-short", "й-short", "mixed"}, decl_type) then
short_title = "possessive"
end
if short_title then
if datedrare == "dated" then
m_table.insertIfNot(dated_short_classes, short_title)
elseif datedrare == "rare" then
m_table.insertIfNot(rare_short_classes, short_title)
else
m_table.insertIfNot(normal_short_classes, short_title)
end
end
if short_stem then
short_stem, short_stemtr = com.split_russian_tr(short_stem)
end
stem, args.allow_unaccented = rsubb(stem, "^%*", "")
if args.allow_unaccented then
track("allow-unaccented")
end
-- Treat suffixes without an accent, and suffixes with an accent on
-- the initial hyphen, as if they were preceded with a *, which
-- overrides all the logic that normally (a) normalizes the accent,
-- and (b) complains about multisyllabic words without an accent.
-- Don't do this if lemma is just -, which is used specially in
-- manual declension tables.
if lemma ~= "-" and (rfind(lemma, "^%-́") or (com.is_unstressed(lemma) and rfind(lemma, "^%-"))) then
args.allow_unaccented = true
end
if not args.allow_unaccented and com.needs_accents(lemma) then
-- Technically we don't need accents in -ой adjectives (which are
-- always ending-stressed) and in -ый adjectives with a monosyllabic
-- stem, such as полный (which are always stem-stressed), but it's
-- better to enforce accents in all multisyllabic words, for
-- consistency with nouns and verbs. Note that we still don't
-- require an accent in monosyllabic adjectives such as злой.
error("Lemma must have an accent in it: " .. lemma)
end
-- Set stem and unstressed version. Also construct end-accented version
-- of stem if unstressed; needed for short forms of adjectives of
-- type -о́й. We do this here before doing the dereduction
-- transformation so that we don't end up stressing an unstressed
-- epenthetic vowel, and so that the previous syllable instead ends
-- up stressed (in type -о́й adjectives the stem won't have any stress).
-- Note that the closest equivalent in nouns is handled in
-- attach_unstressed(), which puts the stress onto the final syllable
-- if the stress pattern calls for ending stress in the genitive
-- plural. This works there because
-- (1) It won't stress an unstressed epenthetic vowel because the
-- cases where the epenthetic vowel is unstressed are precisely
-- those with stem stress in the gen pl, not ending stress;
-- (2) There isn't a need to stress the syllable preceding an
-- unstressed epenthetic vowel because that syllable should
-- already have stress, since we require that the base stem form
-- (parameter 2) have stress in it whenever any case form has
-- stem stress. This isn't the case here in type -о́й adjectives.
-- NOTE: FIXME: I can't actually quote any forms from Zaliznyak
-- (at least not from pages 58-60, where this is discussed) that
-- seem to require last-stem-syllable-stress, i.e. where the stem is
-- multisyllabic. The two examples given are both unusual: дорого́й
-- has stress on the initial syllable до́рог etc. and these forms are
-- marked with a triangle (indicating an apparent irregularity); and
-- голубо́й seems not to have a masculine singular short form, and it's
-- accent pattern b, so the remaining forms are all ending-stressed.
-- In fact it's possible that these examples don't exist: It appears
-- that all the multisyllabic adjectives in -о́й listed in the
-- dictionary have a marking next to them consisting of an X inside of
-- a square, which is glossed p. 69 to something I don't understand,
-- but may be saying that the masculine singular short form is
-- missing. If this is regular, we need to implement it.
args.stem, args.stemtr = stem, stemtr
args.ustem, args.ustemtr = com.make_unstressed_once(stem, stemtr)
local accented_stem, accented_stemtr = stem, stemtr
if not args.allow_unaccented then
if accented_stemtr and com.is_unstressed(accented_stem) ~= com.is_unstressed(accented_stemtr) then
error("Stem " .. accented_stem .. " and translit " .. accented_stemtr .. " must have same accent pattern")
end
if com.is_unstressed(accented_stem) then
accented_stem, accented_stemtr =
com.make_ending_stressed(accented_stem, accented_stemtr)
end
end
local short_forms_allowed = manual and true or decl_type == "ый" or
decl_type == "ой" or decl_type == (old and "ій" or "ий")
overall_short_forms_allowed = overall_short_forms_allowed or
short_forms_allowed
if not short_forms_allowed then
-- FIXME: We might want to allow this in case we have a
-- reducible short, mixed or proper possessive adjective. But
-- in that case we need to reduce rather than dereduce to get
-- the stem.
if short_accent or short_stem then
error("Cannot specify short accent or short stem with declension type " .. decl_type .. ", as short forms aren't allowed")
end
if args.short_m or args.short_f or args.short_n or args.short_p then
error("Cannot specify explicit short forms with declension type " .. decl_type .. ", as short forms aren't allowed")
end
end
local orig_short_accent = short_accent
local short_decl_type
short_accent, short_decl_type = construct_bare_and_short_stem(args,
short_accent, short_stem, short_stemtr, accented_stem,
accented_stemtr, old, decl_type)
local decls = old and declensions_old or declensions
local short_decls = old and short_declensions_old or short_declensions
if not decls[decl_type] then
error("Unrecognized declension type " .. decl_type)
end
if short_accent == "" then
error("Short accent type cannot be blank, should be omitted or given")
end
if short_accent and not short_stress_patterns[short_accent] then
error("Unrecognized short accent type " .. short_accent)
end
tracking_code(decl_type, args, orig_short_accent, short_accent,
short_stem, datedrare, short_forms_allowed)
if not manual and enable_categories then
categorize(decl_type, args, orig_short_accent, short_accent,
short_stem)
end
decline(args, decls[decl_type], m_table.contains({"ой", "ьей", "й-short", "stressed-short", "stressed-proper"}, decl_type))
if short_forms_allowed and short_accent then
decline_short(args, short_decls[short_decl_type],
short_stress_patterns[short_accent], datedrare)
end
local intable = old and internal_notes_table_old or internal_notes_table
local shortintab = old and short_internal_notes_table_old or
short_internal_notes_table
local internal_note = intable[decl_type] or shortintab[short_decl_type]
if internal_note then
m_table.insertIfNot(args.internal_notes, internal_note)
end
end
end
local short_class_titles = {}
local function make_short_class_title(short_classes, datedrare)
local sct = table.concat(short_classes, ",") -- short class title
if sct == "" then
return
end
if not m_table.contains({"surname", "possessive"}, sct) then
-- Convert e.g. a*,a(1) into a*[(1)], either finally or followed by
-- comma.
sct = rsub(sct, "([abc]'*)%*,%1%*?(%([12]%))$", "%1*[%2]")
sct = rsub(sct, "([abc]'*)%*,%1%*?(%([12]%)),", "%1*[%2],")
-- Same for a(1),a*.
sct = rsub(sct, "([abc]'*)%*?(%([12]%)),%1%*$", "%1*[%2]")
sct = rsub(sct, "([abc]'*)%*?(%([12]%)),%1%*,", "%1*[%2],")
-- Convert (1), (2) to ①, ②.
sct = rsub(sct, "%(1%)", "①")
sct = rsub(sct, "%(2%)", "②")
-- Add a * before ①, ②, consistent with Zaliznyak.
sct = rsub(sct, "([abc]'*)([①②])", "%1*%2")
-- Convert ' and '' to ʹ and ʺ, consistent with Zaliznyak.
sct = sct:gsub("''", "ʺ"):gsub("'", "ʹ")
if args.nofull then
sct = "short-form-only class " .. sct
else
sct = "short class " .. sct
end
end
if datedrare then
sct = datedrare .. " " .. sct
end
table.insert(short_class_titles, sct)
end
make_short_class_title(normal_short_classes, nil)
make_short_class_title(rare_short_classes, "rare")
make_short_class_title(dated_short_classes, "dated")
if #short_class_titles == 0 then
if saw_no_short then
args.short_class_title = "no short forms"
else
args.short_class_title = "unknown short forms"
end
else
args.short_class_title = table.concat(short_class_titles, " / ")
end
handle_forms_and_overrides(args, overall_short_forms_allowed)
-- Test code to compare existing module to new one.
if test_new_ru_adjective_module then
local m_new_ru_adjective = require("Module:User:Benwing2/ru-adjective")
local newargs = m_new_ru_adjective.do_generate_forms(orig_args, old, manual)
local difdecl = false
for _, case in ipairs(all_cases) do
local arg = args[case]
local newarg = newargs[case]
if not m_table.deepEquals(arg, newarg) then
-- Uncomment this to display the particular case and
-- differing forms.
--error(case .. " " .. (arg and com.concat_forms(arg) or "nil") .. " || " .. (newarg and com.concat_forms(newarg) or "nil"))
track("different-decl")
difdecl = true
end
break
end
if not difdecl then
track("same-decl")
end
end
return args
end
local function process_params(frame, include_form)
local boolean = {type = "boolean"}
local params = {
[1] = true,
[2] = true,
suffix = true,
title = true,
special = true,
shorttail = true,
shorttailall = true,
notes = true,
old = boolean,
noneuter = boolean,
nofull = boolean,
surname = boolean,
pronoun = boolean
}
if include_form then
params.form = {required = true}
end
for _, case in ipairs(all_cases) do
params[case] = true
params[case .. "_tail"] = true
params[case .. "_tailall"] = true
end
return require("Module:parameters").process(frame:getParent().args, params)
end
-- Implementation of main entry point
local function do_show(frame, old, manual)
local args = export.do_generate_forms(process_params(frame), old, manual)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The main entry point for modern declension tables.
function export.show(frame)
return do_show(frame, false)
end
-- The main entry point for old declension tables.
function export.show_old(frame)
return do_show(frame, true)
end
-- The main entry point for manual declension tables.
function export.show_manual(frame)
return do_show(frame, false, "manual")
end
-- The main entry point for manual old declension tables.
function export.show_manual_old(frame)
return do_show(frame, true, "manual")
end
-- Entry point for use in Module:ru-noun.
function export.get_nominal_decl(decl, gender, old)
local d = old and declensions_old[decl] or declensions[decl]
local n = {}
if gender == "m" then
n.nom_sg = d.nom_m
n.gen_sg = d.gen_m
n.dat_sg = d.dat_m
n.ins_sg = d.ins_m
n.pre_sg = d.pre_m
elseif gender == "f" then
n.nom_sg = d.nom_f
n.gen_sg = d.gen_f
n.dat_sg = d.dat_f
n.acc_sg = d.acc_f
n.ins_sg = d.ins_f
n.pre_sg = d.pre_f
elseif gender == "n" then
n.nom_sg = d.nom_n
n.gen_sg = d.gen_m
n.dat_sg = d.dat_m
n.acc_sg = d.acc_n
n.ins_sg = d.ins_m
n.pre_sg = d.pre_m
else
assert(false, "Unrecognized gender: " .. gender)
end
n.nom_pl = d.nom_p
n.gen_pl = d.gen_p
n.dat_pl = d.dat_p
n.ins_pl = d.ins_p
n.pre_pl = d.pre_p
if gender == "m" and d.nom_mp then
n.nom_pl = d.nom_mp
end
local intable = old and internal_notes_table_old or internal_notes_table
local ingenders = old and internal_notes_genders_old or internal_notes_genders
-- FIXME, what if there are multiple internal notes? See comment in
-- do_generate_forms().
local internal_notes = ingenders[decl] and m_table.contains(ingenders[decl], gender) and intable[decl]
return n, internal_notes
end
local function get_form(forms)
local canon_forms = {}
for _, form in ipairs(forms) do
local ru, tr = form[1], form[2]
local ruentry = m_table_tools.get_notes(ru)
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.get_notes(tr)
end
ruentry = m_links.remove_links(ruentry)
m_table.insertIfNot(canon_forms, {ruentry, trentry})
end
return com.concat_forms(canon_forms)
end
-- The entry point for 'ru-adj-forms' to generate all adjective forms.
function export.generate_forms(frame)
local args = export.do_generate_forms(process_params(frame), false)
local ins_text = {}
for _, case in ipairs(all_cases) do
if args[case] then
table.insert(ins_text, case .. "=" .. get_form(args[case]))
end
end
return table.concat(ins_text, "|")
end
-- The entry point for 'ru-adj-form' to generate a particular adjective form.
function export.generate_form(frame)
local args = process_params(frame, "include form")
local form = args.form
if not m_table.contains(all_cases, form) then
error("Unrecognized form " .. form)
end
args = export.do_generate_forms(args, false)
if not args[form] then
return ""
else
return get_form(args[form])
end
end
--------------------------------------------------------------------------
-- Tracking and categorization --
--------------------------------------------------------------------------
tracking_code = function(decl_class, args, orig_short_accent, short_accent,
short_stem, datedrare)
local hint_types = com.get_stem_trailing_letter_type(args.stem)
local function dotrack(prefix)
if prefix ~= "" then
track(prefix)
prefix = prefix .. "/"
end
track(prefix .. decl_class)
for _, hint_type in ipairs(hint_types) do
track(prefix .. hint_type)
track(prefix .. decl_class .. "/" .. hint_type)
end
end
dotrack("")
if args.short_m or args.short_f or args.short_n or args.short_p then
dotrack("short")
end
if orig_short_accent then
if rfind(orig_short_accent, "%*") then
dotrack("reducible")
dotrack("reducible/" .. short_accent)
end
if rfind(orig_short_accent, "%(1%)") then
dotrack("special-case-1")
dotrack("special-case-1/" .. short_accent)
end
if rfind(orig_short_accent, "%(2%)") then
dotrack("special-case-2")
dotrack("special-case-2/" .. short_accent)
end
end
if short_accent then
dotrack("short-accent/" .. short_accent)
if datedrare == "dated" then
dotrack("short-accent-dated")
dotrack("short-accent-dated/" .. short_accent)
elseif datedrare == "rare" then
dotrack("short-accent-rare")
dotrack("short-accent-rare/" .. short_accent)
end
elseif datedrare == "none" then
dotrack("short-accent/none")
else
dotrack("short-accent/unknown")
end
if short_stem then
dotrack("explicit-short-stem")
dotrack("explicit-short-stem/" .. short_accent)
end
for _, case in ipairs(all_cases) do
if args[case] then
track("irreg/" .. case)
-- questionable use: dotrack("irreg/" .. case)
end
end
end
-- Insert the category CAT (a string) into list CATEGORIES. String will
-- have "Russian " prepended and ~ substituted for the part of speech.
local function insert_category(categories, cat, plpos)
table.insert(categories, "Russian " .. rsub(cat, "~", plpos))
end
categorize = function(decl_type, args, orig_short_accent, short_accent,
short_stem)
local plpos = args.real_surname and "surnames" or args.pronoun and "pronouns" or "adjectives"
-- Insert category CAT into the list of categories in ARGS.
local function insert_cat(cat)
insert_category(args.categories, cat, plpos)
end
-- FIXME: For compatibility with old {{temp|ru-adj7}}, {{temp|ru-adj8}},
-- {{temp|ru-adj9}}; maybe there's a better way.
if m_table.contains({"ьій", "ьий", "ьей", "short", "stressed-short", "й-short", "mixed",
"proper", "stressed-proper"}, decl_type) then
insert_cat("possessive ~")
end
if not args.pronoun then
if m_table.contains({"ьій", "ьий", "ьей"}, decl_type) then
insert_cat("long possessive ~")
elseif m_table.contains({"short", "stressed-short", "й-short"}, decl_type) then
insert_cat("short possessive ~")
elseif m_table.contains({"mixed"}, decl_type) then
insert_cat("mixed possessive ~")
elseif m_table.contains({"proper", "stressed-proper"}, decl_type) then
-- Don't insert a category like [[:Category:Russian proper-name surnames]]. Instead, rely on
-- [[:Category:Russian possessive surnames]] generated above.
if not args.real_surname then
insert_cat("proper-name ~")
end
elseif decl_type == "$" then
insert_cat("indeclinable ~")
else
local hint_types = com.get_stem_trailing_letter_type(args.stem)
-- insert English version of Zaliznyak stem type
local stem_type =
m_table.contains(hint_types, "velar") and "velar-stem" or
m_table.contains(hint_types, "sibilant") and "sibilant-stem" or
m_table.contains(hint_types, "c") and "ц-stem" or
m_table.contains(hint_types, "i") and "i-stem" or
m_table.contains(hint_types, "vowel") and "vowel-stem" or
m_table.contains(hint_types, "soft-cons") and "vowel-stem" or
m_table.contains(hint_types, "palatal") and "vowel-stem" or
decl_type == "ий" and "soft-stem" or
"hard-stem"
if stem_type == "soft-stem" or stem_type == "vowel-stem" then
insert_cat(stem_type .. " ~")
else
insert_cat(stem_type .. " " .. (m_table.contains({"ой", "ьей", "й-short"}, decl_type) and "ending-stressed" or "stem-stressed") .. " ~")
end
end
if m_table.contains({"ой", "ьей", "й-short"}, decl_type) then
insert_cat("ending-stressed ~")
end
end
local short_forms_allowed = m_table.contains({"ый", "ой", "ій", "ий"}, decl_type)
if short_forms_allowed then
local override_m = args.short_m
local override_f = args.short_f
local override_n = args.short_n
local override_p = args.short_p
local has_short = short_accent or override_m or override_f or
override_n or override_p
local missing_short = override_m == "-" or
override_f == "-" or override_n == "-" or override_p == "-" or
not short_accent and (not override_m or not override_f or
not override_n or not override_p)
if has_short then
insert_cat("~ with short forms")
if missing_short then
insert_cat("~ with missing short forms")
end
end
if short_accent then
insert_cat("~ with short accent pattern " .. (short_accent:gsub("''", "ʺ"):gsub("'", "ʹ")))
end
if orig_short_accent then
if rfind(orig_short_accent, "%*") then
insert_cat("~ with reducible short stem")
end
if rfind(orig_short_accent, "%(1%)") then
insert_cat("~ with Zaliznyak short form special case 1")
end
if rfind(orig_short_accent, "%(2%)") then
insert_cat("~ with Zaliznyak short form special case 2")
end
end
if short_stem and short_stem ~= args.stem then
insert_cat("~ with irregular short stem")
end
end
for _, case in ipairs(all_cases) do
if args[case] and args[case] ~= "-" then
local engcase = rsub(case, "^([a-z]*)", {
nom="nominative", gen="genitive", dat="dative",
acc="accusative", ins="instrumental", pre="prepositional",
short="short",
})
engcase = rsub(engcase, "(_[a-z]*)$", {
_m=" masculine singular", _f=" feminine singular",
_n=" neuter singular", _p=" plural",
_mp=" masculine plural"
})
insert_cat("~ with irregular " .. engcase)
end
end
if args.nofull then
insert_cat("short-form-only ~")
end
end
--------------------------------------------------------------------------
-- Autodetection and stem munging --
--------------------------------------------------------------------------
-- Attempt to detect the type of the lemma based on its ending, separating
-- off the base and the ending; also extract the accent type for short
-- adjectives and optional short stem. DECL is the value passed in, and
-- might already specify the ending. Return five values: STEM, STEMTR, DECL,
-- SHORT_ACCENT (accent class of short adjective, or nil for no short
-- adjectives other than specified through overrides), SHORT_STEM (special
-- stem of short adjective, nil if same as long stem). The return value of
-- SHORT_STEM is taken directly from the argument and will include any
-- manual translit.
detect_stem_and_accent_type = function(lemma, tr, decl, args)
local datedrare = false
-- If it looks like a short decl type, canonicalize. [abc] for types
-- a, b, c'', etc.; [d] for dated-*; [r] for rare-*;
-- [-] for - (no short forms); [*(] for * and (1) and (2) modifiers,
-- which might precede the letter.
if rfind(decl, "^[abcdr*(-]") then
decl = ":" .. decl
end
local splitvals = rsplit(decl, ":")
if #splitvals > 3 then
error("Should be at most three colon-separated parts of a declension spec: " .. decl)
end
local short_accent, short_stem
decl, short_accent, short_stem = splitvals[1], splitvals[2], splitvals[3]
-- Check for dated variant of short accent
local dated_short = short_accent and rmatch(short_accent, "^dated%-(.*)$")
if dated_short then
datedrare = "dated"
short_accent = dated_short
else
-- Check for rare variant of short accent
local rare_short = short_accent and rmatch(short_accent, "^rare%-(.*)$")
if rare_short then
datedrare = "rare"
short_accent = rare_short
end
end
decl = ine(decl)
-- Resolve aliases
if decl then
decl = rsub(decl, "^ъ%-", "")
end
if short_accent == "-" then
short_accent = nil
datedrare = "none"
end
short_accent = ine(short_accent)
short_stem = ine(short_stem)
if short_stem and not short_accent then
error("With explicit short stem " .. short_stem .. ", must specify short accent")
end
local base, ending
-- The while loop appears to function solely as a way of allowing a
-- jump to the end of the loop with a 'break' statement. I don't think
-- the loop is ever run more than once.
while true do
if not decl or decl == "ь" then
base, ending = rmatch(lemma, "^(.*)([ыиіое]́?й)$")
if base then
if ending == "ий" and decl == "ь" then
ending = "ьий"
elseif ending == "ій" and decl == "ь" then
ending = "ьій"
elseif ending == "е́й" or (ending == "ей" and com.is_monosyllabic(lemma)) then
ending = "ье́й"
end
tr = com.strip_tr_ending(tr, ending)
-- -ий/-ій will be converted to -ый after velars and sibilants
-- by the caller
decl = com.make_unstressed(ending)
break
else
-- It appears that short, mixed and proper adjectives are always
-- either of the -ов/ев/ёв or -ин/ын types. The former type is
-- always (or almost always?) short, while the latter can be
-- either; apparently mixed is the more "modern" type of
-- declension, and short is "older". However, both -ов/ев/ёв or
-- -ин/ын are of type "proper" (similar to "short") when
-- capitalized.
--
-- NOTE: Following regexp is accented
base = rmatch(lemma, "^([" .. com.uppercase .. "].*[иы]́н)ъ?$")
if base then
decl = "stressed-proper"
break
end
base = rmatch(lemma, "^([" .. com.uppercase .. "].*[еёо]́?в)ъ?$")
if not base then
-- Following regexp is not stressed
base = rmatch(lemma, "^([" .. com.uppercase .. "].*[иы]н)ъ?$")
end
if base then
decl = "proper"
break
end
base = rmatch(lemma, "^(.*[еёо]́?в)ъ?$")
if base then
decl = "short"
break
end
base = rmatch(lemma, "(.*[иы]́н)ъ?$") --accented
if base then
decl = "stressed-short"
break
end
base = rmatch(lemma, "(.*[иы]н)ъ?$") --unaccented
if base then
decl = "mixed"
break
-- error("With -ин/ын adjectives, must specify 'short' or 'mixed':" .. lemma)
end
error("Cannot determine stem type of adjective: " .. lemma)
end
elseif m_table.contains({"short", "stressed-short", "mixed", "proper",
"stressed-proper"}, decl) then
base = rmatch(lemma, "^(.-)ъ?$")
assert(base)
break
elseif decl == "й" or decl == "й-short" then
base = rmatch(lemma, "^(.-)й$")
assert(base)
tr = com.strip_tr_ending(tr, "й")
decl = "й-short"
break
else
base = lemma
break
end
end
if not datedrare and (args.refl or not (decl == "ый" or decl == "ий" and not rfind(base, "[цс]к$"))) then
datedrare = "none"
end
return base, tr, decl, short_accent, short_stem, datedrare
end
-- Add a possible suffix to the bare stem, according to the declension and
-- value of OLD. This may be -ь, -ъ, -й or nothing. We need to do this here
-- because we don't actually attach such a suffix in attach_unstressed() due
-- to situations where we don't want the suffix added, e.g. бескра́йний with
-- dereduced nom sg бескра́ен without expected -ь.
local function add_bare_suffix(bare, baretr, old, decl, dereduced)
if decl == "й-short" then
return bare .. "й", baretr and baretr .. "j" or nil
elseif old and decl ~= "ій" and decl ~= "$" then
return bare .. "ъ", baretr
elseif decl == "ий" or decl == "ій" then
-- This next special case is mentioned in Zaliznyak's 1980 grammatical
-- dictionary for adjectives (footnote, p. 60).
if dereduced and rfind(bare, "[нН]$") then
-- FIXME: What happens in this case old-style? I assume that
-- -ъ is added, but this is a guess.
return bare .. (old and "ъ" or ""), baretr
elseif rfind(bare, "[" .. com.vowel .. "]́?$") then
-- This happens with adjectives like длинноше́ий, short masculine
-- singular длинноше́й.
return bare .. "й", baretr and baretr .. "j" or nil
else
return bare .. "ь", baretr and baretr .. "ʹ" or nil
end
else
return bare, baretr
end
end
-- Construct and set bare and short form in args, and canonicalize
-- short accent spec, handling cases *, (1) and (2). Return canonicalized
-- short accent and the short declension, which is usually the same as
-- the corresponding long one.
construct_bare_and_short_stem = function(args, short_accent, short_stem,
short_stemtr, accented_stem, accented_stemtr, old, decl)
-- Check if short forms allowed; if not, no short-form params can be given.
-- Construct bare version of stem; used for cases where the ending
-- is non-syllabic (i.e. short masculine singular of long adjectives,
-- and masculine singular of short, mixed and proper adjectives). Comes
-- from short masculine or 3rd argument if explicitly given, else from the
-- accented stem, possibly with the dereduction transformation applied
-- (if * occurs in the short accent spec).
local reducible, sc1, sc2
if short_accent then
short_accent, reducible = rsubb(short_accent, "%*", "")
short_accent, sc1 = rsubb(short_accent, "%(1%)", "")
short_accent, sc2 = rsubb(short_accent, "%(2%)", "")
end
if sc1 or sc2 then
-- Reducible isn't compatible with sc1 or sc2, but Zaliznyak's
-- dictionary always seems to notate sc1 and sc2 with reducible *,
-- so ignore it.
reducible = false
end
if sc1 and sc2 then
error("Special cases 1 and 2, i.e. (1) and (2), not compatible")
end
local short_decl = decl
-- Construct short stem. May be explicitly given, else comes from
-- end-accented stem.
if not short_stem then
short_stem, short_stemtr = accented_stem, accented_stemtr
end
-- Try to accent unaccented short stem (happens only when explicitly given),
-- but be conservative -- only if monosyllabic, since otherwise we have
-- no idea where stress should end up; after all, the explicit short stem
-- is for exceptional cases.
if not args.allow_unaccented then
if short_stemtr and com.is_unstressed(short_stem) ~= com.is_unstressed(short_stemtr) then
error("Explicit short stem " .. short_stem .. " and translit " .. short_stemtr .. " must have same accent pattern")
end
if com.is_monosyllabic(short_stem) then
short_stem, short_stemtr = com.make_ending_stressed(short_stem, short_stemtr)
elseif com.needs_accents(short_stem) then
error("Explicit short stem " .. short_stem .. " needs an accent")
end
end
if sc2 then
if not rfind(short_stem, "нн$") then
error("With special case 2, stem needs to end in -нн: " .. short_stem)
end
short_stem = rsub(short_stem, "нн$", "н")
if short_stemtr then
if not rfind(short_stemtr, "nn$") then
error("With special case 2, stem translit needs to end in -nn: " .. short_stemtr)
end
short_stemtr = rsub(short_stemtr, "nn$", "n")
end
end
-- Construct bare form, used for short masculine.
local bare, baretr
if reducible then
bare, baretr = com.dereduce_stem(short_stem, short_stemtr, rfind(short_accent, "^b"))
if not bare then
error("Unable to dereduce stem: " .. short_stem)
end
bare, baretr = add_bare_suffix(bare, baretr, old, decl, true)
else
bare, baretr = short_stem, short_stemtr
if sc1 then
if not rfind(bare, "нн$") then
error("With special case 1, stem needs to end in -нн: " .. bare)
end
bare = rsub(bare, "нн$", "н")
if baretr then
if not rfind(baretr, "nn$") then
error("With special case 1, stem translit needs to end in -nn: " .. baretr)
end
baretr = rsub(baretr, "nn$", "n")
end
end
-- With special case 1 or 2, we don't ever want -ь added, so treat
-- it like a reducible (that may be why these are marked as
-- reducible in Zaliznyak).
bare, baretr = add_bare_suffix(bare, baretr, old, decl, sc1 or sc2)
end
args.short_stem, args.short_stemtr = short_stem, short_stemtr
args.short_ustem, args.short_ustemtr = com.make_unstressed_once(short_stem, short_stemtr)
args.bare, args.baretr = bare, baretr
return short_accent, short_decl
end
--------------------------------------------------------------------------
-- Declensions --
--------------------------------------------------------------------------
declensions["ый"] = {
["nom_m"] = "ый",
["nom_n"] = "ое",
["nom_f"] = "ая",
["nom_p"] = "ые",
["gen_m"] = "ого",
["gen_f"] = "ой",
["gen_p"] = "ых",
["dat_m"] = "ому",
["dat_f"] = "ой",
["dat_p"] = "ым",
["acc_f"] = "ую",
["acc_n"] = "ое",
["ins_m"] = "ым",
["ins_f"] = {"ой", "ою"},
["ins_p"] = "ыми",
["pre_m"] = "ом",
["pre_f"] = "ой",
["pre_p"] = "ых",
}
declensions["ий"] = {
["nom_m"] = "ий",
["nom_n"] = "ее",
["nom_f"] = "яя",
["nom_p"] = "ие",
["gen_m"] = "его",
["gen_f"] = "ей",
["gen_p"] = "их",
["dat_m"] = "ему",
["dat_f"] = "ей",
["dat_p"] = "им",
["acc_f"] = "юю",
["acc_n"] = "ее",
["ins_m"] = "им",
["ins_f"] = {"ей", "ею"},
["ins_p"] = "ими",
["pre_m"] = "ем",
["pre_f"] = "ей",
["pre_p"] = "их",
}
declensions["ой"] = {
["nom_m"] = "о́й",
["nom_n"] = "о́е",
["nom_f"] = "а́я",
["nom_p"] = "ы́е",
["gen_m"] = "о́го",
["gen_f"] = "о́й",
["gen_p"] = "ы́х",
["dat_m"] = "о́му",
["dat_f"] = "о́й",
["dat_p"] = "ы́м",
["acc_f"] = "у́ю",
["acc_n"] = "о́е",
["ins_m"] = "ы́м",
["ins_f"] = {"о́й", "о́ю"},
["ins_p"] = "ы́ми",
["pre_m"] = "о́м",
["pre_f"] = "о́й",
["pre_p"] = "ы́х",
}
-- These can be stressed in ничья́ "draw, tie (in sports)".
declensions["ьий"] = {
["nom_m"] = "и́й",
["nom_n"] = "ье́",
["nom_f"] = "ья́",
["nom_p"] = "ьи́",
["gen_m"] = "ье́го",
["gen_f"] = "ье́й",
["gen_p"] = "ьи́х",
["dat_m"] = "ье́му",
["dat_f"] = "ье́й",
["dat_p"] = "ьи́м",
["acc_f"] = "ью́",
["acc_n"] = "ье́",
["ins_m"] = "ьи́м",
["ins_f"] = {"ье́й", "ье́ю"},
["ins_p"] = "ьи́ми",
["pre_m"] = "ье́м",
["pre_f"] = "ье́й",
["pre_p"] = "ьи́х",
}
declensions["ьей"] = {
["nom_m"] = "е́й",
["nom_n"] = "ьё",
["nom_f"] = "ья́",
["nom_p"] = "ьи́",
["gen_m"] = "ьего́",
["gen_f"] = "ье́й",
["gen_p"] = "ьи́х",
["dat_m"] = "ьему́",
["dat_f"] = "ье́й",
["dat_p"] = "ьи́м",
["acc_f"] = "ью́",
["acc_n"] = "ьё",
["ins_m"] = "ьи́м",
["ins_f"] = {"ье́й", "ье́ю"},
["ins_p"] = "ьи́ми",
["pre_m"] = "ьём",
["pre_f"] = "ье́й",
["pre_p"] = "ьи́х",
}
declensions["short"] = {
["nom_m"] = "",
["nom_n"] = "о́",
["nom_f"] = "а́",
["nom_p"] = "ы́",
["gen_m"] = "а́",
["gen_f"] = "о́й",
["gen_p"] = "ы́х",
["dat_m"] = "у́",
["dat_f"] = "о́й",
["dat_p"] = "ы́м",
["acc_f"] = "у́",
["acc_n"] = "о́",
["ins_m"] = "ы́м",
["ins_f"] = {"о́й", "о́ю"},
["ins_p"] = "ы́ми",
["pre_m"] = "о́м",
["pre_f"] = "о́й",
["pre_p"] = "ы́х",
}
declensions["stressed-short"] = mw.clone(declensions["short"])
declensions["stressed-short"]["pre_m"] = "е́"
declensions["й-short"] = {
["nom_m"] = "й",
["nom_n"] = "ё",
["nom_f"] = "я́",
["nom_p"] = "и́",
["gen_m"] = "его́",
["gen_f"] = "е́й",
["gen_p"] = "и́х",
["dat_m"] = "ему́",
["dat_f"] = "е́й",
["dat_p"] = "и́м",
["acc_f"] = "ю́",
["acc_n"] = "ё",
["ins_m"] = "и́м",
["ins_f"] = {"е́й", "е́ю"},
["ins_p"] = "и́ми",
["pre_m"] = "ём",
["pre_f"] = "е́й",
["pre_p"] = "и́х",
}
declensions["mixed"] = {
["nom_m"] = "",
["nom_n"] = "о",
["nom_f"] = "а",
["nom_p"] = "ы",
["gen_m"] = {"ого", "а2"},
["gen_f"] = "ой",
["gen_p"] = "ых",
["dat_m"] = {"ому", "у2"},
["dat_f"] = "ой",
["dat_p"] = "ым",
["acc_f"] = "у",
["acc_n"] = "о",
["ins_m"] = "ым",
["ins_f"] = {"ой", "ою"},
["ins_p"] = "ыми",
["pre_m"] = "ом",
["pre_f"] = "ой",
["pre_p"] = "ых",
}
internal_notes_table["mixed"] = "<sup>2</sup> Obsolete."
internal_notes_genders["mixed"] = {"m"}
internal_notes_table_old["mixed"] = "<sup>2</sup> Obsolete."
internal_notes_genders_old["mixed"] = {"m"}
declensions["proper"] = {
["nom_m"] = "",
["nom_n"] = nil,
["nom_f"] = "а́",
["nom_p"] = "ы́",
["gen_m"] = "а́",
["gen_f"] = "о́й",
["gen_p"] = "ы́х",
["dat_m"] = "у́",
["dat_f"] = "о́й",
["dat_p"] = "ы́м",
["acc_f"] = "у́",
["acc_n"] = nil,
["ins_m"] = "ы́м",
["ins_f"] = {"о́й", "о́ю1"},
["ins_p"] = "ы́ми",
["pre_m"] = "е́",
["pre_f"] = "о́й",
["pre_p"] = "ы́х",
}
declensions["stressed-proper"] = declensions["proper"]
for _, decl in ipairs({"proper", "stressed-proper"}) do
internal_notes_table[decl] = "<sup>1</sup> Rare."
internal_notes_table_old[decl] = "<sup>1</sup> Rare."
internal_notes_genders[decl] = {"f"}
internal_notes_genders_old[decl] = {"f"}
end
declensions["$"] = {
["nom_m"] = "",
["nom_n"] = "",
["nom_f"] = "",
["nom_p"] = "",
["gen_m"] = "",
["gen_f"] = "",
["gen_p"] = "",
["dat_m"] = "",
["dat_f"] = "",
["dat_p"] = "",
["acc_f"] = "",
-- don't do this; instead we default it to nom_n
-- ["acc_n"] = "",
["ins_m"] = "",
["ins_f"] = "",
["ins_p"] = "",
["pre_m"] = "",
["pre_f"] = "",
["pre_p"] = "",
-- for old-style templates, два, оба, compounds of два
["nom_mp"] = "",
["nom_fp"] = "",
["gen_mp"] = "",
["gen_fp"] = "",
["dat_mp"] = "",
["dat_fp"] = "",
-- don't do this; instead we default them to nom_mp, nom_fp
-- ["acc_mp"] = "",
-- ["acc_fp"] = "",
["ins_mp"] = "",
["ins_fp"] = "",
["pre_mp"] = "",
["pre_fp"] = "",
}
declensions_old["ый"] = {
["nom_m"] = "ый",
["nom_n"] = "ое",
["nom_f"] = "ая",
["nom_mp"] = "ые",
["nom_p"] = "ыя",
["gen_m"] = "аго",
["gen_f"] = "ой",
["gen_p"] = "ыхъ",
["dat_m"] = "ому",
["dat_f"] = "ой",
["dat_p"] = "ымъ",
["acc_f"] = "ую",
["acc_n"] = "ое",
["ins_m"] = "ымъ",
["ins_f"] = {"ой", "ою"},
["ins_p"] = "ыми",
["pre_m"] = "омъ",
["pre_f"] = "ой",
["pre_p"] = "ыхъ",
}
declensions_old["ій"] = {
["nom_m"] = "ій",
["nom_n"] = "ее",
["nom_f"] = "яя",
["nom_mp"] = "іе",
["nom_p"] = "ія",
["gen_m"] = "яго",
["gen_f"] = "ей",
["gen_p"] = "ихъ",
["dat_m"] = "ему",
["dat_f"] = "ей",
["dat_p"] = "имъ",
["acc_f"] = "юю",
["acc_n"] = "ее",
["ins_m"] = "имъ",
["ins_f"] = {"ей", "ею"},
["ins_p"] = "ими",
["pre_m"] = "емъ",
["pre_f"] = "ей",
["pre_p"] = "ихъ",
}
declensions_old["ой"] = {
["nom_m"] = "о́й",
["nom_n"] = "о́е",
["nom_f"] = "а́я",
["nom_mp"] = "ы́е",
["nom_p"] = "ы́я",
["gen_m"] = {"а́го", "о́го"},
["gen_f"] = "о́й",
["gen_p"] = "ы́хъ",
["dat_m"] = "о́му",
["dat_f"] = "о́й",
["dat_p"] = "ы́мъ",
["acc_f"] = "у́ю",
["acc_n"] = "о́е",
["ins_m"] = "ы́мъ",
["ins_f"] = {"о́й", "о́ю"},
["ins_p"] = "ы́ми",
["pre_m"] = "о́мъ",
["pre_f"] = "о́й",
["pre_p"] = "ы́хъ",
}
declensions_old["ьій"] = {
["nom_m"] = "ій",
["nom_n"] = "ье",
["nom_f"] = "ья",
["nom_p"] = "ьи",
["gen_m"] = "ьяго",
["gen_f"] = "ьей",
["gen_p"] = "ьихъ",
["dat_m"] = "ьему",
["dat_f"] = "ьей",
["dat_p"] = "ьимъ",
["acc_f"] = "ью",
["acc_n"] = "ье",
["ins_m"] = "ьимъ",
["ins_f"] = {"ьей", "ьею"},
["ins_p"] = "ьими",
["pre_m"] = "ьемъ",
["pre_f"] = "ьей",
["pre_p"] = "ьихъ",
}
declensions_old["ьей"] = {
["nom_m"] = "е́й",
["nom_n"] = "ьё",
["nom_f"] = "ья́",
["nom_p"] = "ьи́",
["gen_m"] = "ьего́",
["gen_f"] = "ье́й",
["gen_p"] = "ьи́хъ",
["dat_m"] = "ьему́",
["dat_f"] = "ье́й",
["dat_p"] = "ьи́мъ",
["acc_f"] = "ью́",
["acc_n"] = "ьё",
["ins_m"] = "ьи́мъ",
["ins_f"] = {"ье́й", "ье́ю"},
["ins_p"] = "ьи́ми",
["pre_m"] = "ьёмъ",
["pre_f"] = "ье́й",
["pre_p"] = "ьи́хъ",
}
declensions_old["short"] = {
["nom_m"] = "ъ",
["nom_n"] = "о",
["nom_f"] = "а",
["nom_p"] = "ы",
["gen_m"] = "а",
["gen_f"] = "ой",
["gen_p"] = "ыхъ",
["dat_m"] = "у",
["dat_f"] = "ой",
["dat_p"] = "ымъ",
["acc_f"] = "у",
["acc_n"] = "о",
["ins_m"] = "ымъ",
["ins_f"] = {"ой", "ою"},
["ins_p"] = "ыми",
["pre_m"] = "омъ",
["pre_f"] = "ой",
["pre_p"] = "ыхъ",
}
declensions_old["stressed-short"] = mw.clone(declensions_old["short"])
declensions_old["stressed-short"]["pre_m"] = "ѣ́"
declensions_old["й-short"] = {
["nom_m"] = "й",
["nom_n"] = "ё",
["nom_f"] = "я́",
["nom_p"] = "и́",
["gen_m"] = "его́",
["gen_f"] = {"е́й", "ея́"},
["gen_p"] = "и́хъ",
["dat_m"] = "ему́",
["dat_f"] = "е́й",
["dat_p"] = "и́мъ",
["acc_f"] = "ю́",
["acc_n"] = "ё",
["ins_m"] = "и́мъ",
["ins_f"] = {"е́й", "е́ю"},
["ins_p"] = "и́ми",
["pre_m"] = "ёмъ",
["pre_f"] = "е́й",
["pre_p"] = "и́хъ",
}
declensions_old["mixed"] = {
["nom_m"] = "ъ",
["nom_n"] = "о",
["nom_f"] = "а",
["nom_p"] = "ы",
["gen_m"] = {"аго", "а1"},
["gen_f"] = "ой",
["gen_p"] = "ыхъ",
["dat_m"] = {"ому", "у1"},
["dat_f"] = "ой",
["dat_p"] = "ымъ",
["acc_f"] = "у",
["acc_n"] = "о",
["ins_m"] = "ымъ",
["ins_f"] = {"ой", "ою"},
["ins_p"] = "ыми",
["pre_m"] = "омъ",
["pre_f"] = "ой",
["pre_p"] = "ыхъ",
}
declensions_old["proper"] = {
["nom_m"] = "ъ",
["nom_n"] = nil,
["nom_f"] = "а́",
["nom_p"] = "ы́",
["gen_m"] = "а́",
["gen_f"] = "о́й",
["gen_p"] = "ы́хъ",
["dat_m"] = "у́",
["dat_f"] = "о́й",
["dat_p"] = "ы́мъ",
["acc_f"] = "у́",
["acc_n"] = nil,
["ins_m"] = "ы́мъ",
["ins_f"] = {"о́й", "о́ю1"},
["ins_p"] = "ы́ми",
["pre_m"] = "ѣ́",
["pre_f"] = "о́й",
["pre_p"] = "ы́хъ",
}
declensions_old["stressed-proper"] = declensions_old["proper"]
declensions_old["$"] = declensions["$"]
local function frob_genitive_masc(decl)
-- signal to combine_stem_and_suffix() to use the special tr_adj()
-- function so that -го gets transliterated to -vo
if type(decl["gen_m"]) == "table" then
local entries = {}
for _, entry in ipairs(decl["gen_m"]) do
table.insert(entries, rsub(entry, "го$", "го<adj>"))
end
decl["gen_m"] = entries
else
decl["gen_m"] = rsub(decl["gen_m"], "го$", "го<adj>")
end
end
-- Frob declensions, adding <adj> to gen_m forms ending in -го. This is
-- a signal to add manual translit early on that renders -го as -vo.
for _, decl in pairs(declensions_old) do
frob_genitive_masc(decl)
end
for _, decl in pairs(declensions) do
frob_genitive_masc(decl)
end
--------------------------------------------------------------------------
-- Declension functions --
--------------------------------------------------------------------------
local function attach_unstressed(args, suf, short)
local stem, stemtr
if short then
stem, stemtr = args.short_stem, args.short_stemtr
else
stem, stemtr = args.stem, args.stemtr
end
if suf == nil then
return nil
elseif nom.nonsyllabic_suffixes[suf] then
if not args.bare then
return nil
elseif rfind(args.bare, "[йьъ]$") then
return {args.bare, args.baretr}
elseif suf == "ъ" then
return com.concat_russian_tr(args.bare, args.baretr, suf, nil, "dopair")
else
return {args.bare, args.baretr}
end
end
suf = com.make_unstressed(suf)
local rules = nom.unstressed_rules[ulower(usub(stem, -1))]
-- The parens around the return value drop all but the first return value
return (nom.combine_stem_and_suffix(stem, stemtr, suf, rules, args.old))
end
local function attach_stressed(args, suf, short)
local ustem, ustemtr
if short then
ustem, ustemtr = args.short_ustem, args.short_ustemtr
else
ustem, ustemtr = args.ustem, args.ustemtr
end
if suf == nil then
return nil
elseif not rfind(suf, "[ё́]") then -- if suf has no "ё" or accent marks
return attach_unstressed(args, suf, short)
end
local rules = nom.stressed_rules[ulower(usub(ustem, -1))]
-- The parens around the return value drop all but the first return value
return (nom.combine_stem_and_suffix(ustem, ustemtr, suf, rules, args.old))
end
local function attach_both(args, suf, short)
local results = {}
-- Examples with stems with ё on Zaliznyak p. 61 list the
-- ending-stressed forms first, so go with that.
-- NOTE: This assumes we get one value returned, not a list of such values.
table.insert(results, attach_stressed(args, suf, short))
table.insert(results, attach_unstressed(args, suf, short))
return results
end
local function attach_with(args, suf, fun, short)
if type(suf) == "table" then
local all_combineds = {}
for _, x in ipairs(suf) do
local combineds = attach_with(args, x, fun, short)
for _, combined in ipairs(combineds) do
table.insert(all_combineds, combined)
end
end
return all_combineds
else
local funval = fun(args, suf, short)
if funval then
local tbl = {}
assert(type(funval) == "table")
if type(funval[1]) ~= "table" then
funval = {funval}
end
for _, x in ipairs(funval) do
table.insert(tbl,
com.concat_paired_russian_tr(x, args.real_suffix))
end
return tbl
else
return {}
end
end
end
local function gen_form(args, decl, case, fun)
if not args.forms[case] then
args.forms[case] = {}
end
insert_forms_into_existing_forms(args.forms[case],
attach_with(args, decl[case], fun, false))
end
decline = function(args, decl, stressed)
local attacher = stressed and attach_stressed or attach_unstressed
for _, case in ipairs(long_cases) do
gen_form(args, decl, case, attacher)
end
end
canonicalize_override = function(args, case)
local val = args[case]
if not val then
return nil
end
val = rsplit(val, "%s*,%s*")
-- auto-accent/check for necessary accents
local newvals = {}
for _, v in ipairs(val) do
local ru, tr = com.split_russian_tr(v)
if not args.allow_unaccented then
if tr and com.is_unstressed(ru) ~= com.is_unstressed(tr) then
error("Override " .. ru .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(ru) then
ru, tr = com.make_ending_stressed(ru, tr)
elseif com.needs_accents(ru) then
error("Override " .. ru .. " for case " .. case .. " requires an accent")
end
end
table.insert(newvals, {ru, tr})
end
val = newvals
return val
end
handle_forms_and_overrides = function(args, short_forms_allowed)
local f = args.forms
local function append_note_all(case, value)
value = com.split_russian_tr(value, "dopair")
if f[case] then
for i=1,#f[case] do
f[case][i] = com.concat_paired_russian_tr(f[case][i], value)
end
end
end
local function append_note_last(case, value, gt_one)
value = com.split_russian_tr(value, "dopair")
if f[case] then
local lastarg = #f[case]
if lastarg > (gt_one and 1 or 0) then
f[case][lastarg] = com.concat_paired_russian_tr(f[case][lastarg], value)
end
end
end
for _, case in ipairs(long_cases) do
if args[case .. "_tail"] then
append_note_last(case, args[case .. "_tail"])
end
if args[case .. "_tailall"] then
append_note_all(case, args[case .. "_tailall"])
end
args[case] = canonicalize_override(args, case) or f[case]
end
for _, case in ipairs(short_cases) do
if short_forms_allowed then
if args[case .. "_tail"] then
append_note_last(case, args[case .. "_tail"])
end
if args[case .. "_tailall"] then
append_note_all(case, args[case .. "_tailall"])
end
if args.shorttailall then
append_note_all(case, args.shorttailall)
end
if args.shorttail then
append_note_last(case, args.shorttail, ">1")
end
args[case] = canonicalize_override(args, case) or f[case]
else
args[case] = nil
end
end
-- Convert an empty list to nil, so that an mdash is inserted. This happens,
-- for example, with words like голубой where args.bare is set to nil.
for _, case in ipairs(all_cases) do
if args[case] and #args[case] == 0 then
args[case] = nil
end
end
-- default acc_n to nom_n; applies chiefly in the indeclinable declension
-- (used with manual declension tables)
if not args.acc_n then
args.acc_n = args.nom_n
end
-- default inanimate/animate accusative variants as appropriate; this is
-- almost always correct, but not with e.g. два́дцать оди́н, where the
-- masculine animate accusative два́дцать одного́ differs from both the
-- masculine nominative два́дцать оди́н and the masculine genitive
-- двадцати́ одного́.
if not args.acc_m_an then
args.acc_m_an = args.gen_m
end
if not args.acc_m_in then
args.acc_m_in = args.nom_m
end
-- Compounds of два do not have animacy; everything else does.
-- Only copy either the with-animacy or without-animacy variants to avoid
-- having both forms in {{ru-generate-adj-forms}}.
if args.special ~= "cdva" then
if not args.acc_p_an then
args.acc_p_an = args.gen_p
end
if not args.acc_p_in then
args.acc_p_in = args.nom_p
end
if not args.acc_mp_an then
args.acc_mp_an = args.gen_mp
end
if not args.acc_mp_in then
args.acc_mp_in = args.nom_mp
end
if not args.acc_fp_an then
args.acc_fp_an = args.gen_fp
end
if not args.acc_fp_in then
args.acc_fp_in = args.nom_fp
end
else
if not args.acc_mp then
args.acc_mp = args.nom_mp
end
if not args.acc_fp then
args.acc_fp = args.nom_fp
end
end
end
--------------------------------------------------------------------------
-- Short adjective declension --
--------------------------------------------------------------------------
short_declensions["ый"] = { m="", f="а́", n="о́", p="ы́" }
short_declensions["ой"] = short_declensions["ый"]
short_declensions["ий"] = { m="ь", f="я́", n="е́", p="и́" }
short_declensions_old["ый"] = { m="ъ", f="а́", n="о́", p="ы́" }
short_declensions_old["ой"] = short_declensions_old["ый"]
short_declensions_old["ій"] = short_declensions["ий"]
-- Short adjective stress patterns:
-- "-" = stem-stressed
-- "+" = ending-stressed (drawn onto the last syllable of stem in masculine)
-- "-+" = both possibilities
short_stress_patterns["a"] = { m="-", f="-", n="-", p="-" }
short_stress_patterns["a'"] = { m="-", f="-+", n="-", p="-" }
short_stress_patterns["b"] = { m="+", f="+", n="+", p="+" }
short_stress_patterns["b'"] = { m="+", f="+", n="+", p="-+" }
short_stress_patterns["c"] = { m="-", f="+", n="-", p="-" }
short_stress_patterns["c'"] = { m="-", f="+", n="-", p="-+" }
short_stress_patterns["c''"] = { m="-", f="+", n="-+", p="-+" }
local function gen_short_form(args, decl, case, fun, datedrare)
if not args.forms["short_" .. case] then
args.forms["short_" .. case] = {}
end
local inserted = insert_forms_into_existing_forms(
args.forms["short_" .. case], attach_with(args, decl[case], fun, true),
(datedrare == "dated" and datedrare == "rare") and "*" or nil)
if datedrare == "dated" and inserted then
m_table.insertIfNot(args.internal_notes, "<sup>*</sup> Dated.")
elseif datedrare == "rare" and inserted then
m_table.insertIfNot(args.internal_notes, "<sup>*</sup> Rare.")
end
end
local attachers = {
["+"] = attach_stressed,
["-"] = attach_unstressed,
["-+"] = attach_both,
}
decline_short = function(args, decl, stress_pattern, dated)
if stress_pattern then
for _, case in ipairs({"m", "f", "n", "p"}) do
gen_short_form(args, decl, case, attachers[stress_pattern[case]],
dated)
end
end
end
--------------------------------------------------------------------------
-- Create the table --
--------------------------------------------------------------------------
local title_temp = [=[Declension of <b lang="ru" class="Cyrl">{lemma}</b> ({short_class_title})]=]
local old_title_temp = [=[Pre-reform declension of <b lang="ru" class="Cyrl">{lemma}</b> ({short_class_title})]=]
local title_temp_no_short_msg = [=[Declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local old_title_temp_no_short_msg = [=[Pre-reform declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local template = nil
local full_clause = nil
local template_no_neuter = nil
local full_clause_no_neuter = nil
local surname_full_clause = nil
local template_mp = nil
local full_clause_mp = nil
local template_mp_no_neuter = nil
local full_clause_mp_no_neuter = nil
local surname_full_clause_mp = nil
local template_dva = nil
local full_clause_dva = nil
local full_clause_compound_dva = nil
local template_oba = nil
local full_clause_oba = nil
local short_clause_separator = nil
local short_clause = nil
local short_clause_no_neuter_separator = nil
local short_clause_no_neuter = nil
local short_clause_mp_separator = nil
local short_clause_mp = nil
local short_clause_mp_no_neuter_separator = nil
local short_clause_mp_no_neuter = nil
local internal_notes_template = nil
local notes_template = nil
local function get_accel_forms(old, special, has_nom_mp)
return {
-- used with all variants
nom_m = "nom|m|s",
nom_f = "nom|f|s",
nom_n = "nom|n|s",
-- not used with special; applies to all genders normally but only
-- feminine and neuter with old=1
nom_p = old and has_nom_mp and "nom|f//n|p" or "nom|p",
-- only used with old=1 or special; applies to the masculine and neuter if
-- special, but only masculine if old=1
nom_mp = special and "nom|m//n|p" or "nom|m|p",
-- only used with special
nom_fp = "nom|f|p",
-- the remaining singulars and non-gendered plurals used with all variants
-- except special == "oba"
gen_m = "gen|m//n|s",
gen_f = "gen|f|s",
gen_p = "gen|p",
dat_m = "dat|m//n|s",
dat_f = "dat|f|s",
dat_p = "dat|p",
acc_m_an = "an|acc|m|s",
acc_m_in = "in|acc|m|s",
acc_f = "acc|f|s",
acc_n = "acc|n|s",
-- the following two not used with special in ("dva", "oba"); applies to
-- all genders normally but only feminine and neuter with old=1
acc_p_an = old and has_nom_mp and "an|acc|f//n|p" or "an|acc|p",
acc_p_in = old and has_nom_mp and "in|acc|f//n|p" or "in|acc|p",
-- the following two only used with old=1 or special in ("dva|oba");
-- applies to the masculine and neuter if special, but only masculine if
-- old=1
acc_mp_an = special and "an|acc|m//n|p" or "an|acc|m|p",
acc_mp_in = special and "in|acc|m//n|p" or "in|acc|m|p",
-- the following two only used with special in ("dva", "oba")
acc_fp_an = "an|acc|f|p",
acc_fp_in = "in|acc|f|p",
-- the next 6 are used with all variants except special == "oba"
ins_m = "ins|m//n|s",
ins_f = "ins|f|s",
ins_p = "ins|p",
pre_m = "pre|m//n|s",
pre_f = "pre|f|s",
pre_p = "pre|p",
-- the following two gendered plurals are only used with special == "cdva"
acc_mp = "acc|m//n|p",
acc_fp = "acc|f|p",
-- the remaining gendered plurals are only used with special == "oba"
gen_mp = "gen|m//n|p",
gen_fp = "gen|f|p",
dat_mp = "dat|m//n|p",
dat_fp = "dat|f|p",
ins_mp = "ins|m//n|p",
ins_fp = "ins|f|p",
pre_mp = "pre|m//n|p",
pre_fp = "pre|f|p",
-- short forms
short_m = "short|m|s",
short_f = "short|f|s",
short_n = "short|n|s",
short_p = "short|p",
}
end
-- Make the table
make_table = function(args)
local lemma_forms = args.special and args.nom_mp or
args.nofull and args.short_m or args.nom_m
args.lemma = m_links.remove_links(nom.show_form(lemma_forms, true, nil, nil))
args.title = args.title or strutils.format(
(args.special or args.manual) and args.old and old_title_temp_no_short_msg or
(args.special or args.manual) and title_temp_no_short_msg or
args.old and old_title_temp or
title_temp,
args)
local has_nom_mp = args.nom_mp and not (#args.nom_mp == 1 and args.nom_mp[1][1] == "-")
local accel_forms = get_accel_forms(args.old, args.special, has_nom_mp)
for _, case in ipairs(all_cases) do
if args[case] then
local accel_form = accel_forms[case]
if not accel_form then
error("Internal error: Unrecognized case " .. case .. " when looking up accelerator form")
end
if args.noneuter or args.real_surname then
accel_form = rsub(accel_form, "//n", "")
end
if args.real_surname then
accel_form = rsub(accel_form, "an|", "")
end
args[case] = nom.show_form(args[case], false, accel_form, lemma_forms)
else
args[case] = nil
end
end
local temp, fullc
if args.special == "oba" then
temp = template_oba
fullc = full_clause_oba
elseif args.special == "dva" then
temp = template_dva
fullc = full_clause_dva
elseif args.special == "cdva" then
temp = template_dva -- no template_cdva
fullc = full_clause_compound_dva
elseif args.real_surname and args.old and has_nom_mp then
temp = template_mp_no_neuter
fullc = surname_full_clause_mp
elseif args.real_surname then
temp = template_no_neuter
fullc = surname_full_clause
elseif args.noneuter and args.old and has_nom_mp then
temp = template_mp_no_neuter
fullc = full_clause_mp_no_neuter
elseif args.noneuter then
temp = template_no_neuter
fullc = full_clause_no_neuter
elseif args.old and has_nom_mp then
temp = template_mp
fullc = full_clause_mp
else
temp = template
fullc = full_clause
end
if args.old then
if has_nom_mp then
if args.short_m or args.short_n or args.short_f or args.short_p then
args.short_m = args.short_m or "—"
args.short_n = args.short_n or "—"
args.short_f = args.short_f or "—"
args.short_p = args.short_p or "—"
args.shortsep = not args.nofull and (
args.noneuter and short_clause_mp_no_neuter_separator or short_clause_mp_separator
) or ""
args.short_clause = strutils.format(args.noneuter and short_clause_mp_no_neuter or short_clause_mp, args)
else
args.short_clause = ""
end
else
args.short_clause = ""
end
else
if args.short_m or args.short_n or args.short_f or args.short_p then
args.short_m = args.short_m or "—"
args.short_n = args.short_n or "—"
args.short_f = args.short_f or "—"
args.short_p = args.short_p or "—"
args.shortsep = not args.nofull and (
args.noneuter and short_clause_no_neuter_separator or short_clause_separator
) or ""
args.short_clause = strutils.format(args.noneuter and short_clause_no_neuter or short_clause, args)
else
args.short_clause = ""
end
end
if not args.nofull then
args.full_clause = strutils.format(fullc, args)
else
args.full_clause = ""
end
args.internal_notes = table.concat(args.internal_notes, "<br />")
args.internal_notes_clause = #args.internal_notes > 0 and strutils.format(internal_notes_template, args) or ""
args.notes_clause = args.notes and strutils.format(notes_template, args) or ""
return strutils.format(temp, args)
end
-- Used for new-style templates
short_clause_separator = [===[
! style="height:0.2em;background:var(--wikt-palette-lightblue)" colspan=6 |
|-
]===]
-- Used for new-style templates
short_clause = [===[
{shortsep}! style="background:var(--wikt-palette-lighterblue)" colspan=2 | short form
| data-accel-col=1 | {short_m}
| data-accel-col=3 | {short_n}
| data-accel-col=2 | {short_f}
| data-accel-col=4 | {short_p}]===]
-- Used for new-style templates
short_clause_no_neuter_separator = [===[
! style="height:0.2em;background:var(--wikt-palette-lightblue)" colspan=5 |
|-
]===]
-- Used for new-style templates
short_clause_no_neuter = [===[
{shortsep}! style="background:var(--wikt-palette-lighterblue)" colspan=2 | short form
| data-accel-col=1 | {short_m}
| data-accel-col=2 | {short_f}
| data-accel-col=4 | {short_p}]===]
-- Used for old-style templates
short_clause_mp_separator = [===[
! style="height:0.2em;background:var(--wikt-palette-lightblue)" colspan=7 |
|-
]===]
-- Used for old-style templates
short_clause_mp = [===[
{shortsep}! style="background:var(--wikt-palette-lighterblue)" colspan=2 | short form
| data-accel-col=1 | {short_m}
| data-accel-col=3 | {short_n}
| data-accel-col=2 | {short_f}
| data-accel-col=4 colspan=2 | {short_p}]===]
-- Used for old-style templates
short_clause_mp_no_neuter_separator = [===[
! style="height:0.2em;background:var(--wikt-palette-lightblue)" colspan=6 |
|-
]===]
-- Used for old-style templates
short_clause_mp_no_neuter = [===[
{shortsep}! style="background:var(--wikt-palette-lighterblue)" colspan=2 | short form
| data-accel-col=1 | {short_m}
| data-accel-col=2 | {short_f}
| data-accel-col=4 colspan=2 | {short_p}]===]
-- Used for both new-style and old-style templates
notes_template = [===[
<div style="width:100%;text-align:left">
<div style="display:inline-block;text-align:left;padding-left:1em;padding-right:1em">
{notes}
</div></div>
]===]
-- Used for both new-style and old-style templates
internal_notes_template = rsub(notes_template, "notes", "internal_notes")
local function template_prelude(min_width)
min_width = min_width or "70"
return rsub([===[
<div>
<div class="NavFrame">
<div class="NavHead" style="background:var(--wikt-palette-lighterblue)">{title}</div>
<div class="NavContent">
{\op}| style="background:var(--wikt-palette-paleblue);text-align:center" class="inflection-table"
|-
]===], "MINWIDTH", min_width)
end
local function template_postlude()
return [===[{full_clause}|-{short_clause}
|{\cl}{internal_notes_clause}{notes_clause}</div></div></div>]===]
end
-- Used for both new-style and old-style templates
template = template_prelude() .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine
! style="background:var(--wikt-palette-lightblue)" | neuter
! style="background:var(--wikt-palette-lightblue)" | feminine
! style="background:var(--wikt-palette-lightblue)" | plural
]===] .. template_postlude()
-- Used for both new-style and old-style templates
full_clause = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=3 | {nom_n}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 colspan=2 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 colspan=2 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=1 | {acc_m_an}
| data-accel-col=3 rowspan=2 | {acc_n}
| data-accel-col=2 rowspan=2 | {acc_f}
| data-accel-col=4 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=1 | {acc_m_in}
| data-accel-col=4 | {acc_p_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 colspan=2 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 colspan=2 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 | {pre_p}
]===]
-- Used for both new-style and old-style templates
template_no_neuter = template_prelude("55") .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine
! style="background:var(--wikt-palette-lightblue)" | feminine
! style="background:var(--wikt-palette-lightblue)" | plural
]===] .. template_postlude()
-- Used for both new-style and old-style templates
full_clause_no_neuter = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=1 | {acc_m_an}
| data-accel-col=2 rowspan=2 | {acc_f}
| data-accel-col=4 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=1 | {acc_m_in}
| data-accel-col=4 | {acc_p_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 | {pre_p}
]===]
-- Used for both new-style and old-style templates
surname_full_clause = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | accusative
| data-accel-col=1 | {acc_m_an}
| data-accel-col=2 | {acc_f}
| data-accel-col=4 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 | {pre_p}
]===]
-- Used for old-style templates
template_mp = template_prelude() .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine
! style="background:var(--wikt-palette-lightblue)" | neuter
! style="background:var(--wikt-palette-lightblue)" | feminine
! style="background:var(--wikt-palette-lightblue)" | m. plural
! style="background:var(--wikt-palette-lightblue)" | n./f. plural
]===] .. template_postlude()
-- Used for old-style templates
full_clause_mp = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=3 | {nom_n}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 colspan=2 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 colspan=2 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 colspan=2 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 colspan=2 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=1 | {acc_m_an}
| data-accel-col=3 rowspan=2 | {acc_n}
| data-accel-col=2 rowspan=2 | {acc_f}
| data-accel-col=4 colspan=2 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=1 | {acc_m_in}
| data-accel-col=4 | {acc_mp_in}
| data-accel-col=5 | {acc_p_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 colspan=2 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 colspan=2 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 colspan=2 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 colspan=2 | {pre_p}
]===]
-- Used for old-style templates
template_mp_no_neuter = template_prelude("60") .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine
! style="background:var(--wikt-palette-lightblue)" | feminine
! style="background:var(--wikt-palette-lightblue)" | m. plural
! style="background:var(--wikt-palette-lightblue)" | f. plural
]===] .. template_postlude()
-- Used for old-style templates
full_clause_mp_no_neuter = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 colspan=2 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 colspan=2 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=1 | {acc_m_an}
| data-accel-col=2 rowspan=2 | {acc_f}
| data-accel-col=4 colspan=2 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=1 | {acc_m_in}
| data-accel-col=4 | {acc_mp_in}
| data-accel-col=5 | {acc_p_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 colspan=2 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 colspan=2 | {pre_p}
]===]
-- Used for some old-style templates
surname_full_clause_mp = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 colspan=2 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 colspan=2 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | accusative
| data-accel-col=1 | {acc_m_an}
| data-accel-col=2 | {acc_f}
| data-accel-col=4 colspan=2 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 colspan=2 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 colspan=2 | {pre_p}
]===]
-- Used for два and compounds
template_dva = template_prelude("55") .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine/neuter
! style="background:var(--wikt-palette-lightblue)" | feminine
]===] .. template_postlude()
-- Used for both new-style and old-style templates of два (only for два itself,
-- which has an animacy distinction; not for compounds)
full_clause_dva = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=4 colspan=2 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=4 colspan=2 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=4 colspan=2 | {acc_p_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=4 | {acc_mp_in}
| data-accel-col=5 | {acc_fp_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=4 colspan=2 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=4 colspan=2 | {pre_p}
]===]
-- Used for both new-style and old-style templates of compounds of два
full_clause_compound_dva = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=4 colspan=2 | {gen_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=4 colspan=2 | {dat_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | accusative
| data-accel-col=4 | {acc_mp}
| data-accel-col=5 | {acc_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=4 colspan=2 | {ins_p}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=4 colspan=2 | {pre_p}
]===]
-- Used for оба
template_oba = template_prelude() .. [===[
! style="width:20%;background:var(--wikt-palette-lightblue)" colspan=2 |
! style="background:var(--wikt-palette-lightblue)" | masculine
! style="background:var(--wikt-palette-lightblue)" | neuter
! style="background:var(--wikt-palette-lightblue)" | feminine
! style="background:var(--wikt-palette-lightblue)" | m./n. plural
! style="background:var(--wikt-palette-lightblue)" | f. plural
]===] .. template_postlude()
-- Used for both new-style and old-style templates of оба
full_clause_oba = [===[
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | nominative
| data-accel-col=1 | {nom_m}
| data-accel-col=3 | {nom_n}
| data-accel-col=2 | {nom_f}
| data-accel-col=4 | {nom_mp}
| data-accel-col=5 | {nom_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | genitive
| data-accel-col=1 colspan=2 | {gen_m}
| data-accel-col=2 | {gen_f}
| data-accel-col=4 | {gen_mp}
| data-accel-col=5 | {gen_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | dative
| data-accel-col=1 colspan=2 | {dat_m}
| data-accel-col=2 | {dat_f}
| data-accel-col=4 | {dat_mp}
| data-accel-col=5 | {dat_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" rowspan=2 | accusative
! style="background:var(--wikt-palette-lighterblue)" | animate
| data-accel-col=1 | {acc_m_an}
| data-accel-col=3 rowspan=2 | {acc_n}
| data-accel-col=2 rowspan=2 | {acc_f}
| data-accel-col=4 | {acc_mp_an}
| data-accel-col=5 | {acc_fp_an}
|-
! style="background:var(--wikt-palette-lighterblue)" | inanimate
| data-accel-col=1 | {acc_m_in}
| data-accel-col=4 | {acc_mp_in}
| data-accel-col=5 | {acc_fp_in}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | instrumental
| data-accel-col=1 colspan=2 | {ins_m}
| data-accel-col=2 | {ins_f}
| data-accel-col=4 | {ins_mp}
| data-accel-col=5 | {ins_fp}
|-
! style="background:var(--wikt-palette-lighterblue)" colspan=2 | prepositional
| data-accel-col=1 colspan=2 | {pre_m}
| data-accel-col=2 | {pre_f}
| data-accel-col=4 | {pre_mp}
| data-accel-col=5 | {pre_fp}
]===]
return export
ljdyldnivfj65ztu0wa504p25r3fgl3
Modül:User:Benwing2/ru-noun
828
1589239
5669925
2026-06-23T07:33:07Z
MustafaCavlak
59368
imported from en
5669925
Scribunto
text/plain
--[=[
This module contains functions for creating inflection tables for Russian
nouns.
Author: Benwing, rewritten from early version by Wikitiki89
Form of arguments: One of the following:
1. LEMMA|DECL|PLSTEM (all arguments optional)
2. ACCENT|LEMMA|DECL|PLSTEM (all arguments optional)
3. multiple sets of arguments separated by the literal word "or"
Arguments:
ACCENT: Accent pattern (a b c d e f b' d' f' f''). Multiple values can
be specified, separated by commas. If omitted, defaults to a or b
depending on the position of stress on the lemma or
explicitly-specified declension.
LEMMA: Lemma form (i.e. nom sg or nom pl), with appropriately-placed
stress; or the stem, if an explicit declension is specified
(in this case, the declension usually looks like an ending, and
the stem is the portion of the lemma minus the ending). In the
first argument set (i.e. first set of arguments separated
by "or"), defaults to page name; in later sets, defaults to lemma
of previous set. A plural form can be given, and causes argument
n= to default to n=p (plural only). Normally, an accent is
required if multisyllabic, and unaccented monosyllables with
automatically be stressed; prefix with * to override both behaviors.
DECL: Declension field. Normally omitted to autodetect based on the
lemma form; see below.
PLSTEM: special plural stem (defaults to stem of lemma)
Additional named arguments:
a: animacy (a/an/anim = animate, i/in/inan = inanimate, b/bi/both/ai = both
(listing animate first in the headword), ia = both (listing inanimate
first in the headword), otherwise inanimate)
n: number restriction (p = plural only, s = singular only, b = both;
defaults to both unless the lemma is plural, in which case it
defaults to plural only)
CASE_NUM or acc_NUM_ANIM or par/loc/voc: override (or multiple
values separated by commas) for case/number combination;
forms auto-linked; can have raw links in it, can have an
ending "note" (*, +, 1, 2, 3, etc.)
pltail: Specify something (usually a * or similar) to attach to
the end of the last plural form when there's more than one. Used in
conjunction with notes= to indicate that alternative plural forms
are obsolete, poetic, etc.
pltailall: Similar pltail= but attaches to all plural forms.
Typically used in conjunction with notes= to make a comment about
the plural as a whole (e.g. it's mostly hypothetical, rare and
awkward, etc.).
sgtail, sgtailall: Same as pltail=, pltailall= but for the singular.
obltail, obltailall: Same as pltail=, pltailall= but for oblique cases
(not the nominative or accusative).
CASE_NUM_tail: Attach the argument to the end of the last form
(whether there's one or more than one) for the particular
case/number combination. Note that this doesn't work quite like
pltail= or sgtail= in that it doesn't skip adding the argument
when there's only one form.
CASE_NUM_tailall: Attach the argument to the end of all forms
specified for the particular case/number combination. Similar to
pltailall= or sgtailall=.
suffix: Add a suffix such as ся to all forms.
prefix: Add a prefix to all forms.
plhyp, plhypall, CASE_NUM_hyp, etc.: Same as pltail, pltailall,
CASE_NUM_tal, etc. but specify that the marked forms are
mostly hypothetical or rare/awkard. Generally you will want
plhypall=y to mark the plural as hypothetical.
Per word named arguments:
All of the above named arguments have per-word variants, e.g.
a1, a2, ...; n1, n2, ...; CASE_NUM1, CASE_NUM2, ...;
pltail1, pltail2, ...; etc. These apply to the individual words of a
form.
Case abbreviations:
nom: nominative
gen: genitive
dat: dative
acc: accusative
ins: instrumental
pre: prepositional
par: partitive
loc: locative
voc: vocative
Number abbreviations:
sg: singular
pl: plural
Animacy abbreviations:
an: animate
in: inanimate
Declension field:
One of the following for regular nouns:
(blank)
GENDER
-VARIANT
GENDER-VARIANT
$
DECLTYPE
DECLTYPE/DECLTYPE
(also, can append various special-case markers to any of the above)
Or one of the following for adjectival nouns:
+
+ь
+short, +mixed or +proper
+DECLTYPE
GENDER if present is m, f, n or 3f; for regular nouns, required if the
lemma ends in -ь or is plural, ignored otherwise. 3f is the same
as f but in the case of a plural lemma in -и, detects a
third-declension feminine with singular in -ь rather than a
first-declension feminine with singular in -а or -я.
VARIANT is one way of requesting variant declensions (see also special
case (1) for variant nom pls, and special case (2) for variant
gen pls). The currently allowed values are -ья (will select a
mixed declension that has some normal declension in its singular
and the -ья plural declension); -ин (for animate masculine nouns
in -ин with plural in -е -- note that this is autodetected in the
majority of cases where the ending in -янин or -анин); -ишко
(used for inanimate neuter-form diminutive masculine nouns in
-ишко [also сараю́шко] with nom pl -и and colloquial feminine-ending
alternants in some singular cases); -ище (similar to -ишко but
used for *animate* augmentative masculine neuter-form nouns in
-ище). Variants -ишко and -ище must be given with with special
case (1).
$ is for for invariable nouns. It are principally useful in multiword
expressions where some of the words are invariable.
DECLTYPE is an explicit declension type. Normally you shouldn't use
this, and should instead let the declension type be autodetected
based on the ending, supplying the appropriate hint if needed
(gender for regular nouns, +ь for adjectives). If provided, the
declension type is usually the same as the ending, and if present,
the lemma field should be just the stem, without the ending.
Possibilities for regular nouns are (blank) or # for hard-consonant
declension, а, я, о, е or ё, е́, й, ья, ье or ьё, ь-m, ь-f,
ин, ёнок or онок or енок, ёночек or оночек or еночек, мя,
-а or #-а, ь-я, й-я, о-и or о-ы, -ья or #-ья, $ (invariable).
Old-style (pre-reform) declensions use ъ instead of (blank), ъ-а
instead of -а, ъ-ья instead of -ья, and инъ, ёнокъ/онокъ/енокъ,
ёночекъ/оночекъ/еночекъ instead of the same without terminating ъ.
The declensions can also be written with an accent on them; this
chooses the same declension (except for е vs. е́), but causes
ACCENT to default to pattern b instead of a.
For adjectival nouns, you should normally supply just + and let
the ending determine the declension; supply +ь in the case of a
possessive adjectival noun in -ий, which have an extra -ь- in
most endings compared with normal adjectival nouns in -ий, but
which can't be distinguished based on the nominative singular.
You can also supply +short, +mixed or +proper, which constrains
the declension appropriately but still autodetects the
gender-specific and stress-specific variant. If you do supply
a specific declension type, as with regular nouns you need to
omit the ending from the lemma field and supply just the stem.
Possibilities are +ый, +ое, +ая, +ій, +ее, +яя, +ой, +о́е, +а́я,
+ьій, +ье, +ья; +-short or +#-short (masc); +о-short,
+о-stressed-short or +о́-short; +а-short, +а-stressed-short or
+а́-short; and similar for -mixed and -proper (except there aren't
any stressed mixed declensions).
DECLTYPE/DECLTYPE is used for nouns with one declension in the
singular and a different one in the plural, for cases that
PLVARIANT and special case (1) below don't cover.
Special-case markers:
(1) for Zaliznyak-style alternate nominative plural ending:
-а or -я for masculine, -и or -ы for neuter
(2) for Zaliznyak-style alternate genitive plural ending:
-ъ/none for masculine, -ей for feminine, -ов(ъ) for neuter,
-ей for plural variant -ья
* for reducibles (nom sg or gen pl has an extra vowel before the
final consonant as compared with the stem found in other cases)
;ё for Zaliznyak-style alternation between last е in stem and ё
TODO:
1. Multi-word issues:
-- FIXME: Make sure internal_notes handled correctly; we may run into issues
with multiple internal notes from different words, if we're not careful
to use different footnote symbols for each type of footnote (which we
don't do currently). [NOT DONE, MAY NOT DO]
-- Handling default lemma: With multiple words, we should probably split
the page name on spaces and default each word in turn [NOT DONE, MAY
NOT DO]
2a. FIXME: For -ишко diminutives and -ище augmentatives, should add an
appropriate category of some sort (currently marked by colloqfem= in
category).
2b. FIXME: Adding a note to dat_sg also adds it to loc_sg when it exists;
seems wrong. See луг.
2c. FIXME: When you have both d' and f in feminines and you use sgtail=*,
you get two *'s. See User:Benwing2/test-ru-noun-debug.
3. ADJECTIVE FIXMES:
3a. FIXME: Change calls to ru-adj11 to use the new proper name support in
ru-adjective.
3b. FIXME: Test that omitting a manual form in ru-adjective leaves the form as
a big dash.
3c. FIXME: какой-либо and какой-то display genitives with translit -go
instead of -vo. To fix this properly requires implementing real
manual translit for adjectives.
3d. FIXME: Implement real manual translit for adjectives.
5. [FIXME: Consider adding an indicator in the header line when the ё/e
alternation occurs. This is a bit tricky to calculate: If special case
;ё is given, but also if ё occurs in the stem and the accent pattern is
as follows -- for sg-only, b' d' f' f'', also b d f if the noun is masc
or 3rd-decl fem (i.e. nom-sg ending is non-syllabic); for pl-only,
e f f' f'', also b b' c if the gen pl is non-syllabic; for sg/pl,
any but a or b, also b if either nom sg or gen pl is non-syllabic. But
it gets more complicated due to overrides. An alternative is to check
all forms to see if ё is present in some but not all; but this is
tricky also because e.g. reducibles frequently have ё/null alternation,
which doesn't count, and some endings have е or ё in them, which also
doesn't count. If we were to do it this way, we'd have to (a) count
the number of е's in the form(s) with ё and verify that there's at least
one form without ё and with one more е than in the form(s) with ё
(and it gets trickier if different forms with ё have different numbers of
е in them, although that is probably rare); and (b) ignore the appropriate
endings (the best way to do this would probably be to look at the actual
suffixes that were generated in args.suffixes and chop off any matching
ending in the actual form(s), but also chop off final -е/ё, as well as
-ев(ъ)/-ёв(ъ)/-ей/-ёй in the gen pl, which is frequently overridden,
unless perhaps the stem ends in the same way). It'd probably not
possible to do this in a 100% foolproof way but can be "good enough" for
nearly all circumstances.] [MIGHT BE TOO MUCH WORK]
6. HEADWORD FIXMES:
6a. FIXME: In ru-headword, create a category for words whose gender doesn't
match the form. (This is easy to do for ru-noun+ but harder for ru-noun.
We would need to do limited autodetection of the ending: for singulars,
-а/я should be feminine, -е/о/ё should be neuter, -ь should be masculine
or feminine, anything else should be masculine; for plurals, -и/ы should
be masculine or feminine, -а/я should be neuter except that -ія can be
feminine or neuter due to old-style adjectival pluralia tantum nouns,
anything else can be any gender.)
6b. FIXME: Recognize invariable nouns and indicate as indeclinable. Probably
should work by checking the case forms to see if they're the same.
9. FIXME: Change stress-pattern detection and overriding to happen inside of
looping over the two parts of a slash decl. Requires that the loop over
the two parts happen outside of the loop over stress patterns. Requires
that the category code get split into two parts, one to handle combined
singular/plural categories that goes outside the two loops, and one to
handle everything else that goes inside the two loops.
10. FIXME: override_matches_suffix() had a free variable reference to ARGS
in it, which should have triggered an error whenever there was a nom_sg or
nom_pl override but didn't. Is there an error causing this never to be
called? Check.
11a. FIXME: In a multiword lemma, using loc2=+ causes only the second word
to get linked instead of the whole expression. Same for par2=+, voc2=+.
11b. FIXME: Using loc=+ with a multiword lemma should do the right thing, same
as if locN=+ is specified for each individual word. Instead it generates
the locative as a whole from the dative, which fails e.g. if some of the
words are adjectival. Same for par=+, voc=+.
13. Multi-word issues:
-- Setting n=pl when auto-detecting a plural lemma. How does that interact
with multi-word stuff? (DONE)
-- compute_heading() -- what to do with multiple words? I assume we should
display info on the first noun (non-invariable, non-adjectival), and
on the first adjectival otherwise, and finally on an invariable (DONE)
-- args.genders -- it should presumably come from the same word as is used
in compute_heading(); but we should allow the overall gender to be
overridden, at least in ru-noun+ (DONE)
-- Bug in args.suffix: Gets added to every word in attach_with() and then
again at the end, after pltail and such. Needs to be added to the
last word only, before pltail. Need also suffixN for individual words.
(DONE, NEEDS TESTING)
-- Should have ..N versions of pltail and variants. (DONE, NEEDS TESTING)
-- Need to handle overrides of acc_sg, acc_pl (DONE)
-- Overrides of nom_sg/nom_pl should also override acc_sg/acc_pl if it
was originally empty and the animacy is inanimate; similarly for
gen_sg/gen_pl and animates; this needs to work both for per-word and
overall overrides. (DONE)
-- do_generate_forms(_multi) need to run part of make_table(), enough to
combine all per_word_info into single lists of forms and store back
into args[case]. (DONE, NEEDS TESTING)
-- In generate_forms, should probably check if a=="i" and only return
acc_sg_in as acc_sg=; or if a=="a" and only return acc_sg_an as acc_sg=;
in old/new comparison code, do something similar, also when a=="b"
check if acc_sg_in==acc_sg_an and make it acc_sg; when a=="b" and the
_in and _an variants are different, might need to ignore them or check
that acc_sg_in==nom_sg and acc_sg_an==gen_sg; similarly for _pl
(DONE, NEEDS TESTING)
-- Need to test with multiple words! [DONE]
-- Current handling of <adj> won't work properly with multiple words;
will need to translate word-by-word in that case (should be solved by
manual-translit branch) [DONE]
14. In multiple-words branch, fix ru-decl-noun-multi so it recognizes
things like *, (1), (2) and ; without the need for a separator. Consider
using semicolon as a separator, since we already use it to separate ё
from a previous declension. Maybe use $ or ~ for an invariable word; don't
use semicolon. [IMPLEMENTED. NEED TO TEST.]
16. [Consider having ru-noun+ treat par= as a second genitive in
the headword, as is done with край] [WON'T DO]
17. [FIXME: Consider removing slash patterns and instead handling them by
allowing additional declension flags 'sg' and 'pl'. This simplifies the
various special cases caused by slash declensions. It would also be
possible to remove the special plural stem, which would get rid of more
special cases. On the other hand, it makes it more complicated to support
plural variant -ья with all singular types, and the category code that
displays things like "Russian nouns with singular -X and plural -Y"
also gets more complicated, and there's something convenient and intuitive
about plural stems, and slash declensions are also convenient and at least
somewhat intuitive. One possibility is to externally allow slash
declensions and special plural stems and rewrite them internally to
separate stems with 'sg' and 'pl' declension flags; but there are still
the two coding issues mentioned above.]
18. [FIXME: Consider redoing slash patterns so they operate at the outer
level, i.e. things like special cases apply separately in the singular
and plural part of the slash pattern.]
19. In ru-noun, don't recognize -а with m as plural unless (1)
or n=pl is also given, because there are masculine words with the
feminine ending. Check using the test code whether this changes
anything. Also check if there are other similar cases (neuter with
-и isn't parallel because -и is always plural). [IMPLEMENTED. NEED TO TEST.]
19a. Internal notes weren't propagated properly from adjectives.
[IMPLEMENTED. NEED TO TEST.]
19b. Add support for -ишко and -ище variants (p. 74 of Z),
which conversationally and/or colloquially have feminine endings in
certain cases. [IMPLEMENTED. NEED TO TEST. MAKE SURE THE INTERNAL NOTES
APPEAR.]
19d. For masculine animate neuter-form nouns, the accusative singular
ends in -а (-я soft) instead of -о. [IMPLEMENTED. NEED TO TEST.
NOTE: Currently this variant only can be selected using new-style
arguments where the gender can be given. Perhaps we should consider
allowing gender to be specified with old-style explicit declensions.]
21. Put back gender hints for pl adjectival nouns; used by ru-noun+.
[IMPLEMENTED. NEED TO TEST.]
23. Mixed and proper-noun adjectives have built-in notes. We need to
handle those notes with an "internal_notes" section similar to what is used
in the adjective module. [IMPLEMENTED. NEED TO TEST.]
24. Adjective detection code here needs to work the same as for the
adjective module, in particular in the handling of short, stressed-short,
mixed, proper, stressed-proper. [IMPLEMENTED. NEED TO TEST.]
25. Consider simplifying plural-variant code to only allow -ья as a
plural variant [and maybe even change that to be something like (1')].
[IMPLEMENTED REDUCTION OF PLURAL VARIANTS TO -ья; PLURAL-VARIANT CODE
STILL COMPLEX, THOUGH. NEED TO TEST.]
26. Automatically superscript *, numbers and similar things at the
beginning of a note. Also do this in adjective module. [IMPLEMENTED.
NEED TO TEST.]
28. Make the check for multiple stress patterns (categorizing/tracking)
smarter, to keep a list of them and check at the end, so we handle
multiple stress patterns specified through different arg sets.
[IMPLEMENTED; NEED TO TEST.]
29. More sophisticated handling of user-requested plural variant vs. special
case (1) vs. plural-detected variant. [IMPLEMENTED. NEED TO TEST FURTHER.]
30. Solution to ambiguous plural involving gender spec "3f". [IMPLEMENTED;
NEED TO TEST. Use запчасти, новости.]
33. With pluralia tantum adjectival nouns, we don't know the gender.
By default we assume masculine (or feminine for old-style -ія nouns) and
currently this goes into the category, but shouldn't. [IMPLEMENTED.]
39. [Eventually: Even with decl type explicitly given, the full stem with
ending should be included.] [MAY NEVER IMPLEMENT]
40. [Get error "Unable to dereduce" with strange noun ва́йя, what should
happen?] [WILL NOT FIX; USE AN OVERRIDE]
41. In творог, module generates partitive творогу́ when it should copy the
dative творогу́,тво́рогу. (DONE)
42. [[груз 200]] doesn't work. Interprets 200 as a footnote symbol.
43. When converting е -> ё not after cons and with translit, we should
convert e -> o to avoid double j. (DONE)
44. FIXME: In ро́вня/ровня́, similarly with неровня, marks genitive plural
ровня́ as irregular even though it isn't. (NOT OUR ERROR; THE DECLENSIONS
OF THESE NOUNS MARKED THE ENDING-STRESSED VARIANTS WITH (2).)
]=]--
local m_utilities = require("Module:utilities")
local ut = require("Module:utils")
local m_links = require("Module:links")
local com = require("Module:ru-common")
local nom = require("Module:ru-nominal")
local m_ru_adj = require("Module:ru-adjective")
local m_ru_translit = require("Module:ru-translit")
local strutils = require("Module:string utilities")
local scriptutils = require("Module:script utilities")
local m_table_tools = require("Module:table tools")
local m_debug = require("Module:debug")
local export = {}
local lang = require("Module:languages").getByCode("ru")
local Latn = require("Module:scripts").getByCode("Latn")
local u = mw.ustring.char
local rfind = mw.ustring.find
local rsubn = mw.ustring.gsub
local rmatch = mw.ustring.match
local rsplit = mw.text.split
local ulower = mw.ustring.lower
local usub = mw.ustring.sub
local ulen = mw.ustring.len
local AC = u(0x0301) -- acute = ́
local CFLEX = u(0x0302) -- circumflex = ̂
local IRREGMARKER = "△"
local HYPMARKER = "⟐"
-- text class to check lowercase arg against to see if Latin text embedded in it
local latin_text_class = "[a-zščžěáéíóúýàèìòùỳâêîôûŷạẹịọụỵȧėȯẏ]"
-- version of rsubn() that discards all but the first return value
local function rsub(term, foo, bar)
local retval = rsubn(term, foo, bar)
return retval
end
-- version of rsubn() that returns a 2nd argument boolean indicating whether
-- a substitution was made.
local function rsubb(term, foo, bar)
local retval, nsubs = rsubn(term, foo, bar)
return retval, nsubs > 0
end
-- version of rfind() that lowercases its string first, for case-insensitive matching
local function rlfind(term, foo)
return rfind(ulower(term), foo)
end
local function track(page)
m_debug.track("ru-noun/" .. page)
return true
end
-- version of ut.insert_if_not() that makes sure 'false' doesn't get inserted by mistake
local function insert_if_not(foo, bar)
assert(bar ~= false)
ut.insert_if_not(foo, bar)
end
-- Fancy version of ine() (if-not-empty). Converts empty string to nil,
-- but also strips leading/trailing space and then single or double quotes,
-- to allow for embedded spaces.
local function ine(arg)
if not arg then return nil end
arg = rsub(arg, "^%s*(.-)%s*$", "%1")
if arg == "" then return nil end
local inside_quotes = rmatch(arg, '^"(.*)"$')
if inside_quotes then
return inside_quotes
end
inside_quotes = rmatch(arg, "^'(.*)'$")
if inside_quotes then
return inside_quotes
end
return arg
end
-- synthesize a frame so that exported functions meant to be called from
-- templates can be called from the debug console.
local function debug_frame(parargs, args)
return {args = args, getParent = function() return {args = parargs} end}
end
local function rutr_pairs_equal(term1, term2)
local ru1, tr1 = term1[1], term1[2]
local ru2, tr2 = term2[1], term2[2]
local ru1entry, ru1notes = m_table_tools.separate_notes(m_links.remove_links(ru1))
local ru2entry, ru2notes = m_table_tools.separate_notes(m_links.remove_links(ru2))
if ru1entry ~= ru2entry then
return false
end
local tr1entry, tr1notes
local tr2entry, tr2notes
if tr1 then
tr1entry, tr1notes = m_table_tools.separate_notes(tr1)
end
if tr2 then
tr2entry, tr2notes = m_table_tools.separate_notes(tr2)
end
if tr1entry == tr2entry then
return true
elseif type(tr1entry) == type(tr2entry) then
return false
else
tr1entry = tr1entry or nom.translit_no_links(ru1entry)
tr2entry = tr2entry or nom.translit_no_links(ru2entry)
return tr1entry == tr2entry
end
end
local function contains_rutr_pair(list, pair)
for _, item in ipairs(list) do
if rutr_pairs_equal(item, pair) then
return true
end
end
return false
end
-- Clone parent's args while also assigning nil to empty strings.
local function clone_args(frame)
local args = {}
for pname, param in pairs(frame:getParent().args) do
args[pname] = ine(param)
end
return args
end
-- Old-style declensions.
local declensions_old = {}
-- New-style declensions; computed automatically from the old-style ones,
-- for the most part.
local declensions = {}
-- Internal notes for old-style declensions.
local internal_notes_table_old = {}
-- Same for new-style declensions.
local internal_notes_table = {}
-- Category and type information corresponding to declensions: These may
-- contain the following fields: 'singular', 'plural', 'decl', 'hard', 'g',
-- 'suffix', 'gensg', 'irregpl', 'alt_nom_pl', 'cant_reduce', 'ignore_reduce',
-- 'stem_suffix'.
--
-- 'singular' is used to construct a category of the form
-- "Russian nouns SINGULAR". If omitted, a category is constructed of the
-- form "Russian nouns ending in -ENDING", where ENDING is the actual
-- nom sg ending shorn of its acute accents; or "Russian nouns ending
-- in suffix -ENDING", if 'suffix' is true. The value of SINGULAR can be
-- one of the following: a single string, a list of strings, or a function,
-- which is passed one argument (the value of ENDING that would be used to
-- auto-initialize the category), and should return a single string or list
-- of strings. Such a category is only constructed if 'gensg' is true.
--
-- 'plural' is analogous but used to construct a category of the form
-- "Russian nouns with PLURAL", and if omitted, a category is constructed
-- of the form "Russian nouns with plural -ENDING", based on the actual
-- nom pl ending shorn of its acute accents. Currently no plural category
-- is actually constructed.
--
-- In addition, a category may normally constructed from the combination of
-- 'singular' and 'plural', appropriately defaulted; e.g. if both are present,
-- the combined category will be "Russian nouns SINGULAR with PLURAL" and
-- if both are missing, the combined category will be
-- "Russian nouns ending in -SGENDING with plural -PLENDING" (or
-- "Russian nouns ending in suffix -SGENDING with plural -PLENDING" if
-- 'suffix' is true). Note that if either singular or plural or both
-- specifies a list, looping will occur over all combinations. Such a
-- category is constructed only if 'irregpl' or 'alt_nom_pl' or 'suffix'
-- is true or if the declension class is a slash class.
--
-- 'decl' is "1st", "2nd", "3rd" or "invariable"; 'hard' is "hard", "soft"
-- or "none"; 'g' is "m", "f", "n" or "none"; these are all traditional
-- declension categories.
--
-- If 'suffix' is true, the declension type includes a long suffix
-- added to the string that itself undergoes reducibility and such, and so
-- reducibility cannot occur in the stem minus the suffix. Categories will
-- be created for the suffix.
--
-- 'alt_nom_pl' indicates that the declension has an alternative nominative
-- plural (corresponding to Zaliznyak's special case 1; compare special case 2
-- for alternative genitive plural). 'irregpl' indicates that the entire
-- plural is irregular.
--
-- In addition to the above categories, additional more specific categories
-- are constructed based on the final letter of the stem, e.g.
-- "Russian velar-stem 1st-declension hard nouns". See calls to
-- com.get_stem_trailing_letter_type(). 'stem_suffix', if present, is added to
-- the end of the stem when get_stem_trailing_letter_type() is called.
-- This is the only place that 'stem_suffix' is used. This is for use with
-- the '-ья' and '-ье' declension types, so that the trailing letter is
-- 'ь' and not whatever precedes it.
--
-- 'enable_categories' is a special hack for testing, which disables all
-- category insertion if false. Delete this as soon as we've verified the
-- working of the category code and created all the necessary categories.
local enable_categories = true
-- Category/type info corresponding to old-style declensions; see above.
local declensions_old_cat = {}
-- Category/type info corresponding to new-style declensions. Computed
-- automatically from the old-style ones, for the most part. Same format
-- as the old-style ones.
local declensions_cat = {}
-- Table listing aliases of old-style declension classes.
local declensions_old_aliases = {}
-- Table listing aliases of new-style declension classes; computed
-- automatically from the old-style ones.
local declensions_aliases = {}
local stress_patterns = {}
-- Set of patterns with ending-stressed genitive plural.
local ending_stressed_gen_pl_patterns = {}
-- Set of patterns with ending-stressed prepositional singular.
local ending_stressed_pre_sg_patterns = {}
-- Set of patterns with ending-stressed dative singular.
local ending_stressed_dat_sg_patterns = {}
-- Set of patterns with all singular forms ending-stressed.
local ending_stressed_sg_patterns = {}
-- Set of patterns with all plural forms ending-stressed.
local ending_stressed_pl_patterns = {}
-- List of all cases, including those that are declined normally
-- (nom/gen/dat/acc/ins/pre sg and pl), plus animate/inanimate accusative
-- variants (computed automatically as appropriate from the previous cases),
-- plus additional overridable cases (loc/par/voc), plus the "linked" lemma
-- case variants used in ru-noun+ headwords (nom_sg_linked, nom_pl_linked,
-- whose values come from nom_sg and nom_pl but may have additional embedded
-- links if they were given in the lemma).
local all_cases
-- List of all cases that are declined normally.
local decl_cases
-- List of all cases that can be displayed (includes all cases except
-- plain accusatives).
local displayable_cases
-- List of all cases that can be overridden (includes all cases except the
-- "linked" lemma case variants). Also currently the same as the cases
-- returned by export.generate_forms().
local overridable_cases
-- Type of trailing letter, for tracking purposes
local trailing_letter_type
-- If enabled, compare this module with new version of module to make
-- sure all declensions are the same. Eventually consider removing this;
-- but useful as new code is created.
local test_new_ru_noun_module = false
-- Forward functions
local generate_forms_1
local determine_decl
local handle_forms_and_overrides
local handle_overall_forms_and_overrides
local concat_word_forms
local make_table
local detect_adj_type
local detect_stress_pattern
local override_stress_pattern
local determine_stress_variant
local determine_stem_variant
local is_reducible
local is_dereducible
local add_bare_suffix
local attach_stressed
local do_stress_pattern
local canonicalize_override
--------------------------------------------------------------------------
-- Tracking and categorization --
--------------------------------------------------------------------------
-- FIXME! Move below the main code
-- FIXME!! Consider deleting most of this tracking code once we've enabled
-- all the categories. Note that some of the tracking categories aren't
-- completely redundant; e.g. we have tracking pages that combine decl and
-- stress classes, such as "а/a" or "о-и/d'", which are more or less
-- equivalent to stem/gender/stress categories, but we also have the same
-- prefixed by "reducible-stem/" for reducible stems.
local function tracking_code(stress, orig_decl, decl, args, n, islast)
assert(orig_decl)
assert(decl)
local hint_types = com.get_stem_trailing_letter_type(args.stem)
if orig_decl == decl then
orig_decl = nil
end
if args.notes then
track("notes")
end
local function all_pl_irreg()
track("irreg")
for _, case in ipairs(overridable_cases) do
if rfind(case, "_pl") then
track("irreg/" .. case)
end
end
end
local function track_prefix(prefix)
local function dotrack(suf)
track(prefix .. suf)
end
dotrack(stress)
dotrack(decl)
dotrack(decl .. "/" .. stress)
if orig_decl then
dotrack(orig_decl)
dotrack(orig_decl .. "/" .. stress)
end
for _, hint_type in ipairs(hint_types) do
dotrack(hint_type)
dotrack(decl .. "/" .. hint_type)
if orig_decl then
dotrack(orig_decl .. "/" .. hint_type)
end
end
end
track_prefix("")
if args.reducible then
track("reducible-stem")
track_prefix("reducible-stem/")
end
if rlfind(args.stem, "и́?н$") and (decl == "" or decl == "#") then
track("irregular-in")
end
if args.pltail then
track("pltail")
end
if args.sgtail then
track("sgtail")
end
if args.pltailall then
track("pltailall")
end
if args.sgtailall then
track("sgtailall")
end
if args.pl ~= args.stem then
track("irreg-pl-stem")
track("irreg")
end
if args.alt_gen_pl then
track("alt-gen-pl")
track("irreg")
track("irreg/gen_pl")
end
if args.want_sc1 then
track("want-sc1")
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-и$") or rfind(decl, "-а$") or rfind(decl, "-я$") then
track("irreg")
track("irreg/nom_pl")
end
if rfind(decl, "-ья$") then
track("variant-ья")
all_pl_irreg()
end
if rfind(decl, "%(ишк%)$") then
track("variant-ишко")
end
if rfind(decl, "%(ищ%)$") then
track("variant-ище")
end
if args.jo_special then
track("jo-special")
end
if args.manual then
track("manual")
end
if args.explicit_gender then
track("explicit-gender")
track("explicit-gender/" .. args.explicit_gender)
end
for _, case in ipairs(overridable_cases) do
if args[case .. n] and not args.manual then
track("override")
track("override/" .. case .. n)
track("irreg")
track("irreg/" .. case .. n)
-- questionable use: track_prefix("irreg/" .. case .. "/")
-- questionable use: track_prefix("irreg/" .. case .. n .. "/")
end
if islast and args[case] and not args.manual then
track("override")
track("override/" .. case)
track("irreg")
track("irreg/" .. case)
-- questionable use: track_prefix("irreg/" .. case .. "/")
end
if args[case .. "_tail"] then
track("casenum-tail")
track("casenum-tail/" .. case)
end
if args[case .. "_tailall"] then
track("casenum-tailall")
track("casenum-tailall/" .. case)
end
end
end
local gender_to_full = {m="masculine", f="feminine", n="neuter"}
local gender_to_short = {m="masc", f="fem", n="neut"}
-- Insert the category CAT (a string) into list CATEGORIES. String will
-- have "Russian " prepended and ~ substituted for the plural part of speech.
local function insert_category(categories, cat, pos, atbeg)
if enable_categories then
local fullcat = "Russian " .. rsub(cat, "~", pos .. "s")
if atbeg then
table.insert(categories, 1, fullcat)
else
table.insert(categories, fullcat)
end
end
end
-- Insert categories into ARGS.CATEGORIES corresponding to the specified
-- stress and declension classes and to the form of the stem (e.g. velar,
-- sibilant, etc.). Also initialize values used to compute the declension
-- heading that describes similar information. N is the number of the
-- word being processed; ISLAST is true if this is the last word.
local function categorize_and_init_heading(stress, decl, args, n, islast)
local function cat_to_list(cat)
if not cat then
return {}
elseif type(cat) == "string" then
return {cat}
else
assert(type(cat) == "table")
return cat
end
end
-- Insert category CAT into the list of categories in ARGS.
-- CAT may be nil, a single string or a list of strings. We call
-- insert_category() on each string. The strings will have "Russian "
-- prepended and "~" replaced with the plural part of speech.
local function insert_cat(cat)
for _, c in ipairs(cat_to_list(cat)) do
insert_category(args.categories, c, args.pos)
end
end
-- "Resolve" the category spec CATSPEC into the sort of category spec
-- accepted by insert_cat(), i.e. nil, a single string or a list of
-- strings. CATSPEC may be any of these or a function, which takes one
-- argument (SUFFIX) and returns another CATSPEC.
local function resolve_cat(catspec, suffix)
if type(catspec) == "function" then
return resolve_cat(catspec(suffix), suffix)
else
return catspec
end
end
-- Check whether an override for nom_sg or nom_pl still contains the
-- normal suffix (which should already have accents removed) in at least
-- one of its entries. If no override then of course we return true.
local function override_matches_suffix(args, case, n, suffix)
assert(suffix == com.remove_accents(suffix))
-- NOTE: It might not be completely correct to pass in args.forms
-- here when n == ""; this is different behavior from
-- handle_overall_forms_and_overrides(). But args.forms is only used
-- to retrieve the dat_sg for handling loc/par, and we're never
-- called with loc or par as the value of CASE, so it doesn't matter.
-- We add an assert to make sure of this.
assert(case ~= "loc" and case ~= "par")
local override = canonicalize_override(args, case, args.forms, n)
if not override then
return true
end
for _, x in ipairs(override) do
local ru, tr = x[1], x[2]
local entry, notes = m_table_tools.separate_notes(ru)
entry = com.remove_accents(m_links.remove_links(entry))
if rlfind(entry, suffix .. "$") then
return true
end
end
return false
end
if args.manual then
return
end
local h = args.heading_info
assert(decl)
local decl_cats = args.old and declensions_old_cat or declensions_cat
local sgdecl, pldecl
local is_slash_decl = rfind(decl, "/")
if is_slash_decl then
local indiv_decls = rsplit(decl, "/")
sgdecl, pldecl = indiv_decls[1], indiv_decls[2]
else
sgdecl, pldecl = decl, decl
end
local sgdc = decl_cats[sgdecl]
local pldc = decl_cats[pldecl]
assert(sgdc)
assert(pldc)
local sghint_types = com.get_stem_trailing_letter_type(
args.stem .. (sgdc.stem_suffix or ""))
-- insert English version of Zaliznyak stem type
if sgdc.decl == "invariable" then
insert_cat("invariable ~")
insert_if_not(h.stemetc, "invar")
else
local stem_type =
sgdc.decl == "3rd" and "3rd-declension" or
ut.contains(sghint_types, "velar") and "velar-stem" or
ut.contains(sghint_types, "sibilant") and "sibilant-stem" or
ut.contains(sghint_types, "c") and "ц-stem" or
ut.contains(sghint_types, "i") and "i-stem" or
ut.contains(sghint_types, "vowel") and "vowel-stem" or
ut.contains(sghint_types, "soft-cons") and "vowel-stem" or
ut.contains(sghint_types, "palatal") and "vowel-stem" or
sgdc.hard == "soft" and "soft-stem" or
"hard-stem"
local short_stem_type = stem_type == "3rd-declension" and "3rd-decl" or stem_type
if sgdc.adj then
-- Don't include gender for pluralia tantum because it's mostly
-- indeterminate (certainly when specified using a plural lemma,
-- which will be usually; technically it's partly or completely
-- determinate in certain old-style adjectives that distinguish
-- masculine from feminine/neuter, but this is too rare a case
-- to worry about)
local gendertext = args.thisn == "p" and "plural-only" or gender_to_full[sgdc.g]
insert_if_not(h.adjectival, "yes")
if args.thisn ~= "p" then
insert_if_not(h.gender, gender_to_short[sgdc.g])
end
if sgdc.possadj then
insert_cat(sgdc.decl .. " possessive " .. gendertext .. " accent-" .. stress .. " adjectival ~")
insert_if_not(h.stemetc, sgdc.decl .. " poss")
insert_if_not(h.stress, stress)
elseif stem_type == "soft-stem" or stem_type == "vowel-stem" then
insert_cat(stem_type .. " " .. gendertext .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
else
insert_cat(stem_type .. " " .. gendertext .. " accent-" .. stress .. " adjectival ~")
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
else
-- NOTE: There are 8 Zaliznyak-style stem types and 3 genders, but
-- we don't create a category for masculine-form 3rd-declension
-- nouns (there is such a noun, путь, but it mostly behaves
-- like a feminine noun), so there are 23.
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form ~")
-- NOTE: Here we are creating categories for the combination of
-- stem, gender and accent. There are 10 accent patterns and 23
-- combinations of stem and gender, which potentially makes for
-- 10*23 = 230 such categories, which is a lot. Not all such
-- categories should actually exist; there were maybe 75 former
-- declension templates, each of which was essentially categorized
-- by the same three variables, but some of which dealt with
-- ancillary issues like irregular plurals; this amounts to 67
-- actual stem/gender/accent categories, although there are more
-- of them in Zaliznyak (FIXME, how many? See generate_cats.py).
insert_cat(stem_type .. " " .. gender_to_full[sgdc.g] .. "-form accent-" .. stress .. " ~")
insert_if_not(h.adjectival, "no")
insert_if_not(h.gender, gender_to_short[sgdc.g])
insert_if_not(h.stemetc, short_stem_type)
insert_if_not(h.stress, stress)
end
insert_cat("~ with accent pattern " .. stress)
end
local sgsuffix = args.suffixes.nom_sg
if sgsuffix then
assert(#sgsuffix == 1) -- If this ever fails, then implement a loop
sgsuffix = com.remove_accents(sgsuffix[1])
-- If we are plural only or if nom_sg is overridden and has
-- an unusual suffix, then don't create category for sg suffix
if args.thisn == "p" or not override_matches_suffix(args, "nom_sg", n, sgsuffix) or islast and not override_matches_suffix(args, "nom_sg", "", sgsuffix) then
sgsuffix = nil
end
end
local plsuffix = args.suffixes.nom_pl
if plsuffix then
assert(#plsuffix == 1) -- If this ever fails, then implement a loop
plsuffix = com.remove_accents(plsuffix[1])
-- If we are a singulare tantum or if nom_pl is overridden and has
-- an unusual suffix, then don't create category for pl suffix
if args.thisn == "s" or not override_matches_suffix(args, "nom_pl", n, plsuffix) or islast and not override_matches_suffix(args, "nom_pl", "", plsuffix) then
plsuffix = nil
end
end
sgsuffix = sgsuffix and rsub(sgsuffix, "ъ$", "")
plsuffix = plsuffix and rsub(plsuffix, "ъ$", "")
local sgcat = sgsuffix and (resolve_cat(sgdc.singular, sgsuffix) or "ending in " .. (sgsuffix == "" and "a consonant" or (sgdc.suffix and "suffix " or "") .. "-" .. sgsuffix))
local plcat = plsuffix and (resolve_cat(pldc.plural, suffix) or "plural -" .. plsuffix)
if sgcat and sgdc.gensg then
for _, cat in ipairs(cat_to_list(sgcat)) do
insert_cat("~ " .. cat)
end
end
if sgcat and plcat and (sgdc.suffix or sgdc.alt_nom_pl or sgdc.irregpl or
is_slash_decl and plsuffix == "-ья") then
for _, scat in ipairs(cat_to_list(sgcat)) do
for _, pcat in ipairs(cat_to_list(plcat)) do
insert_cat("~ " .. scat .. " with " .. pcat)
end
end
end
if args.pl ~= args.stem then
insert_cat("~ with irregular plural stem")
end
if args.reducible and not sgdc.ignore_reduce then
insert_cat("~ with reducible stem")
insert_if_not(h.reducible, "yes")
else
insert_if_not(h.reducible, "no")
end
if args.alt_gen_pl then
insert_cat("~ with alternate genitive plural")
end
if sgdc.adj then
insert_cat("adjectival ~")
end
end
local function compute_heading(args)
local headings = {}
local h = args.heading_info
table.insert(headings, args.a == "a" and "anim" or args.a == "i" and
"inan" or "bian")
table.insert(headings, args.nonumber and "uncountable" or
args.n == "s" and "sg-only" or args.n == "p" and "pl-only" or nil)
if #h.gender > 0 then
table.insert(headings, table.concat(h.gender, "/") .. "-form")
end
if #h.stemetc > 0 then
table.insert(headings, table.concat(h.stemetc, "/"))
end
if #h.stress > 0 then
local stresses = {}
for _, stress in ipairs(h.stress) do
table.insert(stresses, rsub(stress, "'", "'"))
end
table.insert(headings, "accent-" .. table.concat(stresses, "/"))
end
local function handle_bool(boolvals, text, into)
into = into or headings
if ut.contains(boolvals, "yes") and ut.contains(boolvals, "no") then
table.insert(into, "[" .. text .. "]")
elseif ut.contains(boolvals, "yes") then
table.insert(into, text)
end
end
handle_bool(h.adjectival, "adj")
handle_bool(h.reducible, "reduc")
return headings
end
local function compute_overall_heading_categories_and_genders(args)
local hinfo = args.per_word_heading_info
local index = 0
-- First try for non-adjectival, non-invariable
for i=1,#hinfo do
if not ut.contains(hinfo[i].stemetc, "invar") and not ut.contains(hinfo[i].adjectival, "yes") then
index = i
break
end
end
if index == 0 then
-- Then just non-invariable
for i=1,#hinfo do
if not ut.contains(hinfo[i].stemetc, "invar") then
index = i
break
end
end
end
-- Finally, do anything
if index == 0 then
index = 1
end
-- Compute final heading
local headings = args.per_word_headings[index]
local categories = args.per_word_categories[index]
if args.any_irreg then
table.insert(headings, "irreg")
insert_category(categories, "irregular ~", args.pos)
end
for _, case in ipairs(overridable_cases) do
local is_pl = rfind(case, "_pl")
if args.n == "s" and is_pl or args.n == "p" and not is_pl then
-- Don't create singular categories when plural-only or vice-versa
elseif (case == "loc" or case == "voc" or case == "par") then
if args.any_overridden[case] then
local engcase = rsub(case, "^([a-z]*)", {
par="partitive", loc="locative", voc="vocative"
})
insert_category(categories, "~ with " .. engcase, args.pos)
end
elseif args.any_irreg_case[case] then
local engcase = rsub(case, "^([a-z]*)", {
nom="nominative", gen="genitive", dat="dative",
acc="accusative", ins="instrumental", pre="prepositional",
par="partitive", loc="locative", voc="vocative"
})
engcase = rsub(engcase, "(_[a-z]*)", {
_sg=" singular", _pl=" plural",
_an="", _in="",
--_an=" animate", _in=" inanimate"
})
insert_category(categories, "~ with irregular " .. engcase, args.pos)
end
end
local heading = args.manual and "" or "(<span style=\"font-size: smaller;\">[[Appendix:Russian nouns#Declension tables|" .. table.concat(headings, " ") .. "]]</span>)"
args.heading = heading
args.categories = categories
args.genders = args.per_word_genders[index]
end
--------------------------------------------------------------------------
-- Main code --
--------------------------------------------------------------------------
-- Used by do_generate_forms().
local function arg1_is_stress(arg1)
if not arg1 then return false end
for _, arg in ipairs(rsplit(arg1, ",")) do
if not rfind(arg, "^[a-f]'?'?$") then
return false
end
end
return true
end
-- Used by do_generate_forms(), handling a word joiner argument
-- of the form 'join:JOINER'.
local function extract_word_joiner(spec)
word_joiner = rmatch(spec, "^join:(.*)$")
assert(word_joiner)
return nom.split_russian_tr(word_joiner, "dopair")
end
local function determine_headword_gender(args, sgdc, gender)
-- If gender unspecified, use normal gender of declension, except when
-- adjectival nouns that are pluralia tantum, where the gender is
-- mostly indeterminate (FIXME, not completely with old-style declensions
-- but we don't handle that currently).
if not gender then
if sgdc.adj and args.thisn == "p" then
gender = nil
else
gender = sgdc.g
end
end
-- Determine headword genders
gender = gender and gender ~= "none" and gender .. "-" or ""
local plsuffix = args.n == "p" and "-p" or ""
local hgens
if args.a == "a" then
hgens = {gender .. "an" .. plsuffix}
elseif args.a == "i" then
hgens = {gender .. "in" .. plsuffix}
elseif args.a == "ai" then
hgens = {gender .. "an" .. plsuffix, gender .. "in" .. plsuffix}
else
hgens = {gender .. "in" .. plsuffix, gender .. "an" .. plsuffix}
end
-- Insert into list of genders
for _, hgen in ipairs(hgens) do
insert_if_not(args.genders, hgen)
end
end
function export.do_generate_forms(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SETS and JOINER, where ARG_SETS
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Gather arguments into a list of ARG_SET objects, containing (potentially)
-- elements 1, 2, 3, 4, corresponding to accent pattern, stem, declension
-- type, pl stem and coming from consecutive numbered parameters. Sets of
-- declension parameters are separated by the word "or".
local arg_sets = {}
-- Find maximum-numbered arg, allowing for holes
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Now gather the arguments.
local offset = 0
local arg_set = {}
for i=1,(max_arg + 1) do
local end_arg_set = false
local end_word = false -- FIXME, is this correct?
local word_joiner
if i == max_arg + 1 then
end_arg_set = true
end_word = true
word_joiner = {""}
elseif args[i] == "_" then
end_arg_set = true
end_word = true
word_joiner = {" "}
elseif args[i] == "-" then
end_arg_set = true
end_word = true
word_joiner = {"-"}
elseif args[i] and rfind(args[i], "^join:") then
end_arg_set = true
end_word = true
word_joiner = extract_word_joiner(args[i])
elseif args[i] == "or" then
end_arg_set = true
end
if end_arg_set then
table.insert(arg_sets, arg_set)
arg_set = {}
offset = i
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
else
-- If the first argument isn't stress, that means all arguments
-- have been shifted to the left one. We want to shift them
-- back to the right one, so we change the offset so that we
-- get the same effect of skipping a slot in the arg set.
if i - offset == 1 and not arg1_is_stress(args[i]) then
offset = offset - 1
end
if i - offset > 4 then
error("Too many arguments for argument set: arg " .. i .. " = " .. (args[i] or "(blank)"))
end
arg_set[i - offset] = args[i]
end
end
return generate_forms_1(args, per_word_info)
end
function export.do_generate_forms_multi(args, old)
old = old or args.old
args.old = old
args.pos = args.pos or "noun"
-- This is a list with each element corresponding to a word and
-- consisting of a two-element list, ARG_SET and JOINER, where ARG_SET
-- is a list of ARG_SET objects, one per alternative stem, and JOINER
-- is a string indicating how to join the word to the next one.
local per_word_info = {}
-- Find maximum-numbered arg, allowing for holes (FIXME: Is this needed
-- here? Will there be holes?)
local max_arg = 0
for k, v in pairs(args) do
if type(k) == "number" and k > max_arg then
max_arg = k
end
end
-- Gather arguments into a list of ARG_SET objects, containing
-- (potentially) elements 1, 2, 3, corresponding to accent pattern,
-- lemma+declension spec, pl stem, exactly as with do_generate_forms()
-- and {{ru-noun-table}} except that the values come from a single argument
-- of the form ACCENTPATTERN:LEMMADECL:PL where all but LEMMADECL may (and
-- probably will be) omitted and LEMMADECL may be of the following forms:
-- LEMMA (for a noun with empty decl spec),
-- LEMMADECL (for a noun with non-empty decl spec beginning with a
-- *, left paren or semicolon),
-- LEMMA^DECL (for a noun with non-empty decl spec),
-- LEMMA$ (for an invariable word)
-- LEMMA+ (for an adjective with auto-detected decl class)
-- LEMMA+DECL (for an adjective with explicit decl class)
-- Sets of parameters for the same word are separated by the word "or".
local arg_sets = {}
local continue_arg_sets = true
for i=1,(max_arg + 1) do
local end_word = false
local word_joiner
local process_arg = false
if i == max_arg + 1 then
end_word = true
word_joiner = {""}
elseif args[i] == "-" then
end_word = true
word_joiner = {"-"}
continue_arg_sets = true
elseif rfind(args[i], "^join:") then
end_word = true
word_joiner = extract_word_joiner(args[i])
continue_arg_sets = true
elseif args[i] == "or" then
continue_arg_sets = true
else
if continue_arg_sets then
continue_arg_sets = false
else
end_word = true
word_joiner = {" "}
end
process_arg = true
end
if end_word then
table.insert(per_word_info, {arg_sets, word_joiner})
arg_sets = {}
end
if process_arg then
local vals = rsplit(args[i], ":")
if #vals > 3 then
error("Can't specify more than 3 colon-separated params of param set: " .. args[i])
end
local arg_set = {}
if arg1_is_stress(vals[1]) then
arg_set[1] = vals[1]
arg_set[2] = vals[2]
arg_set[4] = vals[3]
else
arg_set[2] = vals[1]
arg_set[4] = vals[2]
end
-- recognize invariable
local inv_stem = rmatch(arg_set[2], "^(.-)%$$")
if inv_stem then
arg_set[2] = inv_stem
arg_set[3] = "$"
else
-- recognize adjective
local adj_stem, adj_type = rmatch(arg_set[2], "^(.*)(%+.*)$")
if adj_stem then
arg_set[2] = adj_stem
arg_set[3] = adj_type
else
-- recognize noun with ^
local noun_stem, noun_type = rmatch(arg_set[2], "^(.*)%^(.*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- recognize noun without ^ but with decl spec
noun_stem, noun_type = rmatch(arg_set[2], "^(..-)([;*(].*)$")
if noun_stem then
arg_set[2] = noun_stem
arg_set[3] = noun_type
else
-- noun without ^ or decl spec; nothing to do
end
end
end
end
table.insert(arg_sets, arg_set)
end
end
return generate_forms_1(args, per_word_info)
end
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence.
-- Implementation of do_generate_forms() and do_generate_forms_multi(),
-- which have equivalent functionality but different calling sequence,
-- as well as show_z() for template {{ru-decl-noun-z}}, which has a
-- subset of the functionality of the other two.
generate_forms_1 = function(args, per_word_info)
local orig_args
if test_new_ru_noun_module then
orig_args = mw.clone(args)
end
local SUBPAGENAME = mw.loadData("Module:headword/data").pagename
local old = args.old
local function verify_animacy_value(val)
if not val then return nil end
if val == "a" or val == "an" or val == "anim" then
return "a"
elseif val == "i" or val == "in" or val == "inan" then
return "i"
elseif val == "b" or val == "bi" or val == "both" or val == "ai" then
return "ai"
elseif val == "ia" then
return "ia"
end
error("Animacy value " .. val .. " should be empty or a/an/anim (animate), i/in/inan (inanimate), b/bi/both/ai (bianimate, listing animate first), or ia (bianimate, listing inanimate first)")
return nil
end
local function verify_number_value(val, allow_none)
if not val then return nil end
local short = usub(val, 1, 1)
if short == "s" or short == "p" or short == "b" or (allow_none and short == "n" or false) then
return short
end
if allow_none then
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), 'b' (both) or 'n' (none)")
else
error("Number value " .. val .. " should be empty or start with 's' (singular), 'p' (plural), or 'b' (both)")
end
return nil
end
-- Verify and canonicalize animacy, number, prefix, suffix
assert(#per_word_info >= 1)
for i=1,#per_word_info do
args["a" .. i] = verify_animacy_value(args["a" .. i])
args["n" .. i] = verify_number_value(args["n" .. i])
args["prefix" .. i] = nom.split_russian_tr(args["prefix" .. i] or "", "dopair")
args["suffix" .. i] = nom.split_russian_tr(args["suffix" .. i] or "", "dopair")
end
args.a = verify_animacy_value(args.a) or "i"
-- args.ndef, if set, is the default value for args.n; if unset, it defaults
-- to "both". It is set to "singular" in ru-proper noun+. We store the value
-- of args.n in args.orign before defaulting to args.ndef because we may
-- change it to plural-only later on if it was unspecified (this happens if
-- an individual word's lemma is plural), and to determine whether it was
-- unspecified, we need the original value before defaulting.
args.n = verify_number_value(args.n, "allow none")
-- treat n=none like n=sg but set a flag so "singular" isn't displayed
if args.n == "n" then
args.n = "s"
args.nonumber = true
end
args.ndef = verify_number_value(args.ndef)
args.orign = args.n
args.n = args.n or args.ndef
args.prefix = nom.split_russian_tr(args.prefix or "", "dopair")
args.suffix = nom.split_russian_tr(args.suffix or "", "dopair")
-- Attach overall prefix to first per-word prefix, similarly for suffix
args.prefix1 = nom.concat_paired_russian_tr(args.prefix, args.prefix1)
args["suffix" .. #per_word_info] = nom.concat_paired_russian_tr(
args["suffix" .. #per_word_info], args.suffix)
-- Initialize non-word-specific arguments.
--
-- The following is a list of WORD_INFO items, one per word, each of
-- which is a two element list of WORD_FORMS (a table listing the forms for
-- each case) and JOINER (a string, indicating how to join the word with
-- the next one).
args.per_word_info = {}
-- List of HEADING_INFO items, one per word, containing the raw material
-- used to generate the header. Initialized from 'args.heading_info',
-- which is initialized in categorize_and_init_heading().
args.per_word_heading_info = {}
-- List of CATEGORIES items, one per word, containing the categories
-- used to initialize the page categories. Initialized from
-- 'args.categories', which is initialized in categorize_and_init_heading().
args.per_word_categories = {}
-- List of HEADINGS items, one per word, containing the actual words that
-- go into the header if we were to use that word to construct the header.
-- Comes from compute_heading(). We use this to generate the actual header
-- string, which goes into 'args.heading' at the end (done in
-- compute_overall_heading_categories_and_genders()).
args.per_word_headings = {}
-- List of GENDERS items, one per word, containing the headword genders
-- for each word (where "headword gender" is in the format used in
-- headwords and actually includes animacy and number as well, e.g.
-- 'm-in' or 'f-an-p'). We end up selecting one such item and putting
-- it into 'args.genders' at the end
-- (in compute_overall_heading_categories_and_genders()).
args.per_word_genders = {}
args.any_overridden = {}
args.any_non_nil = {}
args.any_irreg = false
args.any_irreg_case = {}
local function insert_cat(cat)
insert_category(args.categories, cat, args.pos)
end
args.internal_notes = {}
-- Superscript footnote marker at beginning of note, similarly to what's
-- done at end of forms.
if args.notes then
local notes, entry = m_table_tools.get_initial_notes(args.notes)
args.notes = notes .. entry
end
local decl_sufs = old and declensions_old or declensions
local decl_cats = old and declensions_old_cat or declensions_cat
local intable = old and internal_notes_table_old or internal_notes_table
-- Default lemma defaults to previous lemma, first one to page name.
-- FIXME: With multiple words, we should probably split the page name
-- on spaces and default each word in turn
local default_lemma
local all_stresses_seen
-- Made into a function to avoid having to indent a lot of code.
-- Process a single arg set of a single word. This inserts the forms
-- for the word into args.forms and sets categories and tracking pages.
local function do_arg_set(arg_set, n, islast)
local stress_arg = arg_set[1]
local decl = arg_set[3] or ""
local pl, pltr
if arg_set[4] then
pl, pltr = nom.split_russian_tr(arg_set[4])
end
-- Extract special markers from declension class.
if decl == "manual" then
decl = "$"
args.manual = true
if #per_word_info > 1 or #per_word_info[1][1] > 1 then
error("Can't specify multiple words or argument sets when manual")
end
if pl then
error("Can't specify optional stem parameters when manual")
end
end
decl, args.jo_special = rsubb(decl, "([^/%a])ё$", "%1")
if not args.jo_special then
decl, args.jo_special = rsubb(decl, "([^/%a])ё([^/%a])", "%1%2")
end
decl, args.want_sc1 = rsubb(decl, "%(1%)", "")
decl, args.alt_gen_pl = rsubb(decl, "%(2%)", "")
decl, args.reducible = rsubb(decl, "%*", "")
decl = rsub(decl, ";", "")
-- Get the lemma.
local lemma = args.manual and "-" or arg_set[2] or default_lemma
if not lemma then
error("Lemma in first argument set must be specified")
end
default_lemma = lemma
local lemmatr
lemma, lemmatr = nom.split_russian_tr(lemma)
args.thisa = args["a" .. n] or args.a
args.thisn = args["n" .. n] or args.n
-- Check for explicit allow-unaccented indication.
lemma, args.allow_unaccented = rsubb(lemma, "^%*", "")
if args.allow_unaccented then
track("allow-unaccented")
end
args.orig_lemma = lemma
lemma = m_links.remove_links(lemma)
args.lemma_no_links = lemma
args.lemmatr = lemmatr
-- Treat suffixes without an accent, and suffixes with an accent on the
-- initial hyphen, as if they were preceded with a *, which overrides
-- all the logic that normally (a) normalizes the accent, and (b)
-- complains about multisyllabic words without an accent. Don't do this
-- if lemma is just -, which is used specially in manual declension
-- tables (e.g. сто, три).
if lemma ~= "-" and (rfind(lemma, "^%-́") or (com.is_unstressed(lemma) and rfind(lemma, "^%-"))) then
args.allow_unaccented = true
end
-- Convert lemma and decl arg into stem and canonicalized decl.
-- This will autodetect the declension from the lemma if an explicit
-- decl isn't given.
local stem, tr, gender, was_accented, was_plural, was_autodetected
if rfind(decl, "^%+") then
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
detect_adj_type(lemma, lemmatr, decl, old)
else
stem, tr, decl, gender, was_accented, was_plural, was_autodetected =
determine_decl(lemma, lemmatr, decl, args)
end
if was_plural then
args.n = args.orign or "p"
args.thisn = args["n" .. n] or args.n
elseif decl ~= "$" then
args.thisn = args.thisn or "b"
end
args.explicit_gender = gender
-- If allow-unaccented not given, maybe check for missing accents.
if not args.allow_unaccented and not stress_arg and was_autodetected and com.needs_accents(lemma) then
-- If user gave the full word and expects us to determine the
-- declension and stress, the word should have an accent on the
-- stem or ending. We have a separate check farther below for
-- an accent on a multisyllabic stem, after stripping off any
-- ending; but this way we get an error if the user e.g. writes
-- "гора" without an accent rather than assuming it's stem
-- stressed, as would otherwise happen.
error("Lemma must have an accent in it: " .. lemma)
end
-- If stress not given, auto-determine; else validate/canonicalize
-- stress arg, override in certain cases and convert to list.
if not stress_arg then
stress_arg = {detect_stress_pattern(stem, decl, decl_cats, args.reducible, was_plural, was_accented)}
else
stress_arg = rsplit(stress_arg, ",")
for i=1,#stress_arg do
local stress = stress_arg[i]
stress = override_stress_pattern(decl, stress)
if not stress_patterns[stress] then
error("Unrecognized accent pattern " .. stress)
end
stress_arg[i] = stress
end
end
-- parse slash decl to list
local sub_decls
if rfind(decl, "/") then
track("mixed-decl")
insert_cat("~ with mixed declension")
local indiv_decls = rsplit(decl, "/")
-- Should have been caught in canonicalize_decl()
assert(#indiv_decls == 2)
sub_decls = {{indiv_decls[1], "sg"}, {indiv_decls[2], "pl"}}
else
sub_decls = {{decl}}
end
-- Get singular declension and corresponding category.
local sgdecl = sub_decls[1][1]
local sgdc = decl_cats[sgdecl]
assert(sgdc)
-- Compute headword gender(s). We base it off the singular declension
-- if we have a slash declension -- it's the best we can do.
determine_headword_gender(args, sgdc, gender)
local original_stem, original_tr = stem, tr
local original_pl, original_pltr = pl, pltr
-- Loop over accent patterns in case more than one given.
for _, stress in ipairs(stress_arg) do
args.suffixes = {}
stem, tr = original_stem, original_tr
local bare, baretr
local stem_for_bare, tr_for_bare
pl, pltr = original_pl, original_pltr
insert_if_not(all_stresses_seen, stress)
local stem_was_unstressed = com.is_unstressed(stem)
-- If special case ;ё was given and stem is unstressed,
-- add ё to the stem now; but don't let this interfere with
-- restressing, to handle cases like железа́ with gen pl желёз
-- but nom pl же́лезы.
if stem_was_unstressed and args.jo_special then
-- Beware, Cyrillic еЕ in first rsub, Latin eE in second
local new_stem = rsub(stem, "([еЕ])([^еЕ]*)$",
function(e, rest)
return (e == "Е" and "Ё" or "ё") .. rest
end
)
if stem == new_stem then
error("No е in stem to replace with ё")
end
stem = new_stem
if tr then
local subbed
-- e after j -> o, e not after j -> jo; don't just convert e -> jo
-- and then map jjo -> jo because we want to preserve double j
tr, subbed = rsubb(tr, "([jJ])([eE])([^eE]*)$",
function(j, e, rest)
return j .. (e == "E" and "O" or "o") .. AC .. rest
end
)
if not subbed then
tr = rsub(tr, "([eE])([^eE]*)$",
function(e, rest)
return (e == "E" and "Jo" or "jo") .. AC .. rest
end
)
end
tr = com.j_correction(tr)
end
-- This is used to handle железа́ with gen pl желёз and nom pl
-- же́лезы. We have two stressed stems, one for the gen pl and
-- one for the remaining pl cases, and the variable 'stem' can
-- handle only one, so we put the second (gen pl) stem in
-- stem_for_bare, which goes into the value of 'bare' (used
-- only for gen pl here).
stem_for_bare, tr_for_bare = stem, tr
end
-- Maybe add stress to the stem, depending on whether the
-- stem was unstressed and the stress pattern. Stem pattern f
-- and variants call for initial stress (голова́ -> го́ловы);
-- stem pattern d and variants call for stem-final stress
-- (сапожо́к -> сапо́жки). Stem patterns b and b' apparently
-- call for stem-final stress as well but it's unlikely to
-- make much of a difference (pattern b' only occurs in 3rd-decl
-- feminines, which should already have stress in the stem,
-- and the only place pattern b gets stem stress is in bare
-- forms, i.e. nom sg and/or gen pl depending on the decl type,
-- and nom sg stress should already be in the lemma while
-- gen pl stress is handled by a different ending-stressing
-- mechanism in attach_unstressed(); however, the user is
-- free to leave a masc or 3rd-decl fem lemma completely
-- unstressed with pattern b, and then the stem-final stress
-- *will* make a difference).
local function restress_stem(stem, tr, stress, stem_unstressed)
-- If the user has indicated they purposely are leaving the
-- word unstressed by putting a * at the beginning of the main
-- stem, leave it unstressed. This might indicate lack of
-- knowledge of the stress or a truly unaccented word
-- (e.g. an unaccented suffix).
if args.allow_unaccented then
return stem, tr
end
if tr and com.is_unstressed(stem) ~= com.is_unstressed(tr) then
error("Stem " .. stem .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(stem) then
stem, tr = com.make_ending_stressed(stem, tr)
-- For those patterns that are ending-stressed in the singular
-- nominative (and hence are likely to be expressed without an
-- accent on the stem) it's safe to put a particular accent on
-- the stem depending on the stress type. Otherwise, give an
-- error if no accent.
elseif stem_unstressed then
if rfind(stress, "^f") then
stem, tr = com.make_beginning_stressed(stem, tr)
elseif (rfind(stress, "^[bd]") or
args.thisn == "p" and ending_stressed_pl_patterns[stress]) then
stem, tr = com.make_ending_stressed(stem, tr)
elseif com.needs_accents(stem) then
error("Stem " .. stem .. " requires an accent")
end
end
return stem, tr
end
stem, tr = restress_stem(stem, tr, stress, stem_was_unstressed)
-- Leave pl unaccented if user wants this; see restress_stem().
if pl and not args.allow_unaccented then
if pltr and com.is_unstressed(pl) ~= com.is_unstressed(pltr) then
error("Plural stem " .. pl .. " and translit " .. pltr .. " must have same accent pattern")
end
if com.is_monosyllabic(pl) then
pl, pltr = com.make_ending_stressed(pl, pltr)
end
-- I think this is safe.
if com.needs_accents(pl) then
if ending_stressed_pl_patterns[stress] then
pl, pltr = com.make_ending_stressed(pl, pltr)
elseif not args.allow_unaccented then
error("Plural stem " .. pl .. " requires an accent")
end
end
end
local resolved_bare, resolved_baretr
-- Handle (de)reducibles
-- FIXME! We are dereducing based on the singular declension.
-- In a slash declension things can get weird and we don't
-- handle that. We are also computing the bare value from the
-- singular stem, and again things can get weird with a plural
-- stem. Note that we don't compute a bare value unless we have
-- to (either (de)reducible or stress pattern f/f'/f'' combined
-- with ё special case); the remaining times we generate the bare
-- value directly from the plural stem.
if args.reducible and not sgdc.ignore_reduce then
-- Zaliznyak treats all nouns in -ье and -ья as being
-- reducible. We handle this automatically and don't require
-- the user to specify this, but ignore it if so for
-- compatibility.
if is_reducible(sgdc) then
-- If we derived the stem from a nom pl form, then
-- it's already reduced, and we need to dereduce it to
-- get a bare form; otherwise the stem comes from the
-- nom sg and we need to reduce it to get the real stem.
if was_plural then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
resolved_bare, resolved_baretr = stem, tr
stem, tr = export.reduce_nom_sg_stem(stem, tr,
sgdecl, "error")
-- Stem will be unstressed if stress was on elided
-- vowel; restress stem the way we did above. (This is
-- needed in at least one word, сапожо́к 3*d(2), with
-- plural stem probably сапо́жк- and gen pl probably
-- сапо́жек.)
stem, tr = restress_stem(stem, tr, stress,
com.is_unstressed(stem))
if stress ~= "a" and stress ~= "b" and args.alt_gen_pl and not pl then
-- Nouns like рожо́к, глазо́к of type 3*d(2) have
-- gen pl's ро́жек, гла́зок; to handle this,
-- dereduce the reduced stem and store in a
-- special place.
args.gen_pl_bare, args.gen_pl_baretr =
export.dereduce_nom_sg_stem(stem, tr,
sgdc, stress, old, "error")
end
end
elseif is_dereducible(sgdc) then
resolved_bare, resolved_baretr =
export.dereduce_nom_sg_stem(stem, tr, sgdc,
stress, old, "error")
else
error("Declension class " .. sgdecl .. " not (de)reducible")
end
elseif stem_for_bare and stem ~= stem_for_bare then
resolved_bare, resolved_baretr =
add_bare_suffix(stem_for_bare, tr_for_bare, old, sgdc, false)
end
-- Leave unaccented if user wants this; see restress_stem().
-- FIXME, we no longer allow the user to specify the bare value
-- so it's unclear if this is needed any more.
if resolved_bare and not args.allow_unaccented then
if resolved_baretr and com.is_unstressed(resolved_bare) ~= com.is_unstressed(resolved_baretr) then
error("Resolved bare stem " .. resolved_bare .. " and translit " .. resolved_baretr .. " must have same accent pattern")
end
if com.is_monosyllabic(resolved_bare) then
resolved_bare, resolved_baretr =
com.make_ending_stressed(resolved_bare, resolved_baretr)
else
if com.needs_accents(resolved_bare) then
error("Resolved bare stem " .. resolved_bare .. " requires an accent")
end
end
end
args.stem, args.stemtr = stem, tr
args.bare, args.baretr = resolved_bare, resolved_baretr
args.ustem, args.ustemtr = com.make_unstressed_once(stem, tr)
if pl then
args.pl, args.pltr = pl, pltr
else
args.pl, args.pltr = stem, tr
end
args.upl, args.upltr = com.make_unstressed_once(args.pl, args.pltr)
-- Special hack for любо́вь and other reducible 3rd-fem nouns,
-- which have the full stem in the ins sg
args.ins_sg_stem = sgdecl == "ь-f" and args.reducible and resolved_bare
args.ins_sg_tr = sgdecl == "ь-f" and args.reducible and resolved_baretr
-- Loop over declension classes (we may have two of them, one for
-- singular and one for plural, in the case of a mixed declension
-- class of the form SGDECL/PLDECL).
for _,decl_spec in ipairs(sub_decls) do
local orig_decl = decl_spec[1]
local number = decl_spec[2]
local real_decl =
determine_stress_variant(orig_decl, stress)
real_decl = determine_stem_variant(real_decl,
number == "pl" and args.pl or args.stem)
-- sanity checking; errors should have been caught in
-- canonicalize_decl()
assert(decl_cats[real_decl], "real_decl " .. real_decl .. " nonexistent")
assert(decl_sufs[real_decl], "real_decl " .. real_decl .. " nonexistent")
tracking_code(stress, orig_decl, real_decl, args, n, islast)
do_stress_pattern(stress, args, real_decl, number, n, islast)
-- handle internal notes
local internal_note = intable[real_decl]
if internal_note then
insert_if_not(args.internal_notes, internal_note)
end
end
categorize_and_init_heading(stress, decl, args, n, islast)
end
end
local n = 0
for _, word_info in ipairs(per_word_info) do
n = n + 1
local islast = n == #per_word_info
local arg_sets, joiner = word_info[1], word_info[2]
args.forms = {}
args.heading_info = {animacy={}, number={}, gender={}, stress={},
stemetc={}, adjectival={}, reducible={}}
args.categories = {}
args.genders = {}
args.this_any_non_nil = {}
if #arg_sets > 1 then
track("multiple-arg-sets")
insert_cat("~ with multiple argument sets")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
default_lemma = SUBPAGENAME
all_stresses_seen = {}
-- Loop over all arg sets.
for _, arg_set in ipairs(arg_sets) do
do_arg_set(arg_set, n, islast)
end
if #all_stresses_seen > 1 then
track("multiple-accent-patterns")
insert_cat("~ with multiple accent patterns")
track("multiple-declensions")
insert_cat("~ with multiple declensions")
end
table.insert(args.per_word_heading_info, args.heading_info)
table.insert(args.per_word_categories, args.categories)
local headings = compute_heading(args)
table.insert(args.per_word_headings, headings)
table.insert(args.per_word_genders, args.genders)
handle_forms_and_overrides(args, n, islast)
table.insert(args.per_word_info, {args.forms, joiner})
end
handle_overall_forms_and_overrides(args)
compute_overall_heading_categories_and_genders(args)
for _, case in ipairs(all_cases) do
if args[case] then
for _, form in ipairs(args[case]) do
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
ruentry = m_links.remove_links(ruentry)
if rfind(ulower(ruentry), latin_text_class) then
--error("Found Latin text " .. ruentry .. " in case " .. case)
track("latin-text")
track("latin-text/" .. case)
end
end
end
end
-- Test code to compare existing module to new one.
if test_new_ru_noun_module then
local m_new_ru_noun = require("Module:User:Benwing2/ru-noun")
local newargs = m_new_ru_noun.do_generate_forms(orig_args, old)
local difdecl = false
for _, case in ipairs(all_cases) do
local arg = args[case]
local newarg = newargs[case]
local is_pl = rfind(case, "_pl")
if args.thisn == "s" and is_pl or args.thisn == "p" and not is_pl then
-- Don't need to check cases that won't be displayed.
elseif not ut.equals(arg, newarg) then
local monosyl_accent_diff = false
-- Differences only in monosyllabic accents. Enable if we
-- change the algorithm for these.
--if arg and newarg and #arg == 1 and #newarg == 1 then
-- local ru1, tr1 = arg[1][1], arg[1][2]
-- local ru2, tr2 = newarg[1][1], newarg[1][2]
-- if com.is_monosyllabic(ru1) and com.is_monosyllabic(ru2) then
-- ru1, tr1 = com.remove_accents(ru1, tr1)
-- ru2, tr2 = com.remove_accents(ru2, tr2)
-- if ru1 == ru2 and tr1 == tr2 then
-- monosyl_accent_diff = true
-- end
-- end
--end
if monosyl_accent_diff then
track("monosyl-accent-diff")
difdecl = true
else
-- Uncomment this to display the particular case and
-- differing forms.
--error(case .. " " .. (arg and nom.concat_forms(arg) or "nil") .. " || " .. (newarg and nom.concat_forms(newarg) or "nil"))
track("different-decl")
difdecl = true
end
break
end
end
if not difdecl then track("same-decl") end
end
return args
end
-- Implementation of main entry point
local function do_show(frame, old)
local args = clone_args(frame)
local args = export.do_generate_forms(args, old)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The main entry point for modern declension tables.
function export.show(frame)
return do_show(frame, false)
end
-- The main entry point for old declension tables.
function export.show_old(frame)
return do_show(frame, true)
end
-- Implementation of new entry point, esp. for multiple words
local function do_show_multi(frame)
local args = clone_args(frame)
local args = export.do_generate_forms_multi(args)
return make_table(args) .. m_utilities.format_categories(args.categories, lang)
end
-- The new entry point, esp. for multiple words (but works fine for
-- single words).
function export.show_multi(frame)
return do_show_multi(frame)
end
local function get_form(forms, preserve_links, raw)
local canon_forms = {}
for _, form in ipairs(forms) do
if raw then
local ru, tr = form[1], form[2]
ru = rsub(ru, "|", "<!>")
if tr then
tr = rsub(tr, "|", "<!>")
end
insert_if_not(canon_forms, {ru, tr})
else
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
-- Skip hypothetical forms (but include in the raw versions)
if not rfind(runotes, HYPMARKER) then
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.separate_notes(tr)
end
if not preserve_links then
ruentry = m_links.remove_links(ruentry)
end
ruentry = rsub(ruentry, "|", "<!>")
if trentry then
trentry = rsub(trentry, "|", "<!>")
end
insert_if_not(canon_forms, {ruentry, trentry})
end
end
end
return nom.concat_forms(canon_forms)
end
local function case_will_be_displayed(args, forms, case)
local ispl = rfind(case, "_pl")
local caseok = true
if args.n == "p" then
caseok = ispl
elseif args.n == "s" then
caseok = not ispl
end
if case == "loc" and not args.any_overridden.loc or
case == "par" and not args.any_overridden.par or
case == "voc" and not args.any_overridden.voc then
caseok = false
end
if args.a == "a" or args.a == "i" then
if rfind(case, "_[ai]n") then
caseok = false
end
else -- bianimate
-- don't include inanimate/animate variants if combined variant exists
-- (typically because inanimate/animate variants are the same);
-- FIXME: This could conceivably be different from how the display
-- code works, which just checks that the inanimate/animate variants
-- are the same when deciding whether to display them, in particular
-- if there is an override. Here we are following the algorithm of
-- handle_overall_forms_and_overrides().
if (case == "acc_sg_in" or case == "acc_sg_an") and args.acc_sg or
(case == "acc_pl_in" or case == "acc_pl_an") and args.acc_pl then
caseok = false
end
end
if not forms[case] then
caseok = false
end
return caseok
end
local function concat_case_args(args, do_all, raw)
local ins_text = {}
for _, case in ipairs(do_all and all_cases or overridable_cases) do
if case_will_be_displayed(args, args, case) then
local forms = get_form(args[case], rfind(case, "_linked"), raw)
if forms ~= "" then
table.insert(ins_text, case .. (raw and "_raw" or "") .. "=" ..
forms)
end
end
end
return table.concat(ins_text, "|")
end
-- The entry point for 'ru-noun-forms' to generate all noun forms.
-- This returns a single string, with | separating arguments and named
-- arguments of the form NAME=VALUE.
function export.generate_forms(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
return concat_case_args(args)
end
-- The entry point to generate multiple sets of noun forms. This is a hack
-- to speed up calling from a bot, where we often want to compare old and new
-- argument results to make sure they're the same. Each set of arguments is
-- jammed together into a single argument with individual values separated by
-- <!>; named arguments are of the form NAME<->VALUE. The return value for
-- each set of arguments is as in export.generate_forms(), and the return
-- values are concatenated with <!> separating them. NOTE: This will fail if
-- the exact sequences <!> or <-> happen to occur in values (which is unlikely,
-- esp. as we don't even use the characters <, ! or > for anything) and aren't
-- HTML-escaped.
function export.generate_multi_forms(frame)
local retvals = {}
for _, argset in ipairs(frame.args) do
local args = {}
local i = 0
local argvals = rsplit(argset, "<!>")
for _, argval in ipairs(argvals) do
local split_arg = rsplit(argval, "<%->")
if #split_arg == 1 then
i = i + 1
args[i] = ine(split_arg[1])
else
assert(#split_arg == 2)
args[split_arg[1]] = ine(split_arg[2])
end
end
args = export.do_generate_forms(args, false)
table.insert(retvals, concat_case_args(args))
end
return table.concat(retvals, "<!>")
end
-- The entry point for 'ru-noun-form' to generate a particular noun form.
function export.generate_form(frame)
local args = clone_args(frame)
if not args.form then
error("Must specify desired form using form=")
end
local form = args.form
if not ut.contains(all_cases, form) then
error("Unrecognized form " .. form)
end
local args = export.do_generate_forms(args, false)
if not args[form] then
return ""
else
return get_form(args[form])
end
end
-- The entry point for generating arguments of various sorts, including
-- the case forms, gender, number and animacy.
function export.generate_args(frame)
local args = clone_args(frame)
args = export.do_generate_forms(args, false)
local retargs = {}
table.insert(retargs, concat_case_args(args, "doall"))
table.insert(retargs, concat_case_args(args, "doall", "raw"))
table.insert(retargs, "g=" .. table.concat(args.genders, ","))
-- The following is correct even with ndef because if ndef is
-- set we will set it in args.n.
table.insert(retargs, "n=" .. (args.n or "b"))
table.insert(retargs, "a=" .. (args.a or "i"))
return table.concat(retargs, "|")
end
-- The entry point for compatibility with {{ru-decl-noun-z}}.
function export.show_z(frame)
local args = clone_args(frame)
local stem = args[1]
local stress = args[2]
local specific = args[4] or ""
-- Parse gender/animacy spec
local gender, anim = rmatch(args[3], "^([mfn])-([a-z]+)")
if not gender then
error("Unrecognized gender/anim spec " .. args[3])
end
if anim ~= "an" and anim ~= "in" then anim = "both" end
args.a = anim
-- Handle specific
specific = rsub(specific, "ё", ";ё")
-- Compute decl; special case for семьянин (perhaps not necessary)
local decl = com.make_unstressed_once(stem) == "семьянин" and "#" .. specific or gender .. specific
-- Handle overrides
args.pre_sg = args.prp_sg
args.pre_pl = args.prp_pl
args.notes = args.note
if args.par then
args.par = "+"
end
if args.loc then
if args.loc == "в" then
args.loc = "в +"
elseif args.loc == "на" then
args.loc = "на +"
else
args.loc = "в +,на +"
end
end
local arg_set = {}
table.insert(arg_set, stress)
table.insert(arg_set, stem)
table.insert(arg_set, decl)
local per_word_info = {{{arg_set}, ""}}
return generate_forms_1(args, per_word_info)
end
local stem_expl = {
["velar-stem"] = "a velar (-к, -г or –x)",
["sibilant-stem"] = "a sibilant (-ш, -ж, -ч or -щ)",
["ц-stem"] = "-ц",
["i-stem"] = "-и (old-style -і)",
["vowel-stem"] = "a vowel other than -и or -і, or -й or -ь",
["soft-stem"] = "a soft consonant",
["hard-stem"] = "a hard consonant",
}
local zaliznyak_stem_type = {
["velar-stem"] = "3",
["sibilant-stem"] = "4",
["ц-stem"] = "5",
["i-stem"] = "7",
["vowel-stem"] = "6",
["soft-stem"] = "2",
["hard-stem"] = "1",
["3rd-declension"] = "8",
}
local stem_gender_endings = {
masculine = {
["hard-stem"] = {"a hard consonant (-ъ old-style)", "-ы"},
["ц-stem"] = {"-ц (-цъ old-style)", "-ы"},
["velar-stem"] = {"a velar (plus -ъ old-style)", "-и"},
["sibilant-stem"] = {"a sibilant (plus -ъ old-style)", "-и"},
["soft-stem"] = {"-ь", "-и"},
["i-stem"] = {"-й", "-и"},
["vowel-stem"] = {"-й", "-и"},
["3rd-declension"] = {"-ь", "-и"},
},
feminine = {
["hard-stem"] = {"-а", "-ы"},
["ц-stem"] = {"-а", "-ы"},
["velar-stem"] = {"-а", "-и"},
["sibilant-stem"] = {"-а", "-и"},
["soft-stem"] = {"-я", "-и"},
["i-stem"] = {"-я", "-и"},
["vowel-stem"] = {"-я", "-и"},
["3rd-declension"] = {"-ь", "-и"},
},
neuter = {
["hard-stem"] = {"-о", "-а"},
["ц-stem"] = {"-е", "-а"},
["velar-stem"] = {"-о", "-а"},
["sibilant-stem"] = {"-е", "-а"},
["soft-stem"] = {"-е", "-я"},
["i-stem"] = {"-е", "-я"},
["vowel-stem"] = {"-е", "-я"},
["3rd-declension"] = {"-мя", "-мена"},
},
}
-- Implementation of template 'runouncatboiler'.
function export.catboiler(frame)
local SUBPAGENAME = mw.loadData("Module:headword/data").pagename
local args = clone_args(frame)
local cats = {}
local function get_stem_gender_text(stem, gender)
if not stem_gender_endings[gender] then
error("Invalid gender " .. gender)
end
local endings = stem_gender_endings[gender][stem]
if not endings then
error("Invalid stem type " .. stem)
end
local sgending, plending = endings[1], endings[2]
local stemtext =
stem == "3rd-declension" and "" or
" The stem ends in " .. stem_expl[stem] .. " and is Zaliznyak's type " .. zaliznyak_stem_type[stem] .. "."
local decltext =
stem == "3rd-declension" and "" or
" This is traditionally considered to belong to the " .. (gender == "feminine" and "1st" or "2nd") .. " declension."
return stem .. ", usually " .. gender .. " ~, normally ending in " .. sgending .. " in the nominative singular " ..
" and " .. plending .. " in the nominative plural." .. stemtext .. decltext
end
local function get_pos()
local pos = rmatch(SUBPAGENAME, "^Russian.- ([^ ]*)s ")
if not pos then
pos = rmatch(SUBPAGENAME, "^Russian.- ([^ ]*)s$")
end
if not pos then
error("Invalid category name, should be e.g. \"Russian nouns with ...\" or \"Russian ... nouns\"")
end
return pos
end
local function get_sort_key()
local pos, sort_key = rmatch(SUBPAGENAME, "^Russian.- ([^ ]*)s with (.*)$")
if sort_key then
return sort_key
end
pos, sort_key = rmatch(SUBPAGENAME, "^Russian ([^ ]*)s (.*)$")
if sort_key then
return sort_key
end
return rsub(SUBPAGENAME, "^Russian ", "")
end
local maintext, pos
if args[1] == "stemgenderstress" then
local stem, gender, stress
stem, gender, stress, pos = rmatch(SUBPAGENAME, "^Russian (.-) (.-)%-form accent%-(.-) (.*)s$")
if not stem then
error("Invalid category name, should be e.g. \"Russian velar-stem masculine-form accent-a nouns\"")
end
local stem_gender_text = get_stem_gender_text(stem, gender)
local accent_text = " This " .. pos .. " is stressed according to accent pattern " .. rsub(stress, "'", "'") .. "."
maintext = stem_gender_text .. accent_text
insert_category(cats, "~ by stem type, gender and accent pattern|" .. get_sort_key(), pos)
elseif args[1] == "stemgender" then
if rfind(SUBPAGENAME, "invariable") then
maintext = "invariable (indeclinable) ~, which normally have the same form for all cases and numbers."
pos = rmatch(SUBPAGENAME, "^Russian invariable (.*)s$")
if not pos then
error("Invalid category name, should be e.g. \"Russian invariable nouns\"")
end
else
local stem, gender
stem, gender, pos = rmatch(SUBPAGENAME, "^Russian (.-) (.-)%-form (.*)s$")
if not stem then
error("Invalid category name, should be e.g. \"Russian velar-stem masculine-form nouns\"")
end
maintext = get_stem_gender_text(stem, gender)
end
insert_category(cats, "~ by stem type and gender|" .. get_sort_key(), pos)
elseif args[1] == "adj" then
local stem, gender, stress
stem, gender, stress, pos = rmatch(SUBPAGENAME, "^Russian (.*) (.-) accent%-(.-) adjectival (.*)s$")
if not stem then
stem, gender, pos = rmatch(SUBPAGENAME, "^Russian (.*) (.-) adjectival (.*)s$")
end
if not stem then
error("Invalid category name, should be e.g. \"Russian velar-stem masculine accent-a adjectival nouns\"")
end
local stemtext
if rfind(stem, "possessive") then
possessive = "possessive "
stem = rsub(stem, " possessive", "")
stemtext = ""
else
if not stem_expl[stem] then
error("Invalid stem type " .. stem)
end
possessive = ""
stemtext = " The stem ends in " .. stem_expl[stem] .. " and is Zaliznyak's type " .. zaliznyak_stem_type[stem] .. "."
end
local stresstext = stress == "a" and
"This " .. pos .. " is stressed according to accent pattern a (stress on the stem)." or
stress == "b" and
"This " .. pos .. " is stressed according to accent pattern b (stress on the ending)." or
"All ~ of this class are stressed according to accent pattern a (stress on the stem)."
maintext = stem .. " " .. gender .. " ~, with " .. possessive .. "adjectival endings, ending in " ..
(gender == "plural-only" and "" or args[2] .. " in the nominative singular and ") ..
args[3] .. " in the nominative plural." .. stemtext .. " " .. stresstext
insert_category(cats, "~ by stem type, gender and accent pattern|" .. get_sort_key(), pos)
else
pos = get_pos()
if args[1] == "sg" then
local sg = rmatch(SUBPAGENAME, "^Russian .* ending in (.*)$")
if not sg then
error("Invalid category name, should be e.g. \"Russian nouns ending in stressed -е\"")
end
maintext = "~ ending in " .. (args[2] or sg) .. " in the nominative singular."
insert_category(cats, "~ by singular ending", pos)
elseif args[1] == "pl" then
local pl = rmatch(SUBPAGENAME, "^Russian .* with plural (.*)$")
if not pl then
error("Invalid category name, should be e.g. \"Russian nouns with plural -е\"")
end
maintext = "~ ending in " .. (args[2] or pl) .. " in the nominative plural."
insert_category(cats, "~ by plural ending|" .. pl, pos)
elseif args[1] == "sgpl" then
local sg, pl = rmatch(SUBPAGENAME, "^Russian .* ending in (.*) with plural (.*)$")
if not sg then
error("Invalid category name, should be e.g. \"Russian nouns ending in -о with plural -и\"")
end
maintext = "~ ending in " .. (args[2] or sg) .. " in the nominative singular, and " .. (args[3] or pl) .. " in the nominative plural."
insert_category(cats, "~ by singular and plural ending|" .. sg .. " " .. pl, pos)
elseif args[1] == "stress" then
local stress = rmatch(SUBPAGENAME, "^Russian .* with accent pattern (.*)$")
if not stress then
error("Invalid category name, should be e.g. \"Russian nouns with accent pattern d\"")
end
maintext = "~ with accent pattern " .. rsub(stress, "'", "'") .. "."
insert_category(cats, "~ by accent pattern|" .. stress, pos)
elseif args[1] == "extracase" then
local case = rmatch(SUBPAGENAME, "^Russian .* with (.*)$")
if not case then
error("Invalid category name, should be e.g. \"Russian nouns with vocative\"")
end
maintext = "~ with a separate " .. case .. " singular case."
insert_category(cats, "~ by case form|" .. case, pos)
elseif args[1] == "irregcase" then
local case = rmatch(SUBPAGENAME, "^Russian .* with irregular (.*)$")
if not case then
error("Invalid category name, should be e.g. \"Russian nouns with irregular nominative plural\"")
end
maintext = "~ with an irregular " .. case .. " case."
insert_category(cats, "~ by case form|" .. case, pos)
else
maintext = "~ " .. args[1]
end
end
insert_category(cats, "~|" .. get_sort_key(), pos, "at beginning")
-- format_categories doesn't work in category space so we need to hack our own
local categories = {}
for _, cat in ipairs(cats) do
table.insert(categories, "[[Category:" .. cat .. "]]")
end
return "This category contains Russian " .. rsub(maintext, "~", pos .. "s")
.. "\n" ..
mw.getCurrentFrame():expandTemplate{title="ru-categoryTOC", args={}}
.. table.concat(categories, "")
end
--------------------------------------------------------------------------
-- Autodetection and lemma munging --
--------------------------------------------------------------------------
-- Attempt to detect the type of the lemma based on its ending, separating
-- off the stem and the ending. GENDER must be present with -ь and plural
-- stems, and is otherwise ignored. Return up to three values: The stem
-- (lemma minus ending), the singular lemma ending, and if the lemma was
-- plural, the plural lemma ending. If the lemma was singular, the singular
-- lemma ending will contain any user-given accents; likewise, if the
-- lemma was plural, the plural ending will contain such accents.
-- VARIANT comes from the declension spec and controls certain declension
-- variants.
local function detect_lemma_type(lemma, tr, gender, args, variant)
local base, ending = rmatch(lemma, "^(.*)([еЕ]́)$") -- accented
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([еЕ])$") -- unaccented
if base then
if variant == "-ище" and not rfind(lemma, "[щЩ][еЕ]$") then
error("With declension variant -ище, lemma should end in -ще: " .. lemma)
end
return base, nom.strip_tr_ending(tr, ending), variant == "-ище" and "(ищ)е-и" or "о"
end
if variant == "-ишко" then
base, ending = rmatch(lemma, "^(.*[шШ][кК])([оО])$") -- unaccented
if not base then
error("With declension variant -ишко, lemma should end in -шко: " .. lemma)
end
return base, nom.strip_tr_ending(tr, ending), "(ишк)о-и"
end
if variant == "-ин" then
base, ending = rmatch(lemma, "^(.*)([иИ]́?[нН][ъЪ]?)$") -- maybe accented
if not base then
error("With declension variant -ин, lemma should end in -ин(ъ): " .. lemma)
end
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
-- Now autodetect -ин; only animate and in -анин/-янин
base, ending = rmatch(lemma, "^(.*[аяАЯ]́?[нН])([иИ]́?[нН][ъЪ]?)$")
-- Need to check the animacy to avoid nouns like маиганин, цианин,
-- меланин, соланин, etc.
if base and args.thisa == "a" then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]́?[нН][оО][кК][ъЪ]?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]́[нН][оО][кК][ъЪ]?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ёЁ]́?[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*[" .. com.sib_c .. "])([оО]́[нН][оО][чЧ][еЕ][кК][ъЪ]?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([мМ][яЯ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
--recognize plural endings
if gender == "n" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ]́?)$")
if base then
-- Don't do this; о/-ья is too rare
-- error("Ambiguous plural lemma " .. lemma .. " in -ья, singular could be -о or -ье/-ьё; specify the singular")
return base, nom.strip_tr_ending(tr, ending), "ье", ending
end
base, ending = rmatch(lemma, "^(.*)([аяАЯ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), rfind(ending, "[аА]") and "о" or "е", ending
end
base, ending = rmatch(lemma, "^(.*)([ыиЫИ]́?)$")
if base then
if rfind(ending, "[ыЫ]") or rfind(base, "[" .. com.sib .. com.velar .. "]$") then
return base, nom.strip_tr_ending(tr, ending), "о-и", ending
else
-- FIXME, should we return a slash declension?
error("No neuter declension е-и available; use a slash declension")
end
end
end
if gender == "f" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][иИ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), "ья", ending
end
end
-- Recognize masculines with irregular plurals, but only if the user
-- either explicitly specified that this noun is plural (n=p) or
-- specifically requested the irregular plural. This is necessary
-- because some masculine nouns have feminine endings, which look
-- like irregular plurals.
if gender == "m" then
if args.thisn == "p" or variant == "-ья" then
base, ending = rmatch(lemma, "^(.*)([ьЬ][яЯ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), (args.old and "ъ-ья" or "-ья"), ending
end
end
if args.thisn == "p" or args.want_sc1 then
base, ending = rmatch(lemma, "^(.*)([аА]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), (args.old and "ъ-а" or "-а"), ending
end
base, ending = rmatch(lemma, "^(.*)([яЯ]́?)$")
if base then
if rfind(base, "[" .. com.vowel .. "]́?$") then
return base, nom.strip_tr_ending(tr, ending), "й-я", ending
else
return base, nom.strip_tr_ending(tr, ending), "ь-я", ending
end
end
end
end
if gender == "m" or gender == "f" then
base, ending = rmatch(lemma, "^(.*[" .. com.sib .. com.velar .. "])([иИ]́?)$")
if not base then
base, ending = rmatch(lemma, "^(.*)([ыЫ]́?)$")
end
if base then
return base, nom.strip_tr_ending(tr, ending), gender == "m" and (args.old and "ъ" or "") or "а", ending
end
base, ending = rmatch(lemma, "^(.*[" .. com.vowel .. "й]́?)([иИ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), gender == "m" and "й" or "я", ending
end
base, ending = rmatch(lemma, "^(.*)([иИ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), gender == "m" and "ь-m" or "я", ending
end
end
if gender == "3f" then
base, ending = rmatch(lemma, "^(.*)([иИ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), "ь-f", ending
end
end
-- end of recognize-plurals code
base, ending = rmatch(lemma, "^(.*)([ьЬ][яеёЯЕЁ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([йаяеоёъЙАЯЕОЁЪ]́?)$")
if base then
return base, nom.strip_tr_ending(tr, ending), ulower(ending)
end
base, ending = rmatch(lemma, "^(.*)([ьЬ])$")
if base then
if gender == "m" or gender == "f" then
return base, nom.strip_tr_ending(tr, ending), "ь-" .. gender
elseif gender == "3f" then
return base, nom.strip_tr_ending(tr, ending), "ь-f"
else
error("Need to specify gender m or f with lemma in -ь: ".. lemma)
end
end
if rfind(lemma, "[ыиЫИ]́?$") then
error("If this is a plural lemma, gender must be specified: " .. lemma)
elseif rfind(lemma, "[уыэюиіѣѵУЫЭЮИІѢѴ]́?$") then
error("Don't know how to decline lemma ending in this type of vowel: " .. lemma)
end
return lemma, tr, ""
end
local plural_variant_detection_map = {
[""] = {["-ья"]="-ья"},
["ъ"] = {["-ья"]="ъ-ья"},
}
local special_case_1_to_plural_variant = {
[""] = "-а",
["ъ"] = "ъ-а",
["й"] = "й-я",
["ь-m"] = "ь-я",
["о"] = "о-и",
-- these last two are here to avoid getting errors in that checks for
-- compatibility with special case 1; the basic variants of these decls
-- don't actually exist
["(ишк)о"] = "(ишк)о-и",
["(ищ)е"] = "(ищ)е-и",
}
local function map_decl(decl, fun)
if rfind(decl, "/") then
local split_decl = rsplit(decl, "/")
if #split_decl ~= 2 then
error("Mixed declensional class " .. decl
.. "needs exactly two classes, singular and plural")
end
return fun(split_decl[1]) .. "/" .. fun(split_decl[2])
else
return fun(decl)
end
end
-- Canonicalize decl class into non-accented and alias-resolved form;
-- but note that some canonical decl class names with an accent in them
-- (e.g. е́, not the same as е, whose accented version is ё; and various
-- adjective declensions).
local function canonicalize_decl(decl, old)
local function do_canon(decl)
-- remove accents, but not from е́ (for adj decls, accent matters
-- as well but we handle that by mapping the accent to a stress pattern
-- and then to the accented version in determine_stress_variant())
if decl ~= "е́" then
decl = com.remove_accents(decl)
end
local decl_aliases = old and declensions_old_aliases or declensions_aliases
local decl_cats = old and declensions_old_cat or declensions_cat
if decl_aliases[decl] then
-- If we find an alias, map it.
decl = decl_aliases[decl]
elseif not decl_cats[decl] then
error("Unrecognized declension class " .. decl)
end
return decl
end
return map_decl(decl, do_canon)
end
-- Attempt to determine the actual declension (including plural variants)
-- based on a combination of the declension the user specified, what can be
-- detected from the lemma, and special case (1), if given in the declension.
-- DECL is the value the user passed for the declension field, after
-- extraneous annotations (special cases (1) and (2), * for reducible,
-- ё for ё/ё alternation and a ; that may precede ё) have been stripped off.
-- What's left is one of the following:
--
-- 1. Blank, meaning to autodetect the declension from the lemma
-- 2. A hyphen followed by a declension variant (-ья, -ин, -ишко, -ище; see
-- long comment at top of file)
-- 3. A gender (m, f, n, 3f)
-- 4. A gender plus declension variant (e.g. f-ья)
-- 5. An actual declension, possibly including a plural variant (e.g. о-и) or
-- a slash declension (e.g. я/-ья, used for the noun дядя).
--
-- Return seven args: stem (lemma minus ending), translit, canonicalized
-- declension, explicitly specified gender if any (m, f, n or nil), whether
-- the specified declension or detected ending was accented, whether the
-- detected ending was pl, and whether the declension was autodetected
-- (corresponds to cases where a full word with ending attached is required
-- in the lemma field). "Canonicalized" means after autodetection, with
-- accents removed, with any aliases mapped to their canonical versions
-- and with any requested declension variants applied. The result is either a
-- declension that will have a categorization entry (in declensions_cat[] or
-- declensions_old_cat[]) or a slash declension where each part similarly has
-- a categorization entry.
--
-- Note that gender is never required when an explicit declension is given,
-- and in connection with stem autodetection is required only when the lemma
-- either ends in -ь or is plural.
determine_decl = function(lemma, tr, decl, args)
-- Assume we're passed a value for DECL of types 1-4 above, and
-- fetch gender and requested declension variant.
local stem
local want_ya_plural, orig_pl_ending, variant
local was_autodetected
local gender = rmatch(decl, "^(3?[mfn]?)$")
if not gender then
gender, variant = rmatch(decl, "^(3?[mfn]?)(%-[^%-]+)$")
-- But be careful with explicit declensions like -а that look like
-- variants without gender (FIXME, eventually we should maybe do
-- something about the potential ambiguity).
if gender == "" and not ut.contains({"-ья", "-ин", "-ишко", "-ище"}, variant) then
gender, variant = nil, nil
end
end
-- If DECL is of type 1-4, handle declension variants and detect
-- the actual declension from the lemma.
if gender then
-- Check for declension variants
if variant then
if variant == "-ья" then
want_ya_plural = "-ья"
else
-- Sanity-check remaining declension variants, which need
-- specific values of animacy, gender and special-case (1)
local sc1_needed
local animate_needed
if variant == "-ишко" then
animate_needed = false
sc1_needed = true
elseif variant == "-ище" then
animate_needed = true
sc1_needed = true
elseif variant == "-ин" then
animate_needed = true
sc1_needed = false
else
-- WARNING: If adding another variant, you need to also
-- add to the list farther above.
error("Unrecognized declension variant " .. variant .. ", should be -ья, -ин, -ишко or -ище")
end
if sc1_needed and not args.want_sc1 then
error("Declension variant " .. variant .. " must be used with special case (1)")
elseif sc1_needed == false and args.want_sc1 then
error("Declension variant " .. variant .. " must not be used with special case (1)")
end
if animate_needed and args.thisa ~= "a" then
error("Declension variant " .. variant .. " must be specified as animate")
elseif animate_needed == false and args.thisa == "a" then
error("Declension variant " .. variant .. " must not be specified as animate")
end
if gender ~= "" and gender ~= "m" then
error("Declension variant " .. variant .. " should be used with the masculine gender")
end
end
end
stem, tr, decl, orig_pl_ending = detect_lemma_type(lemma, tr, gender,
args, variant)
was_autodetected = true
else
stem, tr = lemma, tr
end
-- Now canonicalize gender
if gender == "3f" then
gender = "f"
elseif gender == "" then
gender = nil
end
-- The ending should be treated as accented if either the original singular
-- or plural ending was accented, or if the stem is non-syllabic.
local was_accented = com.is_stressed(decl) or
orig_pl_ending and com.is_stressed(orig_pl_ending) or
com.is_nonsyllabic(stem)
local was_plural = not not orig_pl_ending
decl = canonicalize_decl(decl, args.old)
-- The rest of this code concerns plural variants. It's somewhat
-- complicated because there are potentially four sources of plural
-- variants (not to mention plural variants constructed using slash
-- notation):
--
-- 1. A user-requested plural variant in declension types 2 or 4 above
-- (currently only -ья)
-- 2. An explicit plural variant encoded in an explicit declension of
-- type 5 above
-- 3. An autodetected plural variant (which will happen in some cases
-- when autodetection is performed on a nominative plural)
-- 4. A plural variant derived using special case (1).
--
-- Up to three actual plural variants might exist (e.g. if the user
-- specifies a DECL value or 'm-ья(1)' and a STEM ending in -а,
-- although not all three can ever be compatible because -ья and (1)
-- are never compatible). We can't have all four because if there's
-- an explicit plural variant, there won't be a user-requested or
-- autodetected plural variant.
--
-- The goal below is to do two things: Check that all available plural
-- variants are the same, and generate the actual declension.
-- If we have a type-2 or type-3 variant, we already have the actual
-- declension; else we need to use a table to map the basic declension
-- to the one with the plural variant encoded in it.
--
-- NOTE: The code below was written with a more general plural-variant
-- system. It probably can be simplified a lot now.
-- 1: Handle explicit decl with slash variant
if rfind(decl, "/") then
if want_ya_plural then
-- Don't think this can happen
error("Plural variant " .. want_ya_plural .. " not compatible with slash declension " .. decl)
end
if args.want_sc1 then
error("Special case (1) not compatible with slash declension" .. decl)
end
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- 2: Retrieve explicitly specified or autodetected decl and pl. variant
local basic_decl, detected_or_explicit_plural = rmatch(decl, "^(.*)(%-[^mf]+)$")
if basic_decl == "ь" then
basic_decl = "ь-m"
end
basic_decl = basic_decl or decl
-- 3: Any user-requested plural variant must agree with explicit or
-- autodetected variant.
if want_ya_plural and detected_or_explicit_plural and want_ya_plural ~= detected_or_explicit_plural then
error("Plural variant " .. want_ya_plural .. " requested but plural variant " .. detected_or_explicit_plural .. " detected from plural stem")
end
-- 4: Handle special case (1). Derive the full declension, make sure its
-- plural variant matches any other available plural variants, and
-- return the declension.
if args.want_sc1 then
local sc1_decl = special_case_1_to_plural_variant[basic_decl] or
error("Special case (1) not compatible with declension " .. basic_decl)
local sc1_plural = rsub(sc1_decl, "^.*%-", "-")
local other_plural = want_ya_plural or detected_or_explicit_plural
if other_plural and sc1_plural ~= other_plural then
error("Plural variant " .. other_plural .. " specified or detected, but special case (1) calls for plural variant " .. sc1_plural)
end
return stem, tr, sc1_decl, gender, was_accented, was_plural, was_autodetected
end
-- 5: Handle user-requested plural variant without explicit or detected
-- one. (If an explicit or detected one exists, we've already checked
-- that it agrees with the user-requested one, and so we already have
-- our full declension.)
if want_ya_plural and not detected_or_explicit_plural then
local variant_decl
if plural_variant_detection_map[decl] then
variant_decl = plural_variant_detection_map[decl][want_ya_plural]
end
if variant_decl then
return stem, tr, variant_decl, gender, was_accented, was_plural, was_autodetected
else
return stem, tr, decl .. (args.old and "/ъ-ья" or "/-ья"), gender, was_accented, was_plural, was_autodetected
end
end
-- 6: Just return the full declension, which will include any available
-- plural variant in it.
return stem, tr, decl, gender, was_accented, was_plural, was_autodetected
end
-- Convert soft adjectival declensions into hard ones following certain
-- stem-final consonants. FIXME: We call this in two places, once
-- to handle auto-detection and once to handle explicit declensions; but
-- in the former case we end up calling it twice.
function determine_adj_stem_variant(decl, stem)
local iend = rmatch(decl, "^%+[іи]([йея]?)$")
-- Convert ій/ий to ый after velar or sibilant. This is important for
-- velars; doesn't really matter one way or the other for sibilants as
-- the sibilant rules will convert both sets of endings to the same
-- thing (whereas there will be a difference with о vs. е for velars).
if iend and rfind(stem, "[" .. com.velar .. com.sib .. "]$") then
decl = "+ы" .. iend
-- The following is necessary for -ц, unclear if makes sense for
-- sibilants. (Would be necessary -- I think -- if we were
-- inferring short adjective forms, but we're not.)
elseif decl == "+ее" and rfind(stem, "[" .. com.sib_c .. "]$") then
decl = "+ое"
end
return decl
end
-- Attempt to determine the actual adjective declension based on a
-- combination of the declension the user specified and what can be detected
-- from the stem. DECL is the value the user passed for the declension field,
-- after extraneous annotations have been removed (although none are probably
-- relevant here). What's left is one of the following, which always begins
-- with +:
--
-- 1. +, meaning to autodetect the declension from the stem
-- 2. +ь, same as + but selects +ьий instead of +ий if lemma ends in -ий
-- 3. +short, +mixed or +proper, with the declension partly specified but
-- the particular gender/number-specific short/mixed variant to be
-- autodetected
-- 4. A gender (+m, +f or +n), used only for detecting the singular of
-- plural-form lemmas (this is primarily used in conjunction with template
-- ru-noun+, to explicitly specify the gender; for the actual declension,
-- it doesn't much matter what singular gender we pick since we're a
-- plural only)
-- 5. A gender plus short/mixed/proper/ь (e.g. +f-mixed), again with the gender
-- used only for detecting the singular of plural-form short/mixed lemmas
-- 6. An actual declension, possibly including a slash declension
-- (WARNING: Unclear if slash declensions will work, especially those
-- that are adjective/noun combinations)
--
-- Returns the same seven args as for determine_decl(). The returned
-- declension will always begin with +.
detect_adj_type = function(lemma, tr, decl, old)
local was_autodetected
local base, ending
local basedecl, g = rmatch(decl, "^(%+)([mfn])$")
if not basedecl then
g, basedecl = rmatch(decl, "^%+([mfn])%-([a-zь]+)$")
if basedecl then
basedecl = "+" .. basedecl
end
end
decl = basedecl or decl
if decl == "+" or decl == "+ь" then
base, ending = rmatch(lemma, "^(.*)([ыиіьаяое]́?[йея])$")
if ending == "ий" and decl == "+ь" then
decl = "+ьий"
elseif ending == "ій" and decl == "+ь" then
decl = "+ьій"
elseif ending then
decl = "+" .. ending
else
base, ending = rmatch(lemma, "^(.-)([оаыъ]?́?)$")
assert(base)
local shortmixed = rfind(base, "^[" .. com.uppercase .. "].*[иы]́н$") and "stressed-proper" or -- accented
rfind(base, "^[" .. com.uppercase .. "].*[иы]н$") and "proper" or --not accented
rlfind(base, "[ёео]́?в$") and "short" or
rlfind(base, "[ыи]́н$") and "stressed-short" or -- accented
rlfind(base, "[ыи]н$") and "mixed" --not accented
if not shortmixed then
error("Cannot determine stem type of adjective: " .. lemma)
end
decl = "+" .. ending .. "-" .. shortmixed
end
was_autodetected = true
elseif ut.contains({"+short", "+mixed", "+proper"}, decl) then
base, ending = rmatch(lemma, "^(.-)([оаыъ]?́?)$")
assert(base)
local shortmixed = usub(decl, 2)
if rlfind(base, "[ыи]́н$") then -- accented
if shortmixed == "short" then shortmixed = "stressed-short"
elseif shortmixed == "proper" then shortmixed = "stressed-proper"
end
end
decl = "+" .. ending .. "-" .. shortmixed
was_autodetected = true
else
base = lemma
end
if ending and ending ~= "" then
tr = nom.strip_tr_ending(tr, ending)
end
-- Remove any accents from the declension, but not their presence.
-- We will convert was_accented into stress pattern b, and convert that
-- back to an accented version in determine_stress_variant(). This way
-- we end up with the stressed version whether the user placed an accent
-- in the ending or decl or specified stress pattern b.
-- FIXME, might not work in the presence of slash declensions
local was_accented = com.is_stressed(decl)
decl = com.remove_accents(decl)
decl = map_decl(decl, function(decl)
return determine_adj_stem_variant(decl, base)
end)
local singdecl
if decl == "+ые" then
singdecl = (g == "m" or not g) and (was_accented and "+ой" or "+ый") or not old and g == "f" and "+ая" or not old and g == "n" and "+ое"
elseif decl == "+ыя" and old then
singdecl = (g == "f" or not g) and "+ая" or g == "n" and "+ое"
elseif decl == "+ие" and not old then
singdecl = (g == "m" or not g) and "+ий" or g == "f" and "+яя" or g == "n" and "+ее"
elseif decl == "+іе" and old and (g == "m" or not g) then
singdecl = "+ій"
elseif decl == "+ія" and old then
singdecl = (g == "f" or not g) and "+яя" or g == "n" and "+ее"
elseif decl == "+ьи" then
singdecl = (g == "m" or not g) and (old and "+ьій" or "+ьий") or g == "f" and "+ья" or g == "n" and "+ье"
elseif rfind(decl, "^%+ы%-") then -- decl +ы-mixed or similar
local beg = (g == "m" or not g) and (old and "ъ" or "") or g == "f" and "а" or g == "n" and "о"
singdecl = beg and "+" .. beg .. usub(decl, 3)
end
if singdecl then
was_plural = true
decl = singdecl
end
return base, tr, canonicalize_decl(decl, old), g, was_accented, was_plural, was_autodetected
end
-- If stress pattern omitted, detect it based on whether ending is stressed
-- or the decl class or stem accent calls for inherent stress, defaulting to
-- pattern a. This is run after alias resolution and accent removal of DECL;
-- WAS_ACCENTED indicates whether the ending was originally stressed.
-- FIXME: This is run before splitting slash patterns but should be run after.
detect_stress_pattern = function(stem, decl, decl_cats, reducible,
was_plural, was_accented)
-- ёнок and ёночек always bear stress
if rfind(decl, "ёнокъ?") or rfind(decl, "ёночекъ?") then
return "b"
-- stressed suffix и́н; missing in plural and true endings don't bear stress
-- (except for exceptional господи́н)
elseif rfind(decl, "инъ?") and was_accented then
return "d"
-- Adjectival -ой always bears the stress
elseif rfind(decl, "%+ой") then
return "b"
-- Adjectival stressed-short, stressed-proper bears the stress
elseif rfind(decl, "^%+.*%-stressed") then
return "b"
-- Pattern b if ending was accented by user
elseif was_accented then
return "b"
-- Nonsyllabic stem means pattern b
elseif com.is_nonsyllabic(stem) then
return "b"
-- Accent on reducible vowel in masc nom sg (not plural) means pattern b.
-- Think about whether we want to enable this.
-- elseif reducible and not was_plural then
-- -- FIXME hack. Eliminate plural part of slash declension.
-- decl = rsub(decl, "/.*", "")
-- if decl_cats[decl] and decl_cats[decl].g == "m" then
-- if com.is_ending_stressed(stem) or com.is_monosyllabic(stem) then
-- return "b"
-- end
-- end
end
return "a"
end
-- In certain special cases, depending on the declension, we override the
-- user-specified stress pattern and convert it to something else.
-- NOTE: This function is run after alias resolution and accent removal.
-- FIXME: It's also run before splitting slash patterns but should be run after.
override_stress_pattern = function(decl, stress)
-- ёнок and ёночек always bear stress; if user specified a,
-- convert to b. Don't do this with slash patterns (see FIXME above).
if stress == "a" and (rfind(decl, "^ёнокъ?$") or rfind(decl, "^ёночекъ?$")) then
return "b"
end
return stress
end
-- Canonicalize an adjectival declension to either the stressed or unstressed
-- variant depending on the stress. Ultimately this is what ensures that
-- the user's stress mark on an adjectival ending is respected.
determine_stress_variant = function(decl, stress)
if stress == "b" then
if decl == "+ая" then
return "+а́я"
elseif decl == "+ое" then
return "+о́е"
else
-- Convert +...-short to +...-stressed-short and same for -proper
local b, e = rmatch(decl, "^%+(.*)%-(short)$")
if not b then
b, e = rmatch(decl, "^%+(.*)%-(proper)$")
end
if b and not rfind(b, "%-stressed") then
return "+" .. b .. "-stressed-" .. e
end
end
end
return decl
end
-- Canonicalize a declension based on the final stem consonant, in
-- particular converting soft declensions to hard ones after velars and/or
-- sibilants. FIXME: We also do this canonicalization earlier on during
-- auto-detection (determine_adj_stem_variant() is called by
-- detect_adj_type(), and code in detect_lemma_type() does the equivalent
-- of the first clause below). Doing it here ensures that explicitly
-- specified declensions get handled as well, but it would be nice to not
-- do the same thing twice in the auto-detection case.
determine_stem_variant = function(decl, stem)
if decl == "е" and rfind(stem, "[" .. com.sib_c .. "]$") then
return "о"
end
return determine_adj_stem_variant(decl, stem)
end
is_reducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "3rd" and decl_cat.g == "f" or decl_cat.g == "m" then
return true
else
return false
end
end
-- Reduce nom sg to stem by eliminating the "epenthetic" vowel. Applies to
-- masculine 2nd-declension hard and soft, and 3rd-declension feminine in
-- -ь. STEM and DECL are after determine_decl(), before converting
-- outward-facing declensions to inward ones.
function export.reduce_nom_sg_stem(stem, tr, decl, can_err)
local full_stem = stem .. (decl == "й" and decl or "")
local full_tr = tr and tr .. (decl == "й" and "j" or "")
local ret, rettr = com.reduce_stem(full_stem, full_tr)
if not ret and can_err then
error("Unable to reduce stem " .. stem)
end
return ret, rettr
end
is_dereducible = function(decl_cat)
if decl_cat.suffix or decl_cat.cant_reduce or decl_cat.adj then
return false
elseif decl_cat.decl == "1st" or decl_cat.decl == "2nd" and decl_cat.g == "n" then
return true
else
return false
end
end
-- Add a possible suffix to the bare stem, according to the declension and
-- value of OLD. This may be -ь, -ъ, -й or nothing. We need to do this here
-- because we don't actually attach such a suffix in attach_unstressed() due
-- to situations where we don't want the suffix added, e.g. dereducible nouns
-- in -ня.
add_bare_suffix = function(bare, baretr, old, sgdc, dereduced)
if old and sgdc.hard == "hard" then
-- Final -ъ isn't transliterated
return bare .. "ъ", baretr
elseif sgdc.hard == "soft" or sgdc.hard == "palatal" then
-- This next clause corresponds to a special case in Vitalik's module.
-- It says that nouns in -ня (accent class a) have gen pl without
-- trailing -ь. It appears to apply to most nouns in -ня (possibly
-- all in -льня), but ку́хня (gen pl ку́хонь) and дерéвня (gen pl
-- дереве́нь) is an exception. (Vitalik's module has an extra
-- condition here 'stress == "a"' that would exclude дере́вня but I
-- don't think this condition is in Zaliznyak, as he indicates
-- дере́вня as having an exceptional genitive plural.)
if dereduced and rfind(bare, "[нН]$") and sgdc.decl == "1st" then
-- FIXME: What happens in this case old-style? I assume that
-- -ъ is added, but this is a guess.
-- Final -ъ isn't transliterated
return bare .. (old and "ъ" or ""), baretr
elseif rfind(bare, "[" .. com.vowel .. "]́?$") then
return bare .. "й", baretr and (baretr .. "j")
else
return bare .. "ь", baretr and (baretr .. "ʹ")
end
else
return bare, baretr
end
end
-- Dereduce stem to the form found in the gen pl (and maybe nom sg) by
-- inserting an epenthetic vowel. Applies to 1st declension and 2nd
-- declension neuter, and to 2nd declension masculine when the stem was
-- specified as a plural form (in which case we're deriving the nom sg,
-- and also the gen pl in the alt-gen-pl scenario). STEM and DECL are
-- after determine_decl(), before converting outward-facing declensions
-- to inward ones. STRESS is the stess pattern.
function export.dereduce_nom_sg_stem(stem, tr, sgdc, stress, old, can_err)
local epenthetic_stress = ending_stressed_gen_pl_patterns[stress]
local ret, rettr = com.dereduce_stem(stem, tr, epenthetic_stress)
if not ret then
if can_err then
error("Unable to dereduce stem " .. stem)
else
return nil, nil
end
end
return add_bare_suffix(ret, rettr, old, sgdc, true)
end
--------------------------------------------------------------------------
-- Second-declension masculine --
--------------------------------------------------------------------------
----------------- Masculine hard -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style).
declensions_old["ъ"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and "е́й" or "о́въ"
end,
["alt_gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["ъ"] = { decl="2nd", hard="hard", g="m" }
-- Normal mapping of old ъ would be "" (blank), but we set up "#" as an alias
-- so we have a way of referring to it without defaulting if need be and
-- distinct from auto-detection (e.g. in the second stem of a word, or to
-- override autodetection of -ёнок or -ин -- the latter is necessary in the
-- case of семьянин).
declensions_aliases["#"] = ""
----------------- Masculine hard, irregular plural -------------------
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg nom pl -а.
declensions_old["ъ-а"] = mw.clone(declensions_old["ъ"])
declensions_old["ъ-а"]["nom_pl"] = "а́"
declensions_old_cat["ъ-а"] = { decl="2nd", hard="hard", g="m", alt_nom_pl=true }
declensions_aliases["#-a"] = "-a"
-- Hard-masculine declension, ending in a hard consonant
-- (ending in -ъ, old-style), with irreg soft pl -ья.
-- Differs from the normal declension throughout the plural.
declensions_old["ъ-ья"] = {
["nom_sg"] = "ъ",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = nil,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = "ьёвъ",
["alt_gen_pl"] = "е́й",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ъ-ья"] = { decl="2nd", hard="hard", g="m", irregpl=true }
declensions_aliases["#-ья"] = "-ья"
----------------- Masculine hard, suffixed, irregular plural -------------------
declensions_old["инъ"] = {
["nom_sg"] = "и́нъ",
["gen_sg"] = "и́на",
["dat_sg"] = "и́ну",
["acc_sg"] = nil,
["ins_sg"] = "и́номъ",
["pre_sg"] = "и́нѣ",
["nom_pl"] = "е́",
["gen_pl"] = "ъ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["инъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old["ёнокъ"] = {
["nom_sg"] = "ёнокъ",
["gen_sg"] = "ёнка",
["dat_sg"] = "ёнку",
["acc_sg"] = nil,
["ins_sg"] = "ёнкомъ",
["pre_sg"] = "ёнкѣ",
["nom_pl"] = "я́та",
["gen_pl"] = "я́тъ",
["dat_pl"] = "я́тамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тами",
["pre_pl"] = "я́тахъ",
}
declensions_old_cat["ёнокъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["онокъ"] = "ёнокъ"
declensions_old_aliases["енокъ"] = "ёнокъ"
declensions_old["ёночекъ"] = {
["nom_sg"] = "ёночекъ",
["gen_sg"] = "ёночка",
["dat_sg"] = "ёночку",
["acc_sg"] = nil,
["ins_sg"] = "ёночкомъ",
["pre_sg"] = "ёночкѣ",
["nom_pl"] = "я́тки",
["gen_pl"] = "я́токъ",
["dat_pl"] = "я́ткамъ",
["acc_pl"] = nil,
["ins_pl"] = "я́тками",
["pre_pl"] = "я́ткахъ",
}
declensions_old_cat["ёночекъ"] = { decl="2nd", hard="hard", g="m", suffix=true }
declensions_old_aliases["оночекъ"] = "ёночекъ"
declensions_old_aliases["еночекъ"] = "ёночекъ"
----------------- Masculine soft -------------------
-- Normal soft-masculine declension in -ь
declensions_old["ь-m"] = {
["nom_sg"] = "ь",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["alt_gen_pl"] = "ь",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-m"] = { decl="2nd", hard="soft", g="m" }
-- Soft-masculine declension in -ь with irreg nom pl -я
declensions_old["ь-я"] = mw.clone(declensions_old["ь-m"])
declensions_old["ь-я"]["nom_pl"] = "я́"
declensions_old_cat["ь-я"] = { decl="2nd", hard="soft", g="m", alt_nom_pl=true }
----------------- Masculine palatal -------------------
-- Masculine declension in palatal -й
declensions_old["й"] = {
["nom_sg"] = "й",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = nil,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи]́?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = "ёвъ",
["alt_gen_pl"] = "й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["й"] = { decl="2nd", hard="palatal", g="m" }
declensions_old["й-я"] = mw.clone(declensions_old["й"])
declensions_old["й-я"]["nom_pl"] = "я́"
declensions_old_cat["й-я"] = { decl="2nd", hard="palatal", g="m", alt_nom_pl=true }
--------------------------------------------------------------------------
-- First-declension feminine --
--------------------------------------------------------------------------
----------------- Feminine hard -------------------
-- Hard-feminine declension in -а
declensions_old["а"] = {
["nom_sg"] = "а́",
["gen_sg"] = "ы́",
["dat_sg"] = "ѣ́",
["acc_sg"] = "у́",
["ins_sg"] = {"о́й<insa>", "о́ю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ѣ́",
["nom_pl"] = "ы́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["а"] = { decl="1st", hard="hard", g="f" }
----------------- Feminine soft -------------------
-- Soft-feminine declension in -я
declensions_old["я"] = {
["nom_sg"] = "я́",
["gen_sg"] = "и́",
["dat_sg"] = function(stem, stress)
return rlfind(stem, "[іи]́?$") and not ending_stressed_dat_sg_patterns[stress] and "и" or "ѣ́"
end,
["acc_sg"] = "ю́",
["ins_sg"] = {"ёй<insa>", "ёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи]́?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "и́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "]́?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["я"] = { decl="1st", hard="soft", g="f" }
-- Soft-feminine declension in -ья.
-- Almost like ь + -я endings except for genitive plural.
declensions_old["ья"] = {
["nom_sg"] = "ья́",
["gen_sg"] = "ьи́",
["dat_sg"] = "ьѣ́",
["acc_sg"] = "ью́",
["ins_sg"] = {"ьёй<insa>", "ьёю<insb>"}, -- see concat_word_forms_1()
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ьи́",
["gen_pl"] = function(stem, stress)
-- circumflex accent is a signal that forces stress, particularly
-- in accent pattern d/d'.
return (ending_stressed_gen_pl_patterns[stress] or stress == "d" or stress == "d'") and "е̂й" or "ий"
end,
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ья"] = {
decl="1st", hard="soft", g="f",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
--------------------------------------------------------------------------
-- Second-declension neuter --
--------------------------------------------------------------------------
----------------- Neuter hard -------------------
-- Normal hard-neuter declension in -о
declensions_old["о"] = {
["nom_sg"] = "о́",
["gen_sg"] = "а́",
["dat_sg"] = "у́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "о́" or nil
end,
["ins_sg"] = "о́мъ",
["pre_sg"] = "ѣ́",
["nom_pl"] = "а́",
["gen_pl"] = function(stem, stress)
return nom.sibilant_suffixes[ulower(usub(stem, -1))] and ending_stressed_gen_pl_patterns[stress] and "е́й" or "ъ"
end,
["alt_gen_pl"] = "о́въ",
["dat_pl"] = "а́мъ",
["acc_pl"] = nil,
["ins_pl"] = "а́ми",
["pre_pl"] = "а́хъ",
}
declensions_old_cat["о"] = { decl="2nd", hard="hard", g="n" }
-- Hard-neuter declension in -о with irreg nom pl -и
declensions_old["о-и"] = mw.clone(declensions_old["о"])
declensions_old["о-и"]["nom_pl"] = "ы́"
declensions_old_cat["о-и"] = { decl="2nd", hard="hard", g="n", alt_nom_pl=true }
declensions_old_aliases["о-ы"] = "о-и"
-- Masculine-gender neuter-form declension in -(ишк)о with irreg nom pl -и,
-- with colloquial feminine endings in some of the singular cases
-- (§5 p. 74 of Zaliznyak)
declensions_old["(ишк)о-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ишк)о-и"]["gen_sg"] = {"а́", "ы́1"}
declensions_old["(ишк)о-и"]["dat_sg"] = {"у́", "ѣ́1"}
declensions_old["(ишк)о-и"]["ins_sg"] = {"о́мъ", "о́й1"}
declensions_old_cat["(ишк)о-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ишк)о-и"] = "<sup>1</sup> Colloquial."
-- Masculine-gender animate neuter-form declension in -(ищ)е with irreg
-- nom pl -и, with colloquial feminine endings in some of the singular cases
-- (§4 p. 74 of Zaliznyak)
declensions_old["(ищ)е-и"] = mw.clone(declensions_old["о-и"])
declensions_old["(ищ)е-и"]["acc_sg"] = {"а́", "у́1"}
declensions_old["(ищ)е-и"]["gen_sg"] = {"а́", "ы́2"}
declensions_old["(ищ)е-и"]["dat_sg"] = {"у́", "ѣ́2"}
declensions_old["(ищ)е-и"]["ins_sg"] = {"о́мъ", "о́й2"}
declensions_old_cat["(ищ)е-и"] = { decl="2nd", hard="hard", g="n", colloqfem=true, alt_nom_pl=true }
internal_notes_table_old["(ищ)е-и"] = "<sup>1</sup> Colloquial.<br /><sup>2</sup> Less common, more colloquial."
----------------- Neuter soft -------------------
-- Soft-neuter declension in -е (stressed -ё)
declensions_old["е"] = {
["nom_sg"] = "ё",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ё" or nil
end,
["ins_sg"] = "ёмъ",
["pre_sg"] = function(stem, stress)
return rlfind(stem, "[іи]́?$") and not ending_stressed_pre_sg_patterns[stress] and "и" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and not rlfind(stem, "[" .. com.vowel .. "]́?$") and "е́й" or "й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е"] = {
singular = function(suffix)
if suffix == "ё" then
return "ending in -ё"
else
return {}
end
end,
decl="2nd", hard="soft", g="n", gensg=true
}
-- User-facing declension type "ё" = "е"
declensions_old_aliases["ё"] = "е"
-- Rare soft-neuter declension in stressed -е́ (e.g. муде́, бытие́)
declensions_old["е́"] = {
["nom_sg"] = "е́",
["gen_sg"] = "я́",
["dat_sg"] = "ю́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "е́" or nil
end,
["ins_sg"] = "е́мъ",
["pre_sg"] = function(stem, stress)
-- FIXME!!! Are we sure about this condition? This is what was
-- found in the old template, but the related -е declension has
-- -ие prep sg ending -(и)и only when *not* stressed.
return rlfind(stem, "[іи]́?$") and "и́" or "ѣ́"
end,
["nom_pl"] = "я́",
["gen_pl"] = function(stem, stress)
return rlfind(stem, "[" .. com.vowel .. "]́?$") and "й" or "е́й"
end,
["alt_gen_pl"] = "ёвъ",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["е́"] = {
singular = "ending in stressed -е",
decl="2nd", hard="soft", g="n", gensg=true
}
-- Soft-neuter declension in unstressed -ье (stressed -ьё).
declensions_old["ье"] = {
["nom_sg"] = "ьё",
["gen_sg"] = "ья́",
["dat_sg"] = "ью́",
["acc_sg"] = function(stem, stress, args)
return not (args.explicit_gender == "m" and args.thisa == "a") and "ьё" or nil
end,
["ins_sg"] = "ьёмъ",
["pre_sg"] = "ьѣ́",
["nom_pl"] = "ья́",
["gen_pl"] = function(stem, stress)
return ending_stressed_gen_pl_patterns[stress] and "е́й" or "ий"
end,
["alt_gen_pl"] = "ьёвъ",
["dat_pl"] = "ья́мъ",
["acc_pl"] = nil,
["ins_pl"] = "ья́ми",
["pre_pl"] = "ья́хъ",
}
declensions_old_cat["ье"] = {
decl="2nd", hard="soft", g="n",
stem_suffix="ь", gensg=true,
ignore_reduce=true -- already has dereduced gen pl
}
declensions_old_aliases["ьё"] = "ье"
--------------------------------------------------------------------------
-- Third declension --
--------------------------------------------------------------------------
declensions_old["ь-f"] = {
["nom_sg"] = "ь",
["gen_sg"] = "и́",
["dat_sg"] = "и́",
["acc_sg"] = "ь",
["ins_sg"] = "ью́",
["pre_sg"] = "и́",
["nom_pl"] = "и́",
["gen_pl"] = "е́й",
["dat_pl"] = "я́мъ",
["acc_pl"] = nil,
["ins_pl"] = "я́ми",
["pre_pl"] = "я́хъ",
}
declensions_old_cat["ь-f"] = { decl="3rd", hard="soft", g="f" }
declensions_old["мя"] = {
["nom_sg"] = "мя",
["gen_sg"] = "мени",
["dat_sg"] = "мени",
["acc_sg"] = nil,
["ins_sg"] = "менемъ",
["pre_sg"] = "мени",
["nom_pl"] = "мена́",
["gen_pl"] = "мёнъ",
["dat_pl"] = "мена́мъ",
["acc_pl"] = nil,
["ins_pl"] = "мена́ми",
["pre_pl"] = "мена́хъ",
}
declensions_old_cat["мя"] = { decl="3rd", hard="soft", g="n", cant_reduce=true }
--------------------------------------------------------------------------
-- Invariable --
--------------------------------------------------------------------------
-- Invariable declension; no endings.
declensions_old["$"] = {
["nom_sg"] = "",
["gen_sg"] = "",
["dat_sg"] = "",
["acc_sg"] = nil,
["ins_sg"] = "",
["pre_sg"] = "",
["nom_pl"] = "",
["gen_pl"] = "",
["dat_pl"] = "",
["acc_pl"] = nil,
["ins_pl"] = "",
["pre_pl"] = "",
}
declensions_old_cat["$"] = { decl="invariable", hard="none", g="none" }
--------------------------------------------------------------------------
-- Adjectival --
--------------------------------------------------------------------------
-- This needs to be up here because it is called just below.
local function old_to_new(v)
v = rsub(v, "ъ$", "")
v = rsub(v, "^ъ", "")
v = rsub(v, "(%A)ъ", "%1")
v = rsub(v, "ъ(%A)", "%1")
v = rsub(v, "і", "и")
v = rsub(v, "ѣ", "е")
return v
end
-- Meaning of entry is:
-- 1. The declension name in module ru-adjective
-- 2. The masculine declension name in this module
-- 3. The neuter declension name in this module
-- 4. The feminine declension name in this module
-- 5. The value of hard= for the declensions_cat entry
-- 6. The value of decl= for the declensions_cat entry
-- 7. The value of possadj= for the declensions_cat entry (true if possessive
-- or similar type of adjective)
local adj_decl_map = {
{"ый", "ый", "ое", "ая", "hard", "long", false},
{"ій", "ій", "ее", "яя", "soft", "long", false},
{"ой", "ой", "о́е", "а́я", "hard", "long", false},
{"ьій", "ьій", "ье", "ья", "palatal", "long", true},
{"short", "ъ-short", "о-short", "а-short", "hard", "short", true},
{"mixed", "ъ-mixed", "о-mixed", "а-mixed", "hard", "mixed", true},
{"proper", "ъ-proper", "о-proper", "а-proper", "hard", "proper", true},
{"stressed-short", "ъ-stressed-short", "о-stressed-short", "а-stressed-short", "hard", "short", true},
{"stressed-proper", "ъ-stressed-proper", "о-stressed-proper", "а-stressed-proper", "hard", "proper", true},
}
local function get_adjectival_decl(adjtype, gender, old)
local decl, intnotes = m_ru_adj.get_nominal_decl(adjtype, gender, old)
-- hack fem ins_sg to insert <insa>, <insb>; see concat_word_forms_1()
if gender == "f" and type(decl["ins_sg"]) == "table" and #decl["ins_sg"] == 2 then
decl["ins_sg"][1] = decl["ins_sg"][1] .. "<insa>"
decl["ins_sg"][2] = decl["ins_sg"][2] .. "<insb>"
end
return decl, intnotes
end
for _, declspec in ipairs(adj_decl_map) do
local oadjdecl = declspec[1]
local nadjdecl = old_to_new(oadjdecl)
local odecl_by_gender =
{m="+" .. declspec[2], n="+" .. declspec[3], f="+" .. declspec[4]}
local hard = declspec[5]
local decltype = declspec[6]
local possadj = declspec[7]
for _, g in ipairs({"m", "n", "f"}) do
local odecl = odecl_by_gender[g]
local ndecl = old_to_new(odecl)
declensions_old[odecl], internal_notes_table_old[odecl] =
get_adjectival_decl(oadjdecl, g, true)
declensions[ndecl], internal_notes_table[ndecl] =
get_adjectival_decl(nadjdecl, g, false)
declensions_old_cat[odecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
declensions_cat[ndecl] = {
decl=decltype, hard=hard, g=g, adj=true, possadj=possadj }
end
end
-- Set up some aliases.
declensions_old_aliases["+о́-short"] = "+о-stressed-short"
declensions_old_aliases["+а́-short"] = "+а-stressed-short"
declensions_old_aliases["+о́-proper"] = "+о-stressed-proper"
declensions_old_aliases["+а́-proper"] = "+а-stressed-proper"
declensions_aliases["+#-short"] = "+-short"
declensions_aliases["+#-mixed"] = "+-mixed"
declensions_aliases["+#-proper"] = "+-proper"
declensions_aliases["+#-stressed-short"] = "+-stressed-short"
declensions_aliases["+#-stressed-proper"] = "+-stressed-proper"
--------------------------------------------------------------------------
-- Populate new from old --
--------------------------------------------------------------------------
-- Function to convert an entry in an old declensions table to new.
local function old_decl_entry_to_new(v)
if not v then
return nil
elseif type(v) == "table" then
local new_entry = {}
for _, i in ipairs(v) do
table.insert(new_entry, old_decl_entry_to_new(i))
end
return new_entry
elseif type(v) == "function" then
return function(stem, suffix, args)
return old_decl_entry_to_new(v(stem, suffix, args))
end
else
return old_to_new(v)
end
end
-- Function to convert an old declensions table to new.
local function old_decl_to_new(odecl)
local ndecl = {}
for k, v in pairs(odecl) do
ndecl[k] = old_decl_entry_to_new(v)
end
return ndecl
end
-- Function to convert an entry in an old declensions_cat table to new.
local function old_decl_cat_entry_to_new(odecl_cat_entry)
if not odecl_cat_entry then
return nil
elseif type(odecl_cat_entry) == "function" then
return function(suffix)
return old_decl_cat_entry_to_new(odecl_cat_entry(suffix))
end
elseif type(odecl_cat_entry) == "table" then
local ndecl_cat_entry = {}
for k, v in pairs(odecl_cat_entry) do
ndecl_cat_entry[k] = old_decl_cat_entry_to_new(v)
end
return ndecl_cat_entry
elseif type(odecl_cat_entry) == "boolean" then
return odecl_cat_entry
else
assert(type(odecl_cat_entry) == "string")
return old_to_new(odecl_cat_entry)
end
end
-- Function to convert an old declensions_cat table to new.
local function old_decl_cat_to_new(odeclcat)
local ndeclcat = {}
for k, v in pairs(odeclcat) do
ndeclcat[k] = old_decl_cat_entry_to_new(v)
end
return ndeclcat
end
-- populate declensions[] from declensions_old[]
for odecltype, odecl in pairs(declensions_old) do
local ndecltype = old_to_new(odecltype)
if not declensions[ndecltype] then
declensions[ndecltype] = old_decl_to_new(odecl)
end
end
-- populate declensions_cat[] from declensions_old_cat[]
for odecltype, odeclcat in pairs(declensions_old_cat) do
local ndecltype = old_to_new(odecltype)
if not declensions_cat[ndecltype] then
declensions_cat[ndecltype] = old_decl_cat_to_new(odeclcat)
end
end
-- populate declensions_aliases[] from declensions_old_aliases[]
for ofrom, oto in pairs(declensions_old_aliases) do
local from = old_to_new(ofrom)
if not declensions_aliases[from] then
declensions_aliases[from] = old_to_new(oto)
end
end
-- populate internal_notes_table[] from internal_notes_table_old[]
for odecl, note in pairs(internal_notes_table_old) do
local ndecl = old_to_new(odecl)
if not internal_notes_table[ndecl] then
-- FIXME, should we be calling old_to_new() here?
internal_notes_table[ndecl] = note
end
end
--------------------------------------------------------------------------
-- Inflection functions --
--------------------------------------------------------------------------
-- Attach the stressed stem (or plural stem, or barestem) out of ARGS
-- to the unstressed suffix SUF, modifying the suffix as necessary for the
-- last letter of the stem (e.g. if it is velar, sibilant or ц). CASE is
-- the case form being created and is used to select the plural stem if
-- needed. Returns two values, the combined form and the modified suffix.
local function attach_unstressed(args, case, suf, was_stressed)
if suf == nil then
return nil, nil
elseif rfind(suf, CFLEX) then -- if suf has circumflex accent, it forces stressed
return attach_stressed(args, case, suf)
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.pl, args.pltr
end
if not stem and case == "ins_sg" then
stem, tr = args.ins_sg_stem, args.ins_sg_tr
end
if not stem then
stem, tr = args.stem, args.stemtr
end
if nom.nonsyllabic_suffixes[suf] then
-- If gen_pl, use special args.gen_pl_bare if given, else regular
-- args.bare if there isn't a plural stem. If nom_sg, always use
-- regular args.bare.
local barearg, bareargtr
if case == "gen_pl" then
barearg, bareargtr = args.gen_pl_bare, args.gen_pl_baretr
if not barearg and args.pl == args.stem then
barearg, bareargtr = args.bare, args.baretr
end
else
barearg, bareargtr = args.bare, args.baretr
end
local barestem = barearg or stem
local barestem, baretr
if barearg then
barestem, baretr = barearg, bareargtr
else
barestem, baretr = stem, tr
end
if was_stressed and case == "gen_pl" then
if not barearg then
local gen_pl_stem, gen_pl_tr = com.make_ending_stressed(stem, tr)
barestem, baretr = gen_pl_stem, gen_pl_tr
end
end
if rlfind(barestem, "[йьъ]$") then
suf = ""
else
if suf == "ъ" then
-- OK
elseif suf == "й" or suf == "ь" then
if barearg and case == "gen_pl" then
-- explicit bare or reducible, don't add -ь
suf = ""
elseif rfind(barestem, "[" .. com.vowel .. "]́?$") then
-- not reducible, do add -ь and correct to -й if necessary
suf = "й"
else
suf = "ь"
end
end
end
return nom.concat_russian_tr(barestem, baretr, suf, nil, "dopair"), suf
end
suf = com.make_unstressed(suf)
local rules = nom.unstressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Analogous to attach_unstressed() but for the unstressed stem and a
-- stressed suffix.
attach_stressed = function(args, case, suf)
if suf == nil then
return nil, nil
end
-- circumflex forces stress even when the accent pattern calls for no stress
suf = rsub(suf, "̂", "́")
if not rfind(suf, "[ё́]") then -- if suf has no "ё" or accent marks
return attach_unstressed(args, case, suf, "was stressed")
end
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.upl, args.upltr
end
if not stem then
stem, tr = args.ustem, args.ustemtr
end
local rules = nom.stressed_rules[ulower(usub(stem, -1))]
return nom.combine_stem_and_suffix(stem, tr, suf, rules, args.old)
end
-- Attach the appropriate stressed or unstressed stem (or plural stem as
-- determined by CASE, or barestem) out of ARGS to the suffix SUF, which may
-- be a list of alternative suffixes (e.g. in the inst sg of feminine nouns).
-- Calls FUN (either attach_stressed() or attach_unstressed()) to do the work
-- for an individual suffix. Returns two values, a list of combined forms
-- and a list of the real suffixes used (which may be modified from the
-- passed-in suffixes, e.g. by removing stress marks or modifying vowels in
-- various ways after a stem-final velar, sibilant or ц). Each combined form
-- is a two-element list {stem, tr} (or a one-element list if tr is nil).
-- IRREG is true if this is an irregular form. We are handling the Nth word;
-- ISLAST is true if this is the last one.
local function attach_with(args, case, suf, fun, irreg, n, islast)
if type(suf) == "table" then
local all_combineds = {}
local all_realsufs = {}
for _, x in ipairs(suf) do
local combineds, realsufs =
attach_with(args, case, x, fun, irreg, n, islast)
for _, combined in ipairs(combineds) do
table.insert(all_combineds, combined)
end
for _, realsuf in ipairs(realsufs) do
table.insert(all_realsufs, realsuf)
end
end
return all_combineds, all_realsufs
else
local combined, realsuf = fun(args, case, suf)
local irregsuf = irreg and {IRREGMARKER} or {""}
return {combined and nom.concat_paired_russian_tr(
nom.concat_paired_russian_tr(args["prefix" .. n], combined),
nom.concat_paired_russian_tr(args["suffix" .. n], irregsuf)) or nil},
{realsuf and realsuf .. args["suffix" .. n][1] or nil}
end
end
-- Generate the form(s) and suffix(es) for CASE according to the declension
-- table DECL, using the attachment function FUN (one of attach_stressed()
-- or attach_unstressed()). IS_SLASH is true if this is a slash declension
-- (different declensions for singular and plural). We are handling the Nth
-- word; ISLAST is true if this is the last one.
local function gen_form(args, decl, case, stress, fun, is_slash, n, islast)
local irreg = false
if not args.suffixes[case] then
args.suffixes[case] = {}
end
local decl_sufs = args.old and declensions_old or declensions
decl_sufs = decl_sufs[decl]
local suf = decl_sufs[case]
local decl_cats = args.old and declensions_old_cat or declensions_cat
local ispl = rfind(case, "_pl")
if ispl and (decl_cats[decl].irregpl or args.pl and args.pl ~= args.stem or is_slash) then
irreg = true
end
if case == "nom_pl" and decl_cats[decl].alt_nom_pl then
irreg = true
end
if type(suf) == "function" then
suf = suf(ispl and args.pl or args.stem, stress, args)
end
if case == "gen_pl" and args.alt_gen_pl then
suf = decl_sufs.alt_gen_pl
irreg = true
if not suf then
error("No alternate genitive plural available for this declension class")
end
end
local combineds, realsufs = attach_with(args, case, suf, fun, irreg, n, islast)
for _, realsuf in ipairs(realsufs) do
args.any_non_nil[case] = true
args.this_any_non_nil[case] = true
insert_if_not(args.suffixes[case], realsuf)
end
return combineds
end
local attachers = {
["+"] = attach_stressed,
["-"] = attach_unstressed,
}
do_stress_pattern = function(stress, args, decl, number, n, islast)
local f = {}
for _, case in ipairs(decl_cases) do
if not number or (number == "sg" and rfind(case, "_sg")) or
(number == "pl" and rfind(case, "_pl")) then
f[case] = gen_form(args, decl, case, stress,
attachers[stress_patterns[stress][case]], not not number,
n, islast)
-- Turn empty form lists into nil to facilitate computation of
-- animate/inanimate accusatives below
if f[case] and #f[case] == 0 then
f[case] = nil
end
-- Compute linked versions of potential lemma cases, for use
-- in the ru-noun+ headword. We substitute the original lemma
-- (before removing links) for forms that are the same as the
-- lemma, if the original lemma has links.
if f[case] and (case == "nom_sg" or case == "nom_pl") then
local linked_forms = {}
for _, form in ipairs(f[case]) do
-- Return true if FORM is "close enough" to LEMMA that we can substitute the
-- linked form of the lemma. Currently this means exactly the same except that
-- we ignore acute and grave accent differences in monosyllables, and ignore
-- notes that may have been appended (e.g. the triangle marking irregularity).
local entry, notes = m_table_tools.separate_notes(form[1])
local lemma = args.lemma_no_links
local close_enough_to_lemma =
entry == lemma or (com.is_monosyllabic(entry) and
com.is_monosyllabic(lemma) and
com.remove_accents(entry) == com.remove_accents(lemma))
if close_enough_to_lemma and
rfind(args.orig_lemma, "%[%[") then
table.insert(linked_forms, {args.orig_lemma .. notes, args.lemmatr and args.lemmatr .. notes})
else
table.insert(linked_forms, form)
end
end
f[case .. "_linked"] = linked_forms
end
end
end
-- Set acc an/in variants now as appropriate. We used to do this in
-- handle_forms_and_overrides(), which simplified the handling of
-- nom/gen/acc overrides but caused problems for words like мазло and
-- трепло that had a mixture of nil and non-nil accusative forms.
local an = args.thisa
if not number or number == "sg" then
f.acc_sg_an = f.acc_sg_an or f.acc_sg or an == "i" and f.nom_sg or f.gen_sg
f.acc_sg_in = f.acc_sg_in or f.acc_sg or an == "a" and f.gen_sg or f.nom_sg
end
if not number or number == "pl" then
f.acc_pl_an = f.acc_pl_an or f.acc_pl or an == "i" and f.nom_pl or f.gen_pl
f.acc_pl_in = f.acc_pl_in or f.acc_pl or an == "a" and f.gen_pl or f.nom_pl
end
for case, forms in pairs(f) do
if not args.forms[case] then
args.forms[case] = {}
end
for _, form in ipairs(forms) do
insert_if_not(args.forms[case], form)
end
end
end
stress_patterns["a"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["b"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["b'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["c"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="+", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["d"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["d'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="-", dat_pl="-", acc_pl="-", ins_pl="-", pre_pl="-",
}
stress_patterns["e"] = {
nom_sg="-", gen_sg="-", dat_sg="-", acc_sg="-", ins_sg="-", pre_sg="-",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f'"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="-", ins_sg="+", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
stress_patterns["f''"] = {
nom_sg="+", gen_sg="+", dat_sg="+", acc_sg="+", ins_sg="-", pre_sg="+",
nom_pl="-", gen_pl="+", dat_pl="+", acc_pl="+", ins_pl="+", pre_pl="+",
}
ending_stressed_gen_pl_patterns = ut.list_to_set({"b", "b'", "c", "e", "f", "f'", "f''"})
ending_stressed_pre_sg_patterns = ut.list_to_set({"b", "b'", "d", "d'", "f", "f'", "f''"})
ending_stressed_dat_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_sg_patterns = ending_stressed_pre_sg_patterns
ending_stressed_pl_patterns = ut.list_to_set({"b", "b'", "c"})
local numbers = {
["s"] = "singular",
["p"] = "plural",
}
local old_title_temp = [=[Pre-reform declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local title_temp = [=[Declension of <b lang="ru" class="Cyrl">{lemma}</b>]=]
local partitive = nil
local locative = nil
local vocative = nil
local internal_notes_template = nil
local notes_template = nil
local templates = {}
-- all cases, period
all_cases = {
"nom_sg", "gen_sg", "dat_sg", "acc_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "acc_pl", "ins_pl", "pre_pl",
"nom_sg_linked", "nom_pl_linked",
"acc_sg_an", "acc_sg_in", "acc_pl_an", "acc_pl_in",
"par", "loc", "voc",
}
-- cases that can be declined normally
decl_cases = {
"nom_sg", "gen_sg", "dat_sg", "acc_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "acc_pl", "ins_pl", "pre_pl",
}
-- all cases displayable
displayable_cases = {
"nom_sg", "gen_sg", "dat_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "ins_pl", "pre_pl",
"nom_sg_linked", "nom_pl_linked",
"acc_sg_an", "acc_sg_in", "acc_pl_an", "acc_pl_in",
"par", "loc", "voc",
}
-- all cases that can be overridden
overridable_cases = {
"nom_sg", "gen_sg", "dat_sg", "acc_sg", "ins_sg", "pre_sg",
"nom_pl", "gen_pl", "dat_pl", "acc_pl", "ins_pl", "pre_pl",
"acc_sg_an", "acc_sg_in", "acc_pl_an", "acc_pl_in",
"par", "loc", "voc",
}
-- Convert a raw override into a canonicalized list of individual overrides.
-- If input is nil, so is output. Certain junk (e.g. <br/>) is removed,
-- and ~ and ~~ are substituted appropriately; ARGS and CASE are required for
-- this purpose. FORMS is the list of existing case forms and is currently
-- used only to retrieve the dat_sg for handling + in loc/par (this means
-- that any overrides of the dat_sg have to be handled before handling
-- overrides of loc/par). N is the suffix used to retrieve the override --
-- either a number for word-specific overrides, or an empty string for
-- overall overrides.
--
-- It will still be necessary to call m_table_tools.separate_notes() to separate
-- off any trailing "notes" (asterisks, superscript numbers, etc.), and
-- m_links.remove_links() to remove any links to get the raw override form.
canonicalize_override = function(args, case, forms, n)
local val = args[case .. n]
if not val then
return nil
end
-- clean <br /> that's in many multi-form entries and messes up linking
val = rsub(val, "<br%s*/>", "")
-- substitute ~ and ~~ and split by commas
local stem, tr
if rfind(case, "_pl") then
stem, tr = args.pl, args.pltr
end
if not stem then
stem, tr = args.stem, args.stemtr
end
local ustem, utr = com.make_unstressed_once(stem, tr)
local vals = rsplit(val, "%s*,%s*")
local retvals = {}
for _, val in ipairs(vals) do
local valru, valtr = nom.split_russian_tr(val)
valru = rsub(valru, "~~", ustem)
valru = rsub(valru, "~", com.is_stressed(val) and ustem or stem)
if rfind(valru, "^%*") then
valru = rsub(valru, "^%*", "") .. HYPMARKER
end
if valtr then
tr = tr or nom.translit_no_links(stem)
utr = utr or nom.translit_no_links(ustem)
valtr = rsub(valtr, "~~", utr)
valtr = rsub(valtr, "~", com.is_stressed(val) and utr or tr)
if rfind(valtr, "^%*") then
valtr = rsub(valtr, "^%*", "") .. HYPMARKER
end
end
table.insert(retvals, {valru, valtr})
end
vals = retvals
-- handle + in loc/par meaning "the expected form"; NOTE: This requires
-- that dat_sg has already been processed!
if case == "loc" or case == "par" then
local new_vals = {}
for _, rutr in ipairs(vals) do
-- don't just handle + by itself in case the arg has в or на
-- or whatever attached to it
if rfind(rutr[1], "^%+") or rfind(rutr[1], "[%s%[|]%+") then
for _, dat in ipairs(forms["dat_sg"]) do
local ru, tr = rutr[1], rutr[2]
local datru, dattr = dat[1], dat[2]
local valru, valtr
-- separate off any footnote symbols (which may have been
-- introduced by a *tail or *tailall argument, which we need
-- to process before this stage so overrides don't get
-- automatically marked with footnote symbols); if par,
-- try to preserve them, but with loc this can cause
-- problems in case there are multiple dative forms
-- (stress variants) and the last one is marked with a
-- footnote symbol (occurs in грудь)
local datru_entry, datru_notes = m_table_tools.separate_notes(datru)
local dattr_entry, dattr_notes
if dattr then
dattr_entry, dattr_notes = m_table_tools.separate_notes(dattr)
end
if case == "par" then
valru, valtr = datru_entry, dattr_entry
else
valru, valtr = com.make_ending_stressed(datru_entry, dattr_entry)
datru_notes = ""
dattr_notes = ""
end
-- wrap the word in brackets so it's linked; but not if it
-- appears to already be linked
ru = rsub(ru, "^%+", "[[" .. valru .. "]]")
ru = rsub(ru, "([%[|])%+", "%1" .. valru)
ru = rsub(ru, "(%s)%+", "%1[[" .. valru .. "]]")
ru = ru .. datru_notes
-- do the translit; but it shouldn't have brackets in it
if tr or valtr then
tr = tr or nom.translit_no_links(rutr[1])
valtr = valtr or nom.translit_no_links(valru)
tr = rsub(tr, "^%+", valtr)
tr = rsub(tr, "(%s)%+", "%1" .. valtr)
tr = tr .. dattr_notes
end
table.insert(new_vals, {ru, tr})
end
else
table.insert(new_vals, rutr)
end
end
vals = new_vals
end
-- auto-accent/check for necessary accents
local newvals = {}
for _, v in ipairs(vals) do
local ru, tr = v[1], v[2]
if not args.allow_unaccented then
if tr and com.is_unstressed(ru) ~= com.is_unstressed(tr) then
error("Override " .. ru .. " and translit " .. tr .. " must have same accent pattern")
end
-- it's safe to accent monosyllabic stems
if com.is_monosyllabic(ru) then
ru, tr = com.make_ending_stressed(ru, tr)
elseif com.needs_accents(ru) then
error("Override " .. ru .. " for case " .. case .. n .. " requires an accent")
end
end
table.insert(newvals, {ru, tr})
end
vals = newvals
return vals
end
local function process_overrides(args, f, n)
local function process_override(case)
if args[case .. n] then
local overrides = canonicalize_override(args, case, f, n)
if not f[case] then
f[case] = {}
end
local new_overrides = {}
for _, form in ipairs(overrides) do
-- Don't consider overrides of loc/par/voc irregular since
-- they're only specified through overrides; FIXME: Theoretically
-- we could consider loc/par irregular if they don't follow the
-- expected forms; but we'd have to figure out how to eliminate
-- the preposition that may be specified
if case ~= "loc" and case ~= "par" and case ~= "voc" and
not args.manual and
not contains_rutr_pair(f[case], form) then
local formru, formnotes = m_table_tools.separate_notes(form[1])
if formru ~= "-" then
-- don't mark an override of - as irregular, even if
-- it has an attached footnote symbol
form = nom.concat_paired_russian_tr(form, {IRREGMARKER})
end
end
table.insert(new_overrides, form)
end
f[case] = new_overrides
args.any_overridden[case] = true
end
end
-- do dative singular first because it will be used by loc/par
process_override("dat_sg")
-- now do the rest
for _, case in ipairs(overridable_cases) do
if case ~= "dat_sg" then
process_override(case)
end
end
-- if the nominative is overridden, use it to set the linked version unless
-- that is also specifically overridden
if args["nom_sg" .. n] and not args["nom_sg_linked" .. n] then
f.nom_sg_linked = f.nom_sg
end
if args["nom_pl" .. n] and not args["nom_pl_linked" .. n] then
f.nom_pl_linked = f.nom_pl
end
-- convert empty lists to nil to facilitate computation of accusative
-- case variants below.
for _, case in ipairs(all_cases) do
if f[case] then
if type(f[case]) ~= "table" then
error("Logic error, args[case] should be nil or table")
end
if #f[case] == 0 then
f[case] = nil
end
end
end
end
local function process_tail_args(args, f, n)
local function case_overridden(case)
if args[case .. n] then
return true
end
if case == "acc_sg_an" or case == "acc_sg_in" then
return args["acc_sg" .. n]
end
if case == "acc_pl_an" or case == "acc_pl_in" then
return args["acc_pl" .. n]
end
return false
end
local function append_note(case, value, singlecase, notetype)
value = nom.split_russian_tr(value, "dopair")
local function append1(case)
-- Don't add tail symbol if case overridden
if f[case] and not case_overridden(case) then
-- Clone before side-effecting because some cases may be shared
-- with others (e.g. acc_* with each other and with nom/gen).
f[case] = mw.clone(f[case])
if notetype == "all" then
for i=1,#f[case] do
f[case][i] = nom.concat_paired_russian_tr(f[case][i], value)
end
else
local lastarg = #f[case]
if lastarg > (notetype == ">1" and 1 or 0) then
f[case][lastarg] = nom.concat_paired_russian_tr(f[case][lastarg], value)
end
end
end
end
append1(case)
-- A tail symbol specifically for acc_sg or acc_pl should also affect
-- the animate and inanimate variants, because those are the ones that
-- are displayed. But don't do this if we requested multiple cases at
-- once (e.g. pltail=*), because the list of cases frobbed will
-- already include all accusative variants and we don't want to add
-- the symbol more than once.
if singlecase == "single" then
if case == "acc_sg" then
append1("acc_sg_in")
append1("acc_sg_an")
elseif case == "acc_pl" then
append1("acc_pl_in")
append1("acc_pl_an")
end
end
end
local function handle_tail_hyp(suf)
local arg, val
for _, case in ipairs(all_cases) do
arg = args[case .. "_" .. suf .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "single", "last")
end
arg = args[case .. "_" .. suf .. "all" .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "single", "last")
end
if not rfind(case, "_pl") then
arg = args["sg" .. suf .. "all" .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", "all")
end
arg = args["sg" .. suf .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", ">1")
end
else
arg = args["pl" .. suf .. "all" .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", "all")
end
arg = args["pl" .. suf .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", ">1")
end
end
if not rfind(case, "nom_") and not rfind(case, "acc_") then
arg = args["obl" .. suf .. "all" .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", "all")
end
arg = args["obl" .. suf .. n]
if arg then
append_note(case, suf == "hyp" and HYPMARKER or arg, "multi", ">1")
end
end
end
end
handle_tail_hyp("tail")
handle_tail_hyp("hyp")
end
handle_forms_and_overrides = function(args, n, islast)
local f = args.forms
process_overrides(args, f, n)
local an = args.thisa
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). This logic duplicates logic in handle_overall_forms_and_overrides();
-- see long comment there. We need to duplicate the whole logic here to
-- handle words like мать-одиночка, which has an override of acc_sg1.
if not args["acc_sg_an" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_an = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_an = f.acc_sg or an == "i" and f.nom_sg or f.gen_sg or f.acc_sg_an
end
end
if not args["acc_sg_in" .. n] then
if args["acc_sg" .. n] then
f.acc_sg_in = f.acc_sg
elseif not args.this_any_non_nil.acc_sg then
f.acc_sg_in = f.acc_sg or an == "a" and f.gen_sg or f.nom_sg or f.acc_sg_in
end
end
if not args["acc_pl_an" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_an = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_an = f.acc_pl or an== "i" and f.nom_pl or f.gen_pl or f.acc_pl_an
end
end
if not args["acc_pl_in" .. n] then
if args["acc_pl" .. n] then
f.acc_pl_in = f.acc_pl
elseif not args.this_any_non_nil.acc_pl then
f.acc_pl_in = f.acc_pl or an == "a" and f.gen_pl or f.nom_pl or f.acc_pl_in
end
end
f.loc = f.loc or f.pre_sg
f.par = f.par or f.gen_sg
f.voc = f.voc or f.nom_sg
-- Set these in case we have plural only, in which case the
-- singular will also get set to these same values in case we are
-- a plural-only word in a singular-only expression. NOTE: It's actually
-- unnecessary to do "f.loc_pl = f.loc_pl or f.pre_pl" as f.loc_pl should
-- never previously be set.
f.loc_pl = f.loc_pl or f.pre_pl
f.par_pl = f.par_pl or f.gen_pl
f.voc_pl = f.voc_pl or f.nom_pl
process_tail_args(args, f, n)
local nu = args.thisn
-- If we have a singular-only, set the plural forms to the singular forms,
-- and vice-versa. This is important so that things work in multi-word
-- expressions that combine different number restrictions (e.g.
-- singular-only with singular/plural or singular-only with plural-only,
-- compare "St. Vincent and the Grenadines" [Сент-Винсент и Гренадины]).
if nu == "s" then
f.nom_pl_linked = f.nom_sg_linked
f.nom_pl = f.nom_sg
f.gen_pl = f.gen_sg
f.dat_pl = f.dat_sg
f.acc_pl = f.acc_sg
f.acc_pl_an = f.acc_sg_an
f.acc_pl_in = f.acc_sg_in
f.ins_pl = f.ins_sg
f.pre_pl = f.pre_sg
f.nom_pl = f.nom_sg
-- Unnecessary because only used below to initialize singular
-- loc/par/voc with pluralia tantum.
-- f.loc_pl = f.loc
-- f.par_pl = f.par
-- f.voc_pl = f.voc
elseif nu == "p" then
f.nom_sg_linked = f.nom_pl_linked
f.nom_sg = f.nom_pl
f.gen_sg = f.gen_pl
f.dat_sg = f.dat_pl
f.acc_sg = f.acc_pl
f.acc_sg_an = f.acc_pl_an
f.acc_sg_in = f.acc_pl_in
f.ins_sg = f.ins_pl
f.pre_sg = f.pre_pl
f.nom_sg = f.nom_pl
f.loc = f.loc_pl
f.par = f.par_pl
f.voc = f.voc_pl
end
end
handle_overall_forms_and_overrides = function(args)
local f = {}
for _, case in ipairs(displayable_cases) do
f[case] = concat_word_forms(args.per_word_info, case)
end
process_overrides(args, f, "")
-- if IRREGMARKER is anywhere in text, remove all instances and put
-- at the end before any notes.
local function clean_irreg_marker(case, text)
if rfind(text, IRREGMARKER) then
text = rsub(text, IRREGMARKER, "")
local entry, notes = m_table_tools.separate_notes(text)
if case_will_be_displayed(args, f, case) then
insert_if_not(args.internal_notes, IRREGMARKER .. " Irregular.")
args.any_irreg = true
args.any_irreg_case[case] = true
end
return entry .. IRREGMARKER .. notes
else
return text
end
end
-- set final forms and clean up IRREGMARKER.
for _, case in ipairs(all_cases) do
if f[case] then
local cleaned_forms = {}
for _, form in ipairs(f[case]) do
local ru, tr = form[1], form[2]
ru = clean_irreg_marker(case, ru)
if tr then
tr = clean_irreg_marker(case, tr)
end
table.insert(cleaned_forms, {ru, tr})
end
f[case] = cleaned_forms
end
end
-- Maybe set the value of the animate/inanimate accusative variants based
-- on nom/acc/gen overrides. Don't do this if there was a specific override
-- of this form. Otherwise, always propagate an accusative singular
-- override, and propagate the nom/gen sg if there wasn't a specific
-- accusative suffix anywhere (occurs in the fem sg and sometimes the neut
-- sg). We need to do this somewhat complicated procedure to get overrides
-- to work correctly, e.g. acc_sg overrides of feminine and neuter nouns
-- (such as мать and полслова) and gen_sg/gen_pl overrides of masculine
-- animate nouns. We also ran into an issue with words like мазло that are
-- neuter-form but can be both masculine animate (in which case the
-- acc sg inherits from the genitive) and neuter animate (in which case the
-- acc sg has the fixed ending -о, same as nominative, despite the animacy).
-- Remember also that the animate/inanimate accusative variants are the ones
-- displayed, not the plain acc_sg/acc_pl ones.
if not args.any_overridden.acc_sg_an then
if args.acc_sg then
f.acc_sg_an = f.acc_sg
elseif not args.any_non_nil.acc_sg then
f.acc_sg_an = f.acc_sg or args.a == "i" and f.nom_sg or f.gen_sg or f.acc_sg_an
end
end
if not args.any_overridden.acc_sg_in then
if args.acc_sg then
f.acc_sg_in = f.acc_sg
elseif not args.any_non_nil.acc_sg then
f.acc_sg_in = f.acc_sg or args.a == "a" and f.gen_sg or f.nom_sg or f.acc_sg_in
end
end
if not args.any_overridden.acc_pl_an then
if args.acc_pl then
f.acc_pl_an = f.acc_pl
elseif not args.any_non_nil.acc_pl then
f.acc_pl_an = f.acc_pl or args.a == "i" and f.nom_pl or f.gen_pl or f.acc_pl_an
end
end
if not args.any_overridden.acc_pl_in then
if args.acc_pl then
f.acc_pl_in = f.acc_pl
elseif not args.any_non_nil.acc_pl then
f.acc_pl_in = f.acc_pl or args.a == "a" and f.gen_pl or f.nom_pl or f.acc_pl_in
end
end
-- Try to set the values of acc_sg and acc_pl. The only time we can't is
-- when the noun is bianimate and the anim/inan values are different.
-- This is used primarily for generate_forms() and generate_multi_forms(),
-- since we don't actually display these forms.
if args.a == "a" then
f.acc_sg = f.acc_sg or f.acc_sg_an
f.acc_pl = f.acc_pl or f.acc_pl_an
elseif args.a == "i" then
f.acc_sg = f.acc_sg or f.acc_sg_in
f.acc_pl = f.acc_pl or f.acc_pl_in
else -- bianimate
f.acc_sg = f.acc_sg or ut.equals(f.acc_sg_in, f.acc_sg_an) and f.acc_sg_in or nil
f.acc_pl = f.acc_pl or ut.equals(f.acc_pl_in, f.acc_pl_an) and f.acc_pl_in or nil
end
process_tail_args(args, f, "")
-- set final args[case].
for _, case in ipairs(all_cases) do
args[case] = f[case]
end
end
-- Generate a string to substitute into a particular form in a Wiki-markup
-- table. FORMS is the list of forms, generated by concat_word_forms().
-- OLD is true if we're dealing with a pre-reform declension (in this case,
-- the page we link to has е in place of ё, for reasons I'm not completely
-- sure of). LEMMA is true is we're formatting the entry for use in displaying
-- the lemma in the declension table title. In this case, we don't include
-- the translit, and remove monosyllabic accents from the Cyrillic (but not
-- in multiword expressions).
local function show_form(forms, old, lemma)
local russianvals = {}
local latinvals = {}
local lemmavals = {}
-- Accumulate separately the Russian and transliteration into
-- RUSSIANVALS and LATINVALS, then concatenate each down below.
-- However, if LEMMA, we put each transliteration directly
-- after the corresponding Russian, in parens, and put the results
-- in LEMMAVALS, which get concatenated below. (This is used in the
-- title of the declension table.) (Actually, currently we don't
-- include the translit in the declension table title.)
local is_missing = false
for _, form in ipairs(forms) do
local ru, tr = form[1], form[2]
local ruentry, runotes = m_table_tools.separate_notes(ru)
local trentry, trnotes
if tr then
trentry, trnotes = m_table_tools.separate_notes(tr)
trnotes = rsub(trnotes, HYPMARKER, "")
end
if lemma and com.is_monosyllabic(ruentry) then
ruentry = com.remove_accents(ruentry)
if trentry then
trentry = com.remove_accents(trentry)
end
end
local ishyp = rfind(runotes, HYPMARKER)
if ishyp then
runotes = rsub(runotes, HYPMARKER, "")
end
runotes = m_table_tools.superscript_notes(runotes)
if trnotes then
trnotes = m_table_tools.superscript_notes(trnotes)
end
local ruspan, trspan
if ruentry == "-" and #forms == 1 then
ruspan = "—"
is_missing = true
elseif ishyp then
ruspan = m_links.full_link({lang = lang, term = nil, alt = ruentry, tr = "-"}, "hypothetical")
elseif old then
ruspan = m_links.full_link({lang = lang, term = com.remove_jo(ruentry), alt = not ruentry:find("[[", 1, true) and ruentry, tr = "-"})
else
ruspan = m_links.full_link({lang = lang, term = ruentry, tr = "-"})
end
ruspan = ruspan .. runotes
if not trentry then
trentry = nom.translit_no_links(ruentry)
end
if not trnotes then
trnotes = nom.translit_no_links(runotes)
end
trspan = m_links.remove_links(trentry)
if ishyp then
trspan = scriptutils.tag_text(trspan, lang, Latn, "hypothetical")
end
trspan = scriptutils.tag_translit(trspan .. trnotes, lang, "default", ' style="color: #888"')
if lemma then
-- insert_if_not(lemmavals, ruspan .. " (" .. trspan .. ")")
insert_if_not(lemmavals, ruspan)
else
insert_if_not(russianvals, ruspan)
insert_if_not(latinvals, trspan)
end
end
if lemma then
return table.concat(lemmavals, ", ")
else
local russian_span = table.concat(russianvals, ", ")
if is_missing then
return russian_span
else
local latin_span = table.concat(latinvals, ", ")
return russian_span .. "<br />" .. latin_span
end
end
end
-- Subfunction of concat_word_forms(), used to implement recursively
-- generating all combinations of elements from WORD_FORMS (a list, one
-- element per word, of a list of the forms for a word, each of which is a
-- two-element list of {RUSSIAN, TR}) and TRAILING_FORMS (a list of forms, the
-- accumulated suffixes for trailing words so far in the recursion process,
-- again where each form is a two-element list {RUSSIAN, TR}). Each time we
-- recur we take the last FORMS item off of WORD_FORMS and to each form in
-- FORMS we add all elements in TRAILING_FORMS, passing the newly generated
-- list of items down the next recursion level with the shorter WORD_FORMS.
-- We end up returning a list of concatenated forms, where each list item
-- is a two-element list {RUSSIAN, TR}.
local function concat_word_forms_1(word_forms, trailing_forms)
if #word_forms == 0 then
local retforms = {}
for _, form in ipairs(trailing_forms) do
local ru, tr = form[1], form[2]
-- Remove <insa> and <insb> markers; they've served their purpose.
ru = rsub(ru, "<ins[ab]>", "")
tr = tr and rsub(tr, "<ins[ab]>", "")
table.insert(retforms, {ru, tr})
end
return retforms
else
local last_form_info = table.remove(word_forms)
local last_forms, joiner = last_form_info[1], last_form_info[2]
local new_trailing_forms = {}
for _, form in ipairs(last_forms) do
for _, trailing_form in ipairs(trailing_forms) do
-- If form to prepend is empty, don't add the joiner; this
-- is principally used in overall overrides, where we stuff
-- the entire override into the last word
local full_form = form[1] == "" and trailing_form or
nom.concat_paired_russian_tr(form,
nom.concat_paired_russian_tr(joiner, trailing_form),
"movenotes")
if rfind(full_form[1], "<insa>") and rfind(full_form[1], "<insb>") then
-- REJECT! So we don't get mixtures of the two feminine
-- instrumental singular endings.
else
table.insert(new_trailing_forms, full_form)
end
end
end
return concat_word_forms_1(word_forms, new_trailing_forms)
end
end
-- Generate a list of overall forms by concatenating the per-word forms.
-- PER_WORD_INFO comes from args.per_word_info and is a list of
-- WORD_INFO items, one per word, each of which a two element list of
-- WORD_FORMS (a table listing the forms for each case) and JOINER (a string).
-- We loop over all possible combinations of elements from each word's list
-- of forms for the given case; this requires recursion.
concat_word_forms = function(per_word_info, case)
local word_forms = {}
-- Gather the appropriate word forms. We have to recreate this anew
-- because it will be destructively modified by concat_word_forms_1().
for _, word_info in ipairs(per_word_info) do
table.insert(word_forms, {word_info[1][case], word_info[2]})
end
-- We need to start the recursion with the second parameter containing
-- one blank element rather than no elements, otherwise no elements
-- will be propagated to the next recursion level.
return concat_word_forms_1(word_forms, {{""}})
end
-- Make the table
make_table = function(args)
local data = {}
data.after_title = " " .. args.heading
data.number = args.nonumber and "" or numbers[args.n]
data.lemma = show_form(args[args.n == "p" and "nom_pl_linked" or "nom_sg_linked"], args.old, "lemma")
data.title = args.title or strutils.format(args.old and old_title_temp or title_temp, data)
for _, case in ipairs(displayable_cases) do
data[case] = show_form(args[case], args.old, false)
end
local temp = nil
if args.n == "s" then
data.nom_x = data.nom_sg
data.gen_x = data.gen_sg
data.dat_x = data.dat_sg
data.acc_x_an = data.acc_sg_an
data.acc_x_in = data.acc_sg_in
data.ins_x = data.ins_sg
data.pre_x = data.pre_sg
if data.acc_sg_an == data.acc_sg_in then
temp = "half"
else
temp = "half_a"
end
elseif args.n == "p" then
data.nom_x = data.nom_pl
data.gen_x = data.gen_pl
data.dat_x = data.dat_pl
data.acc_x_an = data.acc_pl_an
data.acc_x_in = data.acc_pl_in
data.ins_x = data.ins_pl
data.pre_x = data.pre_pl
data.par = nil
data.loc = nil
data.voc = nil
if data.acc_pl_an == data.acc_pl_in then
temp = "half"
else
temp = "half_a"
end
else
if data.acc_pl_an == data.acc_pl_in then
temp = "full"
elseif data.acc_sg_an == data.acc_sg_in then
temp = "full_af"
else
temp = "full_a"
end
end
data.par_clause = args.any_overridden.par and strutils.format(partitive, data) or ""
data.loc_clause = args.any_overridden.loc and strutils.format(locative, data) or ""
data.voc_clause = args.any_overridden.voc and strutils.format(vocative, data) or ""
data.notes = args.notes
data.notes_clause = data.notes and strutils.format(notes_template, data) or ""
data.internal_notes = table.concat(args.internal_notes, "<br />")
data.internal_notes_clause = #data.internal_notes > 0 and strutils.format(internal_notes_template, data) or ""
return strutils.format(templates[temp], data)
end
partitive = [===[
! style="background:#eff7ff" | partitive
| {par}
|-]===]
locative = [===[
! style="background:#eff7ff" | locative
| {loc}
|-]===]
vocative = [===[
! style="background:#eff7ff" | vocative
| {voc}
|-]===]
notes_template = [===[
<div style="width:100%;text-align:left;background:#d9ebff">
<div style="display:inline-block;text-align:left;padding-left:1em;padding-right:1em">
{notes}
</div></div>
]===]
internal_notes_template = rsub(notes_template, "notes", "internal_notes")
local function template_prelude(min_width)
min_width = min_width or "70"
return rsub([===[
<div>
<div class="NavFrame" style="display:inline-block; min-width:MINWIDTHem">
<div class="NavHead" style="background:#eff7ff;">{title}<span style="font-weight:normal;">{after_title}</span> </div>
<div class="NavContent">
{\op}| style="background:#F9F9F9; text-align:center; min-width:MINWIDTHem; width:100%;" class="inflection-table"
|-
]===], "MINWIDTH", min_width)
end
local function template_postlude()
return [===[|-{par_clause}{loc_clause}{voc_clause}
|{\cl}{internal_notes_clause}{notes_clause}</div></div></div>]===]
end
templates["full"] = template_prelude("45") .. [===[
! style="width:10em;background:#d9ebff" |
! style="background:#d9ebff" | singular
! style="background:#d9ebff" | plural
|-
! style="background:#eff7ff" | nominative
| {nom_sg}
| {nom_pl}
|-
! style="background:#eff7ff" | genitive
| {gen_sg}
| {gen_pl}
|-
! style="background:#eff7ff" | dative
| {dat_sg}
| {dat_pl}
|-
! style="background:#eff7ff" | accusative
| {acc_sg_an}
| {acc_pl_an}
|-
! style="background:#eff7ff" | instrumental
| {ins_sg}
| {ins_pl}
|-
! style="background:#eff7ff" | prepositional
| {pre_sg}
| {pre_pl}
]===] .. template_postlude()
templates["full_a"] = template_prelude("50") .. [===[
! style="width:15em;background:#d9ebff" |
! style="background:#d9ebff" | singular
! style="background:#d9ebff" | plural
|-
! style="background:#eff7ff" | nominative
| {nom_sg}
| {nom_pl}
|-
! style="background:#eff7ff" | genitive
| {gen_sg}
| {gen_pl}
|-
! style="background:#eff7ff" | dative
| {dat_sg}
| {dat_pl}
|-
! style="background:#eff7ff" rowspan="2" | accusative <span style="padding-left:1em;display:inline-block;vertical-align:middle">animate<br/><br/>inanimate</span>
| {acc_sg_an}
| {acc_pl_an}
|-
| {acc_sg_in}
| {acc_pl_in}
|-
! style="background:#eff7ff" | instrumental
| {ins_sg}
| {ins_pl}
|-
! style="background:#eff7ff" | prepositional
| {pre_sg}
| {pre_pl}
]===] .. template_postlude()
templates["full_af"] = template_prelude("50") .. [===[
! style="width:15em;background:#d9ebff" |
! style="background:#d9ebff" | singular
! style="background:#d9ebff" | plural
|-
! style="background:#eff7ff" | nominative
| {nom_sg}
| {nom_pl}
|-
! style="background:#eff7ff" | genitive
| {gen_sg}
| {gen_pl}
|-
! style="background:#eff7ff" | dative
| {dat_sg}
| {dat_pl}
|-
! style="background:#eff7ff" rowspan="2" | accusative <span style="padding-left:1em;display:inline-block;vertical-align:middle">animate<br/><br/>inanimate</span>
| rowspan="2" | {acc_sg_an}
| {acc_pl_an}
|-
| {acc_pl_in}
|-
! style="background:#eff7ff" | instrumental
| {ins_sg}
| {ins_pl}
|-
! style="background:#eff7ff" | prepositional
| {pre_sg}
| {pre_pl}
]===] .. template_postlude()
templates["half"] = template_prelude("30") .. [===[
! style="width:10em;background:#d9ebff" |
! style="background:#d9ebff" | {number}
|-
! style="background:#eff7ff" | nominative
| {nom_x}
|-
! style="background:#eff7ff" | genitive
| {gen_x}
|-
! style="background:#eff7ff" | dative
| {dat_x}
|-
! style="background:#eff7ff" | accusative
| {acc_x_an}
|-
! style="background:#eff7ff" | instrumental
| {ins_x}
|-
! style="background:#eff7ff" | prepositional
| {pre_x}
]===] .. template_postlude()
templates["half_a"] = template_prelude("35") .. [===[
! style="width:15em;background:#d9ebff" |
! style="background:#d9ebff" | {number}
|-
! style="background:#eff7ff" | nominative
| {nom_x}
|-
! style="background:#eff7ff" | genitive
| {gen_x}
|-
! style="background:#eff7ff" | dative
| {dat_x}
|-
! style="background:#eff7ff" rowspan="2" | accusative <span style="padding-left:1em;display:inline-block;vertical-align:middle">animate<br/><br/>inanimate</span>
| {acc_x_an}
|-
| {acc_x_in}
|-
! style="background:#eff7ff" | instrumental
| {ins_x}
|-
! style="background:#eff7ff" | prepositional
| {pre_x}
]===] .. template_postlude()
return export
-- For Vim, so we get 4-space tabs
-- vim: set ts=4 sw=4 noet:
tkqbtd01uc35r4199pm6kcugpdo71aw
Şablon:ru-ad-cansız-3-D-a
10
1589240
5669927
2026-06-23T08:03:52Z
MustafaCavlak
59368
Yeni sayfa : {{ru-ad-tablo |{{{1}}} |{{#invoke:string|sub|s={{{1}}}|j=-2}}и |{{#invoke:string|sub|s={{{1}}}|j=-2}}и |{{#invoke:string|sub|s={{{1}}}|j=-2}}ей |{{#invoke:string|sub|s={{{1}}}|j=-2}}и |{{#invoke:string|sub|s={{{1}}}|j=-2}}ям |{{{1}}} |{{#invoke:string|sub|s={{{1}}}|j=-2}}и |{{#invoke:string|sub|s={{{1}}}|j=-2}}ью |{{#invoke:string|sub|s={{{1}}}|j=-2}}ями |{{#invoke:string|sub|s={{{1}}}|j=-2}}и |{{#invoke:string|sub|s={{{1}}}|j=-2}}ях }}
5669927
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ей
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ям
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ью
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ями
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ях
}}
k108uqfnzlzv8jwr3sld6msqjkotj0r
5669933
5669927
2026-06-23T08:47:17Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-3da]] sayfasını [[Şablon:ru-ad-cansız-3-D-a]] sayfasına taşıdı
5669927
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ей
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ям
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ью
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ями
|{{#invoke:string|sub|s={{{1}}}|j=-2}}и
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ях
}}
k108uqfnzlzv8jwr3sld6msqjkotj0r
Şablon:ru-ad-cansız-s-D-a
10
1589241
5669929
2026-06-23T08:15:25Z
MustafaCavlak
59368
Yeni sayfa : {{ru-ad-tablo |{{{1}}} |{{#invoke:string|sub|s={{{1}}}|j=-2}}ы |{{#invoke:string|sub|s={{{1}}}|j=-2}}ы |{{#invoke:string|sub|s={{{1}}}|j=-2}} |{{#invoke:string|sub|s={{{1}}}|j=-2}}е |{{{1}}}м |{{#invoke:string|sub|s={{{1}}}|j=-2}}у |{{#invoke:string|sub|s={{{1}}}|j=-2}}ы |[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ой]]<br>[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ою]] |{{{1}}}ми |{{#invoke:string|sub|s={{{1}}}|j=-2}}е |{{{1}}}х }}
5669929
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|{{#invoke:string|sub|s={{{1}}}|j=-2}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}е
|{{{1}}}м
|{{#invoke:string|sub|s={{{1}}}|j=-2}}у
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ой]]<br>[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ою]]
|{{{1}}}ми
|{{#invoke:string|sub|s={{{1}}}|j=-2}}е
|{{{1}}}х
}}
5hl6kl458uj0bwehz2citfmqso1jlq4
5669935
5669929
2026-06-23T08:48:58Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-sda]] sayfasını [[Şablon:ru-ad-cansız-s-D-a]] sayfasına taşıdı
5669929
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|{{#invoke:string|sub|s={{{1}}}|j=-2}}
|{{#invoke:string|sub|s={{{1}}}|j=-2}}е
|{{{1}}}м
|{{#invoke:string|sub|s={{{1}}}|j=-2}}у
|{{#invoke:string|sub|s={{{1}}}|j=-2}}ы
|[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ой]]<br>[[{{#invoke:string|sub|s={{{1}}}|j=-2}}ою]]
|{{{1}}}ми
|{{#invoke:string|sub|s={{{1}}}|j=-2}}е
|{{{1}}}х
}}
5hl6kl458uj0bwehz2citfmqso1jlq4
автомашина
0
1589242
5669930
2026-06-23T08:18:10Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Ad=== {{başlık başı|ru|ad|baş=автомаши́на|c=d}} # [[otomobil]] ====Çekimleme==== {{ru-ad-sda|автомаши́на}}
5669930
wikitext
text/x-wiki
==Rusça==
===Ad===
{{başlık başı|ru|ad|baş=автомаши́на|c=d}}
# [[otomobil]]
====Çekimleme====
{{ru-ad-sda|автомаши́на}}
paqhpa88my8g7i8xwc6swlgrlei3rm2
Şablon:ru-ad-cansız-s-E-a
10
1589243
5669931
2026-06-23T08:24:08Z
MustafaCavlak
59368
Yeni sayfa : {{ru-ad-tablo |{{{1}}} |{{{1}}}ы |{{{1}}}а |{{{1}}}ов |{{{1}}}у |{{{1}}}ам |{{{1}}} |{{{1}}}ы |{{{1}}}ом |{{{1}}}ами |{{{1}}}е |{{{1}}}ах }}
5669931
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{{1}}}ы
|{{{1}}}а
|{{{1}}}ов
|{{{1}}}у
|{{{1}}}ам
|{{{1}}}
|{{{1}}}ы
|{{{1}}}ом
|{{{1}}}ами
|{{{1}}}е
|{{{1}}}ах
}}
shpl2unucsk3t4w7ucthgtf062hch9q
5669937
5669931
2026-06-23T08:49:33Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-sea]] sayfasını [[Şablon:ru-ad-cansız-s-E-a]] sayfasına taşıdı
5669931
wikitext
text/x-wiki
{{ru-ad-tablo
|{{{1}}}
|{{{1}}}ы
|{{{1}}}а
|{{{1}}}ов
|{{{1}}}у
|{{{1}}}ам
|{{{1}}}
|{{{1}}}ы
|{{{1}}}ом
|{{{1}}}ами
|{{{1}}}е
|{{{1}}}ах
}}
shpl2unucsk3t4w7ucthgtf062hch9q
аванс
0
1589244
5669932
2026-06-23T08:28:52Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Köken=== {{nakledilen|ru|fr|avance}}. ===Ad=== {{başlık başı|ru|ad|baş=ава́нс}} # [[avans]] ====Çekimleme==== {{ru-ad-sea|ава́нс}}
5669932
wikitext
text/x-wiki
==Rusça==
===Köken===
{{nakledilen|ru|fr|avance}}.
===Ad===
{{başlık başı|ru|ad|baş=ава́нс}}
# [[avans]]
====Çekimleme====
{{ru-ad-sea|ава́нс}}
cexfnw8wkfh92n2tgdba8ls26bsmltz
Şablon:ru-ad-3da
10
1589245
5669934
2026-06-23T08:47:17Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-3da]] sayfasını [[Şablon:ru-ad-cansız-3-D-a]] sayfasına taşıdı
5669934
wikitext
text/x-wiki
#YÖNLENDİRME [[Şablon:ru-ad-cansız-3-D-a]]
bw9qd33ykz125dyc1ftxrpcff5fa3dt
Şablon:ru-ad-sda
10
1589246
5669936
2026-06-23T08:48:58Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-sda]] sayfasını [[Şablon:ru-ad-cansız-s-D-a]] sayfasına taşıdı
5669936
wikitext
text/x-wiki
#YÖNLENDİRME [[Şablon:ru-ad-cansız-s-D-a]]
130p0ij6rt1duuzee3z60fmyo4y5g09
Şablon:ru-ad-sea
10
1589247
5669938
2026-06-23T08:49:33Z
MustafaCavlak
59368
MustafaCavlak, [[Şablon:ru-ad-sea]] sayfasını [[Şablon:ru-ad-cansız-s-E-a]] sayfasına taşıdı
5669938
wikitext
text/x-wiki
#YÖNLENDİRME [[Şablon:ru-ad-cansız-s-E-a]]
5zqcpj1znxa62iwrv4120jmlx8q83x1
шпинат
0
1589248
5669939
2026-06-23T09:58:08Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Ad=== {{başlık başı|ru|ad|c=e|baş=шпина́т}} # [[ıspanak]] ====Çekimleme==== {{ru-ad-cansız-s-E-a|шпина́т}}
5669939
wikitext
text/x-wiki
==Rusça==
===Ad===
{{başlık başı|ru|ad|c=e|baş=шпина́т}}
# [[ıspanak]]
====Çekimleme====
{{ru-ad-cansız-s-E-a|шпина́т}}
bvs6hsmojpvh5b7i71yf1j9dssuwmxv
лавр
0
1589249
5669940
2026-06-23T10:21:35Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== {{vikipedi|dil=ru}} ===Köken 1=== ====Ad==== {{başlık başı|ru|ad|c=e}} # {{t|dil=ru|ağaçlar}} [[defne]] =====Çekimleme===== {{ru-ad-cansız-s-E-a|ла́вр}} ===Köken 2=== ====Ad==== {{başlık başı|ru|çekimli ad}} # {{çekim|dil=ru|ла́вра||gen|ç}}
5669940
wikitext
text/x-wiki
==Rusça==
{{vikipedi|dil=ru}}
===Köken 1===
====Ad====
{{başlık başı|ru|ad|c=e}}
# {{t|dil=ru|ağaçlar}} [[defne]]
=====Çekimleme=====
{{ru-ad-cansız-s-E-a|ла́вр}}
===Köken 2===
====Ad====
{{başlık başı|ru|çekimli ad}}
# {{çekim|dil=ru|ла́вра||gen|ç}}
rgzyc1dfu8lu7bg5iowz64xugqlrctx
նուռն
0
1589250
5669942
2026-06-23T11:12:51Z
MustafaCavlak
59368
Yeni sayfa : ==Eski Ermenice== [[Dosya:Pomegranate.jpg|küçükresim|220px]] ===Ad=== {{başlık başı|xcl|ad}} # [[nar]]
5669942
wikitext
text/x-wiki
==Eski Ermenice==
[[Dosya:Pomegranate.jpg|küçükresim|220px]]
===Ad===
{{başlık başı|xcl|ad}}
# [[nar]]
foj4pmoe9peg2p9h2pbvw0ywhsw0rei
նռնաքար
0
1589251
5669943
2026-06-23T11:16:37Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Ad=== {{hy-ad}} # [[grena]] ====Çekimleme==== {{hy-ad-ի-եր}}
5669943
wikitext
text/x-wiki
==Ermenice==
===Ad===
{{hy-ad}}
# [[grena]]
====Çekimleme====
{{hy-ad-ի-եր}}
j8o1j49o8qklghn99f9ec81xn3xtaz7
գրեյպֆրուտ
0
1589252
5669944
2026-06-23T11:19:07Z
MustafaCavlak
59368
Yeni sayfa : ==Ermenice== ===Köken=== {{nakledilen|hy|ru|грейпфру́т}}. ===Ad=== {{hy-ad}} # [[greyfurt]] ====Çekimleme==== {{hy-ad-ի-ներ}}
5669944
wikitext
text/x-wiki
==Ermenice==
===Köken===
{{nakledilen|hy|ru|грейпфру́т}}.
===Ad===
{{hy-ad}}
# [[greyfurt]]
====Çekimleme====
{{hy-ad-ի-ներ}}
1fntcs5ae0iypvng1fu4a82oefynwfx
გრეიპფრუტი
0
1589253
5669946
2026-06-23T11:27:20Z
MustafaCavlak
59368
Yeni sayfa : ==Gürcüce== ===Köken=== {{nakledilen|ka|ru|грейпфру́т}}. ===Ad=== {{başlık başı|ka|ad}} # [[greyfurt]] ====Çekimleme==== {{ka-ad-tablo}}
5669946
wikitext
text/x-wiki
==Gürcüce==
===Köken===
{{nakledilen|ka|ru|грейпфру́т}}.
===Ad===
{{başlık başı|ka|ad}}
# [[greyfurt]]
====Çekimleme====
{{ka-ad-tablo}}
l2drnm8c4s7uo0shzautadd7bg7k4bq
Республика Корея
0
1589254
5669947
2026-06-23T11:39:56Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Özel ad=== {{başlık başı|ru|özel ad|c=d|baş=[[республика|Респу́блика]] [[Коре́я]]}} # [[Kore Cumhuriyeti]]
5669947
wikitext
text/x-wiki
==Rusça==
===Özel ad===
{{başlık başı|ru|özel ad|c=d|baş=[[республика|Респу́блика]] [[Коре́я]]}}
# [[Kore Cumhuriyeti]]
icjg3t76dp4928o43dk7ck9mpqjsb1g
Корея
0
1589255
5669948
2026-06-23T11:40:59Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Özel ad=== {{başlık başı|ru|özel ad|c=d|baş=Коре́я}} # [[Kore]]
5669948
wikitext
text/x-wiki
==Rusça==
===Özel ad===
{{başlık başı|ru|özel ad|c=d|baş=Коре́я}}
# [[Kore]]
3uwtl6r1js8diofclyhviw9qdjo5bso
Корейская Народно-Демократическая Республика
0
1589256
5669949
2026-06-23T11:44:53Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Özel ad=== {{başlık başı|ru|özel ad|baş=[[корейский|Коре́йская]] [[народный|Наро́дно]]-[[демократический|Демократи́ческая]] [[республика|Респу́блика]]|c=d}} # [[Kore Demokratik Halk Cumhuriyeti]]
5669949
wikitext
text/x-wiki
==Rusça==
===Özel ad===
{{başlık başı|ru|özel ad|baş=[[корейский|Коре́йская]] [[народный|Наро́дно]]-[[демократический|Демократи́ческая]] [[республика|Респу́блика]]|c=d}}
# [[Kore Demokratik Halk Cumhuriyeti]]
f2n3dv3p742724ygntrzn3gy8ib8nq1
корейский
0
1589257
5669950
2026-06-23T11:46:36Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Köken=== {{ek|dil=ru|Коре́я|ский}} ===Söyleniş=== * {{ses|dil=ru|Ru-корейский.ogg}} ===Ön ad=== {{ru-ön ad|коре́йский}} # [[Kore]] ile ilgili
5669950
wikitext
text/x-wiki
==Rusça==
===Köken===
{{ek|dil=ru|Коре́я|ский}}
===Söyleniş===
* {{ses|dil=ru|Ru-корейский.ogg}}
===Ön ad===
{{ru-ön ad|коре́йский}}
# [[Kore]] ile ilgili
7f6sfbl4v0prgsg3zku5guwdxe5vk25
-ский
0
1589258
5669951
2026-06-23T11:48:45Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Son ek=== {{başlık başı|ru|son ek}} # {{şot|İlişkili ön ad yapım eki.}}
5669951
wikitext
text/x-wiki
==Rusça==
===Son ek===
{{başlık başı|ru|son ek}}
# {{şot|İlişkili ön ad yapım eki.}}
1z6j4x7klzyaiquypoo7xik2g7r0upn
5669952
5669951
2026-06-23T11:49:12Z
MustafaCavlak
59368
/* Son ek */
5669952
wikitext
text/x-wiki
==Rusça==
===Son ek===
{{başlık başı|ru|son ek}}
# {{şot|İlişkili ön ad yapım eki.}}
====Türetilmiş kavramlar====
* {{sonek-t|ru}}
1oin2ezaw6ep5d3d3vy860s68wb28v2
кореец
0
1589259
5669953
2026-06-23T11:53:59Z
MustafaCavlak
59368
Yeni sayfa : {{bak|Кореец}} ==Rusça== ===Köken=== {{ek|dil=ru|Коре́я|ец}} ===Söyleniş=== * {{ses|dil=ru|LL-Q7737 (rus)-DomesticFrog-кореец.wav}} ===Ad=== {{başlık başı|ru|ad|baş=коре́ец|c=e}} # [[Koreli]]
5669953
wikitext
text/x-wiki
{{bak|Кореец}}
==Rusça==
===Köken===
{{ek|dil=ru|Коре́я|ец}}
===Söyleniş===
* {{ses|dil=ru|LL-Q7737 (rus)-DomesticFrog-кореец.wav}}
===Ad===
{{başlık başı|ru|ad|baş=коре́ец|c=e}}
# [[Koreli]]
oe9hhy4veadzpwbyz6rkp7w4ozwrbpi
Kategori:Rusça -ец son ekiyle oluşmuş sözcükler
14
1589260
5669954
2026-06-23T11:54:34Z
MustafaCavlak
59368
Yeni sayfa : {{kategori ek|ru|son|-ец}}
5669954
wikitext
text/x-wiki
{{kategori ek|ru|son|-ец}}
8onevt7vph873z5x89wtzwng5xf3cwe
кореянка
0
1589261
5669955
2026-06-23T11:56:34Z
MustafaCavlak
59368
Yeni sayfa : ==Rusça== ===Ad=== {{başlık başı|ru|ad|baş=корея́нка|c=d}} # {{dişil karşılığı|ru|коре́ец}}
5669955
wikitext
text/x-wiki
==Rusça==
===Ad===
{{başlık başı|ru|ad|baş=корея́нка|c=d}}
# {{dişil karşılığı|ru|коре́ец}}
bd2hxypabsgqg6oxlydrm55oy17mw9l