# Copyright (C) 2009 Google, Inc. All Rights Reserved. # Author: mjansche@google.com (Martin Jansche) # # Japanese (Rōmaji) to Russian (Cyrillic) Polivanov transliteration for ICU. # Can be run in sequence after e.g. Katakana-Latin. # # These rules cannot be used to target Bulgarian, Serbian, Tajik, or Ukrainian. # # TODO: Cyrillization needs to respect morpheme/Kanji boundaries. # 中井 becomes Накаи, but 北海道 becomes Хоккайдо. We need boundary # markup in the input in order to do that properly. # ::NFD(NFC); ::[:Latin:] Lower(); # $lengthMarker = [̂̄]; # # Delete apostrophes. Apostrophes after "n" are consumed below. \' → ; # # Turn long /e:/ into diphthong /ei/. # Note that /ei/ across a morpheme boundary (e.g. 武井 Takei) becomes эи. e $lengthMarker → эй ; # # Turn long /i:/ into two vowels /ii/. i $lengthMarker → | i i ; # # Ignore vowel length everywhere else. $lengthMarker → ; # # Vowels. # # TODO(mjansche): Enable diphthongs once we have Kanji boundaries. ## ai → ай ; a → а ; i\~e → | ye ; i → и ; u\~ → в ; # ウィ etc. ## ui → уй ; u → у ; e → э ; o → о ; # # Consonants. # k → к ; # sh → | sy ; s → с ; # ch → | ty ; c } ch → t ; te\~ → | t ; # テュ to\~ → | t ; # トゥ tsu\~ → | ts ; # ツァ, ツィ, etc. ts → ц ; t → т ; # \~tsu → | tsu ; # n } [bpm] → м ; # 群馬 → Гумма n\' → нъ ; n → н ; # h → х ; fu\~ → | f ; # フュ f → ф ; # m → м ; # ya → я ; yi → и ; # Added for convenience, after sh, ch, j. yu → ю ; ye → е ; # ?? unobserved yo → ё ; # r → р ; # wa → ва ; w → ; # g → г ; # j → | zy ; z → дз ; # de\~ → | d ; # デュ dji\~ → | z ; # ヂャ, ヂュ, etc. dj → | j ; # ヂ do\~ → | d ; # ドゥ dzu\~ → | z ; # ヅァ, ヅィ, etc. dz → | z ; # ヅ d → д ; # b → б ; vu\~ → | v ; # ヴァ, etc. v → в ; # ?? unobserved # p → п ; # ::NFC(NFD);