interscript-maps 2.0.5

data/maps/bgnpcgn-kir-Cyrl-Latn-1979.imp

@@ -0,0 +1,170 @@
+metadata {
+  authority_id: bgnpcgn
+  id: 1979
+  language: iso-639-2:kir
+  source_script: Cyrl
+  destination_script: Latn
+  name: Romanization of Kyrgyz
+  url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/816663/ROMANIZATION_OF_KYRGYZ.pdf
+  creation_date: 1979
+  confirmation_date: 2019
+  description: |
+    The BGN/PCGN system for Kyrgyz Cyrillic was designed for use in romanizing names written
+    in the Kyrgyz Cyrillic alphabet. The Kyrgyz Cyrillic alphabet contains three characters not present in
+    the Russian alphabet: Ң , Ө, and Y.
+  
+  notes:
+    - Both Kyrgyz and Kirghiz may frequently be seen as the language name; both these spellings are used in the
+      ISO 639 Standard on the representation of names for languages.
+    - The character sequence н г may be romanized n·g in order to differentiate that romanization
+      from the digraph ng, which is used to render the character ң.
+    - The character ы may be romanized i (Unicode encoding 0069+0335) instead of y, if so desired.
+    - 'An inventory of letter-diacritic combinations, with their Unicode encoding, in addition to the
+      unmodified letters of the basic Roman script is:
+      All apostrophes appearing in romanization are U+2019
+      Ö (U+00D6)  ö (U+00F6)
+      Ü (U+00DC) ü (U+00FC)'
+    - The Romanization column shows only lowercase forms but, when romanizing, uppercase and
+      lowercase Roman letters as appropriate should be used.
+}
+
+tests {
+  test "Ысык-Көл Облусу", "Ysyk-Köl Oblusu"
+  test "Ысык-Көл", "Ysyk-Köl"
+  test "Шедвик-Сай", "Shedvik-Say"
+  test "Чүй Облусу", "Chüy Oblusu"
+  test "Чүй", "Chüy"
+  test "Чирик-Сай", "Chirik-Say"
+  test "Хребет Джети-Сандал", "Khrebet Djeti-Sandal"
+  test "Узук-Булак", "Uzuk-Bulak"
+  test "Торугарт Ашуу", "Torugart Ashuu"
+  test "Торетал", "Toretal"
+  test "Терек", "Terek"
+  test "Талды-Булак", "Taldy-Bulak"
+  test "Талас Облусу", "Talas Oblusu"
+  test "Талас", "Talas"
+  test "Сарык-Кёль", "Saryk-Kyol’"
+  test "Родник Кара-Суу", "Rodnik Kara-Suu"
+  test "Родник Бейрёк-Булак", "Rodnik Beyryok-Bulak"
+  test "Перевал Сары-Челек", "Pereval Sary-Chelek"
+  test "Перевал Макмал", "Pereval Makmal"
+  test "Перевал Кара-Токой", "Pereval Kara-Tokoy"
+  test "Перевал Ашуу-Тёр", "Pereval Ashuu-Tyor"
+  test "Перевал Ашуу", "Pereval Ashuu"
+  test "Ош Шаары", "Osh Shaary"
+  test "Ош Облусу", "Osh Oblusu"
+  test "Ош", "Osh"
+  test "Ош", "Osh"
+  test "Осоавиахим", "Osoaviakhim"
+  test "Озеро Афлатук", "Ozero Aflatuk"
+  test "Нарын Облусу", "Naryn Oblusu"
+  test "Нарын", "Naryn"
+  test "Метеорологическая Станция Чамкал", "Meteorologicheskaya Stantsiya Chamkal"
+  test "Марза-Булак", "Marza-Bulak"
+  test "Макмал", "Makmal"
+  test "Кыргызстан", "Kyrgyzstan"
+  test "Кыргыз Республикасы", "Kyrgyz Respublikasy"
+  test "Куру-Сай", "Kuru-Say"
+  test "Куру-Сай", "Kuru-Say"
+  test "Кур-Пырылды", "Kur-Pyryldy"
+  test "Кок-Бель-Таш", "Kok-Bel’-Tash"
+  test "Кичи-Сандык", "Kichi-Sandyk"
+  test "Кель-Сай", "Kel’-Say"
+  test "Карагайлы", "Karagayly"
+  test "Кара-Суу", "Kara-Suu"
+  test "Жалал-Абад Облусу", "Jalal-Abad Oblusu"
+  test "Жалал-Абад", "Jalal-Abad"
+  test "Долина Беш-Башат", "Dolina Besh-Bashat"
+  test "Гора Арпа-Турча", "Gora Arpa-Turcha"
+  test "Бишкек Шаары", "Bishkek Shaary"
+  test "Бишкек", "Bishkek"
+  test "Бишкек", "Bishkek"
+  test "Баткен Облусу", "Batken Oblusu"
+  test "Баткен", "Batken"
+  test "Аяк-Терек", "Ayak-Terek"
+  test "Аюу-Чача", "Ayuu-Chacha"
+  test "Арпа", "Arpa"
+  test "Ак-Суу", "Ak-Suu"
+}
+
+stage {
+
+  # CHARACTERS
+  parallel {
+    sub "\u0410", "A" # А
+    sub "\u0411", "B" # Б
+    sub "\u0412", "V" # В
+    sub "\u0413", "G" # Г
+    sub "\u0414", "D" # Д
+    sub "\u0415", "E" # Е
+    sub "\u0401", "Yo" # Ё
+    sub "\u0416", "J" # Ж
+    sub "\u0417", "Z" # З
+    sub "\u0418", "I" # И
+    sub "\u0419", "Y" # Й
+    sub "\u041A", "K" # К
+    sub "\u041B", "L" # Л
+    sub "\u041C", "M" # М
+    sub "\u041D", "N" # Н
+    sub "\u04A2", "Ng" # Ң
+    sub "\u041E", "O" # О
+    sub "\u04E8", "Ö" # Ө
+    sub "\u041F", "P" # П
+    sub "\u0420", "R" # Р
+    sub "\u0421", "S" # С
+    sub "\u0422", "T" # Т
+    sub "\u0423", "U" # У
+    sub "\u04AE", "Ü" # Ү
+    sub "\u0424", "F" # Ф
+    sub "\u0425", "Kh" # Х
+    sub "\u0426", "Ts" # Ц
+    sub "\u0427", "Ch" # Ч
+    sub "\u0428", "Sh" # Ш
+    sub "\u0429", "Shch" # Щ
+    sub "\u042A", "”" # Ъ
+    sub "\u042B", "Y" # Ы
+    sub "\u042C", "’" # Ь
+    sub "\u042D", "E" # Э
+    sub "\u042E", "Yu" # Ю
+    sub "\u042F", "Ya" # Я
+    
+    sub "\u0430", "a" # а
+    sub "\u0431", "b" # б
+    sub "\u0432", "v" # в
+    sub "\u0433", "g" # г
+    sub "\u0434", "d" # д
+    sub "\u0435", "e" # e
+    sub "\u0451", "yo" # ё
+    sub "\u0436", "j" # ж
+    sub "\u0437", "z" # з
+    sub "\u0438", "i" # и
+    sub "\u0439", "y" # й
+    sub "\u043A", "k" # к
+    sub "\u043B", "l" # л
+    sub "\u043C", "m" # м
+    sub "\u043D", "n" # н
+    sub "\u04A3", "ng" # ң
+    sub "\u043E", "o" # о
+    sub "\u04E9", "ö" # ө
+    sub "\u043F", "p" # п
+    sub "\u0440", "r" # р
+    sub "\u0441", "s" # с
+    sub "\u0442", "t" # т
+    sub "\u0443", "u" # у
+    sub "\u04AF", "ü" # ү
+    sub "\u0444", "f" # ф
+    sub "\u0445", "kh" # х
+    sub "\u0446", "ts" # ц
+    sub "\u0447", "ch" # ч
+    sub "\u0448", "sh" # ш
+    sub "\u0449", "shch" # щ
+    sub "\u044A", "”" # ъ
+    sub "\u044B", "y" # ы
+    sub "\u044C", "’" # ь
+    sub "\u044D", "e" # э
+    sub "\u044E", "yu" # ю
+    sub "\u044F", "ya" # я
+  }
+
+}
+

data/maps/bgnpcgn-kor-Hang-Latn-kn-1945.imp

@@ -0,0 +1,177 @@
+metadata {
+  authority_id: bgnpcgn
+  id: kn-1945
+  language: iso-639-2:kor
+  source_script: Hang
+  destination_script: Latn
+  name: BGN/PCGN 1945 Agreement
+  url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/693725/ROMANIZATION_OF_KOREAN-_MR_for_DPRK.pdf
+  creation_date: 1945
+  adoption_date:
+  description:
+
+  notes: "
+
+    1. At the end of a syllable, the character ᄋ should be romanized ng,
+       as in the following example:
+
+       평양 → P’yŏngyang
+
+       At the beginning of a syllable, the character ᄋ is silent and
+       should not be romanized. An example follows:
+
+       용화 → Yonghwa
+
+    2. Syllable boundaries within words are not reflected in romanization.
+       In the different types of syllables shown in the table below, C
+       represents any consonant character, V represents any vowel character
+       and / represents a syllable boundary.
+
+         Han’gŭl              개성      남포      안양
+         Syllable boundaries  CV/CVC   CVC/CV   VC/VC
+         Romanization         Kaesŏng  Namp’o   Anyang
+
+    3. Euphonic changes occurring within a word, including between the
+       specific and generic of a geographical name, should be reflected in
+       romanization. Generic terms are usually seen separated from the name
+       by a hyphen and with a lower case initial letter rather than as a
+       separate word:
+
+         영진리 → Yŏngjil-li
+         덕흥리 → Tŏkhŭng-ni
+         압록강 → Amnok-kang
+         대동강 → Taedong-gang
+
+    4. As a result of 조선말규범집 (‘Standard Korean Language’ guidelines
+       published in North Korea in 1966), unlike the Korean spoken in the
+       Republic of Korea, the language spoken in the Democratic People’s
+       Republic of Korea maintains and pronounces the word-initial ᆯ (‘r’).
+       The use of the word-initial ᄅ ('r') can be seen in official news
+       reports as well as native mapping. Since such examples exist, the
+       word initial ᄅ ('r') is reflected as an option in the tables given above.
+
+    5. The Romanization column shows only lowercase forms but, when romanizing,
+       uppercase and lowercase Roman letters as appropriate should be used.
+    "
+}
+
+tests {
+  test "평양", "P’yŏngyang"
+  test "용화", "Yonghwa"
+  test "개성", "Kaesŏng"
+  test "남포", "Namp’o"
+  test "안양", "Anyang"
+  test "영진-리", "Yŏngjil-li"
+  test "덕흥-리", "Tŏkhŭng-ni"
+  test "압록-강", "Amnok-kang"
+  test "대동-강", "Taedong-gang"
+  test "라선특별시", "Rasŏnt’ŭkpyŏlsi"
+  test "은하-리", "Ŭnha-ri"
+  test "은중-리", "Ŭnjung-ni"
+  test "은장-령", "Ŭnjang-nyŏng"
+  test "은혜-동", "Ŭnhye-dong"
+  test "은호-리", "Ŭnho-ri"
+  test "은행정", "Ŭnhaengjŏng"
+  test "은행-동", "Ŭnhaeng-dong"
+  test "은행-촌", "Ŭnhaeng-ch’on"
+  test "원수", "Wŏnsu"
+  test "원소리-고개", "Wŏnsori-gogae"
+  test "원소참", "Wŏnsoch’am"
+  test "원소-리", "Wŏnso-ri"
+  test "원신-리", "Wŏnsil-li"
+  test "난곡", "Nan’gok"
+  test "난산-리", "Nansal-li"
+  test "난직", "Nanjik"
+  test "영곡", "Yŏnggok"
+  test "윗두밀", "Wittumil"
+  test "윗도심이", "Wittosimi"
+  test "둔지", "Tunji"
+  test "서승", "Sŏsŭng"
+  test "신촌", "Sinch’on"
+  test "비암덕", "Piamdŏk"
+  test "바위안", "Pawian"
+  test "오송평", "Osongp’yŏng"
+  test "그물목", "Kŭmulmok"
+  test "구원정", "Kuwŏnjŏng"
+  test "일하", "Irha"
+  test "황우", "Hwangu"
+  test "자작보", "Chajakpo"
+  test "비파1-동", "Pip’a Il-tong"
+  test "문암 오-동", "Munam O-dong"
+}
+
+dependency "var-kor-Hang-Latn-mr-1939", as: hanglatn
+
+
+stage {
+
+  # RULES
+  # Add Zero-width White-space U+200B after spaces (i.e. before word boundaries)
+  # So that the word-initial conversion rules will be blocked.
+  sub line_start, "\u200B"
+  sub "", "\u200B", before: " "
+
+  # convert numbers to space + Hangul
+  sub capture(any([any("0".."9") + " "])), ref( 1 ) + " ", after: any("0".."9")
+  sub "1", " 일"
+  sub "2", " 이"
+  sub "3", " 삼"
+  sub "4", " 사"
+  sub "5", " 오"
+  sub "6", " 육"
+  sub "7", " 칠"
+  sub "8", " 팔"
+  sub "9", " 구"
+
+  # This is a logic to add hyphen in front of generics
+  # - pattern: "(?<=.)(구역|동|리|도|고개|골|로동자구|사무소|초등학교|중학교|고등학교|강|포|령|역|봉|사|천|교|제|저수지|소류지|재|못|말|면|암|교회|촌|병원|바위|공원|섬|우체국|대학교|보건소|굴|치|대교|지구|폭포|해수욕장|휴게소|중고교|읍|보건진료소|마을|톨게이트|대학|시장|경찰서|학교)$" #to be expanded
+  #   result: "-\\1"
+
+
+  run map.hanglatn.stage.main
+
+
+  # POSTRULES
+
+  # Add space to the two ends of the string for easier word boundary handling
+  sub line_start, " "
+  sub line_end, " "
+
+  # Initial rules in the inherited map were blocked, so that
+  # this set of updated rules (with the onset rules removed) will be used instead.
+  sub "\u200B", ""
+
+  sub "ᄀ", "k", before: " " # HANGUL CHOSEONG KIYEOK
+  sub "ᄂ", "n", before: " " # HANGUL CHOSEONG NIEUN
+  sub "ᄃ", "ch", before: " ", after: any("ᅵᅣᅤᅧᅨᅭᅲ") # HANGUL CHOSEONG TIEUT # T -> Ch before yotized vowels
+  sub "ᄃ", "t", before: " " # HANGUL CHOSEONG TIEUT
+  sub "ᄅ", "r", before: " " # HANGUL CHOSEONG RIEUL
+  sub "ᄆ", "m", before: " " # HANGUL CHOSEONG MIEUM
+  sub "ᄇ", "p", before: " " # HANGUL CHOSEONG PIEUP
+  sub "ᄉ", "sh", before: " ", after: "ᅱ" # HANGUL CHOSEONG SIOS
+  sub "ᄉ", "s", before: " " # HANGUL CHOSEONG SIOS
+  sub "ᄋ", "", before: " " # HANGUL CHOSEONG IEUNG
+  sub "ᄌ", "ch", before: " " # HANGUL CHOSEONG CIEUC
+  sub "ᄎ", "ch’", before: " " # HANGUL CHOSEONG CHIEUCH
+  sub "ᄏ", "k’", before: " " # HANGUL CHOSEONG KHIEUKH
+  sub "ᄐ", "ch’", before: " ", after: any("ᅵᅣᅤᅧᅨᅭᅲ") # HANGUL CHOSEONG THIEUTH + YOTIZED VOWELS
+  sub "ᄐ", "t’", before: " " # HANGUL CHOSEONG THIEUTH
+  sub "ᄑ", "p’", before: " " # HANGUL CHOSEONG PHIEUPH
+  sub "ᄒ", "h", before: " " # HANGUL CHOSEONG HIEUH
+  sub "ᄁ", "kk", before: " " # HANGUL CHOSEONG SSANGKIYEOK
+  sub "ᄭ", "kk", before: " " # HANGUL CHOSEONG SIOS-KIYEOK
+  sub "ᄄ", "tt", before: " " # HANGUL CHOSEONG SSANGTIEUT
+  sub "ᄯ", "tt", before: " " # HANGUL CHOSEONG SIOS-TIEUT
+  sub "ᄈ", "pp", before: " " # HANGUL CHOSEONG SSANGPIEUP
+  sub "ᄲ", "pp", before: " " # HANGUL CHOSEONG SIOS-PIEUP
+  sub "ᄊ", "ss", before: " " # HANGUL CHOSEONG SSANGSIOS
+  sub "ᄍ", "tch", before: " " # HANGUL CHOSEONG SSANGCIEUC
+  sub "ᄶ", "tch", before: " " # HANGUL CHOSEONG SIOS-CIEUC
+
+  # Remove space added
+  sub line_start + " ", ""
+  sub " " + line_end, ""
+
+  title_case
+  compose
+}

data/maps/bgnpcgn-kor-Hang-Latn-rok-2011.imp

@@ -0,0 +1,40 @@
+metadata {
+  authority_id: bgnpcgn
+  id: 2011
+  language: iso-639-2:kor
+  source_script: Hang
+  destination_script: Latn
+  name: Ministry of Culture and Tourism System (2000) BGN/PCGN 2011 Agreement
+  url:
+  creation_date:
+  adoption_date:
+  description:
+
+  notes:
+    BGN/PCGN 2011 Agreement
+}
+
+tests {
+  test "불국사", "Bulguksa"
+  test "묵호", "Mukho"
+  test "울산", "Ulsan"
+  test "독립문", "Dongnimmun"
+  test "강남역", "Gangnamyeok"
+  test "남산리", "Namsan-ri" #Note: no assimilation for -ri even after nasals
+  test "내월리", "Naewol-ri"
+  test "울릉군", "Ulleung-gun"
+  test "설악산", "Seoraksan"
+  test "삼죽면", "Samjuk-myeon"
+  test "평리1동", "Pyeongni Il-dong"
+  test "평리2동", "Pyeongni I-dong"
+  test "탑안이", "Tabani"
+}
+
+dependency "moct-kor-Hang-Latn-2000", as: hanglatn
+
+stage {
+
+  run map.hanglatn.stage.main
+
+  title_case
+}

data/maps/bgnpcgn-kor-Kore-Latn-rok-2011.imp

@@ -0,0 +1,41 @@
+metadata {
+  authority_id: bgnpcgn
+  id: 2011
+  language: iso-639-2:kor
+  source_script: Kore
+  destination_script: Latn
+  name: Ministry of Culture and Tourism System (2000) BGN/PCGN 2011 Agreement
+  url: ""
+  creation_date: ""
+  adoption_date: ""
+  description: |
+    Establishes a system for the transliteration of the characters of Korean script into Latin characters.
+    Intended to provide a means for international communication of written documents.
+
+  notes:
+    - BGN/PCGN 2011 Agreement
+}
+
+tests {
+  test "佛國寺", "Bulguksa"
+  test "묵호", "Mukho"
+  test "蔚山", "Ulsan"
+  test "獨立門", "Dongnimmun"
+  test "江南驛", "Gangnamyeok"
+  test "南山里", "Namsan-ri" #Note: no assimilation for -ri even after nasals
+  test "내월里", "Naewol-ri"
+  test "鬱陵郡", "Ulleung-gun"
+  test "雪嶽山", "Seoraksan"
+  test "三竹面", "Samjuk-myeon"
+  test "坪里1洞", "Pyeongni Il-dong"
+  test "坪里2洞", "Pyeongni I-dong"
+  test "탑안이", "Tabani"
+}
+
+dependency "var-kor-Kore-Hang-2013", as: korehang
+dependency "moct-kor-Hang-Latn-2000", as: hanglatn
+
+stage {
+  run map.korehang.stage.main
+  run map.hanglatn.stage.main
+}

data/maps/bgnpcgn-kur-Arab-Latn-2007.imp

@@ -0,0 +1,240 @@
+metadata {
+  authority_id: bgnpcgn
+  id: 2007
+  language: kur
+  source_script: Arab
+  destination_script: Latn
+  name: ROMANIZATION OF KURDISH -- BGN/PCGN 2007
+  url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/693727/ROMANIZATION_OF_KURDISH.pdf
+  creation_date: 2007
+  confirmation_date: 2017-12
+  description: |
+    The tabulation below is applicable to the Kurdish language as a
+    whole. It is based for the most part on the Hawar Roman alphabet used
+    in the Library of Congress Standard Kurdish Orthography Table, but it
+    also incorporates certain non-Hawar elements found in A Kurdish-English
+    Dictionary (Taufiq Wahby & C J Edmonds, OUP, 1966). The tabulation
+    covers both major varieties of the Kurdish language: Kurmanji and
+    Sorani. Kurmanji is spoken principally in Turkey and in Iraq north of
+    the Great Zab River (Dahūk/Dihok Governorate). It is generally written
+    in Roman script, and usually employs the Roman orthography. Sorani is
+    spoken principally in Iraq south of the Great Zab river (Arbīl/Hewlêr
+    and As Sulaymānīyah/Slêmanî governorates). It is generally written in
+    Perso-Arabic script, and usually employs the Perso-Arabic script
+    orthography.
+
+    Kurdish forms of geographical names in Turkey will usually be found
+    in Roman script, and so no romanization process will be required. The
+    digraph options for consonant letters '\u0686', '\u0634', and '\u063A'
+    will not be encountered for such names. In Iraq, Syria, and Iran,
+    Kurdish will usually be encountered in Perso-Arabic script, in which
+    case it should be romanized into the corresponding Roman script form.
+    Kurdish geographical names for places and features outside Turkey,
+    found in Roman script form, should, where necessary and if possible, be
+    tailored to fit the orthography of the Romanization shown below and
+    should employ the digraph options for consonant letters '\u0686',
+    '\u0634', and '\u063A'.
+
+  notes:
+
+    - In pure Kurdish words hamza is borne by yā’ ( ئ ) and occurs only
+      before initial vowels; it is not romanized. Medial and final hamza in
+      Arabic borrowings are romanized by ’ (apostrophe – Unicode encoding
+      2019).
+
+    - The letters ث ذ ص ض ط ظ do not occur in pure Kurdish words. In Arabic
+      borrowings some writers retain these letters, others substitute س ز س ز
+      ت ز respectively. Only the letters ط ض and ص are catered for in the
+      Library of Congress tabulation, as reflected in lines 16-18 of the
+      above Consonant table. Words of obvious Arabic origin occurring in a
+      Kurdish toponymic environment will be treated as Kurdish rather than
+      Arabic, as will words of other non-Kurdish origins.
+
+    - The digraph options appearing in rows 6, 15 and 20 of the consonants
+      table should be used for Kurdish geographical names in Iraq, Iran, and
+      Syria. The single character options should be used for Kurdish
+      geographical names in Turkey.
+
+    - ڨ is used to represent v in foreign words. Some southern Kurdish
+      writers use it to represent the v in borrowings from northern Kurdish
+      dialects. و is pronounced as a v in the north and as a w elsewhere.
+
+    - Hā’ can be used as a vowel or a consonant. The initial (ه) and medial
+      (forms are used for the consonant h, Consonant table, row 31, while the
+      final (ه) and independent (forms are used to represent the vowel e,
+      Vowel table, row 1. Therefore, when used as a consonant, the final and
+      independent forms of hā’ will be seen as ‘ه’ instead of ‘and ‘ه’,
+      respectively. For example, مهه meh, (“month”). When used as ‘e’, the
+      hā’ behaves like the letters alif (ا) , wāw, dāl (د) , and rā (ر) , in
+      that it never joins to the following letter (i.e., it has no medial
+      form). Consequently, the following letter will display the initial
+      form, e.g. هەولێر Hewlêr (unless there is only one following letter, in
+      which case it will be written in the independent form, e.g. ماوەت
+      Mawet). As with other vowels (see special rules 2 and 3), initial e is
+      preceded by the kursî hamza, yielding initial ئه , e.g. ئهني enî
+      “forehead”.
+
+    - In pure Kurdish words, the vowel ى is always long î, e.g. كانى ماسێ
+      Kanî Masê. When it represents îzafe, it is also romanized î and joined
+      by means of a hyphen to its preceding word e.g. پارێزگاى دهۆك Parêzga-î
+      Dihok.
+
+    - |
+      An inventory of letter-diacritic combinations, used in addition to
+      the unmodified letters of the basic Roman script in the Romanization of
+      Kurdish, with their Unicode encoding, is:
+
+        '‘': '\u2018' , '’': '2019'
+        'Ç': '00C7' , 'ç': '00E7'
+        'Ḍ': '1E0C' , 'ḍ': '1E0D'
+        'Ê': '00CA' , 'ê': '00EA'
+
+        # There is no single Unicode encoding for these letter-diacritic combinations.
+        'Ḧ': '0048+0308' , 'ḧ': '0068+0308'
+        'Î': '00CE' , 'î': '00EE'
+        'Ł': '0141' , 'ł': '0142'
+        'Ö': '00D6' , 'ö': '00F6'
+        'Ṟ': '1E5E' , 'ṟ': '1E5F'
+        'Ş': '015E' , 'ş': '015F'
+        'Ṣ': '1E62' , 'ṣ': '1E63'
+        'Ṭ': '1E6C' , 'ṭ': '1E6D'
+        'Û': '00DB' , 'û': '00FB'
+        'Ü': '00DC' , 'ü': '00FC'
+        'Ẍ': '1E8C' , 'ẍ': '1E8D'
+
+    - The Romanization column shows only lowercase forms but, when
+      romanizing, uppercase and lowercase Roman letters as appropriate should
+      be used.
+
+    # Special Rules
+    - The conjunction و (and) should be rendered u if the
+      preceding word ends in a consonant, and w if the preceding
+      word ends in a vowel. It should be separated by spaces from
+      the preceding and following words.
+
+    - In the Perso-Arabic orthography for Kurdish, all vowels are
+      written, with the exception of the short i, which is
+      expressed with a kasrah under the preceding consonant (ِ).
+      In Perso-Arabic script, the kasrah will rarely be written (
+      e.g., كرن kirin “to do”). Like all Kurdish vowels, the
+      short i will be preceded by a kursî hamza )ئ )if it appears
+      at the beginning of a word (see 3 below; see row 4 of vowel
+      table).
+
+    - In the Perso-Arabic orthography for Kurdish, when a vowel
+      comes at the beginning of a word, or when a vowel directly
+      follows another vowel, a kursî hamza )ئ )precedes it (e.g.,
+      ئاگر agir “fire”).
+
+    - A Kurdish word will never start with alif )ا .)A Kurdish
+      word may begin with a yā’ (ي) or wāw )و ,)but only when
+      they are used as a consonant, when they will be romanized
+      as y and w, respectively.
+
+    - When preceded by a consonant, yā’ (ي )and wāw )و )should be
+      romanized î and u, respectively. When preceded by a vowel (
+      including short i, which is not written), yā’ (ي )and wāw (
+      و )should be romanized y and w, respectively.
+
+    - The Arabic sign shaddah ( ّ ) denoting a doubled consonant
+      is not used in Kurdish; doubled consonants, which are rare,
+      are written twice e.g. موحەممەد Muḧemmed; ننا موسه Musanna.
+      Shaddah might be used in Arabic borrowings but, as in
+      unpointed Arabic, would generally be omitted.
+
+    - Particles such as له le (= at, in, on) and به be (= to,
+      for, by, with) should be written separately from their
+      following word, e.g. كوردستانێ له Le Kurdistanê “in
+      Kurdistan”
+
+    - Occasionally the character sequences چه ,سه and گه occur.
+      They may be romanized c·h, s·h, and g·h in order to
+      differentiate those romanizations from the digraphs ch, sh,
+      and gh.
+}
+
+tests {
+  test "كاني ماسێ", "Kanî Masê"
+  test "كِرِن", "Kirin"
+  test "ئاگِر", "Agir"
+  test "موحەممەد", "Muḧemmed"
+  # - source: موسەننا # issue 604
+  #   expected: Musanna
+  test "لەكوردِستانێ", "Le Kurdistanê"
+}
+
+stage {
+
+  # CHARACTERS
+  parallel {
+
+    sub "\u0650", "i" # ِ kasra special rule 2
+    sub "\u0644\u06d5", "le " # special rule 7
+    sub "\u0628\u06d5", "be " # special rule 7
+    # Note 1
+    sub "\u0621", "’" # ء
+    sub "\u0624", "’" # ؤ
+    sub "\u0626", "’" # ئ
+
+    sub "و", "u", before: any("ء|ب|پ|ت|ج|چ|ح|خ|د|ر|ڕ|ز|ژ|س|ش|ص|ض|ط|ع|غ|́|ڨ|ق|ک|ك|گ|ل|ڵ|م|ن|و|ه|ي") # special note 4/5
+    sub "ي", "î", before: any("ء|ب|پ|ت|ج|چ|ح|خ|د|ر|ڕ|ز|ژ|س|ش|ص|ض|ط|ع|غ|́|ڨ|ق|ک|ك|گ|ل|ڵ|م|ن|و|ه|ي") # special note 4/5
+    sub "\u0621", "’" # ء (see note 1 and 7)
+    sub "\u0628", "b" # ب
+    sub "\u067E", "p" # پ
+    sub "\u062A", "t" # ت (see note 2)
+    sub "\u062C", "c" # ج
+    sub "\u0686", any(["ch", "ç"]) # چ (see notes 3 and 7)
+    sub "\u062D", "ḧ" # ح
+    sub "\u062E", "x" # خ
+    sub "\u062F", "d" # د
+    sub "\u0631", "r" # ر
+    sub "\u0695", "ṟ" # ڕ (Formerly written ڒ ڔ or رر according to typeface available; may vary on older sources. See note 7.)
+    sub "\u0632", "z" # ز (see note 2)
+    sub "\u0698", "j" # ژ
+    sub "\u0633", "s" # س (see note 2)
+    sub "\u0634", any(["sh", "ş"]) # ش (see notes 3 and 7)
+    sub "\u0635", "ṣ" # ص (see notes 2 and 7)
+    sub "\u0636", "ḍ" # ض (see notes 2 and 7)
+    sub "\u0637", "ṭ" # ط  (see notes 2 and 7)
+    sub "\u0639", "‘" # ع  (see note 7)
+    sub "\u063A", any(["gh", "ẍ"]) # غ (see notes 3 and 7)
+    sub "\u0341", "f" # ف
+    sub "\u06A8", "v" # ڨ  (see note 4)
+    sub "\u0642", "q" # ق
+    sub "\u06A9", "k" # ك
+    sub "\u0643", "k" # ك
+    sub "\u06AF", "g" # گ
+    sub "\u0644", "l" # ل
+    sub "\u06B5", "ł" # ڵ  (Formerly written ڶ according to type available; may vary on older sources. See note 7)
+    sub "\u0645", "m" # م
+    sub "\u0646", "n" # ن
+    sub "\u0648", "w" # و  (see note 4)
+    sub "\u0647", "h" # ه  (see note 5)
+    sub "\u064A", "y" # ي
+
+    # VOWELS
+    sub "\u0647" + boundary, "e" # See notes 1 and 5
+    sub "\u06D5", "e" # See notes 1 and 5
+    sub "\u0626\u06D5", "e" # See notes 1 and 5
+    sub "\u0627", "a" # See note 1
+    sub "\u0626\u0627", "a" # See note 1
+    sub "\u064A", "î" # See notes 1, 6 and 7
+    sub "\u0626\u064A", "î" # See notes 1, 6 and 7
+    sub "\u0626", "i"
+    sub "\u06CE", "ê" # See note 7
+    sub "\u0626\u06CE", "ê" # See note 7
+    sub "\u0648", "u"
+    sub "\u0626\u0648", "u"
+    sub "\u0648\u0648", "û" # See note 7
+    sub "\u0626\u0648\u0648", "û" # See note 7
+    sub "\u06C6", "o"
+    sub "\u0626\u06C6", "o"
+    sub "\u0648", "ö" # Rare; previously written وي . See note 7
+    sub "\u06CA", "ü" # Only appearing in some dialects and only in old sources. Often equated to /û/ (row 7 above). Sometimes written يو See note 7.
+  }
+
+  # POSTRULES
+  sub any("\u0061".."\uFFFF"), upcase, before: boundary, not_before: boundary + any("‘’'")
+
+
+}