interscript 0.1.9 → 2.0.5

data/maps/bgnpcgn-bel-Cyrl-Latn-1979.yaml

@@ -1,289 +0,0 @@
----
-authority_id: bgnpcgn
-id: 1979
-language: iso-639-2:bel
-source_script: Cyrl
-destination_script: Latn
-name: United States Board on Geographic Names Foreign Names Committee Staff, 1994. Romanization Systems and Roman-Script Spelling Conventions, p. 23.
-alias:
-  ogc11122:
-    code: bel_Cyrl2Latn_BGN_1979
-    description: Byelorussian US Board on Geographic Names(BGN)/Permanent Committee on Geographical Names for British Official Use(PCGN) 1979 System
-url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/811510/ROMANIZATION_OF_BELARUSIAN.pdf
-creation_date: 1979
-description: |
-  The BGN/PCGN system for Belarusian (formerly referred to as Byelorussian) was designed for use in
-  romanizing names written in the Belarusian Cyrillic alphabet. The Belarusian alphabet contains three
-  characters not present in the Russian alphabet: і, ў, and ’.
-
-notes:
- - The character sequences зг, кг, сг, тс and цг and may be romanized z·h, k·h, s·h, t·s and ts·h in order to differentiate those romanizations from the digraphs zh, kh, sh, ts, and the letter sequence tsh, which are used to render the characters ж, x, ш, ц, and the character sequence тш
- - All apostrophes appearing in romanization are Unicode encoding 2019.
-
-tests:
-  - source: Антон
-    expected: Anton
-  - source: Вілейка
-    expected: Vilyeyka
-  - source: Брэст
-    expected: Brest
-  - source: Дубна
-    expected: Dubna
-  - source: Віцебск
-    expected: Vitsyebsk
-  - source: Асіповічы
-    expected: Asipovichy
-  - source: Гродна
-    expected: Hrodna
-  - source: Брагін
-    expected: Brahin
-  - source: Добруш
-    expected: Dobrush
-  - source: Ліда
-    expected: Lida
-  - source: Гомель
-    expected: Homyel’
-  - source: Беліца
-    expected: Byelitsa
-  - source: Ёдкавічы
-    expected: Yodkavichy
-  - source: Нёман
-    expected: Nyoman
-  - source: Жлобін
-    expected: Zhlobin
-  - source: Ружаны
-    expected: Ruzhany
-  - source: Зоя
-    expected: Zoya
-  - source: князь
-    expected: knyaz’
-  - source: Ігнат
-    expected: Ihnat
-  - source: Мінск
-    expected: Minsk
-  - source: Йосель
-    expected: Yosyel’
-  - source: Койданава
-    expected: Koydanava
-  - source: Крапіўна
-    expected: Krapiwna
-  - source: Менск
-    expected: Myensk
-  - source: Лаўна
-    expected: Lawna
-  - source: Лёсік
-    expected: Lyosik
-  - source: Купала
-    expected: Kupala
-  - source: Вілейка
-    expected: Vilyeyka
-  - source: Міхал
-    expected: Mikhal
-  - source: Вільня
-    expected: Vil’nya
-  - source: Лепель
-    expected: Lyepyel’
-  - source: Магілёў
-    expected: Mahilyow
-  - source: Няміга
-    expected: Nyamiha
-  - source: Наваградак
-    expected: Navahradak
-  - source: Баранавічы
-    expected: Baranavichy
-  - source: Орша
-    expected: Orsha
-  - source: Востраў
-    expected: Vostraw
-  - source: Пінск
-    expected: Pinsk
-  - source: Дняпро
-    expected: Dnyapro
-  - source: Рагачоў
-    expected: Rahachow
-  - source: Сураж
-    expected: Surazh
-  - source: Смаляны
-    expected: Smalyany
-  - source: Арэса
-    expected: Aresa
-  - source: Рось
-    expected: Ros’
-  - source: Талочын
-    expected: Talochyn
-  - source: Масты
-    expected: Masty
-  - source: Уладзімір
-    expected: Uladzimir
-  - source: Бабруйск
-    expected: Babruysk
-  - source: Быхаў
-    expected: Bykhaw
-  - source: Воўпа
-    expected: Vowpa
-  - source: Іўе
-    expected: Iwye
-  - source: Фолюш
-    expected: Folyush
-  - source: фортка
-    expected: fortka
-  - source: Хатынь
-    expected: Khatyn’
-  - source: Быхаў
-    expected: Bykhaw
-  - source: Ганцавічы
-    expected: Hantsavichy
-  - source: Стоўбцы
-    expected: Stowbtsy
-  - source: цьмяны
-    expected: ts’myany
-  - source: мясцовы
-    expected: myastsovy
-  - source: Астравец
-    expected: Astravyets
-  - source: Прыпяць
-    expected: Prypyats’
-  - source: Чэрыкаў
-    expected: Cherykaw
-  - source: Шчара
-    expected: Shchara
-  - source: Нарач
-    expected: Narach
-  - source: Шклоў
-    expected: Shklow
-  - source: Ашмяны
-    expected: Ashmyany
-  - source: Ыттык-Кёль
-    expected: Yttyk-Kyol’
-  - source: Кобрын
-    expected: Kobryn
-  - source: Солы
-    expected: Soly
-  - source: Копысь
-    expected: Kopys’
-  - source: рунь
-    expected: run’
-  - source: Эйсманты
-    expected: Eysmanty
-  - source: Крэва
-    expected: Kreva
-  - source: Юры
-    expected: Yury
-  - source: уюн
-    expected: uyun
-  - source: Язэп
-    expected: Yazep
-  - source: Івянец
-    expected: Ivyanyets
-  - source: з’езд
-    expected: z”yezd
-  - source: Вялiкiя Вераб’евічы
-    expected: Vyalikiya Vyerab”yevichy
-  - source: Дзям’янаўцы
-    expected: Dzyam”yanawtsy
-  - source: Задвор’е
-    expected: Zadvor”ye
-  - source: Гезгалы
-    expected: Hyez·haly
-  - source: Вадасховішча Гезгальскае
-    expected: Vadaskhovishcha Hyez·hal’skaye
-
-map:
-  postrules:
-    - pattern: '\u042C' # Ь
-      result: "\u2019"
-    - pattern: '\u044C' # ь
-      result: "\u2019"
-  # Per documentation those rules are optional
-  rules:
-    - pattern: \u0417\u0413 # ЗГ
-      result: "Z\u00B7H"    # Z·H
-    - pattern: \u0437\u0433 # зг
-      result: "z\u00B7h"    # z·h
-    - pattern: \u041A\u0413 # КГ
-      result: "K\u00B7H"    # K·H
-    - pattern: \u043A\u0433 # кг
-      result: "k\u00B7h"    # k·h
-    - pattern: \u0421\u0413 # СГ
-      result: "S\u00B7H"    # S·H
-    - pattern: \u0441\u0433 # сг
-      result: "s\u00B7h"    # s·h
-    - pattern: \u0422\u0421 # ТС
-      result: "T\u00B7S"    # T·S
-    - pattern: \u0442\u0441 # тс
-      result: "t\u00B7s"    # t·s
-    - pattern: \u0426\u0413 # ЦГ
-      result: "TS\u00B7H"   # TS·H
-    - pattern: \u0446\u0433 # цг
-      result: "ts\u00B7h"   # ts·h
-
-  characters:
-    '\u00B4' : "\u201D" # apostrophe according to spec
-    '\u02BC' : "\u201D" # apostrophe according to spec
-    '\u2019' : "\u201D" # apostrophe in actual examples
-
-    '\u0410' : 'A'  # A
-    '\u0411' : 'B'  # Б
-    '\u0412' : 'V'  # B
-    '\u0413' : 'H'  # Г
-    '\u0414' : 'D'  # Д
-    '\u0415' : 'Ye' # Е
-    '\u0401' : 'Yo' # Ё
-    '\u0416' : 'Zh' # Ж
-    '\u0417' : 'Z'  # З
-    '\u0406' : 'I'  # І
-    '\u0419' : 'Y'  # Й
-    '\u041A' : 'K'  # К
-    '\u041B' : 'L'  # Л
-    '\u041C' : 'M'  # М
-    '\u041D' : 'N'  # Н
-    '\u041E' : 'O'  # О
-    '\u041F' : 'P'  # П
-    '\u0420' : 'R'  # Р
-    '\u0421' : 'S'  # С
-    '\u0422' : 'T'  # Т
-    '\u0423' : 'U'  # У
-    '\U040E' : 'W'  # Ў
-    '\u0424' : 'F'  # Ф
-    '\u0425' : 'Kh' # Х
-    '\u0426' : 'Ts' # Ц
-    '\u0427' : 'Ch' # Ч
-    '\u0428' : 'Sh' # Ш
-    '\u042B' : 'Y'  # Ы
-    '\u042D' : 'E'  # Э
-    '\u042E' : 'Yu' # Ю
-    '\u042F' : 'Ya' # Я
-    '\u0490' : 'G'  # Ґ
-
-    '\u0430' : 'a'  # а
-    '\u0431' : 'b'  # б
-    '\u0432' : 'v'  # в
-    '\u0433' : 'h'  # г
-    '\u0434' : 'd'  # д
-    '\u0435' : 'ye' # е
-    '\u0451' : 'yo' # ё
-    '\u0436' : 'zh'  # ж
-    '\u0437' : 'z'  # з
-    '\u0456' : 'i'  # і
-    '\u0439' : 'y'  # й
-    '\u043A' : 'k'  # к
-    '\u043B' : 'l'  # л
-    '\u043C' : 'm'  # м
-    '\u043D' : 'n'  # н
-    '\u043E' : 'o'  # о
-    '\u043F' : 'p'  # п
-    '\u0440' : 'r'  # р
-    '\u0441' : 's'  # с
-    '\u0442' : 't'  # т
-    '\u0443' : 'u'  # у
-    '\u045E' : 'w'  # ў
-    '\u0444' : 'f'  # ф
-    '\u0445' : 'kh' # х
-    '\u0446' : 'ts' # ц
-    '\u0447' : 'ch' # ч
-    '\u0448' : 'sh' # ш
-    '\u044B' : 'y'  # ы
-    '\u044D' : 'e'  # э
-    '\u044E' : 'yu' # ю
-    '\u044F' : 'ya' # я
-    '\u0491' : 'g'  # ґ

data/maps/bgnpcgn-bul-Cyrl-Latn-1952.yaml

@@ -1,119 +0,0 @@
----
-authority_id: bgnpcgn
-id: 1952
-language: iso-639-2:bul
-source_script: Cyrl
-destination_script: Latn
-name: ROMANIZATION SYSTEM FOR BULGARIAN BGN/PCGN of 1952
-alias:
-  ogc11122:
-    code: bul_Cyrl2Latn_BGN_1952
-    description: Bulgarian US Board on Geographic Names(BGN)/Permanent Committee on Geographical Names for British Official Use(PCGN) Bulgarian 1952 System
-url: https://libraries.ucsd.edu/bib/fed/USBGN_romanization.pdf
-creation_date: 1945
-confirmation_date: 1952
-description: |
-  This system was adopted by the BGN in 1949 and by the PCGN in 1952. It
-  reflects the much simplified Bulgarian orthography as officially revised in
-  February 1945. The Bulgarian alphabet contains all of the characters present
-  in the Russian alphabet with the exception of ё, ы, and э. Notes 1, 2, and 3
-  are applicable to sources predating the orthographic reform of 1945.
-
-notes:
-  - In modern Bulgarian orthography, the character ъ, does not occcur in word-final position. It should be omitted in romanization when found in word-final position on older sources
-  - The obsolete character ѫ, which was replaced by ъ in 1945, should be romanized ŭ
-  - The obsolete character ѣ, replaced in 1945 by е or я according to local pronunciation, should be romanized as e or ya, accordingly, if the pronunciation is known; otherwise, as ye
-  - The character sequence тс may be romanized t·s in order to differentiate that romanization from the regularly occurring digraph ts, which represents the character ц
-
-tests:
-  - source: София
-    expected: Sofiya
-  - source: София-Град
-    expected: Sofiya-Grad
-  - source: България
-    expected: "Bu\u0306lgariya"
-
-map:
-  rules:
-    # note[1]
-    - pattern: (?<=)\u042a(?=\b)
-      result: ""
-    - pattern: (?<=)\u044a(?=\b)
-      result: ""
-    # note[4]
-    - pattern: "\u0422\u0421"
-      result: T·S
-    - pattern: "\u0422\u0441"
-      result: T·s
-    - pattern: "\u0442\u0441"
-      result: t·s
-
-  characters:
-    '\u0410': 'A'
-    '\u0411': 'B'
-    '\u0412': 'V'
-    '\u0413': 'G'
-    '\u0414': 'D'
-    '\u0415': 'E'
-    '\u0416': 'Zh'
-    '\u0417': 'Z'
-    '\u0418': 'I'
-    '\u0419': 'Y'
-    '\u041a': 'K'
-    '\u041b': 'L'
-    '\u041c': 'M'
-    '\u041d': 'N'
-    '\u041e': 'O'
-    '\u041f': 'P'
-    '\u0420': 'R'
-    '\u0421': 'S'
-    '\u0422': 'T'
-    '\u0423': 'U'
-    '\u0424': 'F'
-    '\u0425': 'Kh'
-    '\u0426': 'Ts'
-    '\u0427': 'Ch'
-    '\u0428': 'Sh'
-    '\u0429': 'Sht'
-    '\u042a': "U\u0306"
-    '\u042c': "\\'"
-    '\u042e': 'Yu'
-    '\u042f': 'Ya'
-    '\u0430': 'a'
-    '\u0431': 'b'
-    '\u0432': 'v'
-    '\u0433': 'g'
-    '\u0434': 'd'
-    '\u0435': 'e'
-    '\u0436': 'zh'
-    '\u0437': 'z'
-    '\u0438': 'i'
-    '\u0439': 'y'
-    '\u043a': 'k'
-    '\u043b': 'l'
-    '\u043c': 'm'
-    '\u043d': 'n'
-    '\u043e': 'o'
-    '\u043f': 'p'
-    '\u0440': 'r'
-    '\u0441': 's'
-    '\u0442': 't'
-    '\u0443': 'u'
-    '\u0444': 'f'
-    '\u0445': 'kh'
-    '\u0446': 'ts'
-    '\u0447': 'ch'
-    '\u0448': 'sh'
-    '\u0449': 'sht'
-    '\u044a': "u\u0306"
-    '\u044c': "\\'"
-    '\u044e': 'yu'
-    '\u044f': 'ya'
-
-    # note 2
-    '\u046A': "U\u0306" # Ѫ
-    '\u046B': "u\u0306" # ѫ
-
-    # note[3]
-    '\u0462': "Ye" # Ѣ
-    '\u0463': "ye" # ѣ

data/maps/bgnpcgn-bul-Cyrl-Latn-2013.yaml

@@ -1,42 +0,0 @@
----
-authority_id: bgnpcgn
-id: 2013
-language: iso-639-2:bul
-source_script: Cyrl
-destination_script: Latn
-name: BGN/PCGN 2013 Agreement
-alias:
-  ogc11122:
-    code: bul_Cyrl2Latn_BGN_2013
-    description: Bulgarian US Board on Geographic Names(BGN)/Permanent Committee on Geographical Names(PCGN) Bulgarian 2013 System
-url: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/811509/ROMANIZATION_OF_BULGARIAN.pdf
-creation_date: 2013
-confirmation date: 2019-06
-description: |
-  This system reflects the Bulgarian national system officially adopted
-  by state decree in March 2009. It was adopted by BGN and PCGN in 2013,
-  replacing the BGN/PCGN system of 1952.
-
-notes:
-  - When in final position, “ия” is romanized as “ia” (e.g., София = Sofia; София-Град= Sofia-Grad).
-  - An exception to the romanization system is allowed for the name of the state. Thus, България is romanized as Bulgaria.
-  - The Romanization column shows only lowercase forms but, when romanizing, uppercase and lowercase Roman letters as appropriate should be used.
-
-tests:
-  - source: София
-    expected: Sofia
-  - source: София-Град
-    expected: Sofia-Grad
-  - source: България
-    expected: Bulgaria
-
-map:
-  inherit: bgnpcgn-bul-Cyrl-Latn-1952
-
-  rules:
-    - pattern: България
-      result: Bulgaria
-    - pattern: (?<=\u0418)\u042f(?=\b) # final position, “ИЯ” is romanized as “IA”
-      result: A
-    - pattern: (?<=\u0438)\u044f(?=\b) # final position, “ия” is romanized as “ia”
-      result: a