RubyGems - interscript - Versions diffs - 0.1.1 → 0.1.2 - Mend

interscript 0.1.1 → 0.1.2

Files changed (30) hide show

checksums.yaml +4 -4
data/lib/interscript/version.rb +1 -1
data/maps/bas-rus-Cyrl-Latn-bss.yaml +149 -0
data/maps/bas-rus-Cyrl-Latn-oss.yaml +149 -0
data/maps/bgnpcgn-arm-Armn-Latn-1981.yaml +109 -0
data/maps/bgnpcgn-bul-Cyrl-Latn-2013.yaml +92 -0
data/maps/bgnpcgn-chn-Hans-Latn-pinyin.yaml +7503 -0
data/maps/bgnpcgn-per-Arab-Latn-1956.yaml +93 -0
data/maps/bgnpcgn-rus-Cyrl-Latn-1947.yaml +233 -0
data/maps/bgnpcgn-ukr-Cyrl-Latn-1965.yaml +90 -0
data/maps/cn-chn-Hans-Latn-pinyin.yaml +24760 -0
data/maps/historic-jpn-Hrkt-Latn-hepburn.yaml +336 -0
data/maps/icao-bel-Cyrl-Latn-9303.yaml +125 -0
data/maps/icao-bul-Cyrl-Latn-9303.yaml +123 -0
data/maps/icao-gre-Grek-Latn-9303.yaml +101 -0
data/maps/icao-heb-Hebr-Latn-9303.yaml +157 -0
data/maps/icao-mkd-Cyrl-Latn-9303.yaml +118 -0
data/maps/icao-per-Arab-Latn-9303.yaml +105 -0
data/maps/icao-rus-Cyrl-Latn-9303.yaml +119 -0
data/maps/icao-srp-Cyrl-Latn-9303.yaml +118 -0
data/maps/icao-ukr-Cyrl-Latn-9303.yaml +121 -0
data/maps/iso-rus-Cyrl-Latn-iso9.yaml +273 -0
data/maps/mext-jpn-Hrkt-Latn-hepburn.yaml +330 -0
data/maps/mext-jpn-Hrkt-Latn-kunrei.yaml +308 -0
data/maps/un-jpn-Hrkt-Latn-hepburn.yaml +313 -0
data/maps/un-jpn-Hrkt-Latn-kunrei.yaml +354 -0
data/maps/un-mon-Mong-Latn-2013.yaml +80 -0
data/spec/interscript_spec.rb +11 -0
data/spec/spec_helper.rb +1 -0
metadata +32 -3

data/maps/icao-per-Arab-Latn-9303.yaml ADDED Viewed

@@ -0,0 +1,105 @@
+---
+authority_id: icao
+id: 9303
+language: per
+source_script: Arab
+destination_script: Latn
+name: ICAO Doc 9303 Machine Readable Travel Documents Part 3 -- Persian Arabic to Latin
+url: https://www.icao.int/publications/pages/publication.aspx?docnum=9303
+creation_date: 2015
+description: |
+  Part 3 defines specifications that are common to TD1, TD2 and TD3
+  size machine readable travel documents (MRTDs) including those
+  necessary for global interoperability using visual inspection and
+  machine readable (optical character recognition) means.
+  Since only Latin-alphabet characters are allowed in the VIZ, if
+  mandatory data elements are in a national language that does not use
+  the Latin alphabet, a transcription or transliteration shall also be
+  provided.
+  This document defines the transliteration mappings used to produce
+  this transcription or transliteration.
+tests:
+  - source:
+    expected:
+map:
+  # https://www.icao.int/publications/Documents/9303_p3_cons_en.pdf
+  characters:
+    "\u0027": ""     # '
+    "\u0621": "XE"   # ء
+    "\u0622": "XAA"  # آ
+    "\u0623": 'XAE',  # أ
+    "\u0624": 'U',    # ؤ
+    "\u0625": 'I',    # إ
+    "\u0626": 'XI',   # ئ
+    "\u0627": "A"    # ا
+    "\u0628": "B"    # ب
+    "\u0629": "P"    # ة
+    "\u062A": "T"    # ت
+    "\u062B": "XTH"  # ث
+    "\u062C": "J"    # ج
+    "\u062D": "XH"   # ح
+    "\u062E": "XKH"  # خ
+    "\u062F": "D"    # د
+    "\u0630": "XDH"  # ذ
+    "\u0631": "R"    # ر
+    "\u0632": "Z"    # ز
+    "\u0633": "S"    # س
+    "\u0634": "XSH"  # ش
+    "\u0635": "XSS"  # ص
+    "\u0636": "XDZ"  # ض
+    "\u0637": "XTT"  # ط
+    "\u0638": "XZZ"  # ظ
+    "\u0639": "E"    # ع
+    "\u063A": "G"    # غ
+    "\u0641": "F"    # ف
+    "\u0642": "Q"    # ق
+    "\u0643": "K"    # ك
+    "\u0644": "L"    # ل
+    "\u0645": "M"    # م
+    "\u0646": "N"    # ن
+    "\u0647": "H"    # ه
+    "\u0648": "W"    # و
+    "\u0649": "XAY"  # ى
+    "\u064A": "Y"    # ي
+    "\u064B": "F"    # ً
+    "\u064C": "N"    # ٌ
+    "\u064D": "K"    # ٍ
+    "\u064E": "A"    # َ
+    "\u064F": "U"    # ُ
+    "\u0650": "I"    # ِ
+    "\u0652": "O"    # ْ
+    "\u0670": ""     # '
+    "\u0671": "XXA"  # ٱ
+    "\u0679": "XXT"  # ٹ
+    "\u067C": "XRT"  # ټ
+    "\u067E": "P"    # پ
+    "\u0681": "XKE"  # ځ
+    "\u0685": "XXH"  # څ
+    "\u0686": "XC"   # چ
+    "\u0688": "XXD"  # ڈ
+    "\u0689": "XDR"  # ډ
+    "\u0691": "XXR"  # ڑ
+    "\u0693": "XRR"  # ړ
+    "\u0696": "XRX"  # ږ
+    "\u0698": "XJ"   # ژ
+    "\u069A": "XXS"  # ښ
+    "\u06A9": "XKK"  # ک
+    "\u06AB": "XXK"  # ګ
+    "\u06AD": "XNG"  # ڭ
+    "\u06AF": "XGG"  # گ
+    "\u06BA": "XNN"  # ں
+    "\u06BC": "XXN"  # ڼ
+    "\u06BE": "XDO"  # ھ
+    "\u06C0": "XYH"  # ۀ
+    "\u06C1": "XXG"  # ہ
+    "\u06C2": "XGE"  # ۂ
+    "\u06C3": "XTG"  # ۃ
+    "\u06CC": "XYA"  # ی
+    "\u06CD": "XXY"  # ۍ
+    "\u06D0": "Y"    # ې
+    "\u06D2": "XYB"  # ے
+    "\u06D3": "XBE"  # ۓ

data/maps/icao-rus-Cyrl-Latn-9303.yaml ADDED Viewed

@@ -0,0 +1,119 @@
+---
+authority_id: icao
+id: 9303
+language: rus
+source_script: Cyrl
+destination_script: Latn
+name: ICAO Doc 9303 Machine Readable Travel Documents Part 3 -- Russian Cyrillic to Latin
+url: https://www.icao.int/publications/pages/publication.aspx?docnum=9303
+creation_date: 2015
+description: |
+  Part 3 defines specifications that are common to TD1, TD2 and TD3
+  size machine readable travel documents (MRTDs) including those
+  necessary for global interoperability using visual inspection and
+  machine readable (optical character recognition) means.
+  Since only Latin-alphabet characters are allowed in the VIZ, if
+  mandatory data elements are in a national language that does not use
+  the Latin alphabet, a transcription or transliteration shall also be
+  provided.
+  This document defines the transliteration mappings used to produce
+  this transcription or transliteration.
+tests:
+  - source:
+    expected:
+map:
+  # https://www.icao.int/publications/Documents/9303_p3_cons_en.pdf
+  characters:
+    "\u0027": ""     # '
+    "\u0410": "A"    # А
+    "\u0411": "B"    # Б
+    "\u0414": "D"    # Д
+    "\u0401": "E"    # Ё
+    "\u0415": "E"    # Е
+    "\u042D": "E"    # Э
+    "\u0424": "F"    # Ф
+    "\u0413": "G"    # Г
+    "\u0418": "I"    # И
+    "\u0419": "I"    # 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
+    "\u0412": "V"    # В
+    "\u042B": "Y"    # Ы
+    "\u0417": "Z"    # З
+    "\u0427": "CH"   # Ч
+    "\u042F": "IA"   # Я
+    "\u042E": "IU"   # Ю
+    "\u0425": "KH"   # Х
+    "\u0428": "SH"   # Ш
+    "\u0429": "SHCH" # Щ
+    "\u0426": "TS"   # Ц
+    "\u0416": "ZH"   # Ж
+    "\u0490": "G"    # Ґ
+    "\u040E": "U"    # Ў
+    "\u046A": "U"    # Ѫ
+    "\u0402": "D"    # Ђ
+    "\u0405": "DZ"   # Ѕ
+    "\u0408": "J"    # Ј
+    "\u0409": "LJ"   # Љ
+    "\u040A": "NJ"   # Њ
+    "\u04BA": "C"    # Һ
+    "\u040F": "DZ"   # Џ
+    "\u0404": "IE"   # Є
+    "\u0407": "I"    # Ї
+    "\u0403": "G"    # Ѓ
+    "\u0430": "a"    # а
+    "\u0431": "b"    # б
+    "\u0434": "d"    # д
+    "\u0451": "e"    # ё
+    "\u0435": "e"    # e
+    "\u044D": "e"    # э
+    "\u0444": "f"    # ф
+    "\u0433": "g"    # г
+    "\u0438": "i"    # и
+    "\u0439": "i"    # 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
+    "\u0432": "v"    # в
+    "\u044B": "y"    # ы
+    "\u0437": "z"    # з
+    "\u0447": "ch"   # ч
+    "\u044F": "ia"   # я
+    "\u044E": "iu"   # ю
+    "\u0445": "kh"   # х
+    "\u0448": "sh"   # ш
+    "\u0449": "shch" # щ
+    "\u0446": "ts"   # ц
+    "\u0436": "zh"   # ж
+    "\u0491": "g"    # ґ
+    "\u045E": "u"    # ў
+    "\u046B": "u"    # ѫ
+    "\u0452": "d"    # ђ
+    "\u0455": "dz"   # ѕ
+    "\u0458": "j"    # ј
+    "\u0459": "lj"   # љ
+    "\u045A": "nj"   # њ
+    "\u04BB": "c"    # һ
+    "\u045F": "dz"   # џ
+    "\u0454": "ie"   # є
+    "\u0457": "i"    # ї
+    "\u0453": "g"    # ѓ

data/maps/icao-srp-Cyrl-Latn-9303.yaml ADDED Viewed

@@ -0,0 +1,118 @@
+---
+authority_id: icao
+id: 9303
+language: srp
+source_script: Cyrl
+destination_script: Latn
+name: ICAO Doc 9303 Machine Readable Travel Documents Part 3 -- Serbian Cyrillic to Latin
+url: https://www.icao.int/publications/pages/publication.aspx?docnum=9303
+creation_date: 2015
+description: |
+  Part 3 defines specifications that are common to TD1, TD2 and TD3
+  size machine readable travel documents (MRTDs) including those
+  necessary for global interoperability using visual inspection and
+  machine readable (optical character recognition) means.
+  Since only Latin-alphabet characters are allowed in the VIZ, if
+  mandatory data elements are in a national language that does not use
+  the Latin alphabet, a transcription or transliteration shall also be
+  provided.
+  This document defines the transliteration mappings used to produce
+  this transcription or transliteration.
+tests:
+  - source:
+    expected:
+map:
+  # https://www.icao.int/publications/Documents/9303_p3_cons_en.pdf
+  characters:
+    "\u0027": ""     # '
+    "\u0410": "A"    # А
+    "\u0411": "B"    # Б
+    "\u0414": "D"    # Д
+    "\u0401": "E"    # Ё
+    "\u0415": "E"    # Е
+    "\u042D": "E"    # Э
+    "\u0424": "F"    # Ф
+    "\u0413": "H"    # Г
+    "\u0418": "I"    # И
+    "\u0419": "I"    # 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
+    "\u0412": "V"    # В
+    "\u042B": "Y"    # Ы
+    "\u0417": "Z"    # З
+    "\u0427": "C"    # Ч
+    "\u042F": "IA"   # Я
+    "\u042E": "IU"   # Ю
+    "\u0425": "H"    # Х
+    "\u0428": "S"    # Ш
+    "\u0429": "SHCH" # Щ
+    "\u0426": "C"    # Ц
+    "\u0416": "Z"    # Ж
+    "\u0490": "G"    # Ґ
+    "\u040E": "U"    # Ў
+    "\u046A": "U"    # Ѫ
+    "\u0402": "D"    # Ђ
+    "\u0405": "DZ"   # Ѕ
+    "\u0408": "J"    # Ј
+    "\u0409": "LJ"   # Љ
+    "\u040A": "NJ"   # Њ
+    "\u04BA": "C"    # Һ
+    "\u040F": "DZ"   # Џ
+    "\u0404": "IE"   # Є
+    "\u0407": "I"    # Ї
+    "\u0403": "G"    # Ѓ
+    "\u0430": "a"    # а
+    "\u0431": "b"    # б
+    "\u0434": "d"    # д
+    "\u0451": "e"    # ё
+    "\u0435": "e"    # e
+    "\u044D": "e"    # э
+    "\u0444": "f"    # ф
+    "\u0433": "h"    # г
+    "\u0438": "i"    # и
+    "\u0439": "i"    # 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": ""    # у
+    "\u0432": "v"    # в
+    "\u044B": "y"    # ы
+    "\u0437": "z"    # з
+    "\u0447": "c"    # ч
+    "\u044F": "ia"   # я
+    "\u044E": "i"   # ю
+    "\u0445": "h"    # х
+    "\u0448": "s"    # ш
+    "\u0449": "shch" # щ
+    "\u0446": "c"    # ц
+    "\u0436": "z"    # ж
+    "\u0491": "g"    # ґ
+    "\u045E": ""    # ў
+    "\u046B": ""    # ѫ
+    "\u0452": "d"    # ђ
+    "\u0455": "dz"   # ѕ
+    "\u0458": "j"    # ј
+    "\u0459": "lj"   # љ
+    "\u045A": "nj"   # њ
+    "\u04BB": "c"    # һ
+    "\u045F": "dz"   # џ
+    "\u0454": "ie"   # є
+    "\u0457": "i"    # ї
+    "\u0453": "g"    # ѓ

data/maps/icao-ukr-Cyrl-Latn-9303.yaml ADDED Viewed

@@ -0,0 +1,121 @@
+---
+authority_id: icao
+id: 9303
+language: ukr
+source_script: Cyrl
+destination_script: Latn
+name: ICAO Doc 9303 Machine Readable Travel Documents Part 3 -- Ukranian Cyrillic to Latin
+url: https://www.icao.int/publications/pages/publication.aspx?docnum=9303
+creation_date: 2015
+description: |
+  Part 3 defines specifications that are common to TD1, TD2 and TD3
+  size machine readable travel documents (MRTDs) including those
+  necessary for global interoperability using visual inspection and
+  machine readable (optical character recognition) means.
+  Since only Latin-alphabet characters are allowed in the VIZ, if
+  mandatory data elements are in a national language that does not use
+  the Latin alphabet, a transcription or transliteration shall also be
+  provided.
+  This document defines the transliteration mappings used to produce
+  this transcription or transliteration.
+tests:
+  - source:
+    expected:
+map:
+  # https://www.icao.int/publications/Documents/9303_p3_cons_en.pdf
+  characters:
+    "\u0027": ""     # '
+    "\u0410": "A"    # А
+    "\u0411": "B"    # Б
+    "\u0414": "D"    # Д
+    "\u0401": "E"    # Ё
+    "\u0415": "E"    # Е
+    "\u042D": "E"    # Э
+    "\u0424": "F"    # Ф
+    "\u0413": "G"    # Г
+    "\u0418": "Y"    # И
+    "\u0419": "I"    # 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
+    "\u0412": "V"    # В
+    "\u042B": "Y"    # Ы
+    "\u0417": "Z"    # З
+    "\u0427": "CH"   # Ч
+    "\u042F": "IA"   # Я
+    "\u042E": "IU"   # Ю
+    "\u0425": "KH"   # Х
+    "\u0428": "SH"   # Ш
+    "\u0429": "SHCH" # Щ
+    "\u0426": "TS"   # Ц
+    "\u0416": "ZH"   # Ж
+    "\u0490": "G"    # Ґ
+    "\u040E": "U"    # Ў
+    "\u046A": "U"    # Ѫ
+    "\u0402": "D"    # Ђ
+    "\u0405": "DZ"   # Ѕ
+    "\u0408": "J"    # Ј
+    "\u0409": "LJ"   # Љ
+    "\u040A": "NJ"   # Њ
+    "\u04BA": "C"    # Һ
+    "\u040F": "DZ"   # Џ
+    "\u0404": "IE"   # Є
+    "\u0407": "I"    # Ї
+    "\u0403": "G"    # Ѓ
+    "\u0406": "I"    # І
+    "\u0430": "a"    # а
+    "\u0431": "b"    # б
+    "\u0434": "d"    # д
+    "\u0451": "e"    # ё
+    "\u0435": "e"    # e
+    "\u044D": "e"    # э
+    "\u0444": "f"    # ф
+    "\u0433": "g"    # г
+    "\u0438": "y"    # и
+    "\u0439": "i"    # 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": ""     # у
+    "\u0432": "v"    # в
+    "\u044B": "y"    # ы
+    "\u0437": "z"    # з
+    "\u0447": "ch"   # ч
+    "\u044F": "ia"   # я
+    "\u044E": "i"    # ю
+    "\u0445": "kh"   # х
+    "\u0448": "sh"   # ш
+    "\u0449": "shch" # щщ
+    "\u0446": "ts"   # ц
+    "\u0436": "zh"   # ж
+    "\u0491": "g"    # ґ
+    "\u045E": ""     # ў
+    "\u046B": ""     # ѫ
+    "\u0452": "d"    # ђ
+    "\u0455": "dz"   # ѕ
+    "\u0458": "j"    # ј
+    "\u0459": "lj"   # љ
+    "\u045A": "nj"   # њ
+    "\u04BB": "c"    # һ
+    "\u045F": "dz"   # џ
+    "\u0454": "ie"   # є
+    "\u0457": "i"    # ї
+    "\u0453": "g"    # ѓ