scriptshifter/transliterator/tables/data/armenian.yml

general:
  name: Armenian

roman_to_script:
  ignore:
    - "at head of title"
    - "colophon"
    - "date of publication not identified"
    - "place of publication not identified"
    - "publisher not identified"
    # NOTE There is ambiguity about ignoring these
    # words. Note that the single-character Roman
    # numerals are not included on purpose.
    # Ideally the source editors should use the
    # dedicated U+2160÷U+216F (uppercase Roman
    # numerals) and/or U+2170÷U+217F (lower case Roman
    # numerals) ranges to avoid this ambiguity.
    # TODO implement regular expressions for ignore patterns.
    #- re: "I{2,3}"
    #- re: "I(V|X)"
    #- re: "LI{,3}"
    #- re: "LI?(V|X)"
    #- re: "L(V|X{1,3})I{,3}"
    #- re: "LX{1,3}I?V"
    #- re: "LX{1,3}VI{,3}"
    #- re: "(V|X{1,3})I{,3}"
    #- re: "X{1,3}I{,3}"
    #- re: "X{1,3}I(V|X)"
    #- re: "X{1,3}VI{,3}"
    - "II"
    - "III"
    - "IV"
    - "IX"
    - "LI"
    - "LII"
    - "LIII"
    - "LIV"
    - "LIX"
    - "LV"
    - "LVI"
    - "LVII"
    - "LVIII"
    - "LX"
    - "LXI"
    - "LXII"
    - "LXIII"
    - "LXIV"
    - "LXIX"
    - "LXV"
    - "LXVI"
    - "LXVII"
    - "LXVIII"
    - "LXX"
    - "LXXI"
    - "LXXII"
    - "LXXIII"
    - "LXXIV"
    - "LXXIX"
    - "LXXV"
    - "LXXVI"
    - "LXXVII"
    - "LXXVIII"
    - "LXXX"
    - "LXXXI"
    - "LXXXII"
    - "LXXXIII"
    - "LXXXIV"
    - "LXXXIX"
    - "LXXXV"
    - "LXXXVI"
    - "LXXXVII"
    - "LXXXVIII"
    - "VI"
    - "VII"
    - "VIII"
    - "XI"
    - "XII"
    - "XIII"
    - "XIV"
    - "XIX"
    - "XL"
    - "XLI"
    - "XLII"
    - "XLIII"
    - "XLIV"
    - "XLIX"
    - "XLV"
    - "XLVI"
    - "XLVII"
    - "XLVIII"
    - "XV"
    - "XVI"
    - "XVII"
    - "XVIII"
    - "XX"
    - "XXI"
    - "XXII"
    - "XXIII"
    - "XXIV"
    - "XXIX"
    - "XXV"
    - "XXVI"
    - "XXVII"
    - "XXVIII"
    - "XXX"
    - "XXXI"
    - "XXXII"
    - "XXXIII"
    - "XXXIV"
    - "XXXIX"
    - "XXXV"
    - "XXXVI"
    - "XXXVII"
    - "XXXVIII"
    - "and one other"
    #- re: "and ([a-z0-9]+ )?others"
    - "et al."

  map:
    "A": "\u0531"
    "a": "\u0561"
    "B": "\u0532"
    "b": "\u0562"
    # GH combination
    "GH": "\u0542"
    # Gh combination
    "Gh": "\u0542"
    # gh combination
    "gh": "\u0572"
    "G": "\u0533"
    "g": "\u0563"
    # DZ combination
    "DZ": "\u0541\u0566"
    # Dz combination
    "Dz": "\u0541\u0566"
    # dz combination
    "dz": "\u0571\u0566"
    "D": "\u0534"
    "d": "\u0564"
    # E uppercase with macron
    "E\u0304": "\u0537"
    # e lowercase with macron
    "e\u0304": "\u0567"
    # E uppercase with caron
    "E\u030C": "\u0538"
    # e lowercase with caron
    "e\u030C": "\u0568"
    # EW combination
    "EW": "\u0535\u0582"
    # Ew combination
    "Ew": "\u0535\u0582"
    # ew combination
    "ew": "\u0565\u0582"
    # EV combination
    "EV": "\u0565\u057E"
    # Ev combination
    "Ev": "\u0565\u057E"
    # ev combination
    "ev": "\u0565\u057E"
    "E": "\u0535"
    "e": "\u0565"
    # T uppercase with ayn
    "T\u02BB": "\u0539"
    # t lowercase with ayn
    "t\u02BB": "\u0569"
    # ZH combination
    "ZH": "\u053A"
    # Zh combination
    "Zh": "\u053A"
    # zh combination
    "zh": "\u056A"
    "Z": "\u0536"
    "z": "\u0566"
    "I": "\u053B"
    "i": "\u056B"
    "L": "\u053C"
    "l": "\u056C"
    # KH combination
    "KH": "\u053D"
    # Kh combination
    "Kh": "\u053D"
    # kh combination
    "kh": "\u056D"
    # TS + ayn combination
    "TS\u02BB": "\u0551"
    # Ts + ayn combination
    "Ts\u02BB": "\u0551"
    # ts + ayn combination
    "ts\u02BB": "\u0581"
    # TS combination
    "TS": "\u053E"
    # Ts combination
    "Ts": "\u053E"
    # ts combination
    "ts": "\u056E"
    # K + ayn
    "K\u02BB": "\u0554"
    # k + ayn
    "k\u02BB": "\u0584"
    "K": "\u053F"
    "k": "\u056F"
    # CH + ayn combination 
    "CH\u02BB": "\u0549"
    # Ch + ayn combination
    "Ch\u02BB": "\u0549"
    # ch + ayn combination
    "ch\u02BB": "\u0579"
    # CH combination
    "CH": "\u0543"
    # Ch combination
    "Ch": "\u0543"
    # ch combination
    "ch": "\u0573"
    "M": "\u0544"
    "m": "\u0574"
    "Y": "\u0545"
    "y": "\u0575"
    "N": "\u0546"
    "n": "\u0576"
    # SH combinatiomn
    "SH": "\u0547"
    # Sh combination
    "Sh": "\u0547"
    #sh combination
    "sh": "\u0577"
    "H": "\u0540"
    "h": "\u0570"
    # O uppercase with combining macron
    "O\u0304": "\u0555"
    # o lowercase with combining macron
    "o\u0304": "\u0585"
    "O": "\u0548"
    "o": "\u0578"
    # P uppercase + ayn
    "P\u02BB": "\u0553"
    # p lowercase + ayn
    "p\u02BB": "\u0583"
    "J": "\u054B"
    "j": "\u057B"
    # R uppercase with combining dot below
    "R\u0323": "\u054C"
    # r lowercase with combining dot below
    "r\u0323": "\u057C"
    "S": "\u054D"
    "s": "\u057D"
    "V": "\u054E"
    "v": "\u057E"
    "T": "\u054F"
    "t": "\u057F"
    "R": "\u0550"
    "r": "\u0580"
    "W": "\u0552"
    "w": "\u0582"
    "U": "\u0548\u0582"
    "u": "\u0578\u0582"
    "F": "\u0556"
    "f": "\u0586"
    "\u02B9": ""

script_to_roman:
  map:
    "\u053F\u0540": "K\u02B9H"
    "\u053F\u0570": "K\u02B9h"
    "\u056F\u0570": "k\u02B9h"
    "\u0536\u0540": "Z\u02B9H"
    "\u0536\u0570": "Z\u02B9h"
    "\u0566\u0570": "z\u02B9h"
    "\u054F\u054D": "T\u02B9S"
    "\u054F\u057D": "T\u02B9s"
    "\u057F\u057D": "t\u02B9s"
    "\u0534\u0536": "D\u02B9Z"
    "\u0534\u0566": "D\u02B9z"
    "\u0564\u0566": "d\u02B9z"
    "\u0533\u0540": "G\u02B9H"
    "\u0533\u0570": "G\u02B9h"
    "\u0563\u0570": "g\u02B9h"
    "\u054D\u0540": "S\u02B9H"
    "\u054D\u0570": "S\u02B9h"
    "\u057D\u0570": "s\u02B9h"
    "\u0531": "A"
    "\u0561": "a"
    "\u0532": "B"
    "\u0562": "b"
    # Gh combination
    "\u0542": "Gh"
    # gh combination
    "\u0572": "gh"
    "\u0533": "G"
    "\u0563": "g"
    # Dz combination
    "\u0541\u0566": "Dz"
    # dz combination
    "\u0571\u0566": "dz"
    "\u0534": "D"
    "\u0564": "d"
    # E uppercase with macron
    "\u0537": "E\u0304"
    # e lowercase with macron
    "\u0567": "e\u0304"
    # E uppercase with caron
    "\u0538": "E\u030C"
    # e lowercase with caron
    "\u0568": "e\u030C"
    # Ew combination
    "\u0535\u0582": "Ew"
    # ew combination
    "\u0565\u0582": "ew"
    # Ev combination
    "\u0565\u057E": "Ev"
    # ev combination
    "\u0565\u057E": "ev"
    "\u0535": "E"
    "\u0565": "e"
    # T uppercase with ayn
    "\u0539": "T\u02BB"
    # t lowercase with ayn
    "\u0569": "t\u02BB"
    # Zh combination
    "\u053A": "Zh"
    # zh combination
    "\u056A": "zh"
    "\u0536": "Z"
    "\u0566": "z"
    "\u053B": "I"
    "\u056B": "i"
    "\u053C": "L"
    "\u056C": "l"
    # Kh combination
    "\u053D": "Kh"
    # kh combination
    "\u056D": "kh"
    # Ts + ayn combination
    "\u0551": "Ts\u02BB"
    # ts + ayn combination
    "\u0581": "ts\u02BB"
    # Ts combination
    "\u053E": "Ts"
    # ts combination
    "\u056E": "ts"
    # K + ayn
    "\u0554": "K\u02BB"
    # k + ayn
    "\u0584": "k\u02BB"
    "\u053F": "K"
    "\u056F": "k"
    # Ch + ayn combination
    "\u0549": "Ch\u02BB"
    # ch + ayn combination
    "\u0579": "ch\u02BB"
    # Ch combination
    "\u0543": "Ch"
    # ch combination
    "\u0573": "ch"
    "\u0544": "M"
    "\u0574": "m"
    "\u0545": "Y"
    "\u0575": "y"
    "\u0546": "N"
    "\u0576": "n"
    # Sh combination
    "\u0547": "Sh"
    # sh combination
    "\u0577": "sh"
    "\u0540": "H"
    "\u0570": "h"
    # O uppercase with combining macron
    "\u0555": "O\u0304"
    # o lowercase with combining macron
    "\u0585": "o\u0304"
    "\u0548": "O"
    "\u0578": "o"
    # P uppercase + ayn
    "\u0553": "P\u02BB"
    # p lowercase + ayn
    "\u0583": "p\u02BB"
    "\u054B": "J"
    "\u057B": "j"
    # R uppercase with combining dot below
    "\u054C": "R\u0323"
    # r lowercase with combining dot below
    "\u057C": "r\u0323"
    "\u054D": "S"
    "\u057D": "s"
    "\u054E": "V"
    "\u057E": "v"
    "\u054F": "T"
    "\u057F": "t"
    "\u0550": "R"
    "\u0580": "r"
    "\u0552": "W"
    "\u0582": "w"
    "\u0556": "F"
    "\u0586": "f"