# FOMA script for crh-latn and crh-cyrl conversions
# Generates .att files when run from this directory using:
#   foma -f crh.foma
# Then generate executable .pfst files with:
#   ../tools/build-langconv-fst.js -l crh-latn crh-cyrl
#   ../tools/build-langconv-fst.js -l crh-cyrl crh-latn
echo EXPECT 0

source brackets.foma
source roman.foma
source safety.foma

define WORD'BEGINS [
  "\r" | " " | "\t" | "\n" | "\'" | "(" | ")" | "-" | "<" | ">" |
  "[" | "]" | "/" | "." | "," | ":" | ";" | "!" | "?"
  | {"} # " workaround bad quote-matching
];
define WORD'ENDS WORD'BEGINS;

######## Cyrillic #############

# Crimean Tatar Cyrillic uppercase/lowercase and mappings between
define C'UC'LC [
  {А}:{а} | {Б}:{б} | {В}:{в} | {Г}:{г} | {Д}:{д} | {Е}:{е} | {Ё}:{ё} |
  {Ж}:{ж} | {З}:{з} | {И}:{и} | {Й}:{й} | {К}:{к} | {Л}:{л} | {М}:{м} |
  {Н}:{н} | {О}:{о} | {П}:{п} | {Р}:{р} | {С}:{с} | {Т}:{т} | {У}:{у} |
  {Ф}:{ф} | {Х}:{х} | {Ц}:{ц} | {Ч}:{ч} | {Ш}:{ш} | {Щ}:{щ} | {Ъ}:{ъ} |
  {Ы}:{ы} | {Ь}:{ь} | {Э}:{э} | {Ю}:{ю} | {Я}:{я}
];
define C'LC'UC C'UC'LC.i;
define C'UC    C'UC'LC.u;
define C'LC    C'UC'LC.l;

# Crimean Tatar Cyrillic + CÑ uppercase consonants
define C'CONS'UC [
  {Б}|{В}|{Г}|{Д}|{Ж}|{З}|{Й}|{К}|{Л}|{М}|{Н}|{П}|{Р}|{С}|{Т}|{Ф}|{Х}|{Ц}|
  {Ч}|{Ш}|{Щ}|{C}|{Ñ}
];
# Crimean Tatar Cyrillic + CÑ lowercase consonants
define C'CONS'LC [
  {б}|{в}|{г}|{д}|{ж}|{з}|{й}|{к}|{л}|{м}|{н}|{п}|{р}|{с}|{т}|{ф}|{х}|{ц}|
  {ч}|{ш}|{щ}|{c}|{ñ}
];
# Crimean Tatar Cyrillic M-type consonants
define C'M'CONS [{б}|{г}|{к}|{м}|{ш}|{c}|{Б}|{Г}|{К}|{М}|{Ш}|{C}];

######## Crimean Tatar Cyrillic #############

# Crimean Tatar Cyrillic + CÑ consonants
define C'CONS [
  {б}|{в}|{г}|{д}|{ж}|{з}|{й}|{к}|{л}|{м}|{н}|{п}|{р}|{с}|{т}|{ф}|{х}|{ц}|
  {ч}|{ш}|{щ}|{c}|{ñ}|{Б}|{В}|{Г}|{Д}|{Ж}|{З}|{Й}|{К}|{Л}|{М}|{Н}|{П}|{Р}|
  {С}|{Т}|{Ф}|{Х}|{Ц}|{Ч}|{Ш}|{Щ}|{C}|{Ñ}
];

######## Latin #############

# Crimean Tatar Latin uppercase/lowercase and mappings between
define L'UC'LC [
  {A}:{a} | {Â}:{â} | {B}:{b} | {C}:{c} | {Ç}:{ç} | {D}:{d} | {E}:{e}|
  {F}:{f} | {G}:{g} | {Ğ}:{ğ} | {H}:{h} | {I}:{ı} | {İ}:{i} | {J}:{j}|
  {K}:{k} | {L}:{l} | {M}:{m} | {N}:{n} | {Ñ}:{ñ} | {O}:{o} | {Ö}:{ö}|
  {P}:{p} | {Q}:{q} | {R}:{r} | {S}:{s} | {Ş}:{ş} | {T}:{t} | {U}:{u}|
  {Ü}:{ü} | {V}:{v} | {Y}:{y} | {Z}:{z}
];
define L'LC'UC L'UC'LC.i;
define L'UC    L'UC'LC.u;
define L'LC    L'UC'LC.l;

# Crimean Tatar Latin N-type upper case consonants
define L'N'CONS'UC [{Ç}|{N}|{R}|{S}|{T}|{Z}];
# Crimean Tatar Latin N-type lower case consonants
define L'N'CONS'LC [L'LC'UC .o. L'N'CONS'UC].u;
# Crimean Tatar Latin N-type consonants
define L'N'CONS [{ç}|{n}|{r}|{s}|{t}|{z}|{Ç}|{N}|{R}|{S}|{T}|{Z}];
# Crimean Tatar Latin M-type consonants
define L'M'CONS [{b}|{c}|{g}|{k}|{m}|{p}|{ş}|{B}|{C}|{G}|{K}|{M}|{P}|{Ş}];
# Crimean Tatar Latin uppercase consonants
define L'CONS'UC [
  {B}|{C}|{Ç}|{D}|{F}|{G}|{Ğ}|{H}|{J}|{K}|{L}|{M}|{N}|{Ñ}|
  {P}|{Q}|{R}|{S}|{Ş}|{T}|{V}|{Z}
];
# Crimean Tatar Latin lowercase consonants
define L'CONS'LC [L'LC'UC .o. L'CONS'UC].u;

# Crimean Tatar Latin consonants
define L'CONS [ L'CONS'LC | L'CONS'UC ];

# Crimean Tatar Latin uppercase vowels
define L'VOW'UC [{A}|{Â}|{E}|{I}|{İ}|{O}|{Ö}|{U}|{Ü}];
# Crimean Tatar Latin vowels
define L'VOW L'VOW'UC | [L'LC'UC .o. L'VOW'UC].u;
# Crimean Tatar Latin uppercase front vowels
define L'F'UC [{E}|{İ}|{Ö}|{Ü}];
# Crimean Tatar Latin front vowels
define L'F L'F'UC | [L'LC'UC .o. L'F'UC].u;


# Basic conversions, based on PHP source
define MAP'CYRL'TO'LATN [
  ## these are independent of location in the word, but have
  ## to go first so other transforms don't bleed them
  {гъ} @-> {ğ} ,, {Гъ} @-> {Ğ} ,, {ГЪ} @-> {Ğ} ,,
  {къ} @-> {q} ,, {Къ} @-> {Q} ,, {КЪ} @-> {Q} ,,
  {нъ} @-> {ñ} ,, {Нъ} @-> {Ñ} ,, {НЪ} @-> {Ñ} ,,
  {дж} @-> {c} ,, {Дж} @-> {C} ,, {ДЖ} @-> {C} ,,

  {А} @-> {A} ,, {а} @-> {a} ,, {Б} @-> {B} ,, {б} @-> {b} ,,
  {В} @-> {V} ,, {в} @-> {v} ,, {Г} @-> {G} ,, {г} @-> {g} ,,
  {Д} @-> {D} ,, {д} @-> {d} ,, {Ж} @-> {J} ,, {ж} @-> {j} ,,
  {З} @-> {Z} ,, {з} @-> {z} ,, {И} @-> {İ} ,, {и} @-> {i} ,,
  {Й} @-> {Y} ,, {й} @-> {y} ,, {К} @-> {K} ,, {к} @-> {k} ,,
  {Л} @-> {L} ,, {л} @-> {l} ,, {М} @-> {M} ,, {м} @-> {m} ,,
  {Н} @-> {N} ,, {н} @-> {n} ,, {П} @-> {P} ,, {п} @-> {p} ,,
  {Р} @-> {R} ,, {р} @-> {r} ,, {С} @-> {S} ,, {с} @-> {s} ,,
  {Т} @-> {T} ,, {т} @-> {t} ,, {Ф} @-> {F} ,, {ф} @-> {f} ,,
  {Х} @-> {H} ,, {х} @-> {h} ,, {Ч} @-> {Ç} ,, {ч} @-> {ç} ,,
  {Ш} @-> {Ş} ,, {ш} @-> {ş} ,, {Ы} @-> {I} ,, {ы} @-> {ı} ,,
  {Э} @-> {E} ,, {э} @-> {e} ,, {Е} @-> {E} ,, {е} @-> {e} ,,
  {Я} @-> {Â} ,, {я} @-> {â} ,, {У} @-> {U} ,, {у} @-> {u} ,,
  {О} @-> {O} ,, {о} @-> {o} ,,

  {Ё} @-> {Yo} ,, {ё} @-> {yo} ,, {Ю} @-> {Yu} ,, {ю} @-> {yu} ,,
  {Ц} @-> {Ts} ,, {ц} @-> {ts} ,, {Щ} @-> {Şç} ,, {щ} @-> {şç} ,,
  {Ь} @-> 0 ,, {ь} @-> 0 ,, {Ъ} @-> 0 ,, {ъ} @-> 0
];

define MAP'LATN'TO'CYRL [
  {Â} @-> {Я} ,, {â} @-> {я} ,, {B} @-> {Б} ,, {b} @-> {б} ,,
  {Ç} @-> {Ч} ,, {ç} @-> {ч} ,, {D} @-> {Д} ,, {d} @-> {д} ,,
  {F} @-> {Ф} ,, {f} @-> {ф} ,, {G} @-> {Г} ,, {g} @-> {г} ,,
  {H} @-> {Х} ,, {h} @-> {х} ,, {I} @-> {Ы} ,, {ı} @-> {ы} ,,
  {İ} @-> {И} ,, {i} @-> {и} ,, {J} @-> {Ж} ,, {j} @-> {ж} ,,
  {K} @-> {К} ,, {k} @-> {к} ,, {L} @-> {Л} ,, {l} @-> {л} ,,
  {M} @-> {М} ,, {m} @-> {м} ,, {N} @-> {Н} ,, {n} @-> {н} ,,
  {O} @-> {О} ,, {o} @-> {о} ,, {P} @-> {П} ,, {p} @-> {п} ,,
  {R} @-> {Р} ,, {r} @-> {р} ,, {S} @-> {С} ,, {s} @-> {с} ,,
  {Ş} @-> {Ш} ,, {ş} @-> {ш} ,, {T} @-> {Т} ,, {t} @-> {т} ,,
  {V} @-> {В} ,, {v} @-> {в} ,, {Z} @-> {З} ,, {z} @-> {з} ,,

  {ya} @-> {я} ,, {Ya} @-> {Я} ,, {YA} @-> {Я} ,,
  {ye} @-> {е} ,, {YE} @-> {Е} ,, {Ye} @-> {Е} ,,

  # hack, hack, hack
  {A} @-> {А} ,, {a} @-> {а} ,, {E} @-> {Е} ,, {e} @-> {е} ,,
  {Ö} @-> {Ё} ,, {ö} @-> {ё} ,, {U} @-> {У} ,, {u} @-> {у} ,,
  {Ü} @-> {Ю} ,, {ü} @-> {ю} ,, {Y} @-> {Й} ,, {y} @-> {й} ,,
  {C} @-> {Дж} ,, {c} @-> {дж} ,, {Ğ} @-> {Гъ} ,, {ğ} @-> {гъ} ,,
  {Ñ} @-> {Нъ} ,, {ñ} @-> {нъ} ,, {Q} @-> {Къ} ,, {q} @-> {къ}
];

define QUOTE'LATN'TO'CYRL {“} @-> {«} ,, {”} @-> {»};
define QUOTE'CYRL'TO'LATN {«} @-> {"} ,, {»} @-> {"};

# Now we're in LanguageCrh::translate
define LATN'LETTERS L'UC | L'LC | {'};
define CYRL'LETTERS C'UC | C'LC ;
define PHP'WORD'CHARS LATN'LETTERS | CYRL'LETTERS |
  {_}|{W}|{X}|{w}|{x}|{0}|{1}|{2}|{3}|{4}|{5}|{6}|{7}|{8}|{9};
# Add a few characters which are treated specially by the conversion
define ALL'WORD'CHARS PHP'WORD'CHARS | {-} | {#} | {.};
# In PHP $breaks is [^\w\x80-\xff]
define PHP'BREAK \[PHP'WORD'CHARS | BRACKETS ];
define BREAK \[ALL'WORD'CHARS | BRACKETS ];

## Split input into words
#define BYWORDS(word,break) (word) [break (word)]*;
## ensure that `t` applies only to `word`s, and has a identity fallback.
#define WORDT(word,t) [ word .o. t ] .P. word;

# For crh-cyrl, a "word" is anything other than a bracketed roman numeral
define TRANSLATE'CYRL(regs) [
  NO'BRACKETS .o. BRACKETROMAN'CRH(PHP'BREAK) .o. SWAPBRACKETS .o.
  [ [ [ "[[":0 [\BRACKETS]* "]]":0 ] .o. regs ] | \BRACKETS ]*
];

# For crh-latn, no roman numeral special-casing is done.
define TRANSLATE'LATN(regs) NO'BRACKETS .o. regs;

# LanguageCrh::regsConverter
define REGS'CONVERTER'LATN(latn'exceptions, latn'patterns)
  latn'exceptions .o. latn'patterns .o.
  MAP'CYRL'TO'LATN .o. QUOTE'CYRL'TO'LATN;

define REGS'CONVERTER'CYRL(cyrl'exceptions, cyrl'patterns, cyrl'cleanup)
  cyrl'exceptions .o. cyrl'patterns .o.
  MAP'LATN'TO'CYRL .o. QUOTE'LATN'TO'CYRL .o.
  cyrl'cleanup;

# Exceptions (from CrhExceptions.php)
source crh-exceptions.foma

# To variant: 'crh-latn'
define CONVERT'CRH'LATN TRANSLATE'LATN(REGS'CONVERTER'LATN(
  CRH'LATN'EXCEPTIONS, CRH'LATN'PATTERNS(PHP'BREAK)
));

# To variant: 'crh-cyrl'
define CONVERT'CRH'CYRL TRANSLATE'CYRL(REGS'CONVERTER'CYRL(
  CRH'CYRL'EXCEPTIONS, CRH'CYRL'PATTERNS(PHP'BREAK), CYRL'CLEANUP
));

define CRH'WORD CRH'WORD'HELPER(CONVERT'CRH'CYRL, CONVERT'CRH'LATN);

# Safety approximation functions
define SAFE'STAR(rt) _somenonfinal( [\BRACKETS]* .o. [~NOTSAFE(rt)] );
define APPROX'SAFE(x,rt) [?+ .o. NO'BRACKETS .o. x .o. SAFE'STAR(rt)]+;

source crh-write.foma