မဝ်ဂျူ:lki-translit

Documentation for this module may be created at မဝ်ဂျူ:lki-translit/doc

-- Authors: JavaScript ئاسۆ; Lua Ghybu, Calak

local export = {}

local gsub = mw.ustring.gsub
local U = mw.ustring.char

local mapping = {
	["ا"] = "a", ["ب"] = "b", ["چ"] = "ç", ["ج"] = "c", ["د"] = "d", ["ە"] = "e", ["ێ"] = "ê", ["ف"] = "f", ["گ"] = "g",
	["ھ"] = "h", ["ه"] = "h", ["ح"] = "ḧ", ["ژ"] = "j", ["ک"] = "k", ["ڵ"] = "ll", ["ل"] = "l", ["م"] = "m", ["ن"] = "n",
	["ۆ"] = "o", ["پ"] = "p", ["ق"] = "q", ["ر"] = "r", ["ڕ"] = "r", ["س"] = "s", ["ش"] = "ş", ["ت"] = "t", ["ۊ"] = "ü",
	["ڤ"] = "v", ["خ"] = "x", ["غ"] = "ẍ", ["ز"] = "z", ["ئ"] = "", ["ع"] = "'",
	
	[U(0x200C)] = "", -- ZWNJ (zero-width non-joiner)
	["ـ"] = "", -- kashida, no sound
	
	-- numerals
    ["١"] = "1", ["٢"] = "2", ["٣"] = "3", ["٤"] = "4", ["٥"] = "5",
	["٦"] = "6", ["٧"] = "7", ["٨"] = "8", ["٩"] = "9", ["٠"] = "0",
	-- persian variants to numerals
	["۱"] = "1", ["۲"] = "2", ["۳"] = "3", ["۴"] = "4", ["۵"] = "5",
	["۶"] = "6", ["۷"] = "7", ["۸"] = "8", ["۹"] = "9", ["۰"] = "0",
}

	-- punctuation (leave on separate lines)
local punctuation = {
	["؟"] = "?", -- question mark
	["،"] = ",", -- comma
	["؛"] = ";", -- semicolon
	["«"] = '“', -- quotation mark
	["»"] = '”', -- quotation mark
	["٪"] = "%", -- percent
	["؉"] = "‰", -- per mille
	["٫"] = ".", -- decimals
	["٬"] = ",", -- thousand
}

-- translit
local function tr_word(word)
	
    word = gsub(word, '.', punctuation)
		
	--Remove punctuation at the end of the word.
	if mw.ustring.find(word, '[%.%!،؛»«٪؉٫٬%p]$') then
	   ponct =  mw.ustring.sub(word, -1)
	   word = gsub(word, '[%.%!،؛»«٪؉٫٬%p]$', '')
	else
	   word = word
	   ponct = ''
	end
	
	word = gsub(word, 'ه‌', "ە") --correct unicode for letter ە
	-- U+0647 (Arabic letter heh) + U+200C (zero-width non-joiner) → U+06D5 (Arabic letter ae)
	
	-- diacritics
	word = gsub(word, 'ْ', "i") -- U+0652, Arabic sukun
	word = gsub(word, 'ِ', "i") -- U+0650, Arabic kasra
	 
	--managing 'و' and 'ی'
	word = gsub(word, 'و([iاێۆۊە])', "w%1") --و + vowel => w (e.g. wan)
	word = gsub(word, 'ی([iاێۆۊە])', "y%1") --ی + vowel => y (e.g. yas)
	word = gsub(word, '([iاێۆۊە])و', "%1w") --vowel + و => w (e.g. kew)
	word = gsub(word, '([iاێۆۊە])ی', "%1y") --vowel + ی => y (e.g. bey)
	word = gsub(word, '([iاێۆە])ۊ', "%1ẅ") --vowel + و => ẅ (e.g. taẅ)
	word = gsub(word, '([iاۆۊە])ێ', "%1ÿ") --vowel + ێ => ÿ (e.g.  şeÿtan)
	word = gsub(word, '^و$', "û") --non-letter + 'و' + non-letter => û (=and)
	
	word = gsub(word, '([^ء-يٱ-ەiwẅyÿ])و', "%1w") --non-letter + 'و' => w (e.g. wetar)
	word = gsub(word, '^و', "w") --first 'و' => w (e.g. wetar)
	word = gsub(word, 'یو', "îw") --'ی' + 'و' => îw (e.g. mîwe)
	word = gsub(word, '([^و])یی', "%1îy") --'ی' + 'ی' => îy (e.g. kanîy)
	word = gsub(word, 'وی', "uy") --'و' + 'ی' => uy (e.g. buyn)
	word = gsub(word, 'وو', "û") --'و' + 'و' => û (e.g. nû)
	word = gsub(word, 'ی', "î")
	word = gsub(word, 'و', "u")
	word = gsub(word, 'uu', "û") --'و' + 'و' => û (e.g. nû)
	word = gsub(word, '([ء-يٱ-ەiîuûwẅyÿ])ڕ', "%1rr") --when 'ڕ' not at the beginning of a word => rr
	word = gsub(word, '([ء-يٱ-ەiîuûwẅyÿ])ئ', "%1'") --when 'ئ' not at the beginning of a word => '
	
	word = gsub(word, '.', mapping)
	
	--insert i where applicable
	word = gsub(word, 'll', "Ľ") -- temporary conversion to avoid seeing ll as 2 letters
	word = gsub(word, 'rr', "Ŕ") -- temporary conversion to avoid seeing rr as 2 letters
	
	word = gsub(word, '([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([fjlĽmnrŔsşvwẅxẍyÿz])([fjlĽmnrŔsşvwẅxẍyÿz])([^aeêiîouûüy])', "%1%2i%3%4") --e.g. grft -> grift
	word = gsub(word, '([aeêiîouûü])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])$', "%1%2%3i%4") --e.g. cejnt -> cejnit
	word = gsub(word, '([fjlĽrŔsşwyz])([fjlĽmnrŔsşvwẅxẍyÿz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])', "%1i%2%3") --e.g. wrd -> wird
	
	word = gsub(word, '([bcçdghḧkmnpqtvxẍ])([fjlĽmnrŔsşvwẅxẍyÿz])([^aeêiîouûü])', "%1i%2%3") --e.g. prd -> pird
	word = gsub(word, '([bcçdghḧkmnpqtvxẍ])([fjlĽmnrŔsşvwẅxẍyÿz])$', "%1i%2") --like above
	
	word = gsub(word, '([^aeêiîouûü])([bcçdghḧkmnpqtvxẍ])([fjlĽmnrŔsşvwẅxẍyÿz])([^aeêiîouûü])', "%1%2i%3%4") --repeat the latter expression, in case skipped
	word = gsub(word, '([^aeêiîouûü])([bcçdghḧkmnpqtvxẍ])([fjlĽmnrŔsşvwẅxẍyÿz])$', "%1%2i%3") --repeat the latter expression, in case skipped

	word = gsub(word, '^([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([^aeêiîouûü])', "%1i%2%3") --e.g. ktk -> kitk
	word = gsub(word, '^([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])$', "%1i%2") --e.g. ktk -> kitk
	word = gsub(word, '([^aeêiîouüy])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([^aeêiîouûü])', "%1%2i%3%4") --e.g. ktk -> kitk
	word = gsub(word, '([^aeêiîouüy])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])$', "%1%2i%3") --e.g. ktk -> kitk

	word = gsub(word, '([^a-zçşêîûüĽŔ])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])$', "%1%2i") --e.g. j -> ji
	word = gsub(word, '^([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])$', "%1i") --e.g. j -> ji	


	--word = gsub(word, '([^a-zêîûçş0-9\'’])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvxẍz])', "%1%2i%3") --e.g. bra -> bira
	--word = gsub(word, '^([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvxẍz])', "%1i%2") --e.g. bra -> bira

	--word = gsub(word, '([bcçdfghḧjklmnpqrsştvwẅxẍz][bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])([bcçdfghḧjklĽmnpqrŔsştvwẅxẍz])', "%1i%2") --e.g. aşkra -> aşkira
	
	--word = gsub(word, 'si([tp][aeêiîouû])', "s%1") -- sp, st cluster
	
	word = gsub(word, 'Ľ', "ll") --revert the temporary conversion
	word = gsub(word, 'Ŕ', "rr") --revert the temporary conversion

    -- Add the punctuation who had previously deleted.
    word = word .. ponct
    
	return word
end

function export.tr(text, lang, sc)
    local textTab = {}
    
    -- Create a word table separated by a space (%s).
	for _, word in ipairs(mw.text.split(text, '%s+')) do
		table.insert(textTab, word)
	end
	
	-- Tablo of translit.
	for key, word in ipairs(textTab) do
		textTab[key] = tr_word(word)
	end

	return table.concat(textTab, ' ')
end

return export