Module:sjd-translit

From Wiktionary, the free dictionary
Jump to navigation Jump to search

This module will transliterate Kildin Sami language text. The module should preferably not be called directly from templates or other modules. To use it from a template, use {{xlit}}. Within a module, use Module:languages#Language:transliterate.

For testcases, see Module:sjd-translit/testcases.

Functions

tr(text, lang, sc)
Transliterates a given piece of text written in the script specified by the code sc, and language specified by the code lang.
When the transliteration fails, returns nil.

local export = {}

local u = mw.ustring.char

local macron = u(0x0304)

local tt = {
	["А"]='A', ["а"]='a',
	["Ӓ"]="’a", ["ӓ"]="’a",
	["Б"]='B', ["б"]='b',
	["В"]='V', ["в"]='v',
	["Г"]='G', ["г"]='g',
	["Д"]='D', ["д"]='d',
	["Е"]='Je', ["е"]='je',
	["Ё"]="Jo", ["ё"]="jo",
	["Ж"]='Ž', ["ж"]='ž',
	["З"]='Z', ["з"]='z',
	["Һ"]="H", ["һ"]="h", ["ʼ"]="h",
	["И"]='I', ["и"]='i',
	["Ӣ"]="Ī", ["ӣ"]="ī",
	["Й"]="J", ["й"]="j",
	["Ј"]="J̥", ["ј"]="j̥", ["Ҋ"]="J̥", ["ҋ"]="j̥",
	["К"]='K', ["к"]='k',
	["Л"]='L', ["л"]='l',
	["Ӆ"]='L̥', ["ӆ"]='l̥',
	["М"]='M', ["м"]='m',
	["Ӎ"]="M̥", ["ӎ"]="m̥",
	["Н"]='N', ["н"]='n',
	["Ӊ"]="N̥", ["ӊ"]="n̥",
	["Ӈ"]="Ŋ", ["ӈ"]="ŋ",
	["О"]='O', ["о"]='o',
	["П"]='P', ["п"]='p', 
	["Р"]='R', ["р"]='r',
	["Ҏ"]="R̥", ["ҏ"]="r̥",
	["С"]='S', ["с"]='s',
	["Т"]='T', ["т"]='t',
	["У"]='U', ["у"]='u', 
	["Ӯ"]="Ū", ["ӯ"]="ū",
	["Ф"]='F', ["ф"]='f',
	["Х"]='X', ["х"]='x',
	["Ц"]='C', ["ц"]='c',
	["Ч"]='Č', ["ч"]='č',
	["Ш"]='Š', ["ш"]='š',
	["Щ"]="Šč", ["щ"]="šč",
	["Ы"]="Y", ["ы"]="y",
	["Ъ"]="", ["ъ"]="",
	["Ь"]="’", ["ь"]="’", ["Ҍ"]= "’", ["ҍ"]="’",
	["Э"]="E", ["э"]="e",
	["Ӭ"]="’E", ["ӭ"]="’e",
	["Ю"]="Ju", ["ю"]="ju",
	["Я"]="Ja", ["я"]="ja",
}

local vowel = "аӓеёиӣоуӯыэӭюяАӒЕЁИӢОУӮЫЭӬЮЯ"

function export.tr(text, lang, sc)
	-- digraphs ie and ea
	text = mw.ustring.gsub(text, "([БбВвГгДдЖжЗзКкЛлӅӆМмӍӎНнӉӊӇӈПпРрҎҏТтХхЦцЧчШшЩщ])е" .. macron, "%1иэ")
	text = mw.ustring.gsub(text, "([БбВвГгДдЖжЗзКкЛлӅӆМмӍӎНнӉӊӇӈПпРрҎҏТтХхЦцЧчШшЩщ])я" .. macron, "%1эа")
	
	text = mw.ustring.gsub(text, "([БбВвГгДдЖжЗзКкЛлӅӆМмӍӎНнӉӊӇӈПпРрҎҏТтХхЦцЧчШшЩщ])Е" .. macron, "%1Иэ")
	text = mw.ustring.gsub(text, "([БбВвГгДдЖжЗзКкЛлӅӆМмӍӎНнӉӊӇӈПпРрҎҏТтХхЦцЧчШшЩщ])Я" .. macron, "%1Эа")
	
	-- initial j + vowels
	text = mw.ustring.gsub(text, "^" .. "ю" .. macron, "jū")
	text = mw.ustring.gsub(text, "^" .. "я" .. macron, "jea")
	text = mw.ustring.gsub(text, "^" .. "ӣ", "ji")
	text = mw.ustring.gsub(text, "^" .. "е" .. macron, "jie")
	text = mw.ustring.gsub(text, "^" .. "Ю" .. macron, "Jū")
	text = mw.ustring.gsub(text, "^" .. "Я" .. macron, "Jea")
	text = mw.ustring.gsub(text, "^" .. "Ӣ", "Ji")
	text = mw.ustring.gsub(text, "^" .. "Е" .. macron, "Jie")

	text = mw.ustring.gsub(text, " " .. "ю" .. macron, " jū")
	text = mw.ustring.gsub(text, " " .. "я" .. macron, " jea")
	text = mw.ustring.gsub(text, " " .. "ӣ", " ji")
	text = mw.ustring.gsub(text, " " .. "e" .. macron, " jie")
	text = mw.ustring.gsub(text, " " .. "Ю" .. macron, " jū")
	text = mw.ustring.gsub(text, " " .. "Я" .. macron, " jea")
	text = mw.ustring.gsub(text, " " .. "Ӣ", " ji")
	text = mw.ustring.gsub(text, " " .. "E" .. macron, " jie")

	text = mw.ustring.gsub(text, "йе", "jje")
	
	--palatal н
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])е", "%1ьэ")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])ё", "%1ьо")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])ю", "%1ьу")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])я", "%1ьа")
	
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])Е", "%1ЬЭ")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])Ё", "%1ЬО")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])Ю", "%1ЬУ")
	text = mw.ustring.gsub(text, "([^" .. vowel .. "НнЪъЬьҌҍ])Я", "%1ЬА")
	
	text = mw.ustring.gsub(text, "([нН])ь", "%1й")
	text = mw.ustring.gsub(text, "НЬ", "НЙ")
	
	-- oa digraph
	text = mw.ustring.gsub(text, "ОА" .. macron, "ÅÅ")
	text = mw.ustring.gsub(text, "Оа" .. macron, "Åå")
	text = mw.ustring.gsub(text, "оа" .. macron, "åå")
	text = mw.ustring.gsub(text, "О[Аа]", "Å")
	text = mw.ustring.gsub(text, "оа", "å")
	
	-- ёa digraph
	text = mw.ustring.gsub(text, "ЁА" .. macron, "JÅÅ")
	text = mw.ustring.gsub(text, "Ёа" .. macron, "Jåå")
	text = mw.ustring.gsub(text, "ёа" .. macron, "jåå")
	text = mw.ustring.gsub(text, "ЁА", "JÅ")
	text = mw.ustring.gsub(text, "Ёа", "Jå")
	text = mw.ustring.gsub(text, "ёа", "jå")

	return (mw.ustring.gsub(text, '.', tt))
end

return export