Module:Eve-translit

local export = {}

local u = mw.ustring.char

local MACRON   = u(0x0304) local DOTABOVE = u(0x0307)

local tab = { ["А"]='A', ["а"]='a', ["Б"]='B', ["б"]='ʙ', ["В"]='W', ["в"]='w', ["Е"]='E', ["е"]='e', ["Ё"]='Jo', ["ё"]='jo', ["Г"]='G', ["г"]='g', ["Д"]='D', ["д"]='d', ["И"]='I', ["и"]='i', ["Ӣ"]='Ī', ["ӣ"]='ī', ["Й"]='J', ["й"]='j', ["К"]='K', ["к"]='k', ["Л"]='L', ["л"]='l', ["М"]='M', ["м"]='m', ["Н"]='N', ["н"]='n', ["Ӈ"]='Ŋ', ["ӈ"]='ŋ', ["О"]='O', ["о"]='o', ["Ө"]='Ö', ["ө"]='ö', ["Ӫ"]='Ö', ["ӫ"]='ö', ["П"]='P', ["п"]='p', ["Р"]='R', ["р"]='r',	["С"]='S', ["с"]='s', ["Т"]='T', ["т"]='t', ["У"]='U', ["у"]='u', ["Ӯ"]='Ū', ["ӯ"]='ū', ["Ф"]='F', ["ф"]='f', ["Х"]='H', ["х"]='h', ["Ч"]='C', ["ч"]='c', ["Ы"]='I', ["ы"]='i', ["Э"]='Ə', ["э"]='ə', ["Ю"]='Ju', ["ю"]='ju', ["Я"]='Ẹ', ["я"]='ẹ', -- Not present in the original latinisation -- non-native letters ["Ж"]='Z', ["ж"]='z',	["З"]='Z', ["з"]='z', ["Ц"]='C', ["ц"]='c', ["Ш"]='S', ["ш"]='s', ["Щ"]='S', ["щ"]='s', ['Ъ']='ʺ', ['ъ']='ʺ', ["Ь"]="’", ["ь"]="’", -- non-standard letters ["Ҕ"]='Γ', ["ҕ"]='γ', ["Ҥ"]='Ŋ', ["ҥ"]='ŋ', ["Ү"]='Ü', ["ү"]='ü', ["Һ"]='Ḥ', ["һ"]='ḥ', ["Ӄ"]='Q', ["ӄ"]='q' }

local iotatedTranslit = { ["е"] = "je", ["я"] = "ja", ["и"] = "ji", ["ӣ"] = "jī", ["Е"] = "Je", ["Я"] = "Ja", }

local replacements = { { "Ё", 'Jo' }, { "ё", 'jo' }, { "Ӫ", 'Jö' }, { "ӫ", 'jö' }, { "Ю", 'Ju' }, { "ю", 'ju' }, -- Unfortunately the Cyrillic alphabet doesn't distinguish between ʒe and ʒə { "Де", 'Ʒe' }, { "де", 'ʒe' }, { "Не", 'Ņe' }, { "не", 'ņe' }, { "Ди", 'Ʒi' }, { "ди", 'ʒi' }, { "Ни", 'Ņi' }, { "ни", 'ņi' }, { "Дя", 'Ʒa' }, { "дя", 'ʒa' }, { "Ня", 'Ņa' }, { "ня", 'ņa' }, { "Дj", 'Ʒ' }, { "дj", 'ʒ' }, { "Нj", 'Ņ' }, { "нj", 'ņ' }, -- The following is non-standard but supposedly used in non-standard spelling and dialect forms { "Дь", 'Ʒ' }, { "дь", 'ʒ' }, { "Нь", 'Ņ' }, { "нь", 'ņ' }, { "Нг", 'Ŋ' }, { "нг", 'ŋ' }, } function export.tr(text, lang, sc) local ugsub, str_gsub = mw.ustring.gsub, string.gsub local UTF8char = '[\1-\127\194-\244][\128-\191]*' for i, replacement in ipairs(replacements) do		text = str_gsub(text, unpack(replacement)) end -- е after a vowel or at the beginning of a word becomes ye	-- Again, the Cyrillic alphabet doesn't distinguish between je and jə	text = ugsub(text,		"([АОУЫЯЕИӢЪЬаӣиоуыэяеъьaeioöu]["..MACRON..DOTABOVE.."]?)([еяиӣ])", 		function(preceding, iotated)			return preceding .. iotatedTranslit[iotated]		end) text = ugsub(text, "^[ЕеЯя]", iotatedTranslit) text = ugsub(text, "([^Ѐ-ӿ])([ЕеЯя])",		function(preceding, iotated)			return preceding .. iotatedTranslit[iotated]		end) text = str_gsub(text, UTF8char, tab) return text end

return export