Module:inc-opa-Guru-translit

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

This module will transliterate Old Punjabi 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:inc-opa-Guru-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 conv = {
	--consonants without nukta
	["ਸ"] = "s",
	["ਹ"] = "h",
	["ਕ"] = "k", ["ਖ"] = "kh", ["ਗ"] = "g", ["ਘ"] = "gh", ["ਙ"] = "ṅ",
	["ਚ"] = "c", ["ਛ"] = "ch", ["ਜ"] = "j", ["ਝ"] = "jh", ["ਞ"] = "ñ",
	["ਟ"] = "ṭ", ["ਠ"] = "ṭh", ["ਡ"] = "ḍ", ["ਢ"] = "ḍh", ["ਣ"] = "ṇ",
	["ਤ"] = "t", ["ਥ"] = "th", ["ਦ"] = "d", ["ਧ"] = "dh", ["ਨ"] = "n",
	["ਪ"] = "p", ["ਫ"] = "ph", ["ਬ"] = "b", ["ਭ"] = "bh", ["ਮ"] = "m",
	["ਯ"] = "y", ["ਰ"] = "r", ["ਲ"] = "l", 	["ਵ"] = "v", ["ੜ"] = "ṛ",
	
	-- vowels
	["ਾ"] = "ā",
	["ਿ"] = "i", ["ੀ"] = "ī",
	["ੁ"] = "u", ["ੂ"] = "ū",
	["ੇ"] = "e", ["ੈ"] = "ai",
	["ੋ"] = "o", ["ੌ"] = "au",
	
	-- other diacritics
	["ੰ"] = "ṃ", --ṭippi: nasalize
	["ਂ"] = "ṃ", --bindi: nasalize	
	["੍"] = "", --halant, supresses the inherent vowel "a"
	
	-- independent vowels
	["ਅ"] = "a", ["ਆ"] = "ā",
	["ਇ"] = "i", ["ਈ"] = "ī",
	["ਉ"] = "u", ["ਊ"] = "ū",
	["ਏ"] = "e", ["ਐ"] = "ai",
	["ਓ"] = "o", ["ਔ"] = "au",
	
	-- digits
	["੦"] = "0", ["੧"] = "1", ["੨"] = "2", ["੩"] = "3", ["੪"] = "4",
	["੫"] = "5", ["੬"] = "6", ["੭"] = "7", ["੮"] = "8", ["੯"] = "9",
}

local nasal_assim = {
	["[kg]h?"] = "ṅ",
	["[cj]h?"] = "ñ",
	["[ṭḍ]h?"] = "ṇ",
	["[td]h?"] = "n",
	["[pb]h?"] = "m",
	["n"] = "n",
	["m"] = "m",
    ["s"] = "n",
    ["ñ"] = "ñ",
    ["ṅ"] = "ṅ",
}

-- translit any words or phrases
function export.tr(text, lang, sc)
	local c = "([ਸਹਕਖਗਘਙਚਛਜਝਞਟਠਡਢਣਤਥਦਧਨਪਫਬਭਮਯਰਲਵੜ]਼?)"
	local y = "ਯ"
	local v = "([aਾਿੀੁੂੇੈੋੌ੍])"
	local virama = "੍"
	local n = "([ੰਂ]?)"
	
	local no_virama = mw.ustring.gsub(v,virama,"")
	
	text = text .. " "
	
	text = mw.ustring.gsub(text,c,"%1a")
	text = mw.ustring.gsub(text,"a"..v,"%1")
	
	text = mw.ustring.gsub(text,".",conv)
	
	for key,val in pairs(nasal_assim) do
		text = mw.ustring.gsub(text,"ṃ("..key..")",val.."%1")
	end

	text = mw.ustring.gsub(text," ?[।॥]",".")
	
	text = mw.ustring.gsub(text," $","")
	text = mw.ustring.gsub(text,"hh","h") -- ੍ਹ੍ਹ ਨ੍ਹ੍ਹ
	text = mw.ustring.gsub(text,"yy","y") -- ੍ਯ੍ਯ ਖ੍ਯ੍ਯ
	
	return mw.ustring.toNFC(text)
end
 
return export