Module:User:Santi2222/bsk-translit

The following documentation is generated by Module:documentation/functions/translit.
Useful links: root page • root page’s subpages • links • transclusions • testcases • sandbox
Language code in page name (User:Santi2222/bsk) not recognized.
local U = require("Module:string/char")
local m_string_utilities = require("Module:string utilities")
local gsub = mw.ustring.gsub
local rfind = m_string_utilities.find
local export = {}

local zabar = U(0x64E)
local zer = U(0x650)
local pesh = U(0x64F)
local zwnj = U(0x200C) -- Is this even used in Burushaski?
local tashdid = U(0x651) -- also called tashdid
local jazm = U(0x652)
local ghunna = U(0x658)

local consonants = "تٹثجحڅدذڎرڑزسغفکنملگݣقعظطڞصضشژڈخچپب"
local cons_lat = "mnṅptṭkqbdḍgcćċjʐsśṣhzġlyỵw" --internally we use ċ/ʐ instead of c̣/j̣
local vowels_lat = "aáeéiíoóuú"
local diacritics = "َُِّْٰ"
local ZZP = "َُِ"
local lrm = U(0x200e) -- left-to-right mark
local rlm = U(0x200f) -- right-to-left mark

-- basic mapping
local mapping = {
	 = 'b',  = 'p',  = 't',  = 'ṭ',  = 's',
	 = 'j',  = 'ć',  = 'h',  = "ch",  = 'qh',  = "ċ",
	 = 'd',  = 'ḍ',  = 'z',  = "c",  = 'r',  = "ṛ",  = 'z',  = 'ż',
	 = 's',  = "ṣ",  = 'ś',  = 's',  = 'z',  = "ċh",
	 = 't',  = 'z',  = 'ġ',  = 'f',  = 'q',
	 = 'k',  = "ṅ",  = 'g',  = "h",
	 = 'l',  = 'm',  = 'n',  = "ṇ",
	 = "ỵ",  = "y",  --FIXME: is this correct?
	 = "'", -- FIXME: do we need this?
	
	 = "A",  = "H",  = "W",  = "U",  = "Ú",
	 = "Y",  = "I",  = "Í",
	
	--unambiguous vowels
	 = "a",  = "e",  = "áa",  = "ée",  = "aá",  = "eé",

	-- diacritics
	 = "ă",
	 = "ĭ",
	 = "ŭ",
	 = "@", -- also sukun - no vowel. we use a placeholder
	 = "-", -- ZWNJ (zero-width non-joiner)

	-- kashida
	 = "-", -- kashida, no sound
	
	-- numerals
	 = "1",  = "2",  = "3",  = "4",  = "5",
	 = "6",  = "7",  = "8",  = "9",  = "0",
	
	-- punctuation (leave on separate lines)
	 = "?", -- question mark
	 = ".", -- period
	 = ",", -- comma
	 = ";", -- semicolon
	 = '“', -- quotation mark
	 = '”', -- quotation mark
	 = "%", -- percent
	 = "‰", -- per mille
	 = ".", -- decimals
	 = ",", -- thousand
}

local punctuation = "%-:%(%)%*&٫؛؟،ـ«\".\'!»٪؉۔"

local ye = 'ی'
local vao = "و"
local aspirate = 'ھ'
local aiu = "āīūآ"

-- declared as local above
local function is_valid(text)
	if rfind(text, "") then
		return false
	end
	text = gsub(text, "()h", "%1")
	if rfind(text, "") then
		return false
	end
	return true
end

function export.tr(text, lang, sc)
	if type(text) == "table" then
		local function f(x) return (x ~= "") and x or nil end
		text, lang, sc, omit_i3raab, force_translit =
			f(text.args), f(text.args), f(text.args), f(text.args), f(text.args)
	end

	--define the "end" of a word
	text = gsub(text, " | ", "# | #")
	text = gsub(text, "\n" , "#".."\n" .. "#")
	text = gsub(text, "()" , "#".."%1" .. "#")
	text = "##" .. gsub(text, " ", "# #") .. "##"
	text = gsub(text, zwnj, "#"..zwnj.."#")
	-- hastags now mark the beginning and end of a word
	
	text = gsub(text, 'لا', "ﻻ")
	-- Tashdeed. FIXME: is this used?
	text = gsub(text, '()' .. tashdid, "%1@%1")
	text = gsub(text, '()' .. '()' .. tashdid, "%1@%1%2") --if it gets pushed after diacritics
	text = gsub(text, '()' .. aspirate, aspirate.."%1") 
	text = gsub(text, ye .. '(?)' .. tashdid, "y@y%1")
	text = gsub(text,  vao .. '(?)' .. tashdid, "w@w%1")
    text = gsub(text, "ن" .. ghunna, "ṇ")

    -- convert all characters
    text = string.gsub(text, lrm, "")
	text = string.gsub(text, rlm, "")
    text = gsub(text, '.', mapping)
    
    -- vowel fixes, unambiguous vowels have been processed already
    local function process_vowels(word, init, final, med)
	    for k, v in pairs(init) do
	    	word = gsub(word, "#" .. k, "#" .. v)
	    end
	    for k, v in pairs(final) do
	    	word = gsub(word, k .. "#", v .. "#")
	    end
		for k, v in pairs(med) do
	    	word = gsub(word, k, v)
		end
    	return word
    end
    
    --long, falling
    local init_f = { = "óo",  = "úu",  = "ée",  = "íi"}
    local final_f = { = "áa",  = "úu",  = "óo",  = "íi"}
    local med_f = { = "úu",  = "íi"} --we can't do  A-> áa yet, Y -> ée, W -> óo
    text = process_vowels(text, init_f, final_f, med_f)

	--long, rising
	local init_r = { = "oó",  = "uú",  = "eé",  = "ií"}
    local final_r = { = "uú",  = "oó",  = "ií"}
    local med_r = { = "uú",  = "oó",  = "ií",  = "eé"}
    text = process_vowels(text, init_r, final_r, med_r)

    --short
    local init_s = { = "a",  = "o",  = "u",  = "e",  = "i"}
    local final_s = { = "a",  = "u",  = "o",  = "e",  = "i"}
    local med_s = { = "a",  = "o",  = "u",  = "e",  = "i"}
    text = process_vowels(text, init_s, final_s, med_s)

	--remaining characters
	text = gsub(text, "H", "h")
	text = gsub(text, "A", "áa")
	text = gsub(text, "W@", "w@")
	text = gsub(text, "Y@", "y@")
	text = gsub(text, "W()", "w%1")
	text = gsub(text, "W()", "w%1")
	text = gsub(text, "Y()", "y%1")
	text = gsub(text, "Y()", "y%1")
	text = gsub(text, "W", "óo")
	text = gsub(text, "Y", "ée")
	
	-- get rid of hashtags
    text = gsub(text, "#", "")
    
    if not force_translit and not is_valid(text) then
		--return nil
    end

	--final conversions
	text = gsub(text, "ċ","c̣")
	text = gsub(text, "ʐ", "j̣")
	text = gsub(text, "@", "")
	
	text = mw.ustring.toNFC(text)
	
	return text
end

return export
Module:User:Santi2222/bsk-translit

Wikious

Boobota

Sagapedia
