Module:ne-convert2
Jump to navigation
Jump to search
- This module lacks a documentation subpage. Please create it.
- Useful links: subpage list • links • transclusions • testcases • sandbox
-- Feel free do edit.
-- I haven't been able to solve it yet
local U = require("Module:string/char")
local gsub = mw.ustring.gsub
local export = {}
local zabar = U(0x64E)
local zer = U(0x650)
local pesh = U(0x64F)
local tashdid = U(0x651) -- also called shadda
local jazm = "ْ"
local he = "ہ"
local consonants = "ببپتثجچحخدذرزژسشصضطظعغفقکگلࣇمنݨوہھٹڈڑں"
local consonantS = "ببپتثجچحخدذرزژسشصضطظعغفقکگلࣇمنݨہھٹڈڑں"
local vowels = "ایئےۓوؤ"
local hes = "ہح"
local diacritics = "َُِّْٰ"
local ZZP = "َُِ" -- zzp + kahri zabar
local PZ = "َُ" --Zabar pesh
local mapping = {
["آ"] = 'ā', ["ب"] = 'b', ["پ"] = 'p', ["ت"] = 't', ["ٹ"] = 'ṭ', ["ث"] = 's',
["ج"] = 'j', ["چ"] = 'c', ["ح"] = 'h', ["خ"] = 'x',
["د"] = 'd', ["ڈ"] = 'ḍ', ["ذ"] = 'z', ["ر"] = 'r', ['ڑ'] = "ṛ", ["ز"] = 'z', ["ژ"] = 'ž',
["س"] = 's', ["ش"] = 'ś', ["ص"] = 's', ["ض"] = 'z',
["ط"] = 't', ["ظ"] = 'z', ["غ"] = 'ġ', ["ف"] = 'f', ["ق"] = 'q',
["ک"] = 'k', ["گ"] = 'g', ["ݨ"] = 'ṇ', ["ࣇ"] = 'ḷ',
["ل"] = 'l', ["م"] = 'm', ["ن"] = 'n', ["و"] = 'ō', ["ہ"] = 'h', ["ی"] = 'y', ["ے"] = 'ē', ["۔"] = ".", ["ں"] = '̃',
["ھ"] = "h",
["ع"] = 'a',
["ء"] = '',
["ئ"] = '',
["ؤ"] = 'ō',
["أ"] = '',
-- diacritics
[zabar] = "a",
[zer] = "i",
[pesh] = "u",
[jazm] = "", -- also sukun - no vowel
[U(0x200C)] = "-", -- ZWNJ (zero-width non-joiner)
-- ligatures
["ﻻ"] = "lā",
["ﷲ"] = "allāh",
-- kashida
["ـ"] = "-", -- kashida, no sound
-- 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)
["؟"] = "?", -- question mark
["،"] = ",", -- comma
["؛"] = ";", -- semicolon
["«"] = '“', -- quotation mark
["»"] = '”', -- quotation mark
["٪"] = "%", -- percent
["؉"] = "‰", -- per mille
["٫"] = ".", -- decimals
["٬"] = ",", -- thousand
["ۓ"] = "-ye",
["ۀ"] = "-e" -- he ye (in ezâfe)
}
local ain = 'ع'
local alif = 'ا'
local ye = 'ی'
local ye2 = 'ئ'
local ye3 = "ے"
local vao = "و"
local aspirate = 'ھ'
local ghunna = 'ں'
function export.tr(text, lang, sc)
-- EXCEPTIONS - leave as they are, unless they have been sorted out elsewhere
text = gsub(text, '([' .. consonants .. '])' .. ye .. ghunna, "%1ē̃")
text = gsub(text, zabar .. aspirate .. "(ی)" , "hɛ̄")
text = gsub(text, zabar .. aspirate .. "(و)" , "hɔ̄")
text = gsub(text, "ؤ" .. pesh, "ū")
text = gsub(text, "وہ", "vo")
-- Tashdeed
text = gsub(text, '([' .. consonants .. '])' .. tashdid, "%1%1")
text = gsub(text, '([' .. consonants .. '])' .. tashdid .. '([' .. ZZP .. '])', "%1%1%2")
-- For some reason the tashdeed gets pushed after the other diacritics, so this line is necessary for tashdeed to work with other diacritics
text = gsub(text, '([' .. consonants .. '])' .. '([' .. ZZP .. '])' .. tashdid, "%1%1%2")
-- e, instead of i
text = gsub(text, jazm .. '([' .. consonants .. '])' .. zer .. '([' .. consonants .. '])' .. alif, "%1e%2ā")
-- tanween diacritic
text = gsub(text, '([' .. consonants .. '])' .. 'ً' .. alif, "%1an")
text = gsub(text, alif .. 'ً', "an")
text = gsub(text, '([' .. consonants .. '])' .. 'ً', "%1an")
-- khari zabar --
text = gsub(text, '([' .. vowels .. '])' .. 'ٰ', "ā")
text = gsub(text, '([' .. consonants .. '])' .. 'ٰ' .. '([' .. vowels .. '])', "%1ā")
-- ‘ain
text = gsub(text, alif .. ain .. '([' .. consonants .. '])', "ɛ̄%1")
text = gsub(text, '([' .. consonants .. '])' .. ain .. he, "%1a")
text = gsub(text, '([' .. consonants .. '])' .. ain, "%1")
text = gsub(text, ain .. zer .. '([' .. consonants .. '])', "i%1")
text = gsub(text, ain .. pesh .. '([' .. consonants .. '])', "u%1")
text = gsub(text, ain .. zer .. ye .. '([' .. consonants .. '])', "ī%1")
text = gsub(text, ain .. pesh .. vao .. '([' .. consonantS .. '])', "ū%1")
-- Vao
text = gsub(text, vao .. '([' .. ZZP .. '])', "v%1")
-- یے
text = gsub(text,'([' .. consonants .. '])' .. ye .. ye3, "%1ië")
-- ئے
text = gsub(text, '([' .. consonants .. '])' .. pesh .. ye2 .. ye3, "%1uë")
text = gsub(text, '([' .. consonants .. '])' .. ye2 .. ye3, "%1aë")
-- سیب (example)
text = gsub(text, '([' .. consonants .. '])' .. ye .. '([' .. consonants .. '])', "%1ē%2")
-- کئی (example)
text = gsub(text, '([' .. consonants .. '])' .. pesh .. ye2 .. ye, "%1uï")
text = gsub(text, '([' .. consonants .. alif .. '])' .. ye2 .. ye, "%1aï")
-- اَے
text = gsub(text, alif .. zabar .. ye3, "ɛ̄")
-- َے
text = gsub(text, zabar .. ye3, "ɛ̄")
-- izafat
text = gsub(text, '([' .. consonants .. '])' .. zer .. " ", "%1-e ")
-- Fatha Majhool --
text = gsub(text, "([" .. consonants .. "])" .. zabar .. aspirate .. "([" .. hes .. "])" .. "([" .. ZZP .. "])" .. jazm, "%1hêh%3")
text = gsub(text, "([" .. consonants .. "])" .. zabar .. "([" .. hes .. "])" .. jazm .. "([" .. ZZP .. "])" , "%1êh%3")
text = gsub(text, "([" .. consonants .. "])" .. zabar .. "([" .. hes .. "])" .. jazm , "%1êh")
-- do-chashme-he zabar, zer, pesh
text = gsub(text, zer .. aspirate .. ye .. ghunna, "hï̃")
text = gsub(text, "([" .. consonants .. "])" .. "([" .. ZZP .. "])" .. aspirate, "%1h%2")
text = gsub(text, "([" .. consonants .. "])" .. aspirate .. ye .. "", "%1hī")
-- medial/final consonants
text = gsub(text, zabar .. he .. zer .. ye, "ahī")
text = gsub(text, zabar .. he .. alif, "ahā")
text = gsub(text, zabar .. he .. '([' .. consonants .. vowels .. '])', "ah%1")
text = gsub(text, '([' .. consonants .. '])' .. alif, "%1ā")
text = gsub(text, '([' .. consonants .. '])' .. tashdid .. alif, "%1%1ā")
text = gsub(text, '([' .. consonants .. '])' .. vao, "%1ō")
text = gsub(text, '([' .. consonants .. '])' .. tashdid .. vao, "%1%1ō")
text = gsub(text, alif .. '([' .. consonantS .. '])', "a%1")
text = gsub(text, alif .. zabar .. '([' .. consonantS .. '])', "a%1")
text = gsub(text, alif .. zabar .. vao .. '([' .. consonants .. '])', "ɔ̄%1")
text = gsub(text, alif .. vao .. '([' .. consonants .. '])', "ō%1")
text = gsub(text, alif .. zabar .. ye .. '([' .. consonants .. '])', "ɛ̄%1")
text = gsub(text, alif .. pesh .. '([' .. consonantS .. '])', "u%1")
text = gsub(text, alif .. zer .. '([' .. consonants .. '])', "i%1")
text = gsub(text, pesh .. vao, "ū")
text = gsub(text, alif .. zer .. ye .. '([' .. consonants .. '])', "ī%1")
text = gsub(text, alif .. zer .. ye .. "", "ī")
text = gsub(text, alif .. ye .. '([' .. consonants .. '])', "ē%1")
-- diacritics
text = gsub(text, "([" .. consonants .. "])" .. zabar .. vao, "%1ɔ̄")
text = gsub(text, "([" .. consonants .. "])" .. zabar .. ye, "%1ɛ̄")
text = gsub(text, "([" .. consonants .. "])" .. zabar .. ye3, "%1ɛ̄")
text = gsub(text, "([" .. consonants .. "])" .. ye .. " ", "%1ī ")
text = gsub(text, "([" .. consonants .. "])" .. zer .. ye, "%1ī")
-- final he + short vowel disregards the he and transliterates the vowel
text = gsub(text, ye .. he , "ye")
text = gsub(text, ye .. alif , "yā")
text = gsub(text, "([" .. consonants .. "])" .. he , "%1e")
--
text = gsub(text, zabar .. he .. "([" .. ZZP .. "])" , "ah%1")
text = gsub(text, zabar .. he , "a")
--
text = gsub(text, "ۂ", "a-e")
text = mw.ustring.gsub(text, 'ahē', "hɛ̄")
text = mw.ustring.gsub(text, 'ahō', "hɔ̄")
text = mw.ustring.gsub(text, 'hiē', "hī")
text = mw.ustring.gsub(text, '.', mapping)
text = mw.ustring.gsub(text, 'ōā', "vā")
text = mw.ustring.gsub(text, 'ɔ̄ā', "vā")
text = mw.ustring.gsub(text, 'ōا', "vā")
text = mw.ustring.gsub(text, 'ɔ̄aاa', "avā")
text = mw.ustring.gsub(text, 'ɔ̄ا', "vā")
text = mw.ustring.gsub(text, 'ɔ̄ا', "vā")
-- Changed these to 'iy(*)', because they will be used for with ی, which are normally written as 'iy'
text = mw.ustring.gsub(text, 'īā', "iyā")
text = mw.ustring.gsub(text, 'īa', "iya")
--
text = mw.ustring.gsub(text, 'aا', "ā")
text = mw.ustring.gsub(text, 'اē', "ē")
text = mw.ustring.gsub(text, 'īا', "yā")
text = mw.ustring.gsub(text, 'êha', "êhê")
-- vao as a medial consonant
text = mw.ustring.gsub(text, "ūa", "uva")
text = mw.ustring.gsub(text, "ɔ̄([aiu])", "av%1")
--
-- Final corrections
text = mw.ustring.gsub(text, "āa", "ā")
text = mw.ustring.gsub(text, "āā", "ā")
text = mw.ustring.gsub(text, "aa", "ā")
text = mw.ustring.gsub(text, "ë", "ē")
text = mw.ustring.gsub(text, "ï", "ī")
return text
end
return export