Modül:hi-alfabeçeviri

-- Transliteration for Hindi (possibly other languages using Devanagari script, except for Sanskrit)

local export = {}
local gsub = mw.ustring.gsub
local match = mw.ustring.match

local conv = {
	-- consonants
	 = 'k',  = 'kh',  = 'g',  = 'gh',  = 'ṅ', 
	 = 'c',  = 'ch',  = 'j',  = 'jh',  = 'ñ', 
	 = 'ṭ',  = 'ṭh',  = 'ḍ',  = 'ḍh',  = 'ṇ',
	 = 't',  = 'th',  = 'd',  = 'dh',  = 'n',
	 = 'p',  = 'ph',  = 'b',  = 'bh',  = 'm', 
	 = 'y',  = 'r',  = 'l',  = 'v',  = 'ḷ',
	 = 'ś',  = 'ṣ',  = 's',  = 'h',
	 = 'q',  = 'x',  = 'ġ',  = 'ḻ',
	 = 'z',  = 'ḻ',  = 'ž',  = 'ṛ',  = 'ṛh',
	 = 'f',  = 'θ',  = 'ṉ',  = 'ṟ',
	--  = 'gy',

	-- vowel diacritics
	 = 'i',  = 'u',  = 'e',  = 'o', 
	 = 'ā',  = 'ī',  = 'ū', 
	 = 'ŕ',
	 = 'ai',  = 'au',
	 = 'ŏ',
	 = 'ĕ',

	-- vowel signs
	 = 'a',  = 'i',  = 'u',  = 'e',  = 'o',
	 = 'ā',  = 'ī',  = 'ū', 
	 = 'ŕ', 
	 = 'ai',  = 'au', 
	 = 'ŏ',
	 = 'ĕ',
	
	 = 'om',
	
	-- chandrabindu
	 = '̃',
	
	-- anusvara
	 = 'ṁ',
	
	-- visarga
	 = 'ḥ',
	
	-- virama
	 = '',
	
	-- numerals
	 = '0',  = '1',  = '2',  = '3',  = '4',
	 = '5',  = '6',  = '7',  = '8',  = '9',
	
	-- punctuation
	 = '.', -- danda
	 = '.', -- double danda
	 = '', -- compound separator
	
	-- abbreviation sign
	 = '.',
}

local nasal_assim = {
	 = 'ङ',  = 'ङ',  = 'ङ',  = 'ङ', 
	 = 'ञ',  = 'ञ',  = 'ञ',  = 'ञ',  
	 = 'ण',  = 'ण',  = 'ण',  = 'ण',
	 = 'म',  = 'म',  = 'म',  = 'म',  = 'म',
	 = 'ँ',  = 'ँ',
}

local perm_cl = {
	 = true,  = true,  = true,
	
}

local all_cons, special_cons = 'कखगघङचछजझञटठडढतथदधपफबभशषसयरलवहणनम', 'यरलवहनम'
local vowel, vowel_sign = 'aिुृेोाीूैौॉॅ', 'अइउएओआईऊऋऐऔऑऍ'
local syncope_pattern = '()(़?)a(़?)(?)'

local function rev_string(text)
	local result, length = {}, mw.ustring.len(text)
	for i = length, 1, -1 do
		table.insert(result, mw.ustring.sub(text, i, i))
	end
	return table.concat(result)
end

function export.tr(text, lang, sc)
	text = gsub(text, '(़?)(?)', function(c, d)
		return c .. (d == "" and 'a' or d) end)
	for word in mw.ustring.gmatch(text, "+") do
		local orig_word = word
		word = rev_string(word)
		word = gsub(word, '^a(़?)()(.)(.?)', function(opt, first, second, third)
			return (((match(first, '') and match(second, '्') and not perm_cl)
				or match(first .. second, 'य'))
				and 'a' or "") .. opt .. first .. second .. third end)
		while match(word, syncope_pattern) do
			word = gsub(word, syncope_pattern, '%1%2%3%4')
		end
		word = gsub(word, '(.?)ं(.)', function(succ, prev)
			return succ .. (succ..prev == "a" and "्म" or 
				(succ == "" and match(prev, '') and "̃" or nasal_assim or "n")) .. prev end)
		text = gsub(text, orig_word, rev_string(word))
	end
	text = gsub(text, '.़?', conv)
	text = gsub(text, 'a()̃', 'a͠%1')
    text = gsub(text, 'jñ', 'gy')
    text = gsub(text, 'ñz', 'nz')
	return mw.ustring.toNFC(text)
end

return export