Module:bho-Kthi-translit

-- Transliteration for Bhojpuri

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',  = 'v',  = 'ḷ',
	 = 'ś',  = 'ṣ',  = 's',  = 'h',
	 = 'ṛ',  = 'ṛ',  = 'ṛh',  = 'ṛh',
	--  = 'gy',

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

	-- vowel signs
	 = 'a',  = 'i',  = 'u',  = 'e',  = 'o',
	 = 'ā',  = 'ī',  = 'ū',  
	 = 'ai',  = 'au', 
	
	 = '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 = {
	 = '𑂑',  = '𑂑',  = '𑂑',  = '𑂑',
	 = '𑂖',  = '𑂖',  = '𑂖',  = '𑂖',
	 = '𑂝',  = '𑂝',  = '𑂝',  = '𑂝',
	 = '𑂧',  = '𑂧',  = '𑂧',  = '𑂧',  = '𑂧',
}

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

local all_cons, special_cons = '𑂍𑂎𑂏𑂐𑂑𑂒𑂓𑂔𑂕𑂖𑂗𑂘𑂙𑂚𑂛𑂜𑂞𑂟𑂠𑂡𑂣𑂤𑂥𑂦𑂬𑂭𑂮𑂨𑂩𑂪𑂫𑂯𑂝𑂢𑂧', '𑂨𑂩𑂪𑂥𑂫𑂯𑂢𑂧'
local vowel, vowel_sign = 'a𑂰𑂱𑂲𑂳𑂴𑂵𑂶𑂷𑂸', '𑂃𑂄𑂅𑂆𑂇𑂈𑂉𑂊𑂋𑂌'
local syncope_pattern = '()(𑂺?)a(𑂺?)(?)'

local function rev_string(text)
	local result, length = {}, mw.ustring.len(text)
	for i = 1, length do
		table.insert(result, mw.ustring.sub(text, length - i + 1, length - i + 1))
	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)
				local a = ""
				if match(first, '')
					and match(second, '𑂹')
					and not perm_cl
					or match(first .. second, '𑂨') then
						a = "a"
				end
				
				return a .. 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)
				local mid = nasal_assim or "n"
				if succ..prev == "a" then
					mid = "𑂺𑂧" 
				elseif succ == "" and match(prev, '') then
					mid = "̃" 
				end
				return succ .. mid .. prev
			end
		)
		
		text = gsub(text, orig_word, rev_string(word))
	end
	text = gsub(text, '.𑂺?', conv)
	text = gsub(text, 'a()̃', 'a͠%1')
    text = gsub(text, '𑂔𑂹𑂖', conv)
    text = gsub(text, '%*', 'a')

	return mw.ustring.toNFC(text)
end

return export