Module:User:AmazingJus/sce

local export = {}

local count = require("Module:string utilities").count
local gsub = mw.ustring.gsub
local gsplit = mw.text.gsplit
local lower = mw.ustring.lower
local match = mw.ustring.match
local trim = mw.text.trim

--[[
	Sourced from A grammatical overview of Santa Mongolian by Kenneth Lynn Field (1997),
	东乡语汉语词典 (Dongxiang-Chinese Dictionary) by Ma Guozhong (马国忠) (2001)
]]--
-- List of valid phonemes
local phons = {
	-- list of valid consonant phonemes
	 = {
		 = "p",  = "pʰ",  = "m",  = "f",
		 = "t",  = "tʰ",  = "n",  = "l",  = "r",
		 = "k",  = "kʰ",  = "q",  = "qʰ",
		 = "x",  = "ʁ",  = "h",
		 = "t͡ɕ",  = "t͡ɕʰ",  = "ɕ",
		 = "t͡ʂ",  = "t͡ʂʰ",  = "ʂ",
		 = "t͡s",  = "t͡sʰ",  = "s",
		 = "w",  = "j"
	},
	-- list of valid vowel phomemes
	 = {
		 = "jɑw",  = "wɑj",
		 = "ɑj",  = "əj",  = "ɑw",  = "əw",  = "ɯ",
		 = "jɑ",  = "jə",  = "ju",  = "wəj",  = "wɑ",
		 = "ɑ",  = "o",  = "ə",  = "i",  = "u"
	},
	-- list of phonetic diphthong representations
	 = {
		 = { "i̯ɑu̯", "i̯ou̯" },  = { "u̯ɑi̯", "u̯ɛi̯" },
		 = { "ɑi̯", "ɛi̯" },  = "ə̝i̯",  = { "ɑu̯", "ou̯" },  = { "əu̯", "ɤu̯" },
		 = "i̯ɑ",  = "i̯ɛ",  = "i̯u",  = "u̯(ə̝)i̯",  = "u̯ɑ"
	}
}
-- List of valid graphemes
local graphs = {
	cons_g = "??", -- list of valid consonant graphemes
	vowel_g = "+" -- list of valid vowel graphemes
}

-- Returns the number of syllables in a given string
local function count_syll(text)
	return count(text, "ɚ") + count(text, "+")
end

-- Returns a string with two unique substitutions given a table with two items
local function sub_multiple(text, string, subs)
	local foo, foo_sub = gsub(text, string, subs)
	local bar, bar_sub = gsub(text, string, subs)

	if foo_sub == 0 and bar_sub == 0 then
		return text
	else
		return foo .. " ~ " .. bar
	end
end

-- Handles ambiguous syllable-final n
local function n_final(text)
	local n = match(text, "n+")
	if n and count(text, n) > 1 then
		error("ŋ is ambiguous. Please provide the full respelling.")
	elseif text == "ŋ" then
		text = mw.title.getCurrentTitle().text
		text = gsub(text, "n", "ŋ")
	end

	return text
end

-- Automatically adds stress mark to the final syllable of a word if no stress mark is present
local function stress_word(word)
	if not match(word, "ˈ") and count_syll(word) > 1 then
		if match(word, "ɚ ") then
			word = gsub(word, "(" .. graphs.cons_g .. "ɚ) ", "ˈ%1 ")
		else
			word = gsub(word, "(" .. graphs.cons_g .. graphs.vowel_g .. "*) ", "ˈ%1 ")
		end
	end

	return word
end

-- Adjusts the text including respelling the term
local function adjust(text, etyl)
	text = lower(text) -- make text lowercase
	text = " " .. text .. " " -- make all word borders have a space
	text = gsub(text, "'", "ˈ") -- convert apostrophe to a stress mark
	text = gsub(text, "gvi", "gvii") -- i after gv is pronounced as ii
	text = gsub(text, "(hi)", "%1i") -- same after gh, kh and h
	text = gsub(text, "er()", "ɚ%1") -- "er" is a unique syllable in its own right
	if etyl == "zh" or etyl == "ar" then text = gsub(text, "r", "ʐ") end -- r is pronounced as ʐ in words of Chinese and Arabic origin

	return text
end

-- Convert the text to a phonemic IPA transcription
function export.IPA_m(text, etyl)
	text = n_final(text) -- handle syllable-final n
	text = adjust(text, etyl) -- adjust the text by respelling
	text = gsub(text, "* ", function(word) -- handle stress marks
		return gsub(stress_word(word), graphs.cons_g, phons.cons_m):gsub(graphs.vowel_g, phons.vowel_m) -- and convert graphemes to phonemes
	end)

	return trim(text)
end

-- Convert the text to a phonetic IPA transcription from the phonemic transcription
function export.IPA_t(text, etyl)
	text = n_final(text) -- handle syllable-final n
	text = adjust(text, etyl) -- adjust the text by respelling
	text = gsub(text, "* ", function(word) -- handle stress marks
		return gsub(stress_word(word), graphs.cons_g, phons.cons_m) -- and convert graphemes to phonemes (consonants only)
	end)

	-- diphthongs
	for k, v in pairs(phons.diph_t) do
		if type(v) == "table" then
			text = sub_multiple(text, k, v)
		else
			text = gsub(text, k, v)
		end
	end

	-- a
	text = gsub(text, "a(ˈ?)n", "æ%1n") -- /ɑ/ before /n/ is 
	text = gsub(text, "a", "ɑ") -- otherwise 

	-- e
	text = gsub(text, "e(ˈ?)ŋ", "ɤ%1ŋ") -- stressed /ə/ before /ŋ/ is 
	text = gsub(text, "ˈ(" .. graphs.cons_g .. ")e", "ˈ%1ɛ") -- otherwise stressed /ə/ is 
	text = gsub(text, "e", "ə")

	-- i
	text = gsub(text, "(h)i", "%1ɨ") -- /i/ is  after retroflex consonants
	text = gsub(text, "()i", "%1ɪ") -- /i/ is  after alveolar affricates and fricatives

	-- o
	text = gsub(text, "* ", function(word) -- /o/ is labialised as  only if it appears once in a word
		return count(word, "o") == 1 and gsub(word, "o", "ʷo") or word
	end)

	-- devoiced vowels

	-- vowels before /n, ŋ/ are nasalised
	text = gsub(text, "(.̯?)(ˈ?)", "%1̃%2")

	return trim(text)
end

function export.show(frame)
	local args = frame:getParent().args -- get the arguments
	local etyl = args.etyl -- get the etymology
	local text = mw.title.getCurrentTitle().text -- get current page title
	local results = {} -- store the results

	if args then
		for _, param in ipairs(args) do
			if param ~= "" then
				table.insert(results, {pron = "/" .. export.IPA_m(param, etyl) .. "/"})
			end
		end
	else
		table.insert(results, {pron = "/" .. export.IPA_m(text, etyl) .. "/"})
	end

	return "*" .. require("Module:IPA").format_IPA_full { lang = require("Module:languages").getByCode("sce"), items = results }
end

return export