Module:User:Saph/omv-aro-pron

local export = {}
local pron = {}

local m_para = require("Module:parameters")
local m_lang = require("Module:languages")
local m_IPA = require("Module:IPA")
local m_amf = "Module:amf-utilities"
local PAGENAME = mw.loadData("Module:headword/data").pagename

local gsub = mw.ustring.gsub
local lower = mw.ustring.lower
local split = mw.text.split
local insert = table.insert
local concat = table.concat

local GRAVE = "\204\128"
local ACUTE = "\204\129"
local CIRCUM = "\204\130"
local RETRACT = "\204\160"
local gvowel = "\204??" -- generic vowel

local encode = {
	 = "A",
	 = "B",
	 = "C",
	 = "D",
	 = "E",
	 = "G",
	 = "ɡ",
	 = "I",
	 = "J",
	 = "K",
	 = "N",
	 = "O",
	 = "Q",
	 = "T",
	 = "X",
	 = "Z",
}

function export.encode(text)
	-- universal encoder
	text = mw.ustring.toNFD(text)
	
	mw.log(text:match(""))
	
	return text
		:gsub("", encode)
		:gsub("ʼ", encode)
		:gsub("sh", "S")
end

function export.decode(text)
	local decode = {}
	for m, n in pairs(encode) do decode = m end
	return text:gsub(".", decode)
end

function export.make(frame)
	local lang = frame.args
	
	local params = {
		 = { list = true, default = PAGENAME },
	}
	
	if pron and pron.params then
		for key, val in pairs(pron.params) do
			params = val
		end
	end
	
	local args = m_para.process(frame:getParent().args, params)
	local data = {
		lang = m_lang.getByCode(lang, true),
		items = args
	}
	
	if pron then
		pron.func(args, data)
	else
		pron.func(args, data)
	end
	
	return m_IPA.format_IPA_full(data)
end

--[[

Currently missing:
- Karo (kxh)
- Gayil (gyl)

Gayil's phonology is not well documented, only being briefly touched on in 
Tsuge, 2006; this might still be enough to integrate it into generic, though.

Karo's is undocumented, as far as I can tell; at least, not to any useful
degree.

]]--

pron = {
	-- supports: aiw, dim
	func = function(args, data)
		-- Seyoum, Mulugeta (2008) A Grammar of Dime, LOT Publications
		-- Hayward, Richard J. (1990) "Notes on the Aari Language", in Omotic Language Studies, Routledge, ppg. 425-493 of 666
		local ret = {}
		local correspondences = {
			J = "dʒ",
			K = "tʃ",
			S = "ʃ",
			y = "j",
			Z = "ʒ",
		}
		for _,i in ipairs(args) do
			i = export.encode(lower(i))
			i = i:gsub("(..?)%1", "%1ː")
				:gsub("n()", "n" .. RETRACT .. "%1")
				:gsub("^()", "%1ʰ")
				:gsub("s" .. GRAVE .. "’", "tsʼ")
				:gsub("s" .. GRAVE, "ts")
				:gsub("ds", "dz")
				:gsub("$", {
					 = "χ",
					 = "ɸ",
				})
				:gsub(gvowel .. "" .. gvowel, { 
					 = "ʁ",
					 = "ɸ",
				})
				:gsub("nq", "ɴq")
			insert(ret, { pron = "" })
		end
		data.items = ret
	end
}

pron = {
	func = function(args, data)
		-- Petrollino, Sara (2016) A Grammar of Hamar: A South Omotic language of Ethiopia, Leiden University
		local utilities = require(m_amf)
		local correspondences = {
			c = "tʃ",
			g = "ɡ",
			j = "dʒ",
			y = "j",
			B = "ɓ",
			C = "tʃʼ",
			D = "ɗ",
			E = "ɛ",
			G = "ɠ",
			N = "ɲ",
			O = "ɔ",
			S = "ʃ",
			T = "tʼ",
			Q = "ʔ",
			L = "ː",		-- long
			X = "\204\164",	-- breathy
			H = "\204\165",	-- voiceless
		}
		
		local V = ""
		
		-- custom version of utilities.combine
		local function combine(syllables)
			local a,f = syllables.accent, syllables.falling
			local diacritic = f and "\204\130" or "\204\129"
			local word = "" -- do not use table.concat to avoid modifying input
			for i,syl in ipairs(syllables) do
				if i == a then
					syl = syl:gsub("","%0"..diacritic)
				end
				word = word .. syl
			end
			word = word:gsub("",correspondences)
			return mw.ustring.toNFC(word)
		end
		
		local function pron(text)
			text = utilities.syllabify(text)
			if text.accent == 0 then
				text.accent = 1
			end
			for i, syl in ipairs(text) do
				text = syl:gsub("()%1", "%1L")
					:gsub("cc", "tc")
			end
			text = text:gsub("^()", "ʔ%1")
			return text
		end
		
		local function phon(text,last)
			local C = "rCDB"
			for i=1,#text do
				if i>1 and text:sub(-1,-1):match("") then
					text = text:gsub("^", {r="ɾ", C="c", D="d", B="β"})
						:gsub("^ba", "βa")
				end
				if text:sub(-1,-1) == "n" and i<#text and text:sub(1,1):match("") then
					text = text:gsub("n$", "ŋ")
				end
				text = text:gsub("^T()", "tsʼ%1")
				text = text:gsub("n()$", "ŋ%1")
			end
			text = text:gsub("^Du", "ʔu")
				:gsub("^ha", "a̤")
				:gsub("^q", {qa="qʼa",qe="qʰe",qi="qʰi",qE="qʰE",qO="qʰO",qo="ʔo",qu="ʔu"})
			if not text:find("ʰ") then
				text = text:gsub("^()", "%1ʰ")
			end
			text = text:gsub("tʰsʼ", "tsʼ")
			text = text:gsub("$", {p="p̚",b="b̚",B="ɓ̚",m="m̥",n="n̥",r="r̥",l="l̥"})
			if text.accent ~= #text then
				text = text:gsub("^q()$", "qʰ%1")
			end
			if last then -- make the utterance-final vowel (short/long/diphthong) breathy if stressed, voiceless if unstressed
				text = text:gsub("+R?$", function(vowels)
					vowels = vowels:gsub("","%1"..(text.accent == #text and "X" or "H"))
					return vowels
				end)
			end
			return text
		end
		
		local results = {}
		for _, term in ipairs(args) do
			local phonemic, phonetic = {}, {}
			local words = split(term, " ", true)
			for i,word in ipairs(words) do
				local syllabified = pron(word)
				phonemic = combine(syllabified)
				phonetic = combine(phon(syllabified,i==#words))
			end
			phonemic = concat(phonemic, " ")
			phonetic = concat(phonetic, " ")
			insert(results, { pron = "/" .. phonemic .. "/" })
			if phonemic ~= phonetic then
				insert(results, { pron = "" })
			end
		end
		
		data.items = results
	end
}

return export