Documentation for this module may be created at Module:ja-pron/doc

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

local lang = require("Module:languages").getByCode("ja")
local m_ja = require("Module:ja")
local m_accent = require("Module:a")
local m_ipa = require("Module:IPA")

PAGENAME = PAGENAME or mw.title.getCurrentTitle().text

function export.show(frame)
	local params = {
		[1] = {default = PAGENAME, list = true},
		
		["y"] = {alias_of = "yomi"},
		["yomi"] = {},
		
		["acc"] = {list = true},
		["acc=_loc"] = {list = true},
		["acc=_ref"] = {list = true},
		["acc=_note"] = {list = true},
		
		["accent"] = {list = true},
		["accent=_loc"] = {list = true},
		["accent=_ref"] = {list = true},
		["accent=_note"] = {list = true},
		
		["dev"] = {},
		["dev2"] = {},
		["devm"] = {},
		
		["noipa"] = {},
		
		["a"] = {alias_of = "audio"},
		["audio"] = {}
	}
	
	local args = require("Module:parameters").process(frame:getParent().args, params)
	local yomi, au = args.yomi, args.audio
	local dev, dev2, devm = args.dev, args.dev2, args.devm
	local maxindex = table.getn(args[1])
	local result = ""
	local text
	
	-- Deals with the yomi
	if yomi then
		result = "* "
		if yomi == "o" or yomi == "on" then
			result = result .. "[[音読み#Japanese|On'yomi]]"
		elseif yomi == "go" or yomi == "goon" then
			result = result .. "[[音読み#Japanese|On'yomi]]: [[呉音#Japanese|Goon]]"
		elseif yomi == "ko" or yomi == "kan" or yomi == "kanon" then
			result = result .. "[[音読み#Japanese|On'yomi]]: [[漢音#Japanese|Kan'on]]"
		elseif yomi == "ky" or yomi == "kanyo" or yomi == "kanyoon" then
			result = result .. "[[慣用読み#Japanese|Kan'yōyomi]]"
		elseif yomi == "k" or yomi == "kun" then
			result = result .. "[[訓読み#Japanese|Kun'yomi]]"
		elseif yomi == "j" or yomi == "ju" then
			result = result .. "[[重箱読み#Japanese|Jūbakoyomi]]"
		elseif yomi == "y" or yomi == "yu" then
			result = result .. "[[湯桶読み#Japanese|Yutōyomi]]"
		elseif yomi == "i" or yomi == "irr" or yomi == "irreg" then
			result = result .. '<span class="ib-brac"><span class="qualifier-brac">(</span></span>'
			result = result .. '<span class="ib-content"><span class="qualifier-content">Irregular reading</span></span>'
			result = result .. '<span class="ib-brac"><span class="qualifier-brac">)</span></span>'
		end
	end
	
	-- Deals with the accents
	local a, al, ar, an = args.acc, args.acc_loc, args.acc_ref, args.acc_note
	for i,position in ipairs(a) do
		text = args[1][math.min(maxindex,i)]
		result = result .. "\n*"
		if al[i] then
			result = result .. m_accent.show({al[i]}) .. " "
		else
			result = result .. m_accent.show({"[[w:Tokyo dialect|Tokyo]]"}) .. " "
		end
		result = result .. export.accent(text, position, dev, dev2, devm)
		if ar[i] then
			result = result .. frame:expandTemplate{title = "ja-pron/acc-ref", args = {ar[i]}}
		end
		result = result .. (an[i] and (" " .. an[i]) or "")
	end
	
	-- Alternate parameter "accent"
	a, al, ar, an = args.accent, args.accent_loc, args.accent_ref, args.accent_note
	for i,position in ipairs(a) do
		text = args[1][math.min(maxindex,i)]
		result = result .. "\n*"
		if al[i] then
			result = result .. m_accent.show({al[i]}) .. " "
		else
			result = result .. m_accent.show({"[[w:Tokyo dialect|Tokyo]]"}) .. " "
		end
		result = result .. export.accent(text, position, dev, dev2, devm)
		if ar[i] then
			result = result .. frame:expandTemplate{title = "ja-pron/acc-ref", args = {ar[i]}}
		end
		result = result .. (an[i] and (" " .. an[i]) or "")
	end
	
	-- Deals with the IPA
	if not noipa then
		for i,text in ipairs(args[1]) do
			result = result .. "\n* " .. m_ipa.format_IPA_full(lang, {{pron = "[" .. export.ipa(text, dev, dev2, devm) .. "]"}})
		end
	end
	
	-- Deals with the audio
	if au then
		result = result .. "\n* " .. frame:expandTemplate{title = "audio", args = {au, "Audio", lang = "ja"}}
	end
	
	return result
end

function export.ipa(text, dev, dev2, devm)
	if type(text) == "table" then text, dev, dev2, devm = text.args[1], text.args["dev"], text.args["dev2"], text:getParent().args["devm"] end
	dev = dev or ""
	dev2 = dev2 or ""
	devm = devm or ""
	
	if dev2 ~= "" then error('Please remove parameter dev2 and change paramter dev to \"dev=' .. dev .. ',' .. dev2 .. '"') end
	
	local position_kana = {}
	local position_mora = {}
	
	for i=1,mw.ustring.len(text) do
		if sub(text,i,i) ~= ' ' then
			table.insert(position_kana, i)
			if not mw.ustring.match(sub(text,i,i), "[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ]") then
				table.insert(position_mora, i)
			end
		end
	end
	
	if dev ~= "" then
		for position in mw.text.gsplit(dev,",") do
			position = position_kana[tonumber(position)]
			text = sub(text, 1, position) .. "̥" .. sub(text, position + 1, -1)
			for i=position+1,table.getn(position_kana) do
				position_kana[i] = position_kana[i] + 1
			end
		end
	end
	
	if devm ~= "" then
		for position in mw.text.gsplit(devm,",") do
			position = position_mora[tonumber(position)]
			text = sub(text, 1, position) .. "̥" .. sub(text, position + 1, -1)
			for i=position+1,table.getn(position_mora) do
				position_mora[i] = position_mora[i] + 1
			end
		end
	end
	
	text = m_ja.kana_to_romaji(text, nil, true)
	
	text = gsub(text, "[ptkbjgzsd][ptckbjgzsd][hs]?", {
		["pp"] = "p̚p", ["tch"] = "t̚ch", ["kk"] = "k̚k", ["bb"] = "b̚b̥", ["jj"] = "t̚j",
		["dd"] = "d̚d̥", ["gg"] = "g̚g̊", ["zz"] = "t̚z", ["tt"] = "t̚t", ["tts"] = "t̚ts", ["ssh"] = "ɕː" })
	
	text = gsub(text, "ei", "ē")
	text = gsub(text, "[āēīōūnfjyz]", {
		["ā"] = "aː", ["ē"] = "eː", ["ī"] = "iː", ["ō"] = "oː", ["ū"] = "uː", 
		["n"] = "ɴ", ["f"] = "ɸ", ["j"] = "d͡ʑ", ["y"] = "j", ["z"] = "z" })
	
	text = gsub(text, "[sct][hs]", {
		["sh"] = "ɕ", 
		["ch"] = "t͡ɕ", 
		["ts"] = "t͡s", 
		["ss"] = "sː" })
	
	text = gsub(gsub(text, "([^ ])g([^̊̚])", "%1g̃%2"), "([^ ^ɴ])b", "%1β")
	text = gsub(gsub(text, "ɴ([aeioujw%'])", "n%1"), "([aeiou̥])ɴ", "%1̃ɴ")
	text = gsub(text, "([ɕʑ])([^i^ʲ])", "%1ʲ%2")
	text = gsub(gsub(gsub(text, "ɴ([ ]?)([kg])", "ŋ%1%2"), "ɴ([ ]?)([tdsnzrɕ%(])", "n%1%2"), "ɴ([ ]?)([pbm])", "m%1%2")
	
	text = gsub(text, "h[iju]", {
		["hi"] = "çi", ["hj"] = "çj", 
		["hu"] = "ɸu" })
	
	text = gsub(text, "r[aeiou]", {
		["ra"] = "ɾ̠a", 
		["ri"] = "ɾi", 
		["ru"] = "ɾ̠u", 
		["re"] = "ɺ̠e", 
		["ro"] = "ɺ̠o" })
	
	text = gsub(text, "[aeouw]", {
		["a"] = "a̠", 
		["u"] = "ɯᵝ", 
		["e"] = "e̞", 
		["o"] = "o̞", 
		["w"] = "ɰᵝ"})
	
	text = gsub(text, "ᵝ([̥̃])", "%1ᵝ")
	text = gsub(text, "%.", "")
	text = gsub(text, "'", ".")
	text = gsub(text, "g", "ɡ")
	
	return text
end

function export.rise_and_fall(word, rftype)
	word = m_ja.kana_to_romaji(word)
	
	if rftype == "rise" then
		word = gsub(word, ".", {
			["a"] = "á", ["e"] = "é", ["i"] = "í", ["o"] = "ó", ["u"] = "ú", 
			["ā"] = "áá", ["ē"] = "éé", ["ī"] = "íí", ["ō"] = "óó", ["ū"] = "úú" })
	
		word = gsub(gsub(word, "n([bcdfghjkmnprstvw%'z ])", "ń%1"), "n$", "ń")
		
	elseif rftype == "fall" then
		word = gsub(word, ".", {
			["a"] = "à", ["e"] = "è", ["i"] = "ì", ["o"] = "ò", ["u"] = "ù", 
			["ā"] = "àà", ["ē"] = "èè", ["ī"] = "ìì", ["ō"] = "òò", ["ū"] = "ùù" })
		
		word = gsub(gsub(word, "n([bcdfghjkmnprstvw%'z ])", "ǹ%1"), "n$", "ǹ")
		
	else
		return error("Type not recognised.")
	end
	
	return word
end

function export.accent(text, class, dev, dev2, devm)
	if(type(text)) == "table" then text, class, dev, dev2, devm = text.args[1], text.args[2], text.args["dev"], text.args["dev2"], text.args["devm"] end
	text = gsub(text, "%.", "")
	if dev == "" then dev = false end
	if dev2 == "" then dev2 = false end
	if devm == "" then devm = false end
	
	local down_first = "<span style=\"border-top:1px solid black;position:relative;padding:1px;\">"
	local down_last = "<span style=\"position:absolute;top:0;bottom:67%;right:0%;border-right:1px solid black;\">&#8203;</span></span>"
	local high_first = "<span style=\"border-top:1px solid black\">"
	local start = "<span lang=\"ja\" class=\"Jpan\">"
	local romaji_start = " <span class=\"Latn\"><tt>["
	local romaji_last = "]</tt></span> "
	local last = "</span>"
	
	local position_kana = {}       --position of each kana (ぁ counted), text without space
	local position_mora = {}       --position of each mora (ぁ not counted), text without space
	local position_mora_space = {} --position of each mora (ぁ not counted), text with space
	
	for i=1,mw.ustring.len(text) do
		if not mw.ustring.match(sub(text,i,i), "[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ ]") then
			local extra = mw.ustring.len(mw.ustring.match(sub(text,i+1), "^[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ]*"))
			table.insert(position_mora_space, i+extra)
		end
	end
	local space_removed = mw.ustring.gsub(text," ","")
	for i=1,mw.ustring.len(space_removed) do
		table.insert(position_kana, i)
		if not mw.ustring.match(sub(space_removed,i,i), "[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ]") then
			local extra = mw.ustring.len(mw.ustring.match(sub(space_removed,i+1), "^[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ]*")) mw.log(sub(space_removed,i,i),extra)
			table.insert(position_mora, i+extra)
		end
	end
	
	if match(class, "^[h0]$") then
		acc_type, acc_number = "h", 0
		
	elseif match(class, "^[a1]$") then
		acc_type, acc_number = "a", 1
	
	elseif match(class, "^o$") then
		acc_type = "o"
	end
	
	if match(class, "^[0-9]+$") and not match(class,"^[01]$") then
		class = gsub(class, "[on]", "")
		acc_number = tonumber(class)
		
		morae_count = len(gsub(text, "[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ ]", ""))
			
		if morae_count == acc_number then
			acc_type = "o"
		elseif morae_count < acc_number then
			return error("Mora count is smaller than position of downstep mora.")
		else
			acc_type = "n"
		end
	elseif not acc_number then
		acc_number = class
	end
	
	local start_index = 1
	while match(sub(text, start_index+1, start_index+1), "[ァィゥェォャュョヮぁぃぅぇぉゃゅょゎ]") do
		start_index = start_index + 1
	end
	
	local kanas = mw.text.split(gsub(text, " ", ""), "")
	
	local function kana_devoice(text)
		return '<span style="border:1px dotted gray; border-radius:50%;">' .. text .. "</span>"
	end
	
	if dev then
		for position in mw.text.gsplit(dev, ",") do
			position = tonumber(position)
			kanas[position] = kana_devoice(kanas[position])
		end
	end
	
	if devm then
		for position in mw.text.gsplit(devm, ",") do
			position = position_mora[tonumber(position)]
			kanas[position] = kana_devoice(kanas[position])
		end
	end
	
	romaji_text = gsub(text, "([ょおこそとのほもよろごぞどぼぽオコソトノホモヨロゴゾドボポ])う", "%1お")
	romaji_text = gsub(romaji_text, "([えエけケせセてテねネへヘめメれレゑヱげゲぜゼでデべベぺペ])い", "%1え")
	romaji_text = gsub(romaji_text, "([オコソトノホモヨロヲゴゾドボポョォ])ー", "%1オ")
	romaji_text = gsub(romaji_text, "([エケセテネヘメレヱッゲゼデベペェ])ー", "%1エ")
	romaji_text = gsub(romaji_text, "([ウクスツヌフムユルグズヅブプュゥ])ー", "%1ウ")
	romaji_text = gsub(romaji_text, "([イキシチニヒミリヰギジヂビピィ])ー", "%1イ")
	romaji_text = gsub(romaji_text, "([アカサタナハマヤラワンガザダバパャァ])ー", "%1ア")
	
	local romajis = mw.text.split(romaji_text, "")
	
	local function count_nspaces(text, index)
		local i, sample, nspaces = 0, "", 0
		while len(sample) < index do
			i = i + 1
			sample, nspaces = gsub(sub(text, 1, i), " ", "")
		end
		return nspaces
	end
	
	local function romaji_devoice(text)
		return text .. "̥"
	end
	
	if dev then
		for position in mw.text.gsplit(dev,",") do
			position = tonumber(position)
			romajis[position] = romaji_devoice(romajis[position])
		end
	end
	
	if devm then
		for position in mw.text.gsplit(devm,",") do
			position = position_mora_space[tonumber(position)]
			romajis[position] = romaji_devoice(romajis[position])
		end
	end
	
	if acc_type == "n" then
		r_start_index = start_index + count_nspaces(romaji_text, start_index)
		local r_index = position_mora_space[acc_number]
		local k_index = position_mora[acc_number]
		
		r_parts = {
			[1] = table.concat(romajis, "", 1, r_start_index),
			[2] = table.concat(romajis, "", r_start_index + 1, r_index),
			[3] = table.concat(romajis, "", r_index + 1, #romajis)
		}
				
		k_parts = {
			[1] = table.concat(kanas, "", 1, start_index),
			[2] = table.concat(kanas, "", start_index + 1, k_index),
			[3] = table.concat(kanas, "", k_index + 1, #kanas)
		}
		
		result = start .. 
			k_parts[1] .. 
			down_first .. 
			k_parts[2] .. 
			down_last .. 
			k_parts[3] .. 
			last .. 
			romaji_start .. 
			export.rise_and_fall(r_parts[1], "fall") .. 
			export.rise_and_fall(r_parts[2], "rise") .. 
			"ꜜ" .. 
			export.rise_and_fall(r_parts[3], "fall") .. 
			romaji_last .. 
			"([[中高型|Nakadaka]] - [" .. acc_number .. "])"
		
	else
		r_start_index = start_index + count_nspaces(romaji_text, start_index)
		r_parts = {
			[1] = table.concat(romajis, "", 1, r_start_index),
			[2] = table.concat(romajis, "", r_start_index + 1, #romajis)
		}
		
		k_parts = {
			[1] = table.concat(kanas, "", 1, start_index),
			[2] = table.concat(kanas, "", start_index + 1, #kanas)
		}
		
		if acc_type == "h" then
			result = start .. 
				k_parts[1] .. 
				high_first .. 
				k_parts[2] .. 
				last .. 
				last .. 
				romaji_start .. 
				export.rise_and_fall(r_parts[1], "fall") .. 
				export.rise_and_fall(r_parts[2], "rise") .. 
				romaji_last .. 
				"([[平板型|Heiban]] - [" .. acc_number .. "])"
				
 		elseif acc_type == "a" then
			result = start .. 
				down_first .. 
				k_parts[1] .. 
				down_last .. 
				k_parts[2] .. 
				last .. 
				romaji_start .. 
				export.rise_and_fall(r_parts[1], "rise") .. 
				"ꜜ" .. 
				export.rise_and_fall(r_parts[2], "fall") .. 
				romaji_last .. 
				"([[頭高型|Atamadaka]] - [" .. acc_number .. "])"
				
		elseif acc_type == "o" then
			result = start .. 
				k_parts[1] .. 
				down_first .. 
				k_parts[2] .. 
				down_last .. 
				last .. 
				romaji_start .. 
				export.rise_and_fall(r_parts[1], "fall") .. 
				export.rise_and_fall(r_parts[2], "rise") .. 
				"ꜜ" .. 
				romaji_last .. 
				"([[尾高型|Odaka]] - [" .. acc_number .. "])"
				
		else
			return error("Accent type not recognised.")
		end
		
	end
	
	result = gsub(result, "(.)̥", "<del>%1</del>") .. (dev and "[[Category:Kenny's testing category]]" or "")
	
	return result
end

return export