Modul:IPA

local export = {}
-- ]

local m_data = mw.loadData('Module:IPA/data') -- ]
local m_symbols = mw.loadData('Module:IPA/data/symbols') -- ]
local m_syllables -- ]; loaded below if needed

local sub = mw.ustring.sub
local find = mw.ustring.find
local gsub = mw.ustring.gsub
local match = mw.ustring.match
local gmatch = mw.ustring.gmatch
local U = mw.ustring.char

function export.format_IPA_full(lang, items, err, separator, sortKey, no_count)
	local IPA_key, key_link, err_text, prefix, IPAs, category
	local hasKey = m_data.langs_with_infopages
	local namespace = mw.title.getCurrentTitle().nsText
	
	if err then
		err_text = '<sup>(<span class="error">' .. err .. '</span>)</sup>'
	--[=[
	else
		if hasKey then
			IPA_key = "Appendix:" .. lang:getCanonicalName() .. " pronunciation"
		else
			IPA_key = "wikipedia:" .. lang:getCanonicalName() .. " phonology"
		end
		
		key_link = "]"
	]=]
	end

	
	-- local prefix = "]<sup>(" .. ( key_link or err_text ) .. ")</sup>:&#32;"
	local prefix = ']' .. (err_text or '') .. ':&#32;'
	
	IPAs = export.format_IPA_multiple(lang, items, separator, no_count)
	
	--[=[
	if lang and (namespace == "" or namespace == "Reconstruction") then
		sortKey = sortKey or lang:makeSortKey(mw.title.getCurrentTitle().text)
		sortKey = sortKey and ("|" .. sortKey) or ""
		category = "]"
	else
		category = ""
	end
	]=]
	category = ""

	return prefix .. IPAs .. category
end

local function determine_repr(pron)
	local repr_mark = {}
	local repr, reconstructed
	
	-- remove initial asterisk before representation marks, used on some Reconstruction pages
	if find(pron, "^%*") then
		reconstructed = true
		pron = sub(pron, 2)
	end
	
	local representation_types = {
		 = { right = '/', type = 'phonemic', },
		 = { right = ']', type = 'phonetic', },
		 = { right = '⟩', type = 'orthographic', },
		 = { type = 'rhyme' },
	}
	
	repr_mark.i, repr_mark.f, repr_mark.left, repr_mark.right = find(pron, '^(.).-(.)$')
	
	local representation_type = representation_types
	
	if representation_type then
		if representation_type.right then
			if repr_mark.right == representation_type.right then
				repr = representation_type.type
			end
		else
			repr = representation_type.type
		end
	else
		repr = nil
	end
	
	return repr, reconstructed
end

local function hasInvalidSeparators(transcription)
	if find(transcription, "%.") then
		return true
	else
		return false
	end
end

function export.format_IPA_multiple(lang, items, separator, no_count)
	local categories = {}
	separator = separator or ', '
	
	-- Format
	if not items then
		if mw.title.getCurrentTitle().nsText == "Sablon" then
			table.insert(items, {pron = "/aɪ piː ˈeɪ/"})
		else
			table.insert(categories, "]")
		end
	end
	
	local bits = {}
	
	for _, item in ipairs(items) do
		local bit = export.format_IPA(lang, item.pron)
		
		if item.pretext then
			bit = item.pretext .. bit
		end
		
		if item.posttext then
			bit = bit .. item.posttext
		end
		
		if item.qualifiers and item.qualifiers then
			bit = require("Module:qualifier").format_qualifier(item.qualifiers) .. " " .. bit
		end

		if item.refs or item.note then
			local refspecs
			if item.note then
				-- FIXME: eliminate item.note in favor of item.refs. Use tracking to find places
				-- that use item.note.
				refspecs = {item.note}
				require("Module:debug").track("IPA/note")
			else
				refspecs = item.refs
			end
			local refs = {}
			if #refspecs > 0 then
				for _, refspec in ipairs(refspecs) do
					if type(refspec) ~= "table" then
						refspec = {text = refspec}
					end
					local refargs
					if refspec.name or refspec.group then
						refargs = {name = refspec.name, group = refspec.group}
					end
					table.insert(refs, mw.getCurrentFrame():extensionTag("ref", refspec.text, refargs))
				end
				bit = bit .. table.concat(refs)
			end
		end

		table.insert(bits, bit)
		
		--]
				The length or gemination symbol should not appear after a syllable break or stress symbol.	]=]
		
		if find(item.pron, "") then
			require("Module:debug").track("IPA/syntax-error")
		end
		
		if lang then
			-- Add syllable count if the language's diphthongs are listed in ].
			-- Don't do this if the term has spaces or a liaison mark (‿).
			if not no_count and mw.title.getCurrentTitle().namespace == 0 then
				m_syllables = m_syllables or require('Module:syllables')
				local langcode = lang:getCode()
				if m_data.langs_to_generate_syllable_count_categories then
					local repr = determine_repr(item.pron)
					local use_it
					if m_data.langs_to_use_phonetic_notation then
						use_it = repr == "phonetic"
					else
						use_it = repr == "phonemic"
					end
					if use_it and not find(item.pron, "") then
						local syllable_count = m_syllables.getVowels(item.pron, lang)
						if syllable_count then
							table.insert(categories, "]")
						end
					end
				end
			end

			if lang:getCode() == "en" and hasInvalidSeparators(item.pron) then
				table.insert(categories, "]")
			end
		end
	end

	local styles = mw.getCurrentFrame():extensionTag('templatestyles', '', {src = 'Modul:scripts/style.css'})

	return styles .. table.concat(bits, separator) .. table.concat(categories)
end

-- Takes an IPA pronunciation and formats it and adds cleanup categories.
function export.format_IPA(lang, pron, split_output)
	local err = {}
	local categories = {}
	
	-- Remove wikilinks, so that wikilink brackets are not misinterpreted as
	-- indicating phonemic transcription
	local str_gsub = string.gsub
	local without_links = str_gsub(pron, '%]+|(]+)%]%]', '%1')
	without_links = str_gsub(without_links, '%]+%]%]', '%1')
	
	-- Detect whether this is a phonemic or phonetic transcription
	local repr, reconstructed = determine_repr(without_links)
	
	if reconstructed then
		pron = sub(pron, 2)
	end
	
	-- If valid, strip the representation marks
	if repr == "phonemic" then
		pron = sub(pron, 2, -2)
		without_links = sub(without_links, 2, -2)
	elseif repr == "phonetic" then
		pron = sub(pron, 2, -2)
		without_links = sub(without_links, 2, -2)
	elseif repr == "orthographic" then
		pron = sub(pron, 2, -2)
		without_links = sub(without_links, 2, -2)
	elseif repr == "rhyme" then
		pron = sub(pron, 2)
		without_links = sub(without_links, 2)
	else
		table.insert(categories, "]")
		-- table.insert(err, "invalid representation marks")
		-- Removed because it's annoying when previewing pronunciation pages.
	end
	
	if pron == "" then
		-- table.insert(categories, "]")
	end
	
	-- Check for obsolete and nonstandard symbols
	for i, symbol in ipairs(m_data.nonstandard) do
		local result
		for nonstandard in gmatch(pron, symbol) do
			if not result then
				result = {}
			end
			table.insert(result, nonstandard)
			table.insert(categories, "]")
		end
		
		if result then
			table.insert(err, "elavult vagy nem sztenderd karakterek (" .. table.concat(result) .. ")")
			break
		end
	end
	
	--[[ Check for invalid symbols after removing the following:
			1. wikilinks (handled above)
			2. paired HTML tags
			3. bolding
			4. italics
			5. HTML entity for space
			6. asterisk at beginning of transcription
			7. comma followed by spacing characters
			8. superscripts enclosed in superscript parentheses		]]
	local found_HTML
	local result = str_gsub(without_links, "<(%a+)*>(+)</%1>",
		function(tagName, content)
			found_HTML = true
			return content
		end)
	result = str_gsub(result, "'''(*)'''", "%1")
	result = str_gsub(result, "''(*)''", "%1")
	result = str_gsub(result, "&+;", "") -- This may catch things that are not valid character entities.
	result = str_gsub(result, "^%*", "")
	result = gsub(result, ",%s+", "")
	result = gsub(result, "⁽+⁾", "")
	result = gsub(result, '', '')
	
	-- VS15
	local vs15_class = ""
	if mw.ustring.find(pron, vs15_class) then
		local vs15 = U(0xFE0E)
		if mw.ustring.find(result, vs15) then
			result = gsub(result, vs15, "")
			pron = mw.ustring.gsub(pron, vs15, "")
		end
		pron = mw.ustring.gsub(pron, "(" .. vs15_class .. ")", "%1" .. vs15)
	end

	if result ~= '' then
		local suggestions = {}
		mw.log(pron, result)
		local namespace = mw.title.getCurrentTitle().namespace
		local category
		if namespace == 0 then
			-- main namespace
			category = "IPA-kiejtések érvénytelen IPA-karakterekkel"
		elseif namespace == 118 then
			-- reconstruction namespace
			category = "IPA-kiejtések érvénytelen IPA-karakterekkel"
		else
			category = "IPA-kiejtések érvénytelen IPA-karakterekkel"
		end
		for character in gmatch(result, ".") do
			local suggestion = m_symbols.suggestions
			if suggestion then
				table.insert(suggestions, character .. " with " .. suggestion)
			end
			table.insert(categories, "]")
		end
		table.insert(err, "érvénytelen IPA-karakterek (" .. result .. ")")
		if suggestions then
			table.insert(err, "replace " .. table.concat(suggestions, ", "))
		end
	end
	
	if found_HTML then
		table.insert(categories, "]")
	end
	
	-- Reference inside IPA template usage
	-- FIXME: Doesn't work; you can't put HTML in module output.
	--if mw.ustring.find(pron, '</ref>') then
	--	table.insert(categories, "]")
	--end
	
	if repr == "phonemic" or repr == "rhyme" then
		if lang and m_data.phonemes then
			local valid_phonemes = m_data.phonemes
			local rest = pron
			local phonemes = {}
			
			while mw.ustring.len(rest) > 0 do
				local longestmatch = ""
				
				if sub(rest, 1, 1) == "(" or sub(rest, 1, 1) == ")" then
					longestmatch = sub(rest, 1, 1)
				else
					for _, phoneme in ipairs(valid_phonemes) do
						if mw.ustring.len(phoneme) > mw.ustring.len(longestmatch) and sub(rest, 1, mw.ustring.len(phoneme)) == phoneme then
							longestmatch = phoneme
						end
					end
				end
				
				if mw.ustring.len(longestmatch) > 0 then
					table.insert(phonemes, longestmatch)
					rest = sub(rest, mw.ustring.len(longestmatch) + 1)
				else
					local phoneme = sub(rest, 1, 1)
					table.insert(phonemes, "<span style=\"color: red\">" .. phoneme .. "</span>")
					rest = sub(rest, 2)
					table.insert(categories, "]")
					require("Module:debug").track("IPA/invalid phonemes/" .. phoneme)
				end
			end
			
			pron = table.concat(phonemes)
		end
		
		if repr == "phonemic" then
			pron = "/" .. pron .. "/"
		else
			pron = "-" .. pron
		end
	elseif repr == "phonetic" then
		pron = ""
	elseif repr == "orthographic" then
		pron = "⟨" .. pron .. "⟩"
	end
	
	if reconstructed then
		pron = "*" .. pron
	end
	
	if err then
		err = '<span class="previewonly error" style="font-size: small;>&#32;' .. table.concat(err, ', ') .. '</span>'
	else
		err = ""
	end
	
	if split_output then -- for use of IPA in links 
		return '<span class="IPA">' .. pron .. '</span>', table.concat(categories), err
	else
		return '<span class="IPA">' .. pron .. '</span>' .. err .. table.concat(categories)
	end
end

function export.example(frame)
	local output = {}
	
	local m_links = require('Module:links')
	local m_languages = require('Module:languages')
	
	table.insert(
		output,
[[
{| class="wikitable"
! Term !! IPA !! Generated X-SAMPA !! Regenerated IPA !! Matched?
]]
	)
	local row =
[[
|-
| link || IPA || XSAMPA || regenerated_IPA || matched
]]
	
	local examples = mw.text.split(frame.args, ",%s*")
	
	local m_XSAMPA = require("Module:IPA/X-SAMPA")
	
	for _, example in pairs(examples) do
		local lang, word = match(example, "(%l%l%l?):(.+) ")
		
		if lang then
			lang = m_languages.getByCode(lang) or error('"' .. lang .. '" is not a valid language code.')
		end
		
		local IPA = match(example, "/+/")
			or match(example, "%]+%]")
			or error('No IPA transcription found in "' .. example .. '".')
		local XSAMPA = m_XSAMPA.IPA_to_XSAMPA(IPA)
		local regenerated_IPA = m_XSAMPA.XSAMPA_to_IPA(XSAMPA)
		
		content = {
			link = lang and word and m_links.full_link{ term = word, lang = lang },
			matched = IPA == regenerated_IPA
				and '<span style="color: green;">yes</span>'
				or '<span style="color: red;">no</span>',
			IPA = '<span class="IPA">' .. IPA .. '</span>',
			XSAMPA = '<code>' .. XSAMPA .. '</code>',
			regenerated_IPA = '<span class="IPA">' .. regenerated_IPA .. '</span>'
		}
		
		local function add_content(item)
			return content or ""
		end
		local row = gsub(row, "+", add_content)
		table.insert(output, row)
	end
	
	table.insert(output, "|}")
	
	return table.concat(output)
end

return export