Module:languages

Documentation for this module may be created at Module:languages/doc

local export = {}

-- Temporarily convert various formatting characters to PUA to prevent them from being disrupted by the substitution process.
-- TODO: Handle arbitrary number of capture groups.
local function doTempSubstitutions(text)
	local subbedChars, u = {}, mw.ustring.char
	local patterns = {
		"((" .. u(0x7F) .. "'\"`UNIQ%-%-%l+%-%x+%-QINU`\"'" .. u(0x7F) .. "))", -- Strip marker
		"((<[^<>]+>))", -- HTML tag
		"('*(''').-'*('''))", -- Bold
		"('*('').-'*(''))" -- Italics
	}
	local i, pe = 0, require("Module:utilities").pattern_escape
	for _, pattern in ipairs(patterns) do
		for m1, m2, m3 in text:gmatch(pattern) do
			local m, m1New = {m1, m2, m3}, m1
			for j = 2, #m do
				subbedChars[i+j-1] = m[j]
				m1New = m1New:gsub(pe(m[j]), u(0xFF700+i+j-1), 1)
			end
			text = text:gsub(pe(m1), m1New, 1)
			i = i + #m - 1
		end
	end
	return text, subbedChars
end

-- Reinsert any formatting that was temporarily substituted.
local function undoTempSubstitutions(text, subbedChars)
	local u = mw.ustring.char
	for i = 1, #subbedChars do
		text = text:gsub(u(0xFF700+i), subbedChars[i])
	end
	return text
end

-- Convert any percent encoding and HTML entities.
local function noEncoding(text)
	if text:find("&[^;]+;") or text:find("%%") then
		text = mw.uri.decode(text)
		return require("Module:utilities").get_entities(text)
	else
		return text
	end
end

-- Check if the raw text is an unsupported title, and if so return that. otherwise, remove percent encoding and HTML entities. We do the pre-conversion to avoid loading the unsupported title list unnecessarily.
local function checkNoEncoding(text)
	local textNoEnc = noEncoding(text)
	if textNoEnc ~= text and mw.loadData("Module:links/data").unsupported_titles[text] then
		return text
	else
		return textNoEnc
	end
end

-- If no script object is provided (or if it's invalid or None), get one.
local function checkScript(text, self, sc)
	if type(sc) ~= "table" or sc._type ~= "script object" or sc:getCode() == "None" then
		return self:findBestScript(text)
	else
		return sc
	end
end

local function normalize(text, sc)
	text = sc:fixDiscouragedSequences(text)
	return sc:toFixedNFD(text)
end

local function doSubstitutions(text, self, sc, substitution_data, function_name, recursed)
	-- If there are language-specific substitutes given in the data module, use those.
	if type(substitution_data) == "table" then
		-- If a script is specified, run this function with the script-specific data before continuing.
		local sc_code = sc:getCode()
		if substitution_data[sc_code] then
			text = doSubstitutions(text, self, sc, substitution_data[sc_code], function_name, true)
		-- Hant, Hans and Hani are usually treated the same, so add a special case to avoid having to specify each one separately.
		elseif substitution_data.Han and (sc_code == "Hant" or sc_code == "Hans" or sc_code == "Hani") then
			text = doSubstitutions(text, self, sc, substitution_data.Han, function_name, true)
		-- Substitution data with key 1 in the outer table may be given as a fallback.
		elseif substitution_data[1] then
			text = doSubstitutions(text, self, sc, substitution_data[1], function_name, true)
		end
		-- Iterate over all strings in the "from" subtable, and gsub with the corresponding string in "to". We work with the NFD decomposed forms, as this simplifies many substitutions.
		if substitution_data.from then
			local gsub
			for i, from in ipairs(substitution_data.from) do
				-- We normalize each loop, to ensure multi-stage substitutions work correctly.
				text = sc:toFixedNFD(text)
				-- Check whether specific magic characters are present, as they rely on UTF-8 compatibility. If not, just use string.gsub. In most cases, doing this is faster than using mw.ustring.gsub every time.
				if from:match("[%%.[%]*+%-?]") then gsub = mw.ustring.gsub else gsub = string.gsub end
				text = gsub(text, sc:toFixedNFD(from), substitution_data.to[i] or "")
			end
		end
		
		if substitution_data.remove_diacritics then
			text = sc:toFixedNFD(text)
			 -- Convert any specified exceptions into PUA characters, to avoid having diacritics stripped. Uses the supplemetary PUA planes (U+FXXXX & U+10XXXX), to ensure that any characters in the BMP (U+0XXXX) or SMP (U+1XXXX) can be round-trip converted to PUA.
			-- This will need to be reviewed if any characters in the SIP (U+2XXXX) or TIP (U+3XXXX) need to be processed by it, but as these planes are exclusively CJK characters as of 2022, this is unlikely to happen for the time being. However, it is unwise to start using non-PUA codepoints in the U+4XXXX-U+EXXXX range, as support for these is completely untested, so they may result in unpredictable behaviour.
			if substitution_data.remove_exceptions then
				local u, codepoint, len, substitute = mw.ustring.char, mw.ustring.codepoint, mw.ustring.len
				for _,exception in ipairs(substitution_data.remove_exceptions) do
					exception = sc:toFixedNFD(exception)
					substitute = {codepoint(exception, 1, len(exception))}
					for i, codepoint in ipairs(substitute) do substitute[i] = u(codepoint+0xF0000) end
					text = text:gsub(exception, table.concat(substitute))
				end
			end
			-- Strip diacritics. This must use mw.ustring.gsub, to ensure the character class is UTF-8 compatible.
			text = mw.ustring.gsub(text, "[" .. substitution_data.remove_diacritics .. "]", "")
			-- Convert any exceptions back.
			if substitution_data.remove_exceptions then
				for _,exception in ipairs(substitution_data.remove_exceptions) do
					local u, codepoint, len, substitute = mw.ustring.char, mw.ustring.codepoint, mw.ustring.len
					exception = sc:toFixedNFD(exception)
					substitute = {codepoint(exception, 1, len(exception))}
					for i, codepoint in ipairs(substitute) do substitute[i] = u(codepoint+0xF0000) end
					text = text:gsub(table.concat(substitute), exception)
				end
			end
		end
	elseif type(substitution_data) == "string" then
		-- If there is a dedicated function module, use that.
		local is_module, module = pcall(require, "Module:" .. substitution_data)
		if is_module then
			if function_name == "tr" then
				text = module[function_name](text, self:getCode(), sc:getCode())
			else
				text = module[function_name](sc:toFixedNFD(text), self:getCode(), sc:getCode())
			end
		-- If there is matching shared data, run this function again using that as the substitution data.
		else
			local m_shared = mw.loadData("Module:languages/global").shared
			if m_shared[substitution_data] then
				return doSubstitutions(text, self, sc, m_shared[substitution_data], function_name, recursed)
			else
				error("Substitution data does not match any shared data or an existing module.")
			end
		end
	end
	
	-- Don't normalize to NFC if this is the inner loop or if a module returned nil.
	if recursed or not text then
		return text
	else
		-- Fix any discouraged sequences created during the substitution process, and normalize into the final form.
		text = sc:fixDiscouragedSequences(text)
		return sc:toFixedNFC(text)
	end
end

local Language = {}

function Language:getCode()
	return self._code
end

function Language:getCanonicalName()
	return self._rawData[1] or self._rawData.canonicalName
end

function Language:getDisplayForm()
	return self:getCanonicalName()
end

function Language:getOtherNames(onlyOtherNames)
	self:loadInExtraData()
	return require("Module:language-like").getOtherNames(self, onlyOtherNames)
end

function Language:getAliases()
	self:loadInExtraData()
	return self._extraData.aliases or {}
end

function Language:getVarieties(flatten)
	self:loadInExtraData()
	return require("Module:language-like").getVarieties(self, flatten)
end

function Language:getType()
	return self._rawData.type or "regular"
end

function Language:getWikimediaLanguages()
	if not self._wikimediaLanguageObjects then
		local m_wikimedia_languages = require("Module:wikimedia languages")
		self._wikimediaLanguageObjects = {}
		local wikimedia_codes = self._rawData.wikimedia_codes or {self:getCode()}
		
		for _, wlangcode in ipairs(wikimedia_codes) do
			table.insert(self._wikimediaLanguageObjects, m_wikimedia_languages.getByCode(wlangcode))
		end
	end
	
	return self._wikimediaLanguageObjects
end

function Language:getWikipediaArticle()
	if self._rawData.wikipedia_article then
		return self._rawData.wikipedia_article
	elseif self._wikipedia_article then
		return self._wikipedia_article
	elseif self:getWikidataItem() and mw.wikibase then
		self._wikipedia_article = mw.wikibase.sitelink(self:getWikidataItem(), 'enwiki')
	end
	if not self._wikipedia_article then
		self._wikipedia_article = self:getCategoryName():gsub("Creole language", "Creole")
	end
	
	return self._wikipedia_article
end

function Language:makeWikipediaLink()
	return "[[w:" .. self:getWikipediaArticle() .. "|" .. self:getCanonicalName() .. "]]"
end

function Language:getWikidataItem()
	local item = self._rawData[2]
	
	if type(item) == "number" then
		return "Q" .. item
	else
		return item
	end
end

function Language:getScripts()
	if not self._scriptObjects then
		local m_scripts = require("Module:scripts")
		self._scriptObjects = {}
		
		for _, sc in ipairs(self:getScriptCodes()) do
			table.insert(self._scriptObjects, m_scripts.getByCode(sc))
		end
	end
	
	return self._scriptObjects
end

function Language:getScriptCodes()
	local scripts = self._rawData.scripts or self._rawData[4]
	return type(scripts) == "table" and scripts or type(scripts) == "string" and {scripts} or {"None"}
end

-- Find the best script to use, based on the characters of a string. If forceDetect is set, run the detection algorithm even if there's only one possible script; in that case, if the text isn't in the script, the return value will be None.
function Language:findBestScript(text, forceDetect)
	if not text then
		return require("Module:scripts").getByCode("None")
	end
	
	local scripts = self:getScripts()
	
	if not scripts[2] and not forceDetect then
		return scripts[1]
	end
	
	return require("Module:languages/findBestScript")(export, self, text, scripts, forceDetect)
end

function Language:getFamily()
	if self._familyObject then
		return self._familyObject
	end
		
	local family = self._rawData[3] or self._rawData.family
	if family then
		self._familyObject = require("Module:families").getByCode(family)
	end
	
	return self._familyObject
end

function Language:getAncestors()
	if not self._ancestorObjects then
		self._ancestorObjects = {}
		
		if self._rawData.ancestors then
			for _, ancestor in ipairs(self._rawData.ancestors) do
				table.insert(self._ancestorObjects, export.getByCode(ancestor) or require("Module:etymology languages").getByCode(ancestor))
			end
		else
			local fam = self:getFamily()
			local protoLang = fam and fam:getProtoLanguage() or nil
			
			-- For the case where the current language is the proto-language
			-- of its family, we need to step up a level higher right from the start.
			if protoLang and protoLang:getCode() == self:getCode() then
				fam = fam:getFamily()
				protoLang = fam and fam:getProtoLanguage() or nil
			end
			
			while not protoLang and not (not fam or fam:getCode() == "qfa-not") do
				fam = fam:getFamily()
				protoLang = fam and fam:getProtoLanguage() or nil
			end
			
			table.insert(self._ancestorObjects, protoLang)
		end
	end
	
	return self._ancestorObjects
end

local function iterateOverAncestorTree(node, func)
	for _, ancestor in ipairs(node:getAncestors()) do
		if ancestor then
			local ret = func(ancestor) or iterateOverAncestorTree(ancestor, func)
			if ret then
				return ret
			end
		end
	end
end

function Language:getAncestorChain()
	if not self._ancestorChain then
		self._ancestorChain = {}
		local step = #self:getAncestors() == 1 and self:getAncestors()[1] or nil
		
		while step do
			table.insert(self._ancestorChain, 1, step)
			step = #step:getAncestors() == 1 and step:getAncestors()[1] or nil
		end
	end
	
	return self._ancestorChain
end

function Language:hasAncestor(otherlang)
	local function compare(ancestor)
		return ancestor:getCode() == otherlang:getCode()
	end
	
	return iterateOverAncestorTree(self, compare) or false
end

function Language:getCategoryName(nocap)
	local name = self:getCanonicalName()
	
	-- If the name already has "language" in it, don't add it.
	if not name:find("[Ll]anguage$") then
		name = name .. " language"
	end
	if not nocap then
		name = mw.getContentLanguage():ucfirst(name)
	end
	return name
end

function Language:makeCategoryLink()
	return "[[:Category:" .. self:getCategoryName() .. "|" .. self:getDisplayForm() .. "]]"
end

function Language:getStandardCharacters()
	return self._rawData.standardChars
end

local function processCarets(text, pattern, repl)
	local u = mw.ustring.char
	return text
		:gsub("\\\\^", u(0xE000) .. "^")
		:gsub("\\^", u(0xE001))
		:gsub(pattern or "%^", repl or "")
		:gsub(u(0xE000), "\\")
		:gsub(u(0xE001), "^")
end

-- Make the entry name (i.e. the correct page name).
function Language:makeEntryName(text, sc)
	-- Remove bold, soft hyphens, strip markers and HTML tags.
	text = text
		:gsub("('*)'''(.-'*)'''", "%1%2")
		:gsub("­", "")
	text = mw.text.unstrip(text)
		:gsub("<[^<>]+>", "")
	-- Don't remove italics, as that would allow people to use it instead of {{m}} etc.
	
	-- If this is an interwiki link or there's an initial colon, return what we have.
	if mw.loadData("Module:languages/global").interwikis[text:match("^([^:]*):")] then
		return text:find("&[^;]+;") and require("Module:utilities").get_entities(text) or text
	-- Convert any escaped colons.
	else
		text = text:gsub("\\:", ":")
	end
	
	-- Check if the text is an unsupported title (with and without converting percent encoding/HTML entities).
	local textNoEnc, unsupportedTitles = noEncoding(text), mw.loadData("Module:links/data").unsupported_titles
	if unsupportedTitles[text] or unsupportedTitles[textNoEnc] then
		return "Unsupported titles/" .. (unsupportedTitles[text] or unsupportedTitles[textNoEnc])
	else
		text = textNoEnc
	end
	
	sc = checkScript(text, self, sc)
	
	text = normalize(text, sc)
	text = doSubstitutions(text, self, sc, self._rawData.entry_name, "makeEntryName")
	
	-- Remove carets if they are used to capitalize parts of transliterations (unless they have been escaped).
	if not sc:hasCapitalization() and sc:isTransliterated() and text:find("%^") then
		text = processCarets(text)
	end
	
	text = mw.ustring.match(text, "^[¿¡]?(.-[^%s%p].-)%s*[؟?!;՛՜ ՞ ՟？！︖︕।॥။၊་།]?$") or text
	
	-- Convert risky characters to HTML entities, which minimizes interference once returned (e.g. for the entry "sms:a").
	return mw.text.encode(text, "#%%&+/:<=>?@[\\%]_{|}")
end

-- Generates alternative forms using a specified method, and returns them as a table. If no method is specified, returns a table containing only the input term.
function Language:generateForms(text, sc)
	if self._rawData.generate_forms then
		sc = checkScript(text, self, sc)
		return require("Module:" .. self._rawData.generate_forms).generateForms(text, self:getCode(), sc:getCode())
	else
		return {text}
	end
end

function Language:makeSortKey(text, sc)
	-- Remove strip markers and HTML tags.
	text = mw.text.unstrip(text)
		:gsub("<[^<>]+>", "")
	
	text = checkNoEncoding(text)
	
	-- Remove initial hyphens and *.
	text = mw.ustring.gsub(text, "^[-־ـ᠊*]+(.)", "%1")
	
	sc = checkScript(text, self, sc)
	
	text = normalize(text, sc)
	
	-- Remove carets if they are used to capitalize parts of transliterations (unless they have been escaped).
	if not sc:hasCapitalization() and sc:isTransliterated() and text:find("%^") then
		text = processCarets(text)
	end
	
	-- For languages with dotted dotless i, ensure that "İ" is sorted as "i", and "I" is sorted as "ı".
	if self._rawData.dotted_dotless_i then
		text = text
			:gsub(mw.ustring.toNFD("İ"), "i")
			:gsub("I", "ı")
		text = sc:toFixedNFD(text)
	end
	-- Convert to lowercase, make the sortkey, then convert to uppercase. Where the language has dotted dotless i, it is usually not necessary to convert "i" to "İ" and "ı" to "I" first, because "I" will always be interpreted as conventional "I" (not dotless "İ") by any sorting algorithms, which will have been taken into account by the sortkey substitutions themselves. However, if no sortkey substitutions have been specified, then conversion is necessary so as to prevent "i" and "ı" both being sorted as "I".
	text = mw.ustring.lower(text)
	text = doSubstitutions(text, self, sc, self._rawData.sort_key, "makeSortKey")
	if self._rawData.dotted_dotless_i and not self._rawData.sort_key then
		text = text
			:gsub("ı", "I")
			:gsub("i", "İ")
		text = sc:toFixedNFC(text)
	end
	text = mw.ustring.upper(text)
	
	-- Remove parentheses, as long as they are either preceded or followed by something.
	text = text
		:gsub("(.)[()]+", "%1")
		:gsub("[()]+(.)", "%1")
	
	-- Convert risky characters to HTML entities, which minimizes interference once returned (e.g. for the entry "<!-- -->").
	return mw.text.encode(text, "#%%&+/:<=>?@[\\%]_{|}")
end

-- Create the form used as as a basis for display text and transliteration.
local function processDisplayText(text, self, sc, keepCarets, keepColons)
	local subbedChars; text, subbedChars = doTempSubstitutions(text)
	
	text = checkNoEncoding(text)

	sc = checkScript(text, self, sc)
	
	text = normalize(text, sc)
	text = doSubstitutions(text, self, sc, self._rawData.display_text, "makeDisplayText")
	
	-- Remove carets if they are used to capitalize parts of transliterations (unless they have been escaped).
	if not keepCarets and not sc:hasCapitalization() and sc:isTransliterated() and text:find("%^") then
		text = processCarets(text)
	end
	-- Remove any interwiki link prefixes (unless they have been escaped or this has been disabled).
	if text:find(":") and not keepColons then
		local u = mw.ustring.char
		text = text
			:gsub("\\\\:", u(0xE000) .. ":")
			:gsub("\\:", u(0xE001))
		local prefix = text:match("^([^:]*):") or ""
		if mw.loadData("Module:languages/global").interwikis[prefix] and text:find(":") ~= #text then
			-- Remove prefix plus colon.
			text = text:gsub("^" .. prefix .. ":(.*)", "%1")
			-- If, additionally, there's a language code after the interwiki link, strip that too.
			if text:find(":") ~= #text then text = text:gsub("^[^:]*:(.*)", "%1") end
		end
		text = text
			:gsub(u(0xE000), "\\")
			:gsub(u(0xE001), ":")
	end
	
	return text, subbedChars
end

-- Make the display text (i.e. what is displayed on the page).
function Language:makeDisplayText(text, sc, keepColons)
	local subbedChars; text, subbedChars = processDisplayText(text, self, sc, nil, keepColons)
	
	-- Convert risky characters to HTML entities, which minimizes interference once returned (e.g. for the entry "<!-- -->").
	text = mw.text.encode(text, "#%%&+/:<=>?@[\\%]_{|}")
	
	return undoTempSubstitutions(text, subbedChars)
end

function Language:transliterate(text, sc, module_override)
	-- If there is no text, or the language doesn't have transliteration data and there's no override, return nil.
	if not (text and (self._rawData.translit or module_override)) then
		return nil
	end
	-- If the script is not transliteratable (and no override is given), return nil.
	sc = checkScript(text, self, sc)
	if not (sc:isTransliterated() or module_override) then
		return nil
	end
	
	-- Remove any strip markers.
	text = mw.text.unstrip(text)
	-- Get the display text with the keepCarets flag set.
	local subbedChars; text, subbedChars = processDisplayText(text, self, sc, true)
	
	-- Transliterate (using the module override if applicable).
	if module_override then
		text = require("Module:" .. module_override).tr(text, self:getCode(), sc:getCode())
	else
		text = doSubstitutions(text, self, sc, self._rawData.translit, "tr")
	end
	
	-- Incomplete transliterations return nil.
	-- FIXME: Handle transliterations with characters that are in both Latn/Latinx and a transliteratable script (e.g. U+A700-U+A707 are in Latinx and Hani).
	if not text or sc:countCharacters(text) > 1 then
		return nil
	end
	
	-- If the script does not use capitalization, then capitalize any letters of the transliteration which are immediately preceded by a caret (and remove the caret).
	if text and not sc:hasCapitalization() and text:find("%^") then
		text = processCarets(text, "%^([%z\1-\127\194-\244][\128-\191]*)", mw.ustring.upper)
	end
	
	-- Convert risky characters to HTML entities, which minimizes interference once returned.
	text = mw.text.encode(text, "#%%&+/:<=>?@[\\%]_{|}")
	
	-- Track module overrides.
	if module_override then
		require("Module:debug").track("module_override")
	end
	
	return undoTempSubstitutions(text, subbedChars)
end

function Language:overrideManualTranslit()
	return not not self._rawData.override_translit
end

function Language:hasTranslit()
	return not not self._rawData.translit
end

function Language:link_tr()
	return not not self._rawData.link_tr
end

function Language:toJSON()
	local entryNamePatterns = nil
	local entryNameRemoveDiacritics = nil
	
	if self._rawData.entry_name then
		entryNameRemoveDiacritics = self._rawData.entry_name.remove_diacritics
		if self._rawData.entry_name.from then
			entryNamePatterns = {}
			for i, from in ipairs(self._rawData.entry_name.from) do
				table.insert(entryNamePatterns, {from = from, to = self._rawData.entry_name.to[i] or ""})
			end
		end
	end
	
	local ret = {
		ancestors = self._rawData.ancestors,
		canonicalName = self:getCanonicalName(),
		categoryName = self:getCategoryName("nocap"),
		code = self:getCode(),
		entryNamePatterns = entryNamePatterns,
		entryNameRemoveDiacritics = entryNameRemoveDiacritics,
		family = self._rawData[3] or self._rawData.family,
		otherNames = self:getOtherNames(true),
		aliases = self:getAliases(),
		varieties = self:getVarieties(),
		scripts = self._rawData.scripts or self._rawData[4],
		type = self:getType(),
		wikimediaLanguages = self._rawData.wikimedia_codes,
		wikidataItem = self:getWikidataItem(),
	}
	
	return require("Module:JSON").toJSON(ret)
end

-- Do NOT use these methods!
-- All uses should be pre-approved on the talk page!
function Language:getRawData()
	return self._rawData
end

function Language:getRawExtraData()
	self:loadInExtraData()
	return self._extraData
end

Language.__index = Language

function export.getDataModuleName(code)
	if code:find("^%l%l$") then
		return "languages/data2"
	elseif code:find("^%l%l%l$") then
		local prefix = code:sub(1, 1)
		return "languages/data3/" .. prefix
	elseif code:find("^[%l-]+$") then
		return "languages/datax"
	else
		return nil
	end
end


function export.getExtraDataModuleName(code)
	if code:find("^%l%l$") then
		return "languages/extradata2"
	elseif code:find("^%l%l%l$") then
		local prefix = code:sub(1, 1)
		return "languages/extradata3/" .. prefix
	elseif code:find("^[%l-]+$") then
		return "languages/extradatax"
	else
		return nil
	end
end

local function getRawLanguageData(code)
	local modulename = export.getDataModuleName(code)
	return modulename and mw.loadData("Module:" .. modulename)[code] or nil
end

local function getRawExtraLanguageData(code)
	local modulename = export.getExtraDataModuleName(code)
	return modulename and mw.loadData("Module:" .. modulename)[code] or nil
end

function Language:loadInExtraData()
	if not self._extraData then
		-- load extra data from module and assign to _extraData field
		-- use empty table as a fallback if extra data is nil
		self._extraData = getRawExtraLanguageData(self:getCode()) or {}
	end
end

function export.makeObject(code, data)
	if data and data.deprecated then
		require("Module:debug").track {
			"languages/deprecated",
			"languages/deprecated/" .. code
		}
	end
	
	return data and setmetatable({_rawData = data, _code = code, _type = "language object"}, Language) or nil
end

function export.getByCode(code, paramForError, allowEtymLang, allowFamily)
	if type(code) ~= "string" then
		error("The function getByCode expects a string as its first argument, but received " .. (code == nil and "nil" or "a " .. type(code)) .. ".")
	end
	
	local retval = export.makeObject(code, getRawLanguageData(code))
	if not retval and allowEtymLang then
		retval = require("Module:etymology languages").getByCode(code)
	end
	if not retval and allowFamily then
		retval = require("Module:families").getByCode(code)
	end
	if not retval and paramForError then
		require("Module:languages/errorGetBy").code(code, paramForError, allowEtymLang, allowFamily)
	end
	return retval
end

function export.getByName(name, errorIfInvalid)
	local byName = mw.loadData("Module:languages/by name")
	local code = byName.all and byName.all[name] or byName[name]
	
	if not code then
		if errorIfInvalid then
			error("The language name \"" .. name .. "\" is not valid. See [[Wiktionary:List of languages]].")
		else
			return nil
		end
	end
	
	return export.makeObject(code, getRawLanguageData(code))
end

function export.getByCanonicalName(name, errorIfInvalid, allowEtymLang, allowFamily)
	local byName = mw.loadData("Module:languages/canonical names")
	local code = byName and byName[name]

	local retval = code and export.makeObject(code, getRawLanguageData(code)) or nil
	if not retval and allowEtymLang then
		retval = require("Module:etymology languages").getByCanonicalName(name)
	end
	if not retval and allowFamily then
		local famname = name:match("^(.*) languages$")
		famname = famname or name
		retval = require("Module:families").getByCanonicalName(famname)
	end
	if not retval and errorIfInvalid then
		require("Module:languages/errorGetBy").canonicalName(name, allowEtymLang, allowFamily)
	end
	return retval
end

--[[	If language is an etymology language, iterates through parent languages
		until it finds a non-etymology language. ]]
function export.getNonEtymological(lang)
	while lang:getType() == "etymology language" do
		local parentCode = lang:getParentCode()
		lang = export.getByCode(parentCode)
			or require("Module:etymology languages").getByCode(parentCode)
			or require("Module:families").getByCode(parentCode)
	end
	
	return lang
end

-- for backwards compatibility only; modules should require the /error themselves
function export.err(lang_code, param, code_desc, template_tag, not_real_lang)
	return require("Module:languages/error")(lang_code, param, code_desc, template_tag, not_real_lang)
end

return export