Misplaced Pages

Revision as of 13:05, 5 December 2024

This is the module sandbox page for Module:Wikt-lang (diff). See also the companion subpage for test cases (run).

This module is rated as beta, and is ready for widespread use. It is still new and should be used with some caution to ensure the results are as expected.

This module depends on the following other modules: Module:Wikt-lang/data (sandbox) Module:Unicode data (sandbox)

This module is used by {{Wikt-lang}}. It is inspired by the templates {{m}} and {{l}} and their associated modules on Wiktionary. It has a Wiktionary link function that links to the correct section of the Wiktionary entry, and applies correct language formatting and italics. The language-tagging function does most of what {{Lang}} does, except that italics can't be customized and categories aren't added.

The module uses Module:Wikt-lang/data to retrieve the language name for a language code, and to perform the necessary entry-name replacements (for instance, removing macrons from Latin entry names). These are unfortunately not automatically retrieved from Wiktionary's Wikt-lang data modules. For language codes that do not have a name value in this module, the language name is fetched with mw.language.fetchLanguage. When mw.language.fetchLanguage does not fetch the correct language name (or any language name at all), please add it to Module:Wikt-lang/data, and similarly when the correct entry name is not generated, please add the entry name replacements to the module.

Examples

• {{#invoke:Wikt-lang|wiktlang|en|language}} -> language
• {{#invoke:Wikt-lang|wiktlang|en|language|languages}} -> languages
• {{#invoke:Wikt-lang|wiktlang|fr|bourguignon}} -> bourguignon
• {{#invoke:Wikt-lang|wiktlang|la|homō}} -> homō
• {{#invoke:Wikt-lang|wiktlang|grc|δημοκρατίᾱ}} -> δημοκρατίᾱ
• {{#invoke:Wikt-lang|wiktlang|ru|язы́к}} -> язы́к
• {{#invoke:Wikt-lang|wiktlang|ar|العَرَبِيَّة}} -> ‏العَرَبِيَّة‎
• {{#invoke:Wikt-lang|wiktlang|fa|فارسی}} -> فارسی
• {{#invoke:Wikt-lang|wiktlang|ine-x-proto|*h₂enǵʰ-}} -> *h₂enǵʰ-

Invalid codes

• {{#invoke:Wikt-lang|wiktlang|EN|language}} -> language
• {{#invoke:Wikt-lang|wiktlang|En|language|languages}} -> languages
• {{#invoke:Wikt-lang|wiktlang|La|homō}} -> homō
• {{#invoke:Wikt-lang|wiktlang|Grc|δημοκρατίᾱ}} -> δημοκρατίᾱ
• {{#invoke:Wikt-lang|wiktlang|Ru|язы́к}} -> язы́к
• {{#invoke:Wikt-lang|wiktlang|Ar|العَرَبِيَّة}} -> ‏العَرَبِيَّة‎
• {{#invoke:Wikt-lang|wiktlang|Fa|فارسی}} -> فارسی

Errors

• {{#invoke:Wikt-lang|wiktlang|en-Latin|language}} -> language

Comparison of codes

Tracking categories

• Template:Tracking/wikt-lang/reconstructed with no asterisk

require('strict')
local m_data = mw.loadData("Module:Wikt-lang/data/sandbox")
local langData = m_data.languages or m_data

local p = {}

local error_msgs = {
	internal = {
		 = "Name for the language code %q could not be retrieved. Add it to ].",
		 = "The function get_clean_Wiktionary_page_name requires a link string.",
		 = "The function make_Wiktionary_link needs a Wiktionary link, display text and language code.",
	},
	 = "A Wiktionary entry is required.",
}

local cfg = {
	template = "Wikt-lang",
	error_msg = '<span style="color: #d33;">Error: {{%s}}: %s</span>]',
	category = "Wikt-lang template errors",
	appendix = "Appendix:%s/%s",
	reconstruction = "Reconstruction:%s/%s",
	types = {
		appendix = "appendix",
		reconstructed = "reconstructed",
	},
}

--[[--------------------------< M A K E _ E R R O R >--------------------------------------------------

Creates an error span with the supplied error message and attaches the error category.

]]
local function make_error(msg)
	return string.format(cfg.error_msg, cfg.template, msg, cfg.category)
end

--[[--------------------------< R E P L A C E _ C H A R A C T E R S >--------------------------------------------------

Replaces specific characters as defined in Module:Wikt-lang/data in a language's "replacement" value.

]]
local function replace_characters(data, link_text)
	local replacements = data and data
	if replacements == nil then
		-- No replacements needed; use text as is.
		return link_text
	end

	-- Decompose so that the diacritics of characters such
	-- as á can be removed in one go.
	-- No need to compose at the end, because the MediaWiki software
	-- will handle that.
	if replacements.decompose then
		link_text = mw.ustring.toNFD(link_text)
		for i, from in ipairs(replacements.from) do
			link_text = mw.ustring.gsub(
				link_text,
				from,
				replacements.to and replacements.to or "")
		end
		return link_text
	end

	for regex, replacement in pairs(replacements) do
		return mw.ustring.gsub(link_text, regex, replacement)
	end

end

--[[--------------------------< R E M O V E _ B O L D _ I T A L I C >--------------------------------------------------

Removes bold and italics, so that words that contain bolding or emphasis can be linked without piping.

]]
local function remove_bold_italic(link_text)
	if not link_text then
		return link_text
	end
	link_text = link_text:gsub("\'\'\'", "")
	link_text = link_text:gsub("\'\'", "")
	return link_text
end

--[[--------------------------< G E T _ C L E A N _ W I K T I O N A R Y _ P A G E _ N A M E >--------------------------------------------------

Returns a clean a Wiktionary page name by removing bold and italics, and by replacing specific characters as defined in Module:Wikt-lang/data.

]]
local function get_clean_Wiktionary_page_name(link_text, language_code)
	link_text = tostring(link_text)
	if link_text == nil or link_text == "" then
		return nil, make_error(error_msgs.internal.get_clean_Wiktionary_page_name)
	end

	link_text = remove_bold_italic(link_text)

	local data = langData
	if data == nil then
		-- No language specific data in module; use text as is.
		return link_text
	end

	return replace_characters(data, link_text)
end

--[[--------------------------< G E T _ N A M E S P A C E >--------------------------------------------------

Returns the link_text with a prefix of a Wiktionary namespace, if relevant.
Current namespaces that can be returned: "Appendix:" and "Reconstruction:".
If not one of the above namespaces, returns the unalerted link_text.

]]
local function get_namespace(language_name, link_text, data)
	if link_text:sub(1, 1) == "*" then
		return string.format(cfg.reconstruction, language_name, link_text:sub(2))
	end
	if data and data.type then
		if data.type == cfg.types.reconstructed then
			-- TODO: if this is an error, replace with category
			--return nil, make_error("Reconstructed language without asterisk")
	
			local frame = mw.getCurrentFrame()
			mw.log("Reconstructed language without asterisk:", language_name, link_text)
			pcall(frame.expandTemplate, frame, {title = 'tracking/wikt-lang/reconstructed with no asterisk'})
			return string.format(cfg.reconstruction, language_name, link_text)
		elseif data_type == cfg.types.appendix then
			return string.format(cfg.appendix, language_name, link_text)
		end
	end
	
	-- If for any reason this reaches here, return the unaltered link_text.
	return link_text
end

--[[--------------------------< G E T _ L A N G U A G E _ N A M E >--------------------------------------------------

Retrieves the language name.
A langauge is first searched in Module:Wikt-lang/data and if found and has a language name set, returns it.
That database is used to override the language names produced by Module:Lang/data.
If no language is found or the language does not have a language name set, returns the language name from Module:Lang/data.

]]
local function get_language_name(data, language_name)
	if data and data.name then
		return data.name
	end
	return language_name
end
		
--[[--------------------------< M A K E _ W I K T I O N A R Y _ L I N K >--------------------------------------------------
	
Creates a Wiktionary link.
A link_text, display_text and language_code are always needed. Error if they are missing.

A language name can sometimes be nil if the private code is only listed at Module:Wikt-lang/data and not on Module:Lang/data.
If a language name cannot be retrieved, an erorr is returned.

]]
local function make_Wiktionary_link(link_text, display_text, language_code, language_name)
	if not link_text and not display_text and not language_code then
		return nil, make_error(error_msgs.internal.make_Wiktionary_link)
	end

	local data = langData
	language_name = get_language_name(data, language_name)
	if not language_name then
		return make_error(error_msgs.language_code_missing)
	end

	if link_text:sub(1, 1) == "*" or (data and data.type) then
		link_text = get_namespace(language_name, link_text, data)
	end

	local link = "]"
	return string.format(link, link_text, language_name, display_text)
end

--[[--------------------------< W I K T L A N G >--------------------------------------------------------------------

Entry point for {{Wikt-lang}}.

Parameters are received from the template's frame (parent frame).

* |1= – language code
* |2= – link text
* |3= – display text
* |italic= – "no" to disable

]]
function p.wiktlang(frame)
	local getArgs = require('Module:Arguments').getArgs
	local args = getArgs(frame)

	if not args then
		-- A Wiktionary entry is required.
		return make_error(error_msgs.no_text)
	end

	-- The display text should be the text wrapped in the language tag.
	args = args or args

	-- To allow the errors to be associated with this template.
	args.template = cfg.template
	args.error_category = cfg.category

	-- Handle the display text html tag.
	local lang = require("Module:Lang/sandbox2")
	local result = lang._lang(args)

	-- An error returned, stop here.
	if type(result) == "string" and string.find(result, "Error") then
		return result
	end

	--TODO: we need the result to return without a <span title=""> tag which probably should be removed.

	local entry, error_msg = get_clean_Wiktionary_page_name(args, result.code)
	if error_msg then
		return error_msg
	end

	local link
	link, error_msg = make_Wiktionary_link(entry, result.html, result.code, result.name)

	if error_msg then
		return error_msg
	end

	return link .. result.language_categories .. result.maintenance  
end

return p

• Module sandboxes