Module:Lang

This module exists primarily to provide correct html markup for non-English language text where that text is used in the English Wikipedia. It has a secondary purpose of providing correct visual rendering for this non-English text. The module was developed to consolidate processing for {{Lang}}, the {{lang-??}} (most), and {{Transl}} templates into a single source and to use a clearly defined data set extracted from international standards. It provides error checking and reporting to ensure that the html rendered is correct for browsers and screen readers.

Other templates that get language name support from this module are:

• {{Native name}}

Data set

The data set is specified in Module:Language/name/data and its included data modules:

• language tag-and-name data
  • Module:Language/data/iana languages – from IANA language-subtag-registry file
  • Module:Language/data/ISO 639-3 – from sil.org UTF-8 Complete Code Tables Set
  • Module:Language/data/wp languages – sourcing for the data in this module is unknown
• language script data
  • Module:Language/data/iana scripts – from IANA language-subtag-registry file
• language region data
  • Module:Language/data/iana regions – from IANA language-subtag-registry file
• language variant data
  • Module:Language/data/iana variants – from IANA language-subtag-registry file
• suppressed-script data
  • Module:Language/data/iana suppressed scripts – from IANA language-subtag-registry file

The data set also includes supplementary data modules:

• override and other non-standard data
  • Module:Lang/data – For example, when IANA disagrees with ISO 639-3 or when the standard language name is cumbersome or contains information not suitable for use in article space, the override table can hold that language information
  • Module:Lang/ISO 639 synonyms – maps three character ISO 639-2, -2T, -3 codes to two character ISO 639-1 codes – from Codes for the Representation of Names of Languages
  • Module:Lang/name to tag – reversed language tag-and-name data so that the language name is the key and the language tag is the value

Error messages and categorization

Mainspace articles that have errors are placed in appropriate error categories:

• Category:lang and lang-xx template errors (2)
• Category:transl template errors (0)

Maintenance categories:

• Category:Lang and lang-xx code promoted to ISO 639-1 (0)

See also

MoS on foreign words:

• Wikipedia:Manual of Style/Text formatting § Foreign terms

---

--[=[

Lua support for the {{lang}} and {{lang-xx}} templates and replacement of various supporting templates. 

]=]

require('Module:No globals');
local p = {};

local getArgs = require ('Module:Arguments').getArgs;
local lang_name_table = require ('Module:Language/name/data');

local lang_data =  mw.loadData ('Module:Lang/data');							-- language name override and transliteration tool-tip tables

local namespace = mw.title.getCurrentTitle().namespace;							-- used for categorization


--[[--------------------------< I S _ S E T >------------------------------------------------------------------

Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not an empty string.

]]

local function is_set( var )
	return not (var == nil or var == '');
end


--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------

Whether needle is in haystack

]]

local function in_array( needle, haystack )
	if needle == nil then
		return false;
	end
	for n,v in ipairs( haystack ) do
		if v == needle then
			return n;
		end
	end
	return false;
end


--[[--------------------------< F O R M A T _ I E T F _ T A G >------------------------------------------------

prettify ietf tags to use recommended subtag formats:
	code: lower case
	script: sentence case
	region: upper case
	variant: lower case

]]

local function format_ietf_tag (code, script, region, variant)
	local out = {};
	local c;
	
	table.insert (out, code:lower());
	if is_set (script) then
		c = script:match ('^%a'):upper();										-- make script sentence case
		script = script:lower():gsub ('^%a', c, 1);
		table.insert (out, script);
	end

	if is_set (region) then
		table.insert (out, region:upper());
	end
	
	if is_set (variant) then
		table.insert (out, variant:lower());
	end
	
	return table.concat (out, '-');
end


--[[--------------------------< G E T _ I E T F _ P A R T S >--------------------------------------------------

extracts and returns IETF language tag parts:
	primary language subtag (required) - 2 or 3 character IANA language code
	script subtag - four character IANA script code
	region subtag - two-letter or three digit IANA region code
	variant subtag - four digit or 5-8 alnum variant code

in any one of these forms
	lang					lang-variant
	lang-script				lang-script-variant
	lang-region				lang-region-variant
	lang-script-region		lang-script-region-variant
	
each of lang, script, region, and variant, when used, must be valid

returns four values.  Valid parts are returned as themselves; omitted parts are returned as empty strings, invalid
parts are returned as nil.

see http://www.rfc-editor.org/rfc/bcp/bcp47.txt section 2.1

]]

local function get_ietf_parts (source)
	local code;
	local script = '';
	local region = '';
	local variant = '';
	local c;
	
	if not is_set (source) then
		return nil, nil, nil, nil;
	end
	
	if source:match ('^%a+%-%a%a%a%a%-%a%a%-(%d%d%d%d+$') then												-- ll-Ssss-RR-variant (where variant is 4 digits)
		code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-%a%a%a%a%-%d%d%d%-(%d%d%d%d+$') then										-- ll-Ssss-DDD-variant (where region is 3 digits; variant is 4 digits)
		code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-%a%a%a%a%-%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then						-- ll-Ssss-RR-variant (where variant is 5-8 alnum characters)
		code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
	elseif source:match ('^%a+%-%a%a%a%a%-%d%d%d%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then						-- ll-Ssss-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
		code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');

	elseif source:match ('^%a+%-%a%a%a%a%-(%d%d%d%d+$') then												-- ll-Ssss-variant (where variant is 4 digits)
		code, script, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-%a%a%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then							-- ll-Ssss-variant (where variant is 5-8 alnum characters)
		code, script, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');

	elseif source:match ('^%a+%-%a%a%-(%d%d%d%d+$') then													-- ll-RR-variant (where variant is 4 digits)
		code, region, variant = source:match ('^(%a%a%a?)%-(%a%a)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-%d%d%d%-(%d%d%d%d+$') then													-- ll-DDD-variant (where region is 3 digits; variant is 4 digits)
		code, region, variant = source:match ('^(%a%a%a?)%-(%d%d%d)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then								-- ll-RR-variant (where variant is 5-8 alnum characters)
		code, region, variant = source:match ('^(%a%a%a?)%-(%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
	elseif source:match ('^%a+%-%d%d%d%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then								-- ll-DDD-variant (where region is 3 digits; variant is 4 digits)
		code, region, variant = source:match ('^(%a%a%a?)%-(%d%d%d)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');

	elseif source:match ('^%a+%-(%d%d%d%d+)$') then								-- ll-variant (where variant is 4 digits)
		code, variant = source:match ('^(%a%a%a?)%-(%d%d%d%d)$');
	elseif source:match ('^%a+%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then			-- ll-variant (where variant is 5-8 alnum characters)
		code, variant = source:match ('^(%a%a%a?)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');

	elseif source:match ('^%a+%-%a%a%a%a%-%a%a$') then							-- ll-Ssss-RR
		code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$');
	elseif source:match ('^%a+%-%a%a%a%a%-%d%d%d$') then						-- ll-Ssss-DDD (region is 3 digits)
		code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$');

	elseif source:match ('^%a+%-%a%a%a%a$') then								-- ll-Ssss
		code, script = source:match ('^(%a%a%a?)%-(%a%a%a%a)$');

	elseif source:match ('^%a+%-%a%a$') then									-- ll-RR
		code, region = source:match ('^(%a%a%a?)%-(%a%a)$');
	elseif source:match ('^%a+%-%d%d%d$') then									-- ll-DDD (region is 3 digits)
		code, region = source:match ('^(%a%a%a?)%-(%d%d%d)$');

	elseif source:match ('^%a+$') then											-- ll
		code = source:match ('^(%a%a%a?)$');

	else
		return nil, nil, nil, nil;												-- don't know what we got but it is malformed
	end
	
	code = code:lower();														-- ensure that we use and return lower case version of this
	
	if not (lang_data.override[code] or lang_name_table.lang[code]) then
		return nil, nil, nil, nil;												-- invalid language code, don't know about the others (don't care?)
	end
	
	if is_set (script) then
		if not lang_name_table.script[script:lower()] then
			return code, nil, nil, nil;											-- language code ok, invalid script, don't know about the others (don't care?)
		end
	end
	
	if is_set (region) then
		if not lang_name_table.region[region:lower()] then
			return code, script, nil, nil;
		end
	end
	
	if is_set (variant) then
		if not lang_name_table.variant[variant:lower()] then
			return code, script, region, nil;
		end
		if not in_array (code, lang_name_table.variant[variant:lower()]['prefixes']) then
			return code, script, region, nil;
		end
	end

	return code, script, region, variant;										-- return the good bits
end


--[=[-------------------------< M A K E _ E R R O R _ M S G >--------------------------------------------------

]=]

local function make_error_msg (msg, nocat)
	local out = {};
	
	table.insert (out, '<span style="font-size:100%" class="error">error: ');
	table.insert (out, msg);
	table.insert (out, ' ([[:Category:lang and lang-xx template errors|help]])')
	table.insert (out, '</span>');
	
	if (0 == namespace) and not is_set (nocat) then								-- only categorize in article space
		table.insert (out, '[[Category:lang and lang-xx template errors]]');
	end

	return table.concat (out);
end
	

--[=[-------------------------< M A K E _ W I K I L I N K >----------------------------------------------------

Makes a wikilink; when both link and display text is provided, returns a wikilink in the form [[L|D]]; if only
link is provided, returns a wikilink in the form [[L]]; if neither are provided or link is omitted, returns an
empty string.

]=]

local function make_wikilink (link, display)
	if is_set (link) then
		if is_set (display) then
			return table.concat ({'[[', link, '|', display, ']]'});
		else
			return table.concat ({'[[', link, ']]'});
		end
	else
		return '';
	end
end


--[[--------------------------< M A K E _ T E X T _ S P A N >--------------------------------------------------

]]

local function make_text_span (code, text, rtl, italic, size)
	local span = {};

	table.insert (span, '<span lang="');										-- open <span> tag
	table.insert (span,  code);													-- language attribute
	table.insert (span, '"');
	if 'yes' == rtl then
		table.insert (span, ' dir="rtl"');										-- for right to left languages
	end
	if is_set (size) then														-- {{lang}} only
		table.insert (span, table.concat ({' style="font-size:', size, ';"'}))
	end
	table.insert (span, '>');													-- close the opening span tag
	if 'yes' == italic then
		table.insert (span, table.concat ({"''", text, "''"}));					-- text with italic markup
	else
		table.insert (span, text);												-- DEFAULT: text is not italicized
	end
	table.insert (span, '</span>');												-- close the span
	if 'yes' == rtl then
		table.insert (span, '&lrm;');											-- make sure the browser knows that we're at the end of the rtl
	end
	
	return table.concat (span);													-- put it all together and done
end


--[[--------------------------< M A K E _ C A T E G O R Y >----------------------------------------------------

]]

local function make_category (code, language_name, nocat)
	local cat = {};
	
	if (0 ~= namespace) or nocat then											-- only categorize in article space
		return '';																-- return empty string for concatenation
	end
	
	table.insert (cat, '[[Category:Articles containing ');
	
	if ('en' == code) or ('eng' == code) then
		table.insert (cat, 'explicitly cited English');
	elseif 'art' == code then
		table.insert (cat, 'constructed')
	else
		table.insert (cat, language_name);
	end
	
	table.insert (cat, '-language text]]');

	return table.concat (cat);	
end


--[[--------------------------< M A K E _ T R A N S L I T >----------------------------------------------------

return translit <span>...</span> else return empty string

The value |script= is not used in {{transl}} for this purpose; instead it uses |code.  Because language scripts
are listed in the {{transl}} switches they are included in the data tables.  The script parameter is introduced
at {{Language with name and transliteration}}.  If |script= is set, this function uses it in preference to code.

To avoid confusion, in this module and the templates that use it, the transliteration script parameter is renamed
to be |transl-script= (in this function, tscript)
]]

local function make_translit (code, language_name, translit, std, tscript)
	local title;
	local tout = {};
	local title_table = lang_data.translit_title_table;							-- table of transliteration standards and the language codes and scripts that apply to those standards
	
	table.insert (tout, "''<span title=\"");
	
	if not is_set (std) and not is_set (tscript) then							-- when neither standard nor script specified
		table.insert (tout, language_name);										-- write a generic tool tip
		table.insert (tout, ' transliteration');
	elseif is_set (std) and is_set (tscript) then								-- when both are specified
		if title_table[std][tscript] then										-- and legitimate
			table.insert (tout, title_table[std][tscript]);						-- add the appropriate text to the tool tip
		else
			return '';															-- one or both invalid, set up for an error message
		end
	elseif is_set (std) then													-- transl-script not set, use language code
		if not title_table[std] then return ''; end								-- invalid standard, setupt for error message
		
		if title_table[std][code] then
			table.insert (tout, title_table[std][code]);
		else																	-- code doesn't match
			table.insert (tout, title_table[std]['default']);					-- so use the standard's default
		end
	else																		-- here if transl-script set but std not set
		if title_table['NO_STD'][tscript] then
			table.insert (tout, title_table['NO_STD'][tscript]);				-- use transl-script if set
		elseif title_table['NO_STD'][code] then
			table.insert (tout, title_table['NO_STD'][code]);					-- use language code
		else
			table.insert (tout, language_name);									-- write a generic tool tip
			table.insert (tout, ' transliteration');
		end
	end

	table.insert (tout, '" class="Unicode" style="white-space:normal; text-decoration:none">');
	table.insert (tout, translit);
	table.insert (tout, "</span>''");
	return table.concat (tout);
end


--[[--------------------------< V A L I D A T E _ T E X T >---------------------------------------------------

This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
error message.  The tests are for empty or missing text and for improper or disallowed use of apostrophe markup.

Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
either as ''itself''' or as '''''bold italic'''''.

Also protects single leading and trailing single quote marks from being converted to bold by the addition of
adjacent italic markup.

]]

local function validate_text (template, args)
	if not is_set (args.text) then
		return make_error_msg (table.concat ({'{{', template, '}}: no text'}), args.nocat);
	end

	if args.text:find ("\'\'\'\'\'[\']+") then									-- because we're looking, look for 6+ appostrophes
		return make_error_msg (table.concat ({'{{', template, '}}: text has malformed markup'}), args.nocat);
	end

	if args.text:match ("%f[\']\'\'[^\']+\'\'%f[^\']") or args.text:match ("\'\'\'\'\'[^\']+\'\'\'\'\'") then	-- italic but not bold, or bold italic
		return make_error_msg (table.concat ({'{{', template, '}}: text has italic markup'}), args.nocat);
	end

	if args.text:find ("\'\'\'\'") then											-- because we're looking, look for 4 apostrophes 
		return make_error_msg (table.concat ({'{{', template, '}}: text has malformed markup'}), args.nocat);
	end

	if 'yes' == args.italic then												-- protect single quote marks from being converted to bold markup
		args.text = args.text:gsub ("^\'[^\']+", "<span></span>%1");			-- leading single quote mark
		args.text = args.text:gsub ("[^\']+\'$", "%1<span></span>");			-- trailing single quote mark
	end
end


--[[--------------------------< L A N G >----------------------------------------------------------------------

<includeonly>{{#invoke:lang|lang|code={{{1|}}}|text={{{2|}}}|rtl={{{rtl|}}}|italic={{{italic|}}}|size={{{size|}}}|nocat={{{nocat|}}}}}</includeonly>

|code = the BCP47 language code
|text = the displayed text in language specified by code
|rtl = boolean true identifies the language specified by code as a right-to-left language
|size = css keyword appropriate for use with css font-size:<size>
|nocat = boolean true inhibits normal categorization; error categories are not affected

]]

function p.lang (frame)
	local args = getArgs(frame);
	
	local out = {};
	local language_name;
	local code, script, region, variant = get_ietf_parts (args.code);
	
	if not (code and script and region and variant) then
		return make_error_msg (table.concat ({'{{lang}}: unknown language code: ', args.code or 'missing'}), args.nocat);
	end
	
	if not is_set (args.italic) then
		args.italic = 'no';														-- DEFAULT for {{lang}} templates is to not italicize
	end

	local msg = validate_text ('lang', args);
	if is_set (msg) then
		return msg;
	end
		
	args.code = format_ietf_tag (code, script, region, variant);				-- format to recommended subtag styles

	if lang_data.override[code] then
		language_name = lang_data.override[code][1]
	elseif lang_name_table.lang[code] then
		language_name = lang_name_table.lang[code][1];							-- table entries sometimes have multiple names, always take the first one
	end

	table.insert (out, make_text_span (args.code, args.text, args.rtl, args.italic, args.size));
	table.insert (out, make_category (args.code, language_name, args.nocat));
	return table.concat (out);													-- put it all together and done
	end


--[[--------------------------< L A N G _ X X >----------------------------------------------------------------

<includeonly>{{#invoke:lang|lang_xx|code=<code>|script={{{script|}}}|text={{{1|}}}|link={{{links|{{{link}}}}}}|rtl={{{rtl|}}}|nocat={{{nocat|}}}|italic={{{italic|}}}|lit={{{lit|}}}|translit={{{translit|}}}|transl-script={{{transl-script|}}}|std={{{std|}}}}}</includeonly>

|code = (required) the BCP47 language code
|script = BCP47 script name; especially for use with languages that use multiple writing systems; yields to the script subtag in |code= if present [not currently implemented]
|text = (required) the displayed text in language specified by code
|link = boolean true (default) links language specified by code to associated language article
|rtl = boolean true identifies the language specified by code as a right-to-left language
|nocat = boolean true inhibits normal categorization; error categories are not affected
|italic = boolean true (default) renders displayed text in italic font
|lit = text that is a literal translation of text

for those {{lang-xx}} templates that support transliteration:
|translit = text that is a transliteration of text
|std = the standard that applies to the transliteration
|transl-script = ISO 15924 script name; falls back to code

]]

function p.lang_xx (frame)
	local args = getArgs(frame);
	
	if not is_set (args.italic) then
		args.italic = 'yes';													-- DEFAULT for {{lang-xx}} templates is to italicize
	end
	
	args.size = nil;															-- size not supported in {{lang-xx}}
	
	local out = {};
	local language_name;
	local code, script, region, variant = get_ietf_parts (args.code);

	local translit_script;
	local translit;
	local translit_title;
	
	if not (code and script and region and variant) then
		return make_error_msg (table.concat ({'{{lang-xx}}: unknown language code: ', args.code or 'missing'}), args.nocat);
	end
	
	local msg = validate_text ('lang-xx', args);
	if is_set (msg) then
		return msg;
	end

	args.code = format_ietf_tag (code, script, region, variant);				-- format to recommended subtag styles

	if lang_data.override[code] then
		language_name = lang_data.override[code][1]
	elseif lang_name_table.lang[code] then
		language_name = lang_name_table.lang[code][1];							-- table entries sometimes have multiple names, always take the first one
	end

	translit_script = args['transl-script'] or language_name;					-- for translit prefer |trans-script= over language

	if 'no' == args.link then
		table.insert (out, language_name);										-- language name without wikilink
	else
		table.insert (out, make_wikilink (language_name .. ' language', language_name));	-- language name with wikilink
	end
	table.insert (out, ': ');													-- separator

	table.insert (out, make_text_span (args.code, args.text, args.rtl, args.italic, args.size))
	
	if is_set (args.translit) then												-- transliteration (not supported in {{lang}}); not supported in all {{lang-xx}}
		table.insert (out, ', <small>');
		translit_title = mw.title.makeTitle (0, 'Romanization of ' .. language_name)
		if translit_title.exists and ('no' ~= args.link) then
			table.insert (out, make_wikilink ('Romanization of ' .. translit_script or language_name, 'translit.'));
		else
			table.insert (out, '<abbr title="transliteration">translit.</abbr>');
		end
		table.insert (out, '&nbsp;</small>');
		translit = make_translit (args.code, language_name, args.translit, args.std, args['transl-script'])
		if is_set (translit) then
			table.insert (out, translit);
		else
			return make_error_msg (table.concat ({'{{lang-xx}}: invalid translit std: \'', args.std or 'missing', '\' or script: \'', args['transl-script'] or 'missing', '\''}), args.nocat);
		end
	end
	
	if is_set (args.lit) then													-- translation (not supported in {{lang}})
		table.insert (out, ', <small>');
		if 'no' == args.link then
			table.insert (out, '<abbr title="literal translation">lit.</abbr>');
		else
			table.insert (out, make_wikilink ('Literal translation', 'lit.'));
		end
		table.insert (out, "&nbsp;</small>'");
		table.insert (out, args.lit);
		table.insert (out, "'");
	end
	
	table.insert (out, make_category (args.code, language_name, args.nocat));
	return table.concat (out);													-- put it all together and done
end

return p;