--[=[Lua support for the {{lang}}, {{lang-xx}}, and {{transl}} templates and replacement of various supporting templates.]=]require('strict');localp={};localinitial_style_state;-- set by lang_xx_normal() and lang_xx_italic()localgetArgs=require('Module:Arguments').getArgs;locallang_name_table=mw.loadData('Module:Language/name/data');localsynonym_table=mw.loadData('Module:Lang/ISO 639 synonyms');-- ISO 639-2/639-2T code translation to 639-1 codelocallang_data=mw.loadData('Module:Lang/data');-- language name override and transliteration tool-tip tableslocalnamespace=mw.title.getCurrentTitle().namespace;-- used for categorizationlocalmaint_cats={};-- maintenance categories go herelocalmaint_msgs={};-- and their messages go here--[[--------------------------< 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.]]localfunctionis_set(var)returnnot(var==nilorvar=='');end--[[--------------------------< I S _ L A T N >----------------------------------------------------------------Returns true if all of text argument is written using Latn script for letters, numbers and punctuationset; false else.]]p.is_latn=require('Module:Unicode data').is_Latin--[[--------------------------< I N V E R T _ I T A L I C S >-------------------------------------------------This function attempts to invert the italic markup a args.text by adding/removing leading/trailing italic markupin args.text. Like |italic=unset, |italic=invert disables automatic italic markup. Individual leading/trailingapostrophes are converted to their html numeric entity equivalent so that the new italic markup doesn't becomebold markup inadvertently.Leading and trailing wiki markup is extracted from args.text into separate table elements. Addition, removal,replacement of wiki markup is handled by a string.gsub() replacement table operating only on these separate elements.In the string.gsub() matching pattern, '.*' matches empty string as well as the three expected wiki markup patterns.This function expects that markup in args.text is complete and correct; if it is not, oddness may result.]]localfunctioninvert_italics(source)localinvert_pattern_table={-- leading/trailing markup add/remove/replace patterns[""]="\'\'",-- empty string becomes italic markup["\'\'"]="",-- italic markup becomes empty string["\'\'\'"]="\'\'\'\'\'",-- bold becomes bold italic["\'\'\'\'\'"]="\'\'\'",-- bold italic become bold};localseg={};source=source:gsub("%f[\']\'%f[^\']",'&');-- protect single quote marks from being interpreted as bold markupseg[1]=source:match('^(\'\'+%f[^\']).+')or'';-- get leading markup, if any; ignore single quoteseg[3]=source:match('.+(%f[\']\'\'+)$')or'';-- get trailing markup, if any; ignore single quoteif''~=seg[1]and''~=seg[3]then-- extract the 'text'seg[2]=source:match('^\'\'+%f[^\'](.+)%f[\']\'\'+$')-- from between leading and trailing markupelseif''~=seg[1]thenseg[2]=source:match('^\'\'+%f[^\'](.+)')-- following leading markupelseif''~=seg[3]thenseg[2]=source:match('(.+)%f[\']\'\'+$')-- preceding trailing markupelseseg[2]=source-- when there is no markupendseg[1]=seg[1]:gsub(".*",invert_pattern_table,1);-- replace leading markup according to pattern tableseg[3]=seg[3]:gsub(".*",invert_pattern_table,1);-- replace leading markup according to pattern tablereturntable.concat(seg);-- put it all back together and doneend--[[--------------------------< V A L I D A T E _ I T A L I C >------------------------------------------------validates |italic= or |italics= assigned values.When |italic= is set and has an acceptible assigned value, return the matching css font-style property value or,for the special case 'default', return nil.When |italic= is not set, or has an unacceptible assigned value, return nil and a nil error message.When both |italic= and |italics= are set, returns nil and a 'conflicting' error message.The return value nil causes the calling lang, lang_xx, or transl function to set args.italic according to the template'sdefined default ('inherit' for {{lang}}, 'inherit' or 'italic' for {{lang-xx}} depending onthe individual template's requirements, 'italic' for {{transl}}) or to the value appropriate to |script=, if set ({{lang}}and {{lang-xx}} only).Accepted values and the values that this function returns are are:nil-when |italic= absent or not set; returns nildefault-for completeness, should rarely if ever be used; returns nilyes-force args.text to be rendered in italic font; returns 'italic'no-force args.text to be rendered in normal font; returns 'normal'unset-disables font control so that font-style applied to text is dictated by markup inside or outside the template; returns 'inherit'invert-disables font control so that font-style applied to text is dictated by markup outside or inverted inside the template; returns 'invert']]localfunctionvalidate_italic(args)localproperties={['yes']='italic',['no']='normal',['unset']='inherit',['invert']='invert',['default']=nil};localcount=0for_,arginpairs{'italic','italics','i'}doifargs[arg]thencount=count+1endendifcount>1then-- return nil and an error message if more than one is setreturnnil,'only one of |italic=, |italics=, or |i= can be specified';endreturnproperties[args.italicorargs.italicsorargs.i],nil;-- return an appropriate value and a nil error messageend--[=[--------------------------< V A L I D A T E _ C A T _ A R G S >----------------------------------------------------------Default behavior of the {{lang}} and {{lang-xx}} templates is to add categorization when the templates are used in mainspace.This default functionality may be suppressed by setting |nocat=yes or |cat=no. This function selects one of these two parametersto control categorization.Because having two parameters with 'opposite' names and 'opposite' values is confusing, this function accepts only affirmativevalues for |nocat= and only negative values for |cat=; in both cases the 'other' sense (and non-sense) is not accepted and theparameter is treated as if it were not set in the template.Sets args.nocat to true if categorization is to be turned off; to nil if the default behavior should apply.Accepted values for |nocat= are the text strings:'yes', 'y', 'true', 't', on, '1'-- [[Module:Yesno]] returns logical true for all of these; false or nil elsefor |cat='no', 'n', 'false', 'f', 'off', '0'-- [[Module:Yesno]] returns logical false for all of these; true or nil else]=]localfunctionvalidate_cat_args(args)ifnot(args.nocatorargs.cat)then-- both are nil, so categorizereturn;endlocalyesno=require"Module:Yesno";iffalse==yesno(args.cat)ortrue==yesno(args.nocat)thenargs.nocat=true;-- set to true when args.nocat is affirmative; nil else (as if the parameter were not set in the template)else-- args.nocat is the parameter actually used.args.nocat=nil;endend--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------Whether needle is in haystack]]localfunctionin_array(needle,haystack)ifneedle==nilthenreturnfalse;endforn,vinipairs(haystack)doifv==needlethenreturnn;endendreturnfalse;end--[[--------------------------< F O R M A T _ I E T F _ T A G >------------------------------------------------prettify ietf tags to use recommended subtag formats:code: lower casescript: sentence caseregion: upper casevariant: lower case]]localfunctionformat_ietf_tag(code,script,region,variant)localout={};table.insert(out,code:lower());ifis_set(script)thenscript=script:lower():gsub('^%a',string.upper);table.insert(out,script);endifis_set(region)thentable.insert(out,region:upper());endifis_set(variant)thentable.insert(out,variant:lower());endreturntable.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 codescript subtag - four character IANA script coderegion subtag - two-letter or three digit IANA region codevariant subtag - four digit or 5-8 alnum variant codeprivate subtag - x- followed by 1-8 alnum private code; only supported with the primary language tagin any one of these formslanglang-variantlang-scriptlang-script-variantlang-regionlang-region-variantlang-script-regionlang-script-region-variantlang-x-privateeach of lang, script, region, variant, and private, when used, must be validLanguages with both two- and three-character code synonyms are promoted to the two-character synonym becausethe IANA registry file omits the synonymous three-character code; we cannot depend on browsers understandingthe synonymous three-character codes in the lang= attribute.For {{lang-xx}} templates, the parameters |script=, |region=, and |variant= are supported (not supported in {{lang}}because those parameters are superfluous to the IETF subtags in |code=)returns six values; all lower case. Valid parts are returned as themselves; omitted parts are returned as empty strings, invalidparts are returned as nil; the sixth returned item is an error message (if an error detected) or nil.see http://www.rfc-editor.org/rfc/bcp/bcp47.txt section 2.1]]localfunctionget_ietf_parts(source,args_script,args_region,args_variant)localcode;localscript='';localregion='';localvariant='';localprivate='';ifnotis_set(source)thenreturnnil,nil,nil,nil,nil,'missing language tag';endifsource:match('^%a%a%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)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%d%d%d%-%d%d%d%d$')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)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%a%a%-%w%w%w%w%w+$')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)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%d%d%d%-%w%w%w%w%w+$')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)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%d%d%d%d$')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)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%w%w%w%w%w+$')then-- ll-Ssss-variant (where variant is 5-8 alnum characters)code,script,variant=source:match('^(%a%a%a?)%-(%a%a%a%a)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%a%a%-%d%d%d%d$')then-- ll-RR-variant (where variant is 4 digits)code,region,variant=source:match('^(%a%a%a?)%-(%a%a)%-(%d%d%d%d)$');elseifsource:match('^%a%a%a?%-%d%d%d%-%d%d%d%d$')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)$');elseifsource:match('^%a%a%a?%-%a%a%-%w%w%w%w%w+$')then-- ll-RR-variant (where variant is 5-8 alnum characters)code,region,variant=source:match('^(%a%a%a?)%-(%a%a)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%d%d%d%-%w%w%w%w%w+$')then-- ll-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)code,region,variant=source:match('^(%a%a%a?)%-(%d%d%d)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%d%d%d%d$')then-- ll-variant (where variant is 4 digits)code,variant=source:match('^(%a%a%a?)%-(%d%d%d%d)$');elseifsource:match('^%a%a%a?%-%w%w%w%w%w+$')then-- ll-variant (where variant is 5-8 alnum characters)code,variant=source:match('^(%a%a%a?)%-(%w%w%w%w%w%w?%w?%w?)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%a%a$')then-- ll-Ssss-RRcode,script,region=source:match('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$');elseifsource:match('^%a%a%a?%-%a%a%a%a%-%d%d%d$')then-- ll-Ssss-DDD (region is 3 digits)code,script,region=source:match('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$');elseifsource:match('^%a%a%a?%-%a%a%a%a$')then-- ll-Sssscode,script=source:match('^(%a%a%a?)%-(%a%a%a%a)$');elseifsource:match('^%a%a%a?%-%a%a$')then-- ll-RRcode,region=source:match('^(%a%a%a?)%-(%a%a)$');elseifsource:match('^%a%a%a?%-%d%d%d$')then-- ll-DDD (region is 3 digits)code,region=source:match('^(%a%a%a?)%-(%d%d%d)$');elseifsource:match('^%a%a%a?$')then-- llcode=source:match('^(%a%a%a?)$');elseifsource:match('^%a%a%a?%-x%-%w%w?%w?%w?%w?%w?%w?%w?$')then-- ll-x-pppppppp)code,private=source:match('^(%a%a%a?)%-x%-(%w+)$');elsereturnnil,nil,nil,nil,nil,table.concat({'unrecognized language tag: ',source});-- don't know what we got but it is malformedendcode=code:lower();-- ensure that we use and return lower case version of thisifnot(lang_data.override[code]orlang_name_table.lang[code])thenreturnnil,nil,nil,nil,nil,table.concat({'unrecognized language code: ',code});-- invalid language code, don't know about the others (don't care?)endifsynonym_table[code]then-- if 639-2/639-2T code has a 639-1 synonymtable.insert(maint_cats,table.concat({'Lang and lang-xx code promoted to ISO 639-1|',code}));table.insert(maint_msgs,table.concat({'code: ',code,' promoted to code: ',synonym_table[code]}));code=synonym_table[code];-- use the synonymendifis_set(script)thenifis_set(args_script)thenreturncode,nil,nil,nil,nil,'redundant script tag';-- both code with script and |script= not allowedendelsescript=args_scriptor'';-- use args.script if providedendifis_set(script)thenscript=script:lower();-- ensure that we use and return lower case version of thisifnotlang_name_table.script[script]thenreturncode,nil,nil,nil,nil,table.concat({'unrecognized script: ',script,' for code: ',code});-- language code ok, invalid script, don't know about the others (don't care?)endendiflang_name_table.suppressed[script]then-- ensure that code-script does not use a suppressed scriptifin_array(code,lang_name_table.suppressed[script])thenreturncode,nil,nil,nil,nil,table.concat({'script: ',script,' not supported for code: ',code});-- language code ok, script is suppressed for this codeendendifis_set(region)thenifis_set(args_region)thenreturncode,nil,nil,nil,nil,'redundant region tag';-- both code with region and |region= not allowedendelseregion=args_regionor'';-- use args.region if providedendifis_set(region)thenregion=region:lower();-- ensure that we use and return lower case version of thisifnotlang_name_table.region[region]thenreturncode,script,nil,nil,nil,table.concat({'unrecognized region: ',region,' for code: ',code});endendifis_set(variant)thenifis_set(args_variant)thenreturncode,nil,nil,nil,nil,'redundant variant tag';-- both code with variant and |variant= not allowedendelsevariant=args_variantor'';-- use args.variant if providedendifis_set(variant)thenvariant=variant:lower();-- ensure that we use and return lower case version of thisifnotlang_name_table.variant[variant]then-- make sure variant is validreturncode,script,region,nil,nil,table.concat({'unrecognized variant: ',variant});end-- does this duplicate/replace tests in lang() and lang_xx()?ifis_set(script)then-- if script set it must be part of the 'prefix'ifnotin_array(table.concat({code,'-',script}),lang_name_table.variant[variant]['prefixes'])thenreturncode,script,region,nil,nil,table.concat({'unrecognized variant: ',variant,' for code-script pair: ',code,'-',script});endelseifnotin_array(code,lang_name_table.variant[variant]['prefixes'])thenreturncode,script,region,nil,nil,table.concat({'unrecognized variant: ',variant,' for code: ',code});endendendifis_set(private)thenprivate=private:lower();-- ensure that we use and return lower case version of thisifnotlang_data.override[table.concat({code,'-x-',private})]then-- make sure private tag is valid; note that indexreturncode,script,region,nil,nil,table.concat({'unrecognized private tag: ',private});endendreturncode,script,region,variant,private,nil;-- return the good bits; make sure that msg is nilend--[[--------------------------< M A K E _ E R R O R _ M S G >--------------------------------------------------assembles an error message from template name, message text, help link, and error category.]]localfunctionmake_error_msg(msg,args,template)localout={};localcategory;if'transl'==templatethencategory='transl';elsecategory='lang and lang-xx'endtable.insert(out,table.concat({'[',args.textor'undefined','] '}));-- for error messages output args.text if availabletable.insert(out,table.concat({'<span style=\"font-size:100%; font-style:normal;\" class=\"error\">error: {{',template,'}}: '}));table.insert(out,msg);table.insert(out,table.concat({' ([[:Category:',category,' template errors|help]])'}));table.insert(out,'</span>');if(0==namespace)andnotargs.nocatthen-- only categorize in article spacetable.insert(out,table.concat({'[[Category:',category,' template errors]]'}));endreturntable.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 onlylink is provided, returns a wikilink in the form [[L]]; if neither are provided or link is omitted, returns anempty string.]=]localfunctionmake_wikilink(link,display)ifis_set(link)thenifis_set(display)thenreturntable.concat({'[[',link,'|',display,']]'});elsereturntable.concat({'[[',link,']]'});endelsereturn'';endend--[[--------------------------< M A K E _ T E X T _ S P A N >--------------------------------------------------TODO: add support for block: div tags instead of span tags; would need some sort of proper parameter to control the switchFor italic style, can't do ''{{lang|xx|text}}'' without using <span/> tags when text is italic because of -Latn, |italic=yes,or auto-italics because the wrapping wikimarkup produces this:<i><i lang="xx">text</i></i>which is later reduced to this:<i>text</i>This reduction happens in some sort of cleanup process outside the scope of this template/module.Until or unless this is fixed italic text must be:<i><span lang="xx">text</span></i>]]localfunctionmake_text_span(code,text,rtl,style,size,language)localspan={};localstyle_added='';iftext:match('^%*')thentable.insert(span,'*');-- move proto language text prefix outside of italic markup if any; use numeric entity because plan splat confuses MediaWikitext=text:gsub('^%*','');-- remove the splat from the textendif'italic'==stylethentable.insert(span,'<i>');-- open italic style tagendtable.insert(span,table.concat({'<span lang="'}));-- open <span> tagtable.insert(span,table.concat({code,'\"'}));-- add language attributeifrtlthentable.insert(span,' dir="rtl"');-- add direction attribute for right to left languagesendif'normal'==stylethen-- when |italic=notable.insert(span,' style=\"font-style:normal;');-- override external markup, if anystyle_added='\"';-- remember that style attribute added and is not yet closedendifis_set(size)then-- when |size=<something>ifis_set(style_added)thentable.insert(span,table.concat({' font-size:',size,';'}));-- add when style attribute already insertedelsetable.insert(span,table.concat({' style=\"font-size:',size,';'}));-- create style attributestyle_added='\"';-- remember that style attribute added and is not yet closedendendifis_set(language)thentable.insert(span,table.concat({style_added,' title=\"',language}));--start the title textiflanguage:find('languages')thentable.insert(span,' collective text\"');-- for collective languageselsetable.insert(span,' language text\"');-- for individual languagesendtable.insert(span,'>');-- close the opening span tagelsetable.insert(span,table.concat({style_added,'>'}));-- close the style attribute and close opening span tagendtable.insert(span,text);-- insert the texttable.insert(span,'</span>');-- close the <span> tagif'italic'==stylethentable.insert(span,'</i>');-- close italic style tagendifrtlthen-- legacy; shouldn't be necessary because all of the rtl text is wrapped in <span dir="rtl">text</span>table.insert(span,'‎');-- make sure the browser knows that we're at the end of the rtlendreturntable.concat(span);-- put it all together and doneend--[=[-------------------------< M A K E _ C A T E G O R Y >----------------------------------------------------For individual language, <language>, returns:[[Category:Articles containing <language>-language text]]for English:[[Category:Articles containing explicitly cited English-language text]]for artificial languages (code: art)[[Category:Articles containing constructed-language text]]for ISO 639-2 collective languages (and for 639-1 bh):[[Category:Articles with text from the <language> languages collective]]]=]localfunctionmake_category(code,language_name,nocat)localcat={};if(0~=namespace)ornocatthen-- only categorize in article spacereturn'';-- return empty string for concatenationendiflanguage_name:find('languages')thenreturntable.concat({'[[Category:Articles with text from the ',language_name,' collective]]'});endtable.insert(cat,'[[Category:Articles containing ');if'en'==codethentable.insert(cat,'explicitly cited English');elseif'art'==codethentable.insert(cat,'constructed')elsetable.insert(cat,language_name);endtable.insert(cat,'-language text]]');returntable.concat(cat);end--[[--------------------------< M A K E _ T R A N S L I T >----------------------------------------------------return translit <i lang=xx-Latn>...</i> where xx is the language code; else return empty stringThe value |script= is not used in {{transl}} for this purpose; instead it uses |code. Because language scriptsare listed in the {{transl}} switches they are included in the data tables. The script parameter is introducedat {{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 renamedto be |translit-script= (in this function, tscript)This function is used by both lang_xx() and transl()lang_xx() always provides code, language_name, and translit; may provide tscript; never provides styletransl() always provides language_name, translit, and one of code or tscript, never both; always provides styleFor {{transl}}, style only applies when a language code is provided]]localfunctionmake_translit(code,language_name,translit,std,tscript,style)localtitle;localtout={};localtitle_table=lang_data.translit_title_table;-- table of transliteration standards and the language codes and scripts that apply to those standardsifis_set(code)then-- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})ifnotstylethen-- nil for is the default italic styletable.insert(tout,"<i lang=\"");-- so use <i> tagelsetable.insert(tout,table.concat({'<span style=\"font-style:',style,'\" lang=\"'}));-- non-standard style, construct a span tag for itendtable.insert(tout,code);table.insert(tout,"-Latn\" title=\"");-- transliterations are always Latin scriptelsetable.insert(tout,"<span title=\"");-- when no language code: no lang= attribute, not italic ({{transl}} only)endstd=stdandstd:lower();-- lower case for table indexingifnotis_set(std)andnotis_set(tscript)then-- when neither standard nor script specifiedtable.insert(tout,language_name);-- write a generic tool tipifnotlanguage_name:find('languages')then-- collective language names (plural 'languages' is part of the name)table.insert(tout,'-language')-- skip this text (individual and macro languages only)endtable.insert(tout,' transliteration');-- finish the tool tipelseifis_set(std)andis_set(tscript)then-- when both are specifiediftitle_table[std]then-- and if standard is legitimateiftitle_table[std][tscript]then-- and if script for that standard is legitimatetable.insert(tout,table.concat({title_table[std][tscript:lower()],' (',lang_name_table.script[tscript][1],' script) transliteration'}));-- add the appropriate text to the tool tipelsetable.insert(tout,title_table[std]['default']);-- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?endelsereturn'';-- invalid standard, setup for error messageendelseifis_set(std)then-- translit-script not set, use language codeifnottitle_table[std]thenreturn'';end-- invalid standard, setup for error messageiftitle_table[std][code]then-- if language code is in the table (transl may not provide a language code)table.insert(tout,table.concat({title_table[std][code:lower()],' (',lang_name_table.lang[code][1],' language) transliteration'}));-- add the appropriate text to the tool tipelse-- code doesn't matchtable.insert(tout,title_table[std]['default']);-- so use the standard's defaultendelse-- here if translit-script set but translit-std not setiftitle_table['no_std'][tscript]thentable.insert(tout,title_table['no_std'][tscript]);-- use translit-script if setelseiftitle_table['no_std'][code]thentable.insert(tout,title_table['no_std'][code]);-- use language codeelseifis_set(tscript)thentable.insert(tout,table.concat({language_name,'-script transliteration'}));-- write a script tool tipelseifis_set(code)thenifnotlanguage_name:find('languages')then-- collective language names (plural 'languages' is part of the name)table.insert(tout,'-language')-- skip this text (individual and macro languages only)endtable.insert(tout,' transliteration');-- finish the tool tipelsetable.insert(tout,' transliteration');-- generic tool tip (can we ever get here?)endendendtable.insert(tout,'">');table.insert(tout,translit);ifis_set(code)andnotstylethen-- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})table.insert(tout,"</i>");-- close the italic tagelsetable.insert(tout,"</span>");-- no language code so close the span tagendreturntable.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 anerror 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.texteither as ''itself''' or as '''''bold italic''''' unless |italic=unset or |italic=invert.]=]localfunctionvalidate_text(template,args)ifnotis_set(args.text)thenreturnmake_error_msg('no text',args,template);endifargs.text:find("%f[\']\'\'\'\'%f[^\']")orargs.text:find("\'\'\'\'\'[\']+")then-- because we're looking, look for 4 appostrophes or 6+ appostrophesreturnmake_error_msg('text has malformed markup',args,template);endlocalstyle=args.italic;if('unset'~=style)and('invert'~=style)thenifargs.text:find("%f[\']\'\'%f[^\']")orargs.text:find("%f[\']\'\'\'\'\'%f[^\']")then-- italic but not bold, or bold italicreturnmake_error_msg('text has italic markup',args,template);endendend--[[--------------------------< R E N D E R _ M A I N T >------------------------------------------------------render mainenance messages and categories]]localfunctionrender_maint(nocat)localmaint={};if0<#maint_msgsthen-- when there are maintenance messagestable.insert(maint,table.concat({'<span class="lang-comment" style="font-style:normal; display:none; color:#33aa33; margin-left:0.3em">'}));-- opening <span> tagfor_,msginipairs(maint_msgs)dotable.insert(maint,table.concat({msg,' '}));-- add message stringsendtable.insert(maint,'</span>');-- close the spanendif(0<#maint_cats)and(0==namespace)andnotnocatthen-- when there are maintenance categories; article namespace onlyfor_,catinipairs(maint_cats)dotable.insert(maint,table.concat({'[[Category:',cat,']]'}));-- format and add the categoriesendendreturntable.concat(maint);end--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------for proto languages, text is prefixed with a splat. We do that here as a flag for make_text_span() so that a splatwill be rendered outside of italic markup (if used). If the first character in text here is already a splat, wedo nothing]]localfunctionproto_prefix(text,language_name)iflanguage_name:find('^Proto%-')andnottext:find('^*')then-- language is a proto and text does not already have leading splatreturntable.concat({'*',text});-- prefix proto language text with a splatendreturntext;end--[[--------------------------< L A N G >----------------------------------------------------------------------entry point for {{lang}}there should be no reason to set parameters in the {{lang}} {{#invoke:}}<includeonly>{{#invoke:lang|lang}}</includeonly>parameters are recieved from the template's frame (parent frame)]]functionp.lang(frame)localargs=getArgs(frame);localout={};locallanguage_name;-- used to make category nameslocalsubtags={};-- IETF subtags script, region, variant, and privatelocalcode;-- the language codelocalmsg;-- for error messagesifargs[1]andargs.codethenreturnmake_error_msg('conflicting: {{{1}}} and |code=',args,'lang');elseargs.code=args[1]orargs.code;-- prefer args.codeendifargs[2]andargs.textthenreturnmake_error_msg('conflicting: {{{2}}} and |text=',args,'lang');elseargs.text=args[2]orargs.text;-- prefer args.textendmsg=validate_text('lang',args);-- ensure that |text= is set (italic test disabled for the time being)ifis_set(msg)then-- msg is an already-formatted error messagereturnmsg;endvalidate_cat_args(args);-- determine if categorization should be suppressedargs.rtl=args.rtl=='yes';-- convert to boolean: 'yes' -> true, other values -> falsecode,subtags.script,subtags.region,subtags.variant,subtags.private,msg=get_ietf_parts(args.code);-- |script=, |region=, |variant= not supported because they should be part of args.code ({{{1}}} in {{lang}})ifmsgthenreturnmake_error_msg(msg,args,'lang');endargs.italic,msg=validate_italic(args);ifmsgthenreturnmake_error_msg(msg,args,'lang');endifnil==args.italicthen-- nil when |italic= absent or not set or |italic=default; args.italic controlsif('latn'==subtags.script)or(p.is_latn(args.text)and'en'~=code)then-- script set to latn or text is wholly latn script but not rendering Englishargs.italic='italic';-- DEFAULT for {{lang}} templates is upright; but if latn script set for font-style:italicelseargs.italic='inherit';-- italic not set; script not latn; inherit current styleendendifis_set(subtags.script)then-- if script set, override rtl settingifin_array(subtags.script,lang_data.rtl_scripts)thenargs.rtl=true;-- script is an rtl scriptelseargs.rtl=false;-- script is not an rtl scriptendendargs.code=format_ietf_tag(code,subtags.script,subtags.region,subtags.variant);-- format to recommended subtag styles; private omitted because privateifis_set(subtags.private)andlang_data.override[table.concat({code,'-x-',subtags.private})]then-- look for private use tags; done this way because ...language_name=lang_data.override[table.concat({code,'-x-',subtags.private})][1];-- ... args.code does not get private subtagelseiflang_data.override[code]then-- get the language name for categorizationlanguage_name=lang_data.override[code][1]-- prefer language names taken from the override tableelseiflang_name_table.lang[code]thenlanguage_name=lang_name_table.lang[code][1];-- table entries sometimes have multiple names, always take the first oneendif'invert'==args.italicthenargs.text=invert_italics(args.text)endargs.text=proto_prefix(args.text,language_name);-- prefix proto-language text with a splattable.insert(out,make_text_span(args.code,args.text,args.rtl,args.italic,args.size,language_name));table.insert(out,make_category(code,language_name,args.nocat));table.insert(out,render_maint(args.nocat));-- maintenance messages and categoriesreturntable.concat(out);-- put it all together and doneend--[[--------------------------< L A N G _ X X >----------------------------------------------------------------For the {{lang-xx}} templates, the only parameter required to be set in the template is the language code. Allother parameters can, usually should, be written in the template call. For {{lang-xx}} templates for languagesthat can have multiple writing systems, it may be appropriate to set |script= as well.For each {{lang-xx}} template choose the appropriate entry-point function so that this function know the defaultstyling that should be applied to text.For normal, upright style:<includeonly>{{#invoke:lang|lang_xx_inherit|code=xx}}</includeonly>For italic style:<includeonly>{{#invoke:lang|lang_xx_italic|code=xx}}</includeonly>All other parameters should be received from the template's frame (parent frame)Supported parameters are:|code = (required) the IANA language code|script = IANA script code; especially for use with languages that use multiple writing systems; yields to the script subtag in |code= if present [not yet implemented]|region = IANA region code|variant = IANA variant code|text = (required) the displayed text in language specified by code|link = boolean false ('no') unlinks language specified by code to associated language article|rtl = boolean true ('yes') identifies the language specified by code as a right-to-left language|nocat = boolean true ('yes') inhibits normal categorization; error categories are not affected|italic = boolean true ('yes') renders displayed text in italic font; boolean false ('no') renders displayed text in normal font; not set renders according to initial_style_state|lit = text that is a literal translation of text|label ='none' to suppress all labeling (language name, 'translit.', 'lit.')any other text replaces language-name label - automatic wikilinking disabledfor those {{lang-xx}} templates that support transliteration (those template where |text= is entirely latn script):|translit = text that is a transliteration of text|translit-std = the standard that applies to the transliteration|translit-script = ISO 15924 script name; falls back to codeFor {{lang-xx}}, the positional parameters are:{{{1}}}text{{{2}}}transliterated text{{{3}}}literal translation textno other positional parameters are allowed]]localfunction_lang_xx(frame)localargs=getArgs(frame,{parentFirst=true});-- parameters in the template override parameters set in the {{#invoke:}}localout={};locallanguage_name;-- used to make display text, article linkslocalcategory_name;-- same as language_name except that it retains any parenthetical disambiguators (if any) from the data setlocalsubtags={};-- IETF subtags script, region, and variantlocalcode;-- the language codelocaltranslit_script_name;-- name associated with IANA (ISO 15924) script codelocaltranslit;localtranslit_title;localmsg;-- for error messagesifargs[1]andargs.textthenreturnmake_error_msg('conflicting: {{{1}}} and |text=',args,'lang-xx');elseargs.text=args[1]orargs.text;-- prefer args.textendmsg=validate_text('lang-xx',args);-- ensure that |text= is set, does not contain italic markup and is protected from improper boldingifis_set(msg)thenreturnmsg;endifargs[2]andargs.translitthenreturnmake_error_msg('conflicting: {{{2}}} and |translit=',args,'lang-xx');elseargs.translit=args[2]orargs.translit-- prefer args.translitendifargs[3]and(args.translationorargs.lit)thenreturnmake_error_msg('conflicting: {{{3}}} and |lit= or |translation=',args,'lang-xx');elseifargs.translationandargs.litthenreturnmake_error_msg('conflicting: |lit= and |translation=',args,'lang-xx');elseargs.translation=args[3]orargs.translationorargs.lit;-- prefer args.translationendifargs.linksandargs.linkthenreturnmake_error_msg('conflicting: |links= and |link=',args,'lang-xx');elseargs.link=args.linkorargs.links;-- prefer args.linkendvalidate_cat_args(args);-- determine if categorization should be suppressedargs.rtl=args.rtl=='yes';-- convert to boolean: 'yes' -> true, other values -> falsecode,subtags.script,subtags.region,subtags.variant,subtags.private,msg=get_ietf_parts(args.code,args.script,args.region,args.variant);-- private omitted because privateifmsgthen-- if an error detected then there is an error messagereturnmake_error_msg(msg,args,'lang-xx');endargs.italic,msg=validate_italic(args);ifmsgthenreturnmake_error_msg(msg,args,'lang-xx');endifnil==args.italicthen-- args.italic controlsifis_set(subtags.script)thenif'latn'==subtags.scriptthenargs.italic='italic';-- |script=Latn; set for font-style:italicelseargs.italic=initial_style_state;-- italic not set; script is not latn; set for font-style:<initial_style_state>endelseargs.italic=initial_style_state;-- here when |italic= and |script= not set; set for font-style:<initial_style_state>endendifis_set(subtags.script)then-- if script set override rtl settingifin_array(subtags.script,lang_data.rtl_scripts)thenargs.rtl=true;-- script is an rtl scriptelseargs.rtl=false;-- script is not an rtl scriptendendargs.code=format_ietf_tag(code,subtags.script,subtags.region,subtags.variant);-- format to recommended subtag stylesifis_set(subtags.private)andlang_data.override[table.concat({code,'-x-',subtags.private})]then-- look for private use tags; done this way because ...language_name=lang_data.override[table.concat({code,'-x-',subtags.private})][1];-- ... args.code does not get private subtagelseiflang_data.override[args.code:lower()]then-- look for whole IETF tag in override tablelanguage_name=lang_data.override[args.code:lower()][1];-- args.code:lower() because format_ietf_tag() returns mixed caseelseiflang_data.override[code]then-- not there so try basic language codelanguage_name=lang_data.override[code][1];elseifnotis_set(subtags.variant)theniflang_name_table.lang[code]thenlanguage_name=lang_name_table.lang[code][1];-- table entries sometimes have multiple names, always take the first oneendelse-- TODO: is this the right thing to do: take language display name from variants table?iflang_name_table.variant[subtags.variant]then-- TODO: there is some discussion at Template talk:Lang about having a label parameter for use when variant name is not desired among other thingslanguage_name=lang_name_table.variant[subtags.variant]['descriptions'][1];-- table entries sometimes have multiple names, always take the first oneendendcategory_name=language_name;-- category names retain IANA parenthetical diambiguators (if any)language_name=language_name:gsub('%s+%b()','');-- remove IANA parenthetical disambiguators or qualifiers from names that have themifargs.labelthenif'none'~=args.labelthentable.insert(out,table.concat({args.label,': '}));-- custom labelendelseif'no'==args.linkthentable.insert(out,language_name);-- language name without wikilinkelseiflanguage_name:find('languages')thentable.insert(out,make_wikilink(language_name));-- collective language name uses simple wikilinkelsetable.insert(out,make_wikilink(language_name..' language',language_name));-- language name with wikilinkendendtable.insert(out,': ');-- separatorendif'invert'==args.italicthenargs.text=invert_italics(args.text)endargs.text=proto_prefix(args.text,language_name);-- prefix proto-language text with a splattable.insert(out,make_text_span(args.code,args.text,args.rtl,args.italic,args.size))ifis_set(args.translit)andnotp.is_latn(args.text)then-- transliteration (not supported in {{lang}}); not supported when args.text is wholly latn text (this is an imperfect test)table.insert(out,', ');-- comma to separate text from translitif'none'~=args.labelthentable.insert(out,'<small>');iflang_name_table.script[args['translit-script']]then-- when |translit-script= is set, try to use the script's nametranslit_script_name=lang_name_table.script[args['translit-script'][1]];elsetranslit_script_name=language_name;-- fall back on language nameendtranslit_title=mw.title.makeTitle(0,table.concat({'Romanization of ',translit_script_name}));-- make a title objectiftranslit_title.existsand('no'~=args.link)thentable.insert(out,make_wikilink('Romanization of '..translit_script_nameorlanguage_name,'translit.'));-- make a wikilink if there is an article to link toelsetable.insert(out,'<abbr title="transliteration">translit.</abbr>');-- else define the abbreviationendtable.insert(out,' </small>');-- close the small tagendtranslit=make_translit(args.code,language_name,args.translit,args['translit-std'],args['translit-script'])ifis_set(translit)thentable.insert(out,translit);elsereturnmake_error_msg(table.concat({'invalid translit-std:\'',args['translit-std']or'[missing]'}),args,'lang-xx');endendifis_set(args.translation)then-- translation (not supported in {{lang}})table.insert(out,', ');if'none'~=args.labelthentable.insert(out,'<small>');if'no'==args.linkthentable.insert(out,'<abbr title="literal translation">lit.</abbr>');elsetable.insert(out,make_wikilink('Literal translation','lit.'));endtable.insert(out," </small>");endtable.insert(out,table.concat({''',args.translation,'''}));-- use html entities to avoid wiki markup confusionendtable.insert(out,make_category(code,category_name,args.nocat));table.insert(out,render_maint(args.nocat));-- maintenance messages and categoriesreturntable.concat(out);-- put it all together and doneend--[[--------------------------< L A N G _ X X _ I T A L I C >--------------------------------------------------Entry point for those {{lang-xx}} templates that call lang_xx_italic(). Sets the initial style state to italic.]]functionp.lang_xx_italic(frame)initial_style_state='italic';return_lang_xx(frame);end--[[--------------------------< L A N G _ X X _ I N H E R I T >------------------------------------------------Entry point for those {{lang-xx}} templates that call lang_xx_inherit(). Sets the initial style state to inherit.]]functionp.lang_xx_inherit(frame)initial_style_state='inherit';return_lang_xx(frame);end--[[--------------------------< N A M E _ F R O M _ C O D E >--------------------------------------------------Returns language name associated with IETF language tag if valid; empty string else.All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.Set invoke's |link= parameter to yes to get wikilinked version of the language name.]]functionp.name_from_code(frame)localsubtags={};-- IETF subtags script, region, variant, and privatelocalraw_code=frame.args[1];-- save a copy of the input IETF subtaglocallink='yes'==frame.args['link'];-- save a copy of the link-enable positional parameter (value can be anything)localcode;-- the language codelocalmsg;-- gets an error message if IETF language tag is malformed or invalidlocallanguage_name='';code,subtags.script,subtags.region,subtags.variant,subtags.private,msg=get_ietf_parts(raw_code);ifmsgthenlocaltemplate=(frame.args['template']andtable.concat({'{{',frame.args['template'],'}}: '}))or'';-- make template name (if provided by the template)returntable.concat({'<span style=\"font-size:100%; font-style:normal;\" class=\"error\">error: ',template,msg,'</span>'});endiflang_data.override[raw_code:lower()]then-- look for whole IETF tag in override table (force lower case)language_name=lang_data.override[raw_code:lower()][1];elseiflang_data.override[code]then-- not there so try basic language code in override tablelanguage_name=lang_data.override[code][1];elseifnotis_set(subtags.variant)theniflang_name_table.lang[code]thenlanguage_name=lang_name_table.lang[code][1];-- table entries sometimes have multiple names, always take the first oneendelse-- TODO: is this the right thing to do: take language display name from variants table?iflang_name_table.variant[subtags.variant]then-- TODO: there is some discussion at Template talk:Lang about having a label parameter for use when variant name is not desired among other thingslanguage_name=lang_name_table.variant[subtags.variant]['descriptions'][1];-- table entries sometimes have multiple names, always take the first oneendendlanguage_name=language_name:gsub('%s+%b()','');-- remove IANA parenthetical disambiguators or qualifiers from names that have themiflinkthen-- when |link=yes, wikilink the language nameiflanguage_name:find('languages')thenreturnmake_wikilink(language_name);-- collective language name uses simple wikilinkelsereturnmake_wikilink(language_name..' language',language_name);-- language name with wikilinkendendreturnlanguage_name;end--[[--------------------------< T R A N S L >------------------------------------------------------------------Entry point for {{transl}}]]functionp.transl(frame)localargs=getArgs(frame);-- no {{#invoke:}} parameterslocaltitle_table=lang_data.translit_title_table;-- table of transliteration standards and the language codes and scripts that apply to those standardslocallanguage_name;-- language name that matches language code; used for tool tiplocaltranslit;-- translitterated text to displaylocalscript;-- IANA scriptlocalmsg;-- for when called functions return an error messageifis_set(args[3])then-- [3] set when {{transl|code|standard|text}}args.text=args[3];-- get the transliterated textargs.translit_std=args[2]andargs[2]:lower();-- get the standard; lower case for table indexingifnottitle_table[args.translit_std]thenreturnmake_error_msg(table.concat({'unrecognized transliteration standard: ',args.translit_std}),args,'transl');endelseifis_set(args[2])then-- [2] set when {{transl|code|text}}args.text=args[2];-- get the transliterated textelseifargs[1]andargs[1]:match('^%a%a%a?%a?$')then-- args[2] missing; is args[1] a code or its it the transliterated text?returnmake_error_msg('no text',args,'transl');-- args[1] is a code so we're missing textelseargs.text=args[1];-- args[1] is not a code so we're missing that; assign args.text for error messagereturnmake_error_msg('missing language / script code',args,'transl');endendendifis_set(args[1])then-- IANA language code used for html lang= attribute; or ISO 15924 script codeifargs[1]:match('^%a%a%a?%a?$')then-- args[1] has correct form?args.code=args[1]:lower();-- use the language/script code; only (2, 3, or 4 alpha characters); lower case because table indexes are lower caseelsereturnmake_error_msg(table.concat({'unrecognized language / script code: ',args[1]}),args,'transl');-- invalid language / script codeendelsereturnmake_error_msg('missing language / script code',args,'transl');-- missing language / script code so quitendargs.italic,msg=validate_italic(args);ifmsgthenreturnmake_error_msg(msg,args,'transl');endif'italic'==args.italicthen-- 'italic' when |italic=yes; because that is same as absent or not set and |italic=defaultargs.italic=nil;-- set to nil;endiflang_data.override[args.code]then-- is code a language code defined in the override table?language_name=lang_data.override[args.code][1];elseiflang_name_table.lang[args.code]then-- is code a language code defined in the standard language code tables?language_name=lang_name_table.lang[args.code][1];elseiflang_name_table.script[args.code]then-- if here, code is not a language code; is it a script code?language_name=lang_name_table.script[args.code][1];script=args.code;-- code was an ISO 15924 script so use that insteadargs.code='';-- unset because not a language codeelsereturnmake_error_msg(table.concat({'unrecognized language / script code: ',args.code}),args,'transl');-- invalid language / script codeend-- here only when all parameters passed to make_translit() are validreturnmake_translit(args.code,language_name,args.text,args.translit_std,script,args.italic);endreturnp;
