2 * @author Neil Kandalgaonkar, 2010
3 * @author Timo Tijhof, 2011-2013
7 /* eslint-disable no-use-before-define */
11 * Parse titles into an object structure. Note that when using the constructor
12 * directly, passing invalid titles will result in an exception. Use #newFromText to use the
13 * logic directly and get null for invalid titles which is easier to work with.
15 * Note that in the constructor and #newFromText method, `namespace` is the **default** namespace
16 * only, and can be overridden by a namespace prefix in `title`. If you do not want this behavior,
17 * use #makeTitle. Compare:
19 * new mw.Title( 'Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Template:Foo'
20 * mw.Title.newFromText( 'Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Template:Foo'
21 * mw.Title.makeTitle( NS_TEMPLATE, 'Foo' ).getPrefixedText(); // => 'Template:Foo'
23 * new mw.Title( 'Category:Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Category:Foo'
24 * mw.Title.newFromText( 'Category:Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Category:Foo'
25 * mw.Title.makeTitle( NS_TEMPLATE, 'Category:Foo' ).getPrefixedText(); // => 'Template:Category:Foo'
27 * new mw.Title( 'Template:Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Template:Foo'
28 * mw.Title.newFromText( 'Template:Foo', NS_TEMPLATE ).getPrefixedText(); // => 'Template:Foo'
29 * mw.Title.makeTitle( NS_TEMPLATE, 'Template:Foo' ).getPrefixedText(); // => 'Template:Template:Foo'
33 * @param {string} title Title of the page. If no second argument given,
34 * this will be searched for a namespace
35 * @param {number} [namespace=NS_MAIN] If given, will used as default namespace for the given title
36 * @throws {Error} When the title is invalid
38 function Title( title, namespace ) {
39 var parsed = parse( title, namespace );
41 throw new Error( 'Unable to parse title' );
44 this.namespace = parsed.namespace;
45 this.title = parsed.title;
46 this.ext = parsed.ext;
47 this.fragment = parsed.fragment;
52 // eslint-disable-next-line vars-on-top
54 namespaceIds = mw.config.get( 'wgNamespaceIds' ),
61 NS_MAIN = namespaceIds[ '' ],
68 NS_TALK = namespaceIds.talk,
73 * @property NS_SPECIAL
75 NS_SPECIAL = namespaceIds.special,
82 NS_MEDIA = namespaceIds.media,
89 NS_FILE = namespaceIds.file,
94 * @property FILENAME_MAX_BYTES
96 FILENAME_MAX_BYTES = 240,
101 * @property TITLE_MAX_BYTES
103 TITLE_MAX_BYTES = 255,
106 * Get the namespace id from a namespace name (either from the localized, canonical or alias
109 * Example: On a German wiki this would return 6 for any of 'File', 'Datei', 'Image' or
114 * @method getNsIdByName
115 * @param {string} ns Namespace name (case insensitive, leading/trailing space ignored)
116 * @return {number|boolean} Namespace id or boolean false
118 getNsIdByName = function ( ns ) {
121 // Don't cast non-strings to strings, because null or undefined should not result in
122 // returning the id of a potential namespace called "Null:" (e.g. on null.example.org/wiki)
123 // Also, toLowerCase throws exception on null/undefined, because it is a String method.
124 if ( typeof ns !== 'string' ) {
127 // TODO: Should just use local var namespaceIds here but it
128 // breaks test which modify the config
129 id = mw.config.get( 'wgNamespaceIds' )[ ns.toLowerCase() ];
130 if ( id === undefined ) {
138 * @method getNamespacePrefix_
139 * @param {number} namespace
142 getNamespacePrefix = function ( namespace ) {
143 return namespace === NS_MAIN ?
145 ( mw.config.get( 'wgFormattedNamespaces' )[ namespace ].replace( / /g, '_' ) + ':' );
148 rUnderscoreTrim = /^_+|_+$/g,
150 rSplit = /^(.+?)_*:_*(.*)$/,
152 // See MediaWikiTitleCodec.php#getTitleInvalidRegex
153 rInvalid = new RegExp(
154 '[^' + mw.config.get( 'wgLegalTitleChars' ) + ']' +
155 // URL percent encoding sequences interfere with the ability
156 // to round-trip titles -- you can't link to them consistently.
158 // XML/HTML character references produce similar issues.
159 '|&[A-Za-z0-9\u0080-\uFFFF]+;' +
164 // From MediaWikiTitleCodec::splitTitleString() in PHP
165 // Note that this is not equivalent to /\s/, e.g. underscore is included, tab is not included.
166 rWhitespace = /[ _\u00A0\u1680\u180E\u2000-\u200A\u2028\u2029\u202F\u205F\u3000]+/g,
168 // From MediaWikiTitleCodec::splitTitleString() in PHP
169 rUnicodeBidi = /[\u200E\u200F\u202A-\u202E]/g,
172 * Slightly modified from Flinfo. Credit goes to Lupo and Flominator.
175 * @property sanitationRules
184 // control characters
186 // eslint-disable-next-line no-control-regex
187 pattern: /[\x00-\x1f\x7f]/g,
191 // URL encoding (possibly)
193 pattern: /%([0-9A-Fa-f]{2})/g,
197 // HTML-character-entities
199 pattern: /&(([A-Za-z0-9\x80-\xff]+|#[0-9]+|#x[0-9A-Fa-f]+);)/g,
203 // slash, colon (not supported by file systems like NTFS/Windows, Mac OS 9 [:], ext4 [/])
205 pattern: new RegExp( '[' + mw.config.get( 'wgIllegalFileChars', '' ) + ']', 'g' ),
209 // brackets, greater than
215 // brackets, lower than
221 // everything that wasn't covered yet
223 pattern: new RegExp( rInvalid.source, 'g' ),
227 // directory structures
229 pattern: /^(\.|\.\.|\.\/.*|\.\.\/.*|.*\/\.\/.*|.*\/\.\.\/.*|.*\/\.|.*\/\.\.)$/g,
236 * Internal helper for #constructor and #newFromText.
238 * Based on Title.php#secureAndSplit
243 * @param {string} title
244 * @param {number} [defaultNamespace=NS_MAIN]
245 * @return {Object|boolean}
247 parse = function ( title, defaultNamespace ) {
248 var namespace, m, id, i, fragment, ext;
250 namespace = defaultNamespace === undefined ? NS_MAIN : defaultNamespace;
253 // Strip Unicode bidi override characters
254 .replace( rUnicodeBidi, '' )
255 // Normalise whitespace to underscores and remove duplicates
256 .replace( rWhitespace, '_' )
258 .replace( rUnderscoreTrim, '' );
260 // Process initial colon
261 if ( title !== '' && title[ 0 ] === ':' ) {
262 // Initial colon means main namespace instead of specified default
268 .replace( rUnderscoreTrim, '' );
271 if ( title === '' ) {
275 // Process namespace prefix (if any)
276 m = title.match( rSplit );
278 id = getNsIdByName( m[ 1 ] );
279 if ( id !== false ) {
280 // Ordinary namespace
284 // For Talk:X pages, make sure X has no "namespace" prefix
285 if ( namespace === NS_TALK && ( m = title.match( rSplit ) ) ) {
286 // Disallow titles like Talk:File:x (subject should roundtrip: talk:file:x -> file:x -> file_talk:x)
287 if ( getNsIdByName( m[ 1 ] ) !== false ) {
295 i = title.indexOf( '#' );
300 // Get segment starting after the hash
303 // NB: Must not be trimmed ("Example#_foo" is not the same as "Example#foo")
304 .replace( /_/g, ' ' );
309 // Trim underscores, again (strips "_" from "bar" in "Foo_bar_#quux")
310 .replace( rUnderscoreTrim, '' );
313 // Reject illegal characters
314 if ( title.match( rInvalid ) ) {
318 // Disallow titles that browsers or servers might resolve as directory navigation
320 title.indexOf( '.' ) !== -1 && (
321 title === '.' || title === '..' ||
322 title.indexOf( './' ) === 0 ||
323 title.indexOf( '../' ) === 0 ||
324 title.indexOf( '/./' ) !== -1 ||
325 title.indexOf( '/../' ) !== -1 ||
326 title.slice( -2 ) === '/.' ||
327 title.slice( -3 ) === '/..'
333 // Disallow magic tilde sequence
334 if ( title.indexOf( '~~~' ) !== -1 ) {
338 // Disallow titles exceeding the TITLE_MAX_BYTES byte size limit (size of underlying database field)
339 // Except for special pages, e.g. [[Special:Block/Long name]]
340 // Note: The PHP implementation also asserts that even in NS_SPECIAL, the title should
341 // be less than 512 bytes.
342 if ( namespace !== NS_SPECIAL && $.byteLength( title ) > TITLE_MAX_BYTES ) {
346 // Can't make a link to a namespace alone.
347 if ( title === '' && namespace !== NS_MAIN ) {
351 // Any remaining initial :s are illegal.
352 if ( title[ 0 ] === ':' ) {
356 // For backwards-compatibility with old mw.Title, we separate the extension from the
357 // rest of the title.
358 i = title.lastIndexOf( '.' );
359 if ( i === -1 || title.length <= i + 1 ) {
360 // Extensions are the non-empty segment after the last dot
363 ext = title.slice( i + 1 );
364 title = title.slice( 0, i );
368 namespace: namespace,
376 * Convert db-key to readable text.
384 text = function ( s ) {
385 if ( s !== null && s !== undefined ) {
386 return s.replace( /_/g, ' ' );
393 * Sanitizes a string based on a rule set and a filter
399 * @param {Array} filter
402 sanitize = function ( s, filter ) {
403 var i, ruleLength, rule, m, filterLength,
404 rules = sanitationRules;
406 for ( i = 0, ruleLength = rules.length; i < ruleLength; ++i ) {
408 for ( m = 0, filterLength = filter.length; m < filterLength; ++m ) {
409 if ( rule[ filter[ m ] ] ) {
410 s = s.replace( rule.pattern, rule.replace );
418 * Cuts a string to a specific byte length, assuming UTF-8
419 * or less, if the last character is a multi-byte one
423 * @method trimToByteLength
425 * @param {number} length
428 trimToByteLength = function ( s, length ) {
429 var byteLength, chopOffChars, chopOffBytes;
431 // bytelength is always greater or equal to the length in characters
432 s = s.substr( 0, length );
433 while ( ( byteLength = $.byteLength( s ) ) > length ) {
434 // Calculate how many characters can be safely removed
435 // First, we need to know how many bytes the string exceeds the threshold
436 chopOffBytes = byteLength - length;
437 // A character in UTF-8 is at most 4 bytes
438 // One character must be removed in any case because the
439 // string is too long
440 chopOffChars = Math.max( 1, Math.floor( chopOffBytes / 4 ) );
441 s = s.substr( 0, s.length - chopOffChars );
447 * Cuts a file name to a specific byte length
451 * @method trimFileNameToByteLength
452 * @param {string} name without extension
453 * @param {string} extension file extension
454 * @return {string} The full name, including extension
456 trimFileNameToByteLength = function ( name, extension ) {
457 // There is a special byte limit for file names and ... remember the dot
458 return trimToByteLength( name, FILENAME_MAX_BYTES - extension.length - 1 ) + '.' + extension;
461 // Polyfill for ES5 Object.create
462 createObject = Object.create || ( function () {
463 return function ( o ) {
465 if ( o !== Object( o ) ) {
466 throw new Error( 'Cannot inherit from a non-object' );
476 * Constructor for Title objects with a null return instead of an exception for invalid titles.
478 * Note that `namespace` is the **default** namespace only, and can be overridden by a namespace
479 * prefix in `title`. If you do not want this behavior, use #makeTitle. See #constructor for
483 * @param {string} title
484 * @param {number} [namespace=NS_MAIN] Default namespace
485 * @return {mw.Title|null} A valid Title object or null if the title is invalid
487 Title.newFromText = function ( title, namespace ) {
488 var t, parsed = parse( title, namespace );
493 t = createObject( Title.prototype );
494 t.namespace = parsed.namespace;
495 t.title = parsed.title;
497 t.fragment = parsed.fragment;
503 * Constructor for Title objects with predefined namespace.
505 * Unlike #newFromText or #constructor, this function doesn't allow the given `namespace` to be
506 * overridden by a namespace prefix in `title`. See #constructor for details about this behavior.
508 * The single exception to this is when `namespace` is 0, indicating the main namespace. The
509 * function behaves like #newFromText in that case.
512 * @param {number} namespace Namespace to use for the title
513 * @param {string} title
514 * @return {mw.Title|null} A valid Title object or null if the title is invalid
516 Title.makeTitle = function ( namespace, title ) {
517 return mw.Title.newFromText( getNamespacePrefix( namespace ) + title );
521 * Constructor for Title objects from user input altering that input to
522 * produce a title that MediaWiki will accept as legal
525 * @param {string} title
526 * @param {number} [defaultNamespace=NS_MAIN]
527 * If given, will used as default namespace for the given title.
528 * @param {Object} [options] additional options
529 * @param {boolean} [options.forUploading=true]
530 * Makes sure that a file is uploadable under the title returned.
531 * There are pages in the file namespace under which file upload is impossible.
532 * Automatically assumed if the title is created in the Media namespace.
533 * @return {mw.Title|null} A valid Title object or null if the input cannot be turned into a valid title
535 Title.newFromUserInput = function ( title, defaultNamespace, options ) {
536 var namespace, m, id, ext, parts;
538 // defaultNamespace is optional; check whether options moves up
539 if ( arguments.length < 3 && $.type( defaultNamespace ) === 'object' ) {
540 options = defaultNamespace;
541 defaultNamespace = undefined;
544 // merge options into defaults
545 options = $.extend( {
549 namespace = defaultNamespace === undefined ? NS_MAIN : defaultNamespace;
551 // Normalise additional whitespace
552 title = $.trim( title.replace( /\s/g, ' ' ) );
554 // Process initial colon
555 if ( title !== '' && title[ 0 ] === ':' ) {
556 // Initial colon means main namespace instead of specified default
562 .replace( rUnderscoreTrim, '' );
565 // Process namespace prefix (if any)
566 m = title.match( rSplit );
568 id = getNsIdByName( m[ 1 ] );
569 if ( id !== false ) {
570 // Ordinary namespace
577 namespace === NS_MEDIA ||
578 ( options.forUploading && ( namespace === NS_FILE ) )
581 title = sanitize( title, [ 'generalRule', 'fileRule' ] );
583 // Operate on the file extension
584 // Although it is possible having spaces between the name and the ".ext" this isn't nice for
585 // operating systems hiding file extensions -> strip them later on
586 parts = title.split( '.' );
588 if ( parts.length > 1 ) {
590 // Get the last part, which is supposed to be the file extension
593 // Remove whitespace of the name part (that W/O extension)
594 title = $.trim( parts.join( '.' ) );
596 // Cut, if too long and append file extension
597 title = trimFileNameToByteLength( title, ext );
601 // Missing file extension
602 title = $.trim( parts.join( '.' ) );
604 // Name has no file extension and a fallback wasn't provided either
609 title = sanitize( title, [ 'generalRule' ] );
611 // Cut titles exceeding the TITLE_MAX_BYTES byte size limit
612 // (size of underlying database field)
613 if ( namespace !== NS_SPECIAL ) {
614 title = trimToByteLength( title, TITLE_MAX_BYTES );
618 // Any remaining initial :s are illegal.
619 title = title.replace( /^\:+/, '' );
621 return Title.newFromText( title, namespace );
625 * Sanitizes a file name as supplied by the user, originating in the user's file system
626 * so it is most likely a valid MediaWiki title and file name after processing.
627 * Returns null on fatal errors.
630 * @param {string} uncleanName The unclean file name including file extension but
632 * @return {mw.Title|null} A valid Title object or null if the title is invalid
634 Title.newFromFileName = function ( uncleanName ) {
636 return Title.newFromUserInput( 'File:' + uncleanName, {
642 * Get the file title from an image element
644 * var title = mw.Title.newFromImg( $( 'img:first' ) );
647 * @param {HTMLElement|jQuery} img The image to use as a base
648 * @return {mw.Title|null} The file title or null if unsuccessful
650 Title.newFromImg = function ( img ) {
651 var matches, i, regex, src, decodedSrc,
653 // thumb.php-generated thumbnails
654 thumbPhpRegex = /thumb\.php/,
657 /\/[a-f0-9]\/[a-f0-9]{2}\/([^\s\/]+)\/[^\s\/]+-[^\s\/]*$/,
660 /\/[a-f0-9]\/[a-f0-9]{2}\/([^\s\/]+)$/,
662 // Thumbnails in non-hashed upload directories
663 /\/([^\s\/]+)\/[^\s\/]+-(?:\1|thumbnail)[^\s\/]*$/,
665 // Full-size images in non-hashed upload directories
669 recount = regexes.length;
671 src = img.jquery ? img[ 0 ].src : img.src;
673 matches = src.match( thumbPhpRegex );
676 return mw.Title.newFromText( 'File:' + mw.util.getParamValue( 'f', src ) );
679 decodedSrc = decodeURIComponent( src );
681 for ( i = 0; i < recount; i++ ) {
682 regex = regexes[ i ];
683 matches = decodedSrc.match( regex );
685 if ( matches && matches[ 1 ] ) {
686 return mw.Title.newFromText( 'File:' + matches[ 1 ] );
694 * Whether this title exists on the wiki.
697 * @param {string|mw.Title} title prefixed db-key name (string) or instance of Title
698 * @return {boolean|null} Boolean if the information is available, otherwise null
700 Title.exists = function ( title ) {
702 type = $.type( title ),
703 obj = Title.exist.pages;
705 if ( type === 'string' ) {
706 match = obj[ title ];
707 } else if ( type === 'object' && title instanceof Title ) {
708 match = obj[ title.toString() ];
710 throw new Error( 'mw.Title.exists: title must be a string or an instance of Title' );
713 if ( typeof match === 'boolean' ) {
721 * Store page existence
724 * @property {Object} exist
725 * @property {Object} exist.pages Keyed by title. Boolean true value indicates page does exist.
727 * @property {Function} exist.set The setter function.
729 * Example to declare existing titles:
731 * Title.exist.set( ['User:John_Doe', ...] );
733 * Example to declare titles nonexistent:
735 * Title.exist.set( ['File:Foo_bar.jpg', ...], false );
737 * @property {string|Array} exist.set.titles Title(s) in strict prefixedDb title form
738 * @property {boolean} [exist.set.state=true] State of the given titles
744 set: function ( titles, state ) {
748 titles = $.isArray( titles ) ? titles : [ titles ];
749 state = state === undefined ? true : !!state;
751 for ( i = 0, len = titles.length; i < len; i++ ) {
752 pages[ titles[ i ] ] = state;
759 * Normalize a file extension to the common form, making it lowercase and checking some synonyms,
760 * and ensure it's clean. Extensions with non-alphanumeric characters will be discarded.
761 * Keep in sync with File::normalizeExtension() in PHP.
763 * @param {string} extension File extension (without the leading dot)
764 * @return {string} File extension in canonical form
766 Title.normalizeExtension = function ( extension ) {
768 lower = extension.toLowerCase(),
776 if ( squish.hasOwnProperty( lower ) ) {
777 return squish[ lower ];
778 } else if ( /^[0-9a-z]+$/.test( lower ) ) {
791 * Get the namespace number
793 * Example: 6 for "File:Example_image.svg".
797 getNamespaceId: function () {
798 return this.namespace;
802 * Get the namespace prefix (in the content language)
804 * Example: "File:" for "File:Example_image.svg".
805 * In #NS_MAIN this is '', otherwise namespace name plus ':'
809 getNamespacePrefix: function () {
810 return getNamespacePrefix( this.namespace );
814 * Get the page name without extension or namespace prefix
816 * Example: "Example_image" for "File:Example_image.svg".
818 * For the page title (full page name without namespace prefix), see #getMain.
822 getName: function () {
824 $.inArray( this.namespace, mw.config.get( 'wgCaseSensitiveNamespaces' ) ) !== -1 ||
829 // PHP's strtoupper differs from String.toUpperCase in a number of cases
831 return mw.Title.phpCharToUpper( this.title[ 0 ] ) + this.title.slice( 1 );
835 * Get the page name (transformed by #text)
837 * Example: "Example image" for "File:Example_image.svg".
839 * For the page title (full page name without namespace prefix), see #getMainText.
843 getNameText: function () {
844 return text( this.getName() );
848 * Get the extension of the page name (if any)
850 * @return {string|null} Name extension or null if there is none
852 getExtension: function () {
857 * Shortcut for appendable string to form the main page name.
859 * Returns a string like ".json", or "" if no extension.
863 getDotExtension: function () {
864 return this.ext === null ? '' : '.' + this.ext;
868 * Get the main page name
870 * Example: "Example_image.svg" for "File:Example_image.svg".
874 getMain: function () {
875 return this.getName() + this.getDotExtension();
879 * Get the main page name (transformed by #text)
881 * Example: "Example image.svg" for "File:Example_image.svg".
885 getMainText: function () {
886 return text( this.getMain() );
890 * Get the full page name
892 * Example: "File:Example_image.svg".
893 * Most useful for API calls, anything that must identify the "title".
897 getPrefixedDb: function () {
898 return this.getNamespacePrefix() + this.getMain();
902 * Get the full page name (transformed by #text)
904 * Example: "File:Example image.svg" for "File:Example_image.svg".
908 getPrefixedText: function () {
909 return text( this.getPrefixedDb() );
913 * Get the page name relative to a namespace
917 * - "Foo:Bar" relative to the Foo namespace becomes "Bar".
918 * - "Bar" relative to any non-main namespace becomes ":Bar".
919 * - "Foo:Bar" relative to any namespace other than Foo stays "Foo:Bar".
921 * @param {number} namespace The namespace to be relative to
924 getRelativeText: function ( namespace ) {
925 if ( this.getNamespaceId() === namespace ) {
926 return this.getMainText();
927 } else if ( this.getNamespaceId() === NS_MAIN ) {
928 return ':' + this.getPrefixedText();
930 return this.getPrefixedText();
935 * Get the fragment (if any).
937 * Note that this method (by design) does not include the hash character and
938 * the value is not url encoded.
940 * @return {string|null}
942 getFragment: function () {
943 return this.fragment;
947 * Get the URL to this title
949 * @see mw.util#getUrl
950 * @param {Object} [params] A mapping of query parameter names to values,
951 * e.g. `{ action: 'edit' }`.
954 getUrl: function ( params ) {
955 var fragment = this.getFragment();
957 return mw.util.getUrl( this.toString() + '#' + fragment, params );
959 return mw.util.getUrl( this.toString(), params );
964 * Whether this title exists on the wiki.
966 * @see #static-method-exists
967 * @return {boolean|null} Boolean if the information is available, otherwise null
969 exists: function () {
970 return Title.exists( this );
975 * @alias #getPrefixedDb
978 Title.prototype.toString = Title.prototype.getPrefixedDb;
981 * @alias #getPrefixedText
984 Title.prototype.toText = Title.prototype.getPrefixedText;
989 }( mediaWiki, jQuery ) );