| blob | 1b926c468fe066f164e5a4c93040433455ddd6c3 |
1 <?php
2 /**
3 * PHP parser that converts wiki markup to HTML.
4 *
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
9 *
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
19 *
20 * @file
21 * @ingroup Parser
22 */
24 /**
25 * @defgroup Parser Parser
26 */
28 /**
29 * PHP Parser - Processes wiki markup (which uses a more user-friendly
30 * syntax, such as "[[link]]" for making links), and provides a one-way
31 * transformation of that wiki markup it into (X)HTML output / markup
32 * (which in turn the browser understands, and can display).
33 *
34 * There are seven main entry points into the Parser class:
35 *
36 * - Parser::parse()
37 * produces HTML output
38 * - Parser::preSaveTransform()
39 * produces altered wiki markup
40 * - Parser::preprocess()
41 * removes HTML comments and expands templates
42 * - Parser::cleanSig() and Parser::cleanSigInSig()
43 * cleans a signature before saving it to preferences
44 * - Parser::getSection()
45 * return the content of a section from an article for section editing
46 * - Parser::replaceSection()
47 * replaces a section by number inside an article
48 * - Parser::getPreloadText()
49 * removes <noinclude> sections and <includeonly> tags
50 *
51 * Globals used:
52 * object: $wgContLang
53 *
54 * @warning $wgUser or $wgTitle or $wgRequest or $wgLang. Keep them away!
55 *
56 * @par Settings:
57 * $wgNamespacesWithSubpages
58 *
59 * @par Settings only within ParserOptions:
60 * $wgAllowExternalImages
61 * $wgAllowSpecialInclusion
62 * $wgInterwikiMagic
63 * $wgMaxArticleSize
64 *
65 * @ingroup Parser
66 */
68 /**
69 * Update this version number when the ParserOutput format
70 * changes in an incompatible way, so the parser cache
71 * can automatically discard old data.
72 */
75 /**
76 * Update this version number when the output of serialiseHalfParsedText()
77 * changes in an incompatible way
78 */
81 # Flags for Parser::setFunctionHook
85 # Constants needed for external link processing
86 # Everything except bracket, space, or control characters
87 # \p{Zs} is unicode 'separator, space' category. It covers the space 0x20
88 # as well as U+3000 is IDEOGRAPHIC SPACE for bug 19052
93 # State constants for the definition list colon extraction
103 # Flags for preprocessToDom
106 # Allowed values for $this->mOutputType
107 # Parameter to startExternalParse().
114 # Marker Suffix needs to be accessible staticly.
117 # Markers used for wrapping the table of contents
121 # Persistent:
135 # Initialised by initialiseVariables()
137 /**
138 * @var MagicWordArray
139 */
142 /**
143 * @var MagicWordArray
144 */
146 public $mConf, $mPreprocessor, $mExtLinkBracketedRegex, $mUrlProtocols; # Initialised in constructor
148 # Cleared with clearState():
149 /**
150 * @var ParserOutput
151 */
155 /**
156 * @var StripState
157 */
161 /**
162 * @var LinkHolderArray
163 */
173 /**
174 * @var User
175 */
178 # Temporary
179 # These are variables reset at least once per parse regardless of $clearState
181 /**
182 * @var ParserOptions
183 */
186 /**
187 * @var Title
188 */
200 /**
201 * @var string
202 */
205 /**
206 * @var array Array with the language name of each language link (i.e. the
207 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
208 * duplicate language links to the ParserOutput.
209 */
212 /**
213 * @var MapCacheLRU|null
214 * @since 1.24
215 *
216 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
217 */
220 /**
221 * @var bool Recursive call protection.
222 * This variable should be treated as if it were private.
223 */
226 /** @var SectionProfiler */
229 /**
230 * @param array $conf
231 */
240 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
243 # PECL extension that conflicts with the core DOM extension (bug 13770)
250 }
252 }
254 /**
255 * Reduce memory usage to reduce the impact of circular references
256 */
260 }
263 }
264 }
266 /**
267 * Allow extensions to clean up when the parser is cloned
268 */
272 // Bug 56226: When you create a reference "to" an object field, that
273 // makes the object field itself be a reference too (until the other
274 // reference goes out of scope). When cloning, any field that's a
275 // reference is copied as a reference in the new object. Both of these
276 // are defined PHP5 behaviors, as inconvenient as it is for us when old
277 // hooks from PHP4 days are passing fields by reference.
279 // Make a non-reference copy of the field, then rebind the field to
280 // reference the new copy.
284 }
287 }
289 /**
290 * Do various kinds of initialisation on the first call of the parser
291 */
295 }
306 }
308 /**
309 * Clear Parser state
310 *
311 * @private
312 */
317 }
335 /**
336 * Prefix for temporary replacement strings for the multipass parser.
337 * \x07 should never appear in input as it's disallowed in XML.
338 * Using it at the front also gives us a little extra robustness
339 * since it shouldn't match when butted up against identifier-like
340 * string constructs.
341 *
342 * Must not consist of all title characters, or else it will change
343 * the behavior of <nowiki> in a link.
344 */
348 # Clear these on every parse, bug 4549
356 );
365 # Fix cloning
368 }
374 }
376 /**
377 * Convert wikitext to HTML
378 * Do not call this function recursively.
379 *
380 * @param string $text Text we want to parse
381 * @param Title $title
382 * @param ParserOptions $options
383 * @param bool $linestart
384 * @param bool $clearState
385 * @param int $revid Number to pass in {{REVISIONID}}
386 * @return ParserOutput A ParserOutput
387 */
390 ) {
391 /**
392 * First pass--just handle <nowiki> sections, pass the rest off
393 * to internalParse() which does all the real work.
394 */
403 }
411 }
413 # Remove the strip marker tag prefix from the input, if present.
416 }
429 }
432 # No more strip!
439 /**
440 * A converted title will be provided in the output object if title and
441 * content conversion are enabled, the article text does not contain
442 * a conversion-suppressing double-underscore tag, and no
443 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
444 * automatic link conversion.
445 */
450 ) {
457 }
458 }
464 );
465 }
467 # Information on include size limits, for the benefit of users who try to skirt them
475 );
476 }
481 );
485 );
488 );
491 );
494 );
497 );
500 );
506 }
510 ) ) {
516 }
520 }
521 }
522 }
523 // Since we're not really outputting HTML, decode the entities and
524 // then re-encode the things that need hiding inside HTML comments.
528 // Sanitize for comment. Note '‐' in the replacement is U+2010,
529 // which looks much like the problematic '-'.
533 // Add on template profiling data
537 } );
543 }
549 }
550 }
564 }
566 /**
567 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
568 * can be called from an extension tag hook.
569 *
570 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
571 * instead, which means that lists and links have not been fully parsed yet,
572 * and strip markers are still present.
573 *
574 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
575 *
576 * Use this function if you're a parser tag hook and you want to parse
577 * wikitext before or after applying additional transformations, and you
578 * intend to *return the result as hook output*, which will cause it to go
579 * through the rest of parsing process automatically.
580 *
581 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
582 * $text are not expanded
583 *
584 * @param string $text Text extension wants to have parsed
585 * @param bool|PPFrame $frame The frame to use for expanding any template variables
586 * @return string UNSAFE half-parsed HTML
587 */
595 }
597 /**
598 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
599 * point can be called from an extension tag hook.
600 *
601 * The output of this function is fully-parsed HTML that is safe for output.
602 * If you're a parser tag hook, you might want to use recursiveTagParse()
603 * instead.
604 *
605 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
606 * $text are not expanded
607 *
608 * @since 1.25
609 *
610 * @param string $text Text extension wants to have parsed
611 * @param bool|PPFrame $frame The frame to use for expanding any template variables
612 * @return string Fully parsed HTML
613 */
620 }
622 /**
623 * Expand templates and variables in the text, producing valid, static wikitext.
624 * Also removes comments.
625 * Do not call this function recursively.
626 * @param string $text
627 * @param Title $title
628 * @param ParserOptions $options
629 * @param int|null $revid
630 * @param bool|PPFrame $frame
631 * @return mixed|string
632 */
633 public function preprocess( $text, Title $title = null, ParserOptions $options, $revid = null, $frame = false ) {
639 }
646 }
648 /**
649 * Recursive parser entry point that can be called from an extension tag
650 * hook.
651 *
652 * @param string $text Text to be expanded
653 * @param bool|PPFrame $frame The frame to use for expanding any template variables
654 * @return string
655 * @since 1.19
656 */
663 }
665 /**
666 * Process the wikitext for the "?preload=" feature. (bug 5210)
667 *
668 * "<noinclude>", "<includeonly>" etc. are parsed as for template
669 * transclusion, comments, templates, arguments, tags hooks and parser
670 * functions are untouched.
671 *
672 * @param string $text
673 * @param Title $title
674 * @param ParserOptions $options
675 * @param array $params
676 * @return string
677 */
678 public function getPreloadText( $text, Title $title, ParserOptions $options, $params = array() ) {
682 # Parser (re)initialisation
691 }
693 /**
694 * Get a random string
695 *
696 * @return string
697 */
700 }
702 /**
703 * Set the current user.
704 * Should only be used when doing pre-save transform.
705 *
706 * @param User|null $user User object or null (to reset)
707 */
710 }
712 /**
713 * Accessor for mUniqPrefix.
714 *
715 * @return string
716 */
719 # @todo FIXME: This is probably *horribly wrong*
720 # LanguageConverter seems to want $wgParser's uniqPrefix, however
721 # if this is called for a parser cache hit, the parser may not
722 # have ever been initialized in the first place.
723 # Not really sure what the heck is supposed to be going on here.
725 # throw new MWException( "Accessing uninitialized mUniqPrefix" );
726 }
728 }
730 /**
731 * Set the context title
732 *
733 * @param Title $t
734 */
738 }
741 # Strip the fragment to avoid various odd effects
746 }
747 }
749 /**
750 * Accessor for the Title object
751 *
752 * @return Title
753 */
756 }
758 /**
759 * Accessor/mutator for the Title object
760 *
761 * @param Title $x Title object or null to just get the current one
762 * @return Title
763 */
766 }
768 /**
769 * Set the output type
770 *
771 * @param int $ot New value
772 */
775 # Shortcut alias
781 );
782 }
784 /**
785 * Accessor/mutator for the output type
786 *
787 * @param int|null $x New value or null to just get the current one
788 * @return int
789 */
792 }
794 /**
795 * Get the ParserOutput object
796 *
797 * @return ParserOutput
798 */
801 }
803 /**
804 * Get the ParserOptions object
805 *
806 * @return ParserOptions
807 */
810 }
812 /**
813 * Accessor/mutator for the ParserOptions object
814 *
815 * @param ParserOptions $x New value or null to just get the current one
816 * @return ParserOptions Current ParserOptions object
817 */
820 }
822 /**
823 * @return int
824 */
827 }
829 /**
830 * @param int $id
831 */
834 }
836 /**
837 * Get a language object for use in parser functions such as {{FORMATNUM:}}
838 * @return Language
839 */
842 }
844 /**
845 * Get the target language for the content being parsed. This is usually the
846 * language that the content is in.
847 *
848 * @since 1.19
849 *
850 * @throws MWException
851 * @return Language
852 */
862 }
865 }
867 /**
868 * Get the language object for language conversion
869 * @return Language|null
870 */
873 }
875 /**
876 * Get a User object either from $this->mUser, if set, or from the
877 * ParserOptions object otherwise
878 *
879 * @return User
880 */
884 }
886 }
888 /**
889 * Get a preprocessor object
890 *
891 * @return Preprocessor
892 */
897 }
899 }
901 /**
902 * Replaces all occurrences of HTML-style comments and the given tags
903 * in the text with a random marker and returns the next text. The output
904 * parameter $matches will be an associative array filled with data in
905 * the form:
906 *
907 * @code
908 * 'UNIQ-xxxxx' => array(
909 * 'element',
910 * 'tag content',
911 * array( 'param' => 'x' ),
912 * '<element param="x">tag content</element>' ) )
913 * @endcode
914 *
915 * @param array $elements List of element names. Comments are always extracted.
916 * @param string $text Source text string.
917 * @param array $matches Out parameter, Array: extracted tags
918 * @param string $uniq_prefix
919 * @return string Stripped text
920 */
934 }
936 # comment
942 # tag
947 }
953 # Empty element tag, <tag />
962 }
966 # No end tag -- let it run out to the end of the text.
972 }
973 }
979 }
981 }
983 /**
984 * Get a list of strippable XML-like elements
985 *
986 * @return array
987 */
990 }
992 /**
993 * Add an item to the strip state
994 * Returns the unique tag which must be inserted into the stripped text
995 * The tag will be replaced with the original text in unstrip()
996 *
997 * @param string $text
998 *
999 * @return string
1000 */
1006 }
1008 /**
1009 * parse the wiki syntax used to render tables
1010 *
1011 * @private
1012 * @param string $text
1013 * @return string
1014 */
1033 }
1039 # First check if we are starting a new table
1052 # Don't do any of the following
1056 # We are ending a table
1062 }
1066 }
1070 }
1074 # Now we have a table row
1077 # Whats after the tag is now only attributes
1090 }
1094 }
1103 ) {
1104 # This might be cell elements, td, th or captions
1108 }
1114 }
1116 # Split up multiple cells on the same line.
1117 # FIXME : This can result in improper nesting of tags processed
1118 # by earlier parser steps, but should avoid splitting up eg
1119 # attribute values containing literal "||".
1124 # Loop through each table cell
1131 }
1136 }
1142 }
1152 }
1156 # A cell could contain both parameters and data
1159 # Bug 553: Note that a '|' inside an invalid link should not
1160 # be mistaken as delimiting cell parameters
1169 }
1173 }
1174 }
1176 }
1178 # Closing open td, tr && table
1182 }
1185 }
1188 }
1191 }
1193 # Remove trailing line-ending (b/c)
1196 }
1198 # special case: don't return empty table
1201 }
1206 }
1208 /**
1209 * Helper function for parse() that transforms wiki markup into half-parsed
1210 * HTML. Only called for $mOutputType == self::OT_HTML.
1211 *
1212 * @private
1213 *
1214 * @param string $text
1215 * @param bool $isMain
1216 * @param bool $frame
1217 *
1218 * @return string
1219 */
1225 # Hook to suspend the parser in this state
1226 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState ) ) ) {
1229 }
1231 # if $frame is provided, then use $frame for replacing any variables
1233 # use frame depth to infer how include/noinclude tags should be handled
1234 # depth=0 means this is the top-level document; otherwise it's an included document
1239 }
1243 # if $frame is not provided, then use old-style replaceVariables
1245 }
1253 );
1256 # Tables need to come after variable replacement for things to work
1257 # properly; putting them before other transformations should keep
1258 # exciting things like link expansions from showing up in surprising
1259 # places.
1271 # replaceInternalLinks may sometimes leave behind
1272 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1280 }
1282 /**
1283 * Helper function for parse() that transforms half-parsed HTML into fully
1284 * parsed HTML.
1285 *
1286 * @param string $text
1287 * @param bool $isMain
1288 * @param bool $linestart
1289 * @return string
1290 */
1296 # Clean up special characters, only run once, next-to-last before doBlockLevels
1298 # french spaces, last one Guillemet-left
1299 # only if there is something before the space
1301 # french spaces, Guillemet-right
1304 );
1311 /**
1312 * The input doesn't get language converted if
1313 * a) It's disabled
1314 * b) Content isn't converted
1315 * c) It's a conversion table
1316 * d) it is an interface message (which is in the user language)
1317 */
1320 ) {
1322 # The position of the convert() call should not be changed. it
1323 # assumes that the links are all replaced and the only thing left
1324 # is the <nowiki> mark.
1326 }
1327 }
1333 }
1343 # attempt to sanitize at least some nesting problems
1344 # (bug #2702 and quite a few others)
1346 # ''Something [http://www.cool.com cool''] -->
1347 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1350 # fix up an anchor inside another anchor, only
1351 # at least for a single single nested link (bug 3695)
1354 # fix div inside inline elements- doBlockLevels won't wrap a line which
1355 # contains a div, so fix it up here; replace
1356 # div with escaped text
1359 # remove empty italic or bold tag pairs, some
1360 # introduced by rules above
1362 );
1368 }
1372 }
1375 }
1377 /**
1378 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1379 * magic external links.
1380 *
1381 * DML
1382 * @private
1383 *
1384 * @param string $text
1385 *
1386 * @return string
1387 */
1393 '!(?: # Start cases
1397 (?:RFC|PMID)\s+([0-9]+) | # m[4]: RFC or PMID, capture number
1399 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
1400 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
1401 [0-9Xx] # check digit
1406 }
1408 /**
1409 * @throws MWException
1410 * @param array $m
1411 * @return HTML|string
1412 */
1415 # Skip anchor
1418 # Skip HTML element
1421 # Free external link
1424 # RFC or PMID
1438 }
1442 # ISBN
1448 ));
1455 }
1456 }
1458 /**
1459 * Make a free external link, given a user-supplied URL
1460 *
1461 * @param string $url
1462 *
1463 * @return string HTML
1464 * @private
1465 */
1471 # The characters '<' and '>' (which were escaped by
1472 # removeHTMLtags()) should not be included in
1473 # URLs, per RFC 2396.
1478 }
1480 # Move trailing punctuation to $trail
1482 # If there is no left bracket, then consider right brackets fair game too
1485 }
1491 }
1495 # Is this an external image?
1498 # Not an image, make a link
1503 # Register it in the output object...
1504 # Replace unnecessary URL escape codes with their equivalent characters
1507 }
1510 }
1512 /**
1513 * Parse headers and return html
1514 *
1515 * @private
1516 *
1517 * @param string $text
1518 *
1519 * @return string
1520 */
1526 }
1529 }
1531 /**
1532 * Replace single quotes with HTML markup
1533 * @private
1534 *
1535 * @param string $text
1536 *
1537 * @return string The altered text
1538 */
1545 }
1549 }
1551 /**
1552 * Helper function for doAllQuotes()
1553 *
1554 * @param string $text
1555 *
1556 * @return string
1557 */
1563 }
1565 // First, do some preliminary work. This may shift some apostrophes from
1566 // being mark-up to being text. It also counts the number of occurrences
1567 // of bold and italics mark-ups.
1572 // If there are ever four apostrophes, assume the first is supposed to
1573 // be text, and the remaining three constitute mark-up for bold text.
1574 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1580 // If there are more than 5 apostrophes in a row, assume they're all
1581 // text except for the last 5.
1582 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1586 }
1587 // Count the number of occurrences of bold and italics mark-ups.
1595 }
1596 }
1598 // If there is an odd number of both bold and italics, it is likely
1599 // that one of the bold ones was meant to be an apostrophe followed
1600 // by italics. Which one we cannot know for certain, but it is more
1601 // likely to be one that has a single-letter word before it.
1613 }
1617 // if $firstsingleletterword is set, we don't
1618 // look at the other options, so we can bail early.
1620 }
1624 }
1625 }
1626 }
1627 }
1629 // If there is a single-letter word, use it!
1634 // If not, but there's a multi-letter word, use that one.
1638 // ... otherwise use the first one that has neither.
1639 // (notice that it is possible for all three to be -1 if, for example,
1640 // there is only one pentuple-apostrophe in the line)
1643 }
1644 }
1646 // Now let's actually convert our apostrophic mush to HTML!
1657 }
1676 }
1693 }
1713 }
1714 }
1715 }
1717 }
1718 // Now close all remaining tags. Notice that the order is important.
1721 }
1724 }
1727 }
1728 // There might be lonely ''''', so make sure we have a buffer
1731 }
1733 }
1735 /**
1736 * Replace external links (REL)
1737 *
1738 * Note: this is all very hackish and the order of execution matters a lot.
1739 * Make sure to run tests/parserTests.php if you change this code.
1740 *
1741 * @private
1742 *
1743 * @param string $text
1744 *
1745 * @throws MWException
1746 * @return string
1747 */
1756 }
1766 # The characters '<' and '>' (which were escaped by
1767 # removeHTMLtags()) should not be included in
1768 # URLs, per RFC 2396.
1773 }
1775 # If the link text is an image URL, replace it with an <img> tag
1776 # This happened by accident in the original parser, but some people used it extensively
1780 }
1784 # Set linktype for CSS - if URL==text, link is essentially free
1787 # No link text, e.g. [http://domain.tld/some.link]
1789 # Autonumber
1794 # Have link text, e.g. [http://domain.tld/some.link text]s
1795 # Check for trail
1797 }
1803 # Use the encoded URL
1804 # This means that users can paste URLs directly into the text
1805 # Funny characters like ö aren't valid in URLs anyway
1806 # This was changed in August 2004
1810 # Register link in the output object.
1811 # Replace unnecessary URL escape codes with the referenced character
1812 # This prevents spammers from hiding links from the filters
1815 }
1819 }
1821 /**
1822 * Get the rel attribute for a particular external link.
1823 *
1824 * @since 1.21
1825 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1826 * nofollow if appropriate
1827 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1828 * @return string|null Rel attribute for $url
1829 */
1835 ) {
1837 }
1839 }
1841 /**
1842 * Get an associative array of additional HTML attributes appropriate for a
1843 * particular external link. This currently may include rel => nofollow
1844 * (depending on configuration, namespace, and the URL's domain) and/or a
1845 * target attribute (depending on configuration).
1846 *
1847 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1848 * nofollow if appropriate
1849 * @return array Associative array of HTML attributes
1850 */
1857 }
1859 }
1861 /**
1862 * Replace unusual escape codes in a URL with their equivalent characters
1863 *
1864 * @deprecated since 1.24, use normalizeLinkUrl
1865 * @param string $url
1866 * @return string
1867 */
1871 }
1873 /**
1874 * Replace unusual escape codes in a URL with their equivalent characters
1875 *
1876 * This generally follows the syntax defined in RFC 3986, with special
1877 * consideration for HTTP query strings.
1878 *
1879 * @param string $url
1880 * @return string
1881 */
1883 # First, make sure unsafe characters are encoded
1887 },
1888 $url
1889 );
1894 # Fragment part - 'fragment'
1900 }
1902 # Query part - 'query' minus &=+;
1908 }
1910 # Scheme and path part - 'pchar'
1911 # (we assume no userinfo or encoded colons in the host)
1916 }
1923 # Unescape it
1926 # Leave it escaped, but use uppercase for a-f
1928 }
1929 };
1931 }
1933 /**
1934 * make an image if it's allowed, either through the global
1935 * option, through the exception, or through the on-wiki whitelist
1936 *
1937 * @param string $url
1938 *
1939 * @return string
1940 */
1945 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1952 }
1953 }
1958 }
1962 ) {
1964 # Image found
1966 }
1967 }
1970 ) {
1974 );
1977 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1980 }
1982 # Image matches a whitelist entry
1985 }
1986 }
1987 }
1989 }
1991 /**
1992 * Process [[ ]] wikilinks
1993 *
1994 * @param string $s
1995 *
1996 * @return string Processed text
1997 *
1998 * @private
1999 */
2003 }
2005 /**
2006 * Process [[ ]] wikilinks (RIL)
2007 * @param string $s
2008 * @throws MWException
2009 * @return LinkHolderArray
2010 *
2011 * @private
2012 */
2019 # the % is needed to support urlencoded titles as well
2022 # Match a link having the form [[namespace:link|alternate]]trail
2024 # Match cases where there is no "]]", which might still be images
2026 }
2030 # split the entire text string on occurrences of [[
2032 # get the first element (all text up to first [[), and remove the space we added
2041 # Match the end of a line for a word that's not followed by whitespace,
2042 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2046 }
2052 }
2061 }
2064 }
2069 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2070 # Loop for each link
2072 // @codingStandardsIgnoreStart
2074 # Check for excessive memory usage
2076 # Too big
2077 # Do the existence check, replace the link holders and clear the array
2080 }
2089 }
2090 # first link
2094 }
2096 }
2103 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2104 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2105 # the real problem is with the $e1 regex
2106 # See bug 1300.
2107 #
2108 # Still some problems for cases where the ] is meant to be outside punctuation,
2109 # and no image is in sight. See bug 2095.
2110 #
2114 ) {
2117 }
2118 # fix up urlencoded title texts
2120 # Should anchors '#' also be rejected?
2122 }
2125 # Invalid, but might be an image with a link in its caption
2130 }
2136 }
2142 # Don't allow internal links to pages containing
2143 # PROTO: where PROTO is a valid URL protocol; these
2144 # should be external links.
2149 }
2151 # Make subpage if necessary
2156 }
2160 # Strip off leading ':'
2162 }
2171 }
2182 # look at the next 'line' to see if we can close it there
2187 }
2190 # the first ]] closes the inner link, the second the image
2196 # if there's exactly one ]] that's fine, we'll keep looking
2199 # if $next_line is invalid too, we need look no further
2202 }
2203 }
2205 # we couldn't find the end of this imageLink, so output it raw
2206 # but don't ignore what might be perfectly normal links in the text we've examined
2209 # note: no $trail, because without an end, there *is* no trail
2212 }
2215 # note: no $trail, because without an end, there *is* no trail
2218 }
2220 }
2226 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2227 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2228 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2229 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2231 }
2233 # Link not escaped by : , create the various objects
2235 # Interwikis
2241 )
2242 ) {
2243 # Bug 24502: filter duplicates
2247 }
2253 }
2260 # if no parameters were passed, $text
2261 # becomes something like "File:Foo.png",
2262 # which we don't want to pass on to the
2263 # image generator
2266 # recursively parse links inside the image caption
2267 # actually, this will parse them in any other parameters, too,
2268 # but it might be hard to fix that, and it doesn't matter ATM
2271 }
2272 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2277 }
2280 }
2290 }
2296 /**
2297 * Strip the whitespace Category links produce, see bug 87
2298 */
2303 }
2304 }
2306 # Self-link checking. For some languages, variants of the title are checked in
2307 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2308 # for linking to a different variant.
2312 }
2314 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2315 # @todo FIXME: Should do batch file existence checks, see comment below
2318 # Give extensions a chance to select the file revision for us
2323 # Fetch and register the file (file title may be different via hooks)
2325 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2330 }
2333 # Some titles, such as valid special pages or files in foreign repos, should
2334 # be shown as bluelinks even though they're not included in the page table
2335 #
2336 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2337 # batch file existence checks for NS_FILE and NS_MEDIA
2342 # Links will be added to the output link list after checking
2344 }
2346 }
2349 }
2351 /**
2352 * Render a forced-blue link inline; protect against double expansion of
2353 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2354 * Since this little disaster has to split off the trail text to avoid
2355 * breaking URLs in the following text without breaking trails on the
2356 * wiki links, it's been made into a horrible function.
2357 *
2358 * @param Title $nt
2359 * @param string $text
2360 * @param array|string $query
2361 * @param string $trail
2362 * @param string $prefix
2363 * @return string HTML-wikitext mix oh yuck
2364 */
2365 public function makeKnownLinkHolder( $nt, $text = '', $query = array(), $trail = '', $prefix = '' ) {
2370 }
2373 }
2378 }
2380 /**
2381 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2382 * going to go through further parsing steps before inline URL expansion.
2383 *
2384 * Not needed quite as much as it used to be since free links are a bit
2385 * more sensible these days. But bracketed links are still an issue.
2386 *
2387 * @param string $text More-or-less HTML
2388 * @return string Less-or-more HTML with NOPARSE bits
2389 */
2393 }
2395 /**
2396 * Return true if subpage links should be expanded on this page.
2397 * @return bool
2398 */
2400 # Some namespaces don't allow subpages
2402 }
2404 /**
2405 * Handle link to subpage if necessary
2406 *
2407 * @param string $target The source of the link
2408 * @param string &$text The link text, modified as necessary
2409 * @return string The full name of the link
2410 * @private
2411 */
2414 }
2416 /**#@+
2417 * Used by doBlockLevels()
2418 * @private
2419 *
2420 * @return string
2421 */
2426 }
2430 }
2432 /**
2433 * getCommon() returns the length of the longest common substring
2434 * of both arguments, starting at the beginning of both.
2435 * @private
2436 *
2437 * @param string $st1
2438 * @param string $st2
2439 *
2440 * @return int
2441 */
2447 }
2452 }
2453 }
2455 }
2457 /**
2458 * These next three functions open, continue, and close the list
2459 * element appropriate to the prefix character passed into them.
2460 * @private
2461 *
2462 * @param string $char
2463 *
2464 * @return string
2465 */
2480 }
2483 }
2485 /**
2486 * TODO: document
2487 * @param string $char
2488 * @private
2489 *
2490 * @return string
2491 */
2499 }
2506 }
2507 }
2509 }
2511 /**
2512 * @todo Document
2513 * @param string $char
2514 * @private
2515 *
2516 * @return string
2517 */
2529 }
2532 }
2534 }
2535 /**#@-*/
2537 /**
2538 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2539 *
2540 * @param string $text
2541 * @param bool $linestart Whether or not this is at the start of a line.
2542 * @private
2543 * @return string The lists rendered as HTML
2544 */
2548 # Parsing through the text line by line. The main thing
2549 # happening here is handling of block-level elements p, pre,
2550 # and making lists from lines starting with * # : etc.
2551 #
2561 # Fix up $linestart
2566 }
2567 # * = ul
2568 # # = ol
2569 # ; = dt
2570 # : = dd
2575 # If not in a <pre> element, scan for and figure out what prefixes are there.
2577 # Multiple prefixes may abut each other for nested lists.
2581 # eh?
2582 # ; and : are both from definition-lists, so they're equivalent
2583 # for the purposes of determining whether or not we need to open/close
2584 # elements.
2589 # Don't interpret any other prefixes in preformatted text
2593 }
2595 # List generation
2597 # Same as the last item, so no need to deal with nesting or opening stuff
2602 # The one nasty exception: definition lists work like this:
2603 # ; title : definition text
2604 # So we check for : in the remainder text to split up the
2605 # title and definition, without b0rking links.
2610 }
2611 }
2613 # We need to open or close prefixes, or both.
2615 # Either open or close a level...
2619 # Close all the prefixes which aren't shared.
2623 }
2625 # Continue the current prefix if appropriate.
2628 }
2630 # Open prefixes where appropriate.
2633 }
2639 # @todo FIXME: This is dupe of code above
2643 }
2644 }
2646 }
2649 }
2651 }
2653 # If we have no prefixes, go to paragraph mode.
2656 # No prefix (not in list)--go to paragraph mode
2657 # XXX: use a stack for nestable elements like span, table and div
2659 '/(?:<table|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|'
2661 $t
2662 );
2668 $t
2669 );
2673 # @todo bug 5718: paragraph closed
2677 }
2679 while ( preg_match( '/<(\\/?)blockquote[\s>]/i', $t, $bqMatch, PREG_OFFSET_CAPTURE, $bqOffset ) ) {
2682 }
2688 ) {
2689 # pre
2694 }
2697 # paragraph
2710 }
2711 }
2720 }
2721 }
2722 }
2723 }
2725 }
2726 # somewhere above we forget to get out of pre block (bug 785)
2729 }
2734 }
2735 }
2736 }
2742 }
2743 }
2747 }
2751 }
2753 /**
2754 * Split up a string on ':', ignoring any occurrences inside tags
2755 * to prevent illegal overlapping.
2756 *
2757 * @param string $str The string to split
2758 * @param string &$before Set to everything before the ':'
2759 * @param string &$after Set to everything after the ':'
2760 * @throws MWException
2761 * @return string The position of the ':', or false if none found
2762 */
2768 # Nothing to find!
2771 }
2775 # Easy; no tag nesting to worry about
2780 }
2782 # Ugly state machine to walk through avoiding tags.
2790 # (Using the number is a performance hack for common cases)
2794 # Could be either a <start> tag or an </end> tag
2799 # We found it!
2804 }
2805 # Embedded in a tag; don't break it.
2808 # Skip ahead looking for something interesting
2811 # Nothing else interesting
2814 }
2818 # We found it!
2823 }
2824 }
2826 # Nothing else interesting to find; abort!
2827 # We're nested, but there's no close tags left. Abort!
2829 }
2830 # Skip ahead to next tag start
2833 }
2836 # In a <tag>
2843 # Slash may be followed by >?
2847 # ignore
2848 }
2859 # Illegal early close? This shouldn't happen D:
2864 }
2867 # In a </tag>
2874 }
2876 }
2880 # Yes, a self-closed tag <blah/>
2883 # Probably we're jumping the gun, and this is an attribute
2885 }
2890 }
2897 }
2904 }
2909 }
2910 }
2912 wfDebug( __METHOD__ . ": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2915 }
2918 }
2920 /**
2921 * Return value of a magic variable (like PAGENAME)
2922 *
2923 * @private
2924 *
2925 * @param int $index
2926 * @param bool|PPFrame $frame
2927 *
2928 * @throws MWException
2929 * @return string
2930 */
2936 // If no title set, bad things are going to happen
2937 // later. Title should always be set since this
2938 // should only be called in the middle of a parse
2939 // operation (but the unit-tests do funky stuff)
2942 }
2944 /**
2945 * Some of these require message or data lookups and can be
2946 * expensive to check many times.
2947 */
2951 }
2952 }
2997 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
3031 ) ) );
3041 ) ) );
3049 }
3057 }
3070 # 0 means the page doesn't exist in the database,
3071 # which means the user is previewing a new page.
3072 # The vary-revision flag must be set, because the magic word
3073 # will have a different value once the page is saved.
3076 }
3080 # Let the edit saving system know we should parse the page
3081 # *after* a revision ID has been assigned.
3087 # Let the edit saving system know we should parse the page
3088 # *after* a revision ID has been assigned. This is for null edits.
3094 # Let the edit saving system know we should parse the page
3095 # *after* a revision ID has been assigned. This is for null edits.
3101 # Let the edit saving system know we should parse the page
3102 # *after* a revision ID has been assigned. This is for null edits.
3108 # Let the edit saving system know we should parse the page
3109 # *after* a revision ID has been assigned. This is for null edits.
3115 # Let the edit saving system know we should parse the page
3116 # *after* a revision ID has been assigned. This is for null edits.
3122 # Let the edit saving system know we should parse the page
3123 # *after* a revision ID has been assigned. This is for null edits.
3129 # Let the edit saving system know we should parse the page
3130 # *after* a revision ID has been assigned. This is for null edits.
3136 # Let the edit saving system know we should parse the page
3137 # *after* a revision ID has been assigned. This is for null edits.
3178 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3179 # int to remove the padding
3188 );
3197 false
3198 );
3204 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3205 # int to remove the padding
3266 );
3269 }
3273 }
3276 }
3278 /**
3279 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
3280 *
3281 * @private
3282 */
3291 }
3293 /**
3294 * Preprocess some wikitext and return the document tree.
3295 * This is the ghost of replace_variables().
3296 *
3297 * @param string $text The text to parse
3298 * @param int $flags Bitwise combination of:
3299 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
3300 * included. Default is to assume a direct page view.
3301 *
3302 * The generated DOM tree must depend only on the input text and the flags.
3303 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
3304 *
3305 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
3306 * change in the DOM tree for a given text, must be passed through the section identifier
3307 * in the section edit link and thus back to extractSections().
3308 *
3309 * The output of this function is currently only cached in process memory, but a persistent
3310 * cache may be implemented at a later date which takes further advantage of these strict
3311 * dependency requirements.
3312 *
3313 * @return PPNode
3314 */
3318 }
3320 /**
3321 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3322 *
3323 * @param string $s
3324 *
3325 * @return array
3326 */
3336 }
3338 }
3340 /**
3341 * Replace magic variables, templates, and template arguments
3342 * with the appropriate text. Templates are substituted recursively,
3343 * taking care to avoid infinite loops.
3344 *
3345 * Note that the substitution depends on value of $mOutputType:
3346 * self::OT_WIKI: only {{subst:}} templates
3347 * self::OT_PREPROCESS: templates but not extension tags
3348 * self::OT_HTML: all templates and extension tags
3349 *
3350 * @param string $text The text to transform
3351 * @param bool|PPFrame $frame Object describing the arguments passed to the
3352 * template. Arguments may also be provided as an associative array, as
3353 * was the usual case before MW1.12. Providing arguments this way may be
3354 * useful for extensions wishing to perform variable replacement
3355 * explicitly.
3356 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
3357 * double-brace expansion.
3358 * @return string
3359 */
3361 # Is there any text? Also, Prevent too big inclusions!
3364 }
3373 }
3381 }
3383 /**
3384 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3385 *
3386 * @param array $args
3387 *
3388 * @return array
3389 */
3402 }
3405 }
3406 }
3407 }
3410 }
3412 /**
3413 * Warn the user when a parser limitation is reached
3414 * Will warn at most once the user per limitation type
3415 *
3416 * @param string $limitationType Should be one of:
3417 * 'expensive-parserfunction' (corresponding messages:
3418 * 'expensive-parserfunction-warning',
3419 * 'expensive-parserfunction-category')
3420 * 'post-expand-template-argument' (corresponding messages:
3421 * 'post-expand-template-argument-warning',
3422 * 'post-expand-template-argument-category')
3423 * 'post-expand-template-inclusion' (corresponding messages:
3424 * 'post-expand-template-inclusion-warning',
3425 * 'post-expand-template-inclusion-category')
3426 * 'node-count-exceeded' (corresponding messages:
3427 * 'node-count-exceeded-warning',
3428 * 'node-count-exceeded-category')
3429 * 'expansion-depth-exceeded' (corresponding messages:
3430 * 'expansion-depth-exceeded-warning',
3431 * 'expansion-depth-exceeded-category')
3432 * @param string|int|null $current Current value
3433 * @param string|int|null $max Maximum allowed, when an explicit limit has been
3434 * exceeded, provide the values (optional)
3435 */
3437 # does no harm if $current and $max are present but are unnecessary for the message
3442 }
3444 /**
3445 * Return the text of a template, after recursively
3446 * replacing any variables or templates within the template.
3447 *
3448 * @param array $piece The parts of the template
3449 * $piece['title']: the title, i.e. the part before the |
3450 * $piece['parts']: the parameter array
3451 * $piece['lineStart']: whether the brace was at the start of a line
3452 * @param PPFrame $frame The current frame, contains template arguments
3453 * @throws Exception
3454 * @return string The text of the template
3455 */
3460 // Flags
3462 // $text has been filled
3464 // wiki markup in $text should be escaped
3466 // $text is HTML, armour it against wikitext transformation
3468 // Force interwiki transclusion to be done in raw mode not rendered
3470 // $text is a DOM node needing expansion in a child frame
3472 // $text is a DOM node needing expansion in the current frame
3475 # Title object, where $text came from
3478 # $part1 is the bit before the first |, and must contain only title characters.
3479 # Various prefixes will be stripped from it later.
3484 # Original title text preserved for various purposes
3487 # $args is a list of argument nodes, starting from index 0, not including $part1
3488 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3489 # below won't work b/c this $args isn't an object
3495 # SUBST
3501 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3502 # Decide whether to expand template or keep wikitext as-is.
3508 }
3514 }
3515 }
3520 }
3521 }
3523 # Variables
3530 }
3532 }
3533 }
3535 # MSG, MSGNW and RAW
3537 # Check for MSGNW:
3542 # Remove obsolete MSG:
3545 }
3547 # Check for RAW:
3551 }
3552 }
3555 # Parser functions
3565 }
3572 }
3574 # The interface for parser functions allows for extracting
3575 # flags into the local scope. Extract any forwarded flags
3576 # here.
3578 }
3580 }
3582 # Finish mangling title and then check for loops.
3583 # Set $title to a Title object and $titleText to the PDBK
3586 # Split the title into page and subpage
3592 }
3596 # Check for language variants if the template is not found
3599 }
3600 # Do recursion depth check
3608 }
3609 }
3610 }
3612 # Load from database
3620 ) {
3621 // Pass the template arguments as URL parameters.
3622 // "uselang" will have no effect since the Language object
3623 // is forced to the one defined in ParserOptions.
3632 }
3633 }
3635 // Create a new context to execute the special page
3648 }
3658 }
3659 }
3661 # If the title is valid but undisplayable, make a link to it
3665 }
3667 # Interwiki transclusion
3673 # Preprocess it like a template
3676 }
3678 }
3680 # Do infinite loop check
3681 # This has to be done after redirect resolution to avoid infinite loops via redirects
3688 }
3690 }
3692 # If we haven't found text to substitute by now, we're done
3693 # Recover the source wikitext and return it
3698 }
3701 }
3703 # Expand DOM-style return values in a child frame
3705 # Clean up argument array
3711 # Expansion is eligible for the empty-frame cache
3714 # Uncached expansion
3716 }
3717 }
3721 }
3725 }
3727 # Replace raw HTML by a placeholder
3731 # Escape nowiki-style return values
3736 ) {
3737 # Bug 529: if the template begins with a table or block-level
3738 # element, it should be treated as beginning a new line.
3739 # This behavior is somewhat controversial.
3741 }
3744 # Error, oversize inclusion
3746 # Make a working, properly escaped link if possible (bug 23588)
3749 # This will probably not be a working link, but at least it may
3750 # provide some hint of where the problem is
3753 }
3757 }
3763 }
3767 }
3769 /**
3770 * Call a parser function and return an array with text and flags.
3771 *
3772 * The returned array will always contain a boolean 'found', indicating
3773 * whether the parser function was found or not. It may also contain the
3774 * following:
3775 * text: string|object, resulting wikitext or PP DOM object
3776 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3777 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3778 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3779 * nowiki: bool, wiki markup in $text should be escaped
3780 *
3781 * @since 1.21
3782 * @param PPFrame $frame The current frame, contains template arguments
3783 * @param string $function Function name
3784 * @param array $args Arguments to the function
3785 * @throws MWException
3786 * @return array
3787 */
3793 # Case sensitive functions
3797 # Case insensitive functions
3804 }
3805 }
3810 # Workaround for PHP bug 35229 and similar
3815 }
3819 # Convert arguments to PPNodes and collect for appending to $allArgs
3826 }
3827 }
3829 # Add a frame parameter, and pass the arguments as an array
3833 # Convert arguments to plain text and append to $allArgs
3841 }
3842 }
3843 }
3847 # The interface for function hooks allows them to return a wikitext
3848 # string or an array containing the string and any flags. This mungs
3849 # things around to match what this method should return.
3854 );
3858 }
3862 );
3863 }
3869 }
3872 }
3877 }
3882 }
3884 /**
3885 * Get the semi-parsed DOM representation of a template with a given title,
3886 * and its redirect destination title. Cached.
3887 *
3888 * @param Title $title
3889 *
3890 * @return array
3891 */
3900 }
3903 }
3905 # Cache miss, go to the database
3911 }
3919 }
3922 }
3924 /**
3925 * Fetch the current revision of a given title. Note that the revision
3926 * (and even the title) may not exist in the database, so everything
3927 * contributing to the output of the parser should use this method
3928 * where possible, rather than getting the revisions themselves. This
3929 * method also caches its results, so using it benefits performance.
3930 *
3931 * @since 1.24
3932 * @param Title $title
3933 * @return Revision
3934 */
3939 }
3942 // Defaults to Parser::statelessFetchRevision()
3944 );
3945 }
3947 }
3949 /**
3950 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3951 * without passing them on to it.
3952 *
3953 * @since 1.24
3954 * @param Title $title
3955 * @param Parser|bool $parser
3956 * @return Revision
3957 */
3960 }
3962 /**
3963 * Fetch the unparsed text of a template and register a reference to it.
3964 * @param Title $title
3965 * @return array ( string or false, Title )
3966 */
3968 // Defaults to Parser::statelessFetchTemplate()
3977 // If we transclude ourselves, the final result
3978 // will change based on the new version of the page
3980 }
3981 }
3982 }
3984 }
3986 /**
3987 * Fetch the unparsed text of a template and register a reference to it.
3988 * @param Title $title
3989 * @return string|bool
3990 */
3994 }
3996 /**
3997 * Static function to get a template
3998 * Can be overridden via ParserOptions::setTemplateCallback().
3999 *
4000 * @param Title $title
4001 * @param bool|Parser $parser
4002 *
4003 * @return array
4004 */
4010 # Loop to fetch the article, with up to 1 redirect
4012 # Give extensions a chance to select the revision instead
4023 );
4025 }
4026 # Get the revision
4033 }
4035 # If there is no current revision, there is no page
4039 }
4046 # We fetched a rev from a different title; register it too...
4051 }
4060 }
4067 }
4072 }
4075 }
4076 # Redirect?
4079 }
4084 }
4086 /**
4087 * Fetch a file and its title and register a reference to it.
4088 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4089 * @param Title $title
4090 * @param array $options Array of options to RepoGroup::findFile
4091 * @return File|bool
4092 */
4096 }
4098 /**
4099 * Fetch a file and its title and register a reference to it.
4100 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4101 * @param Title $title
4102 * @param array $options Array of options to RepoGroup::findFile
4103 * @return array ( File or false, Title of file )
4104 */
4110 # Register the file as a dependency...
4113 # Update fetched file title
4116 }
4118 }
4120 /**
4121 * Helper function for fetchFileAndTitle.
4122 *
4123 * Also useful if you need to fetch a file but not use it yet,
4124 * for example to get the file's handler.
4125 *
4126 * @param Title $title
4127 * @param array $options Array of options to RepoGroup::findFile
4128 * @return File|bool
4129 */
4137 }
4139 }
4141 /**
4142 * Transclude an interwiki link.
4143 *
4144 * @param Title $title
4145 * @param string $action
4146 *
4147 * @return string
4148 */
4154 }
4160 }
4162 }
4164 /**
4165 * @param string $url
4166 * @return mixed|string
4167 */
4176 }
4183 // Though we failed to fetch the content, this status is useless.
4188 }
4195 ) );
4197 }
4199 /**
4200 * Triple brace replacement -- used for template arguments
4201 * @private
4202 *
4203 * @param array $piece
4204 * @param PPFrame $frame
4205 *
4206 * @return array
4207 */
4221 )
4222 ) {
4223 # No match in frame, use the supplied default
4225 }
4229 }
4232 # No match anywhere
4234 }
4237 }
4242 }
4246 }
4248 /**
4249 * Return the text to be used for a given extension tag.
4250 * This is the ghost of strip().
4251 *
4252 * @param array $params Associative array of parameters:
4253 * name PPNode for the tag name
4254 * attr PPNode for unparsed text where tag attributes are thought to be
4255 * attributes Optional associative array of parsed attributes
4256 * inner Contents of extension element
4257 * noClose Original text did not have a close tag
4258 * @param PPFrame $frame
4259 *
4260 * @throws MWException
4261 * @return string
4262 */
4276 }
4282 }
4285 # Workaround for PHP bug 35229 and similar
4288 }
4295 }
4301 }
4304 # Extract flags to local scope (to override $markerType)
4309 }
4313 }
4318 }
4319 }
4325 }
4326 }
4336 }
4338 }
4340 /**
4341 * Increment an include size counter
4342 *
4343 * @param string $type The type of expansion
4344 * @param int $size The size of the text
4345 * @return bool False if this inclusion would take it over the maximum, true otherwise
4346 */
4353 }
4354 }
4356 /**
4357 * Increment the expensive function count
4358 *
4359 * @return bool False if the limit has been exceeded
4360 */
4364 }
4366 /**
4367 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
4368 * Fills $this->mDoubleUnderscores, returns the modified text
4369 *
4370 * @param string $text
4371 *
4372 * @return string
4373 */
4377 # The position of __TOC__ needs to be recorded
4383 # Set a placeholder. At the end we'll fill it in with the TOC.
4386 # Only keep the first one.
4388 }
4390 # Now match and remove the rest of them
4396 }
4399 }
4402 ) {
4404 }
4405 # (bug 8068) Allow control over whether robots index a page.
4406 #
4407 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
4408 # is not desirable, the last one on the page should win.
4412 }
4416 }
4418 # Cache all double underscores in the database
4421 }
4425 }
4427 /**
4428 * @see ParserOutput::addTrackingCategory()
4429 * @param string $msg Message key
4430 * @return bool Whether the addition was successful
4431 */
4434 }
4436 /**
4437 * This function accomplishes several tasks:
4438 * 1) Auto-number headings if that option is enabled
4439 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
4440 * 3) Add a Table of contents on the top for users who have enabled the option
4441 * 4) Auto-anchor headings
4442 *
4443 * It loops through all headlines, collects the necessary data, then splits up the
4444 * string and re-inserts the newly formatted headlines.
4445 *
4446 * @param string $text
4447 * @param string $origText Original, untouched wikitext
4448 * @param bool $isMain
4449 * @return mixed|string
4450 * @private
4451 */
4455 # Inhibit editsection links if requested in the page
4461 }
4464 }
4466 # Get all headlines for numbering them and adding funky stuff like [edit]
4467 # links - this is for later, but we need the number of headlines right now
4472 $matches
4473 );
4475 # if there are fewer than 4 headlines in the article, do not show TOC
4476 # unless it's been explicitly enabled.
4480 # Allow user to stipulate that a page should have a "new section"
4481 # link added via __NEWSECTIONLINK__
4484 }
4486 # Allow user to remove the "new section"
4487 # link via __NONEWSECTIONLINK__
4490 }
4492 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4493 # override above conditions and always show TOC above first header
4497 }
4499 # headline counter
4503 # Ugh .. the TOC should have neat indentation levels which can be
4504 # passed to the skin functions. These are determined here
4536 }
4540 }
4544 # Increase TOC level
4551 }
4553 # Decrease TOC level, find level to jump to
4557 # Found last matching level
4561 # Found first matching level below current level
4564 }
4565 }
4568 }
4571 # Unindent only if the previous toc level was shown :p
4576 }
4577 }
4579 # No change in level, end TOC line
4582 }
4583 }
4587 # count number of headlines for each level
4594 }
4597 }
4598 }
4600 # The safe header is a version of the header text safe to use for links
4602 # Remove link placeholders by the link text.
4603 # <!--LINK number-->
4604 # turns into
4605 # link text with suffix
4606 # Do this before unstrip since link text can contain strip markers
4609 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4612 # Strip out HTML (first regex removes any tag not allowed)
4613 # Allowed tags are:
4614 # * <sup> and <sub> (bug 8393)
4615 # * <i> (bug 26375)
4616 # * <b> (r105284)
4617 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4618 #
4619 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4620 # to allow setting directionality in toc items.
4625 ),
4627 $safeHeadline
4628 );
4631 # For the anchor, strip out HTML-y stuff period
4635 # Save headline for section edit hint before it's escaped
4639 # For reverse compatibility, provide an id that's
4640 # HTML4-compatible, like we used to.
4641 #
4642 # It may be worth noting, academically, that it's possible for
4643 # the legacy anchor to conflict with a non-legacy headline
4644 # anchor on the page. In this case likely the "correct" thing
4645 # would be to either drop the legacy anchors or make sure
4646 # they're numbered first. However, this would require people
4647 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4648 # manually, so let's not bother worrying about it.
4654 # No reason to have both (in fact, we can't)
4656 }
4661 }
4663 # HTML names must be case-insensitively unique (bug 10721).
4664 # This does not apply to Unicode characters per
4665 # http://dev.w3.org/html5/spec/infrastructure.html#case-sensitivity-and-string-comparison
4666 # @todo FIXME: We may be changing them depending on the current locale.
4672 }
4674 # count how many in assoc. array so we can track dupes in anchors
4679 }
4684 }
4686 # Don't number the heading if it is the only one (looks silly)
4688 # the two are different if the line contains a link
4692 $numbering
4694 }
4696 # Create the anchor for linking from the TOC to the section
4701 }
4704 }
4708 }
4710 # Add the section to the section tree
4711 # Find the DOM node for this header
4718 }
4719 }
4723 }
4733 );
4735 # give headline the correct <h#> tag
4737 // Output edit section links as markers with styles that can be customized by skins
4739 # Put a T flag in the section identifier, to indicate to extractSections()
4740 # that sections inside <includeonly> should be counted.
4748 }
4749 // We use a bit of pesudo-xml for editsection markers. The
4750 // language converter is run later on. Using a UNIQ style marker
4751 // leads to the converter screwing up the tokens when it
4752 // converts stuff. And trying to insert strip tags fails too. At
4753 // this point all real inputted tags have already been escaped,
4754 // so we don't have to worry about a user trying to input one of
4755 // these markers directly. We use a page and section attribute
4756 // to stop the language converter from converting these
4757 // important bits of data, but put the headline hint inside a
4758 // content block because the language converter is supposed to
4759 // be able to convert that piece of data.
4760 // Gets replaced with html in ParserOutput::getText
4767 }
4770 }
4776 }
4780 # Never ever show TOC if no headers
4783 }
4788 }
4793 }
4797 }
4799 # split up and insert constructed headlines
4803 // build an array of document sections
4806 // $head is zero-based, sections aren't.
4811 }
4813 /**
4814 * Send a hook, one per section.
4815 * The idea here is to be able to make section-level DIVs, but to do so in a
4816 * lower-impact, more correct way than r50769
4817 *
4818 * $this : caller
4819 * $section : the section number
4820 * &$sectionContent : ref to the content of the section
4821 * $showEditLinks : boolean describing whether this section has an edit link
4822 */
4826 }
4829 // append the TOC at the beginning
4830 // Top anchor now in skin
4832 }
4840 }
4841 }
4843 /**
4844 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4845 * conversion, substituting signatures, {{subst:}} templates, etc.
4846 *
4847 * @param string $text The text to transform
4848 * @param Title $title The Title object for the current article
4849 * @param User $user The User object describing the current user
4850 * @param ParserOptions $options Parsing options
4851 * @param bool $clearState Whether to clear the parser state first
4852 * @return string The altered wiki markup
4853 */
4856 ) {
4859 }
4865 );
4869 }
4875 }
4877 /**
4878 * Pre-save transform helper function
4879 *
4880 * @param string $text
4881 * @param User $user
4882 *
4883 * @return string
4884 */
4888 # Note: This is the timestamp saved as hardcoded wikitext to
4889 # the database, we use $wgContLang here in order to give
4890 # everyone the same signature and use the default one rather
4891 # than the one selected in each user's preferences.
4892 # (see also bug 12815)
4898 # Allow translation of timezones through wiki. format() can return
4899 # whatever crap the system uses, localised or not, so we cannot
4900 # ship premade translations.
4905 }
4909 # Variable replacement
4910 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4913 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4914 # which may corrupt this parser instance via its wfMessage()->text() call-
4916 # Signatures
4922 ) );
4924 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4928 // [[ns:page (context)|]]
4930 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4932 // [[ns:page (context), context|]] (using either single or double-width comma)
4934 // [[|page]] (reverse pipe trick: add context from page title)
4937 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4949 # if there's no context, don't bother duplicating the title
4951 }
4953 # Trim trailing whitespace
4957 }
4959 /**
4960 * Fetch the user's signature text, if any, and normalize to
4961 * validated, ready-to-insert wikitext.
4962 * If you have pre-fetched the nickname or the fancySig option, you can
4963 * specify them here to save a database query.
4964 * Do not reuse this parser instance after calling getUserSig(),
4965 * as it may have changed if it's the $wgParser.
4966 *
4967 * @param User $user
4968 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4969 * @param bool|null $fancySig whether the nicknname is the complete signature
4970 * or null to use default value
4971 * @return string
4972 */
4978 # If not given, retrieve from the user object.
4981 }
4985 }
4993 # Sig. might contain markup; validate this
4995 # Validated; clean up (if needed) and return it
4998 # Failed to validate; fall back to the default
5001 }
5002 }
5004 # Make sure nickname doesnt get a sig in a sig
5007 # If we're still here, make it a link to the user page
5014 }
5016 /**
5017 * Check that the user's signature contains no bad XML
5018 *
5019 * @param string $text
5020 * @return string|bool An expanded string, or false if invalid.
5021 */
5024 }
5026 /**
5027 * Clean up signature text
5028 *
5029 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
5030 * 2) Substitute all transclusions
5031 *
5032 * @param string $text
5033 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
5034 * @return string Signature text
5035 */
5041 }
5043 # Option to disable this feature
5046 }
5048 # @todo FIXME: Regex doesn't respect extension tags or nowiki
5049 # => Move this logic to braceSubstitution()
5051 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
5062 }
5065 }
5067 /**
5068 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
5069 *
5070 * @param string $text
5071 * @return string Signature text with /~{3,5}/ removed
5072 */
5076 }
5078 /**
5079 * Set up some variables which are usually set up in parse()
5080 * so that an external function can call some class members with confidence
5081 *
5082 * @param Title|null $title
5083 * @param ParserOptions $options
5084 * @param int $outputType
5085 * @param bool $clearState
5086 */
5089 ) {
5091 }
5093 /**
5094 * @param Title|null $title
5095 * @param ParserOptions $options
5096 * @param int $outputType
5097 * @param bool $clearState
5098 */
5101 ) {
5107 }
5108 }
5110 /**
5111 * Wrapper for preprocess()
5112 *
5113 * @param string $text The text to preprocess
5114 * @param ParserOptions $options Options
5115 * @param Title|null $title Title object or null to use $wgTitle
5116 * @return string
5117 */
5121 # Guard against infinite recursion
5124 }
5131 }
5138 }
5140 /**
5141 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
5142 * The callback should have the following form:
5143 * function myParserHook( $text, $params, $parser, $frame ) { ... }
5144 *
5145 * Transform and return $text. Use $parser for any required context, e.g. use
5146 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
5147 *
5148 * Hooks may return extended information by returning an array, of which the
5149 * first numbered element (index 0) must be the return string, and all other
5150 * entries are extracted into local variables within an internal function
5151 * in the Parser class.
5152 *
5153 * This interface (introduced r61913) appears to be undocumented, but
5154 * 'markerName' is used by some core tag hooks to override which strip
5155 * array their results are placed in. **Use great caution if attempting
5156 * this interface, as it is not documented and injudicious use could smash
5157 * private variables.**
5158 *
5159 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5160 * @param callable $callback The callback function (and object) to use for the tag
5161 * @throws MWException
5162 * @return callable|null The old value of the mTagHooks array associated with the hook
5163 */
5168 }
5173 }
5176 }
5178 /**
5179 * As setHook(), but letting the contents be parsed.
5180 *
5181 * Transparent tag hooks are like regular XML-style tag hooks, except they
5182 * operate late in the transformation sequence, on HTML instead of wikitext.
5183 *
5184 * This is probably obsoleted by things dealing with parser frames?
5185 * The only extension currently using it is geoserver.
5186 *
5187 * @since 1.10
5188 * @todo better document or deprecate this
5189 *
5190 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5191 * @param callable $callback The callback function (and object) to use for the tag
5192 * @throws MWException
5193 * @return callable|null The old value of the mTagHooks array associated with the hook
5194 */
5199 }
5200 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
5204 }
5206 /**
5207 * Remove all tag hooks
5208 */
5213 }
5215 /**
5216 * Create a function, e.g. {{sum:1|2|3}}
5217 * The callback function should have the form:
5218 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
5219 *
5220 * Or with Parser::SFH_OBJECT_ARGS:
5221 * function myParserFunction( $parser, $frame, $args ) { ... }
5222 *
5223 * The callback may either return the text result of the function, or an array with the text
5224 * in element 0, and a number of flags in the other elements. The names of the flags are
5225 * specified in the keys. Valid flags are:
5226 * found The text returned is valid, stop processing the template. This
5227 * is on by default.
5228 * nowiki Wiki markup in the return value should be escaped
5229 * isHTML The returned text is HTML, armour it against wikitext transformation
5230 *
5231 * @param string $id The magic word ID
5232 * @param callable $callback The callback function (and object) to use
5233 * @param int $flags A combination of the following flags:
5234 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
5235 *
5236 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
5237 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
5238 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
5239 * the arguments, and to control the way they are expanded.
5240 *
5241 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
5242 * arguments, for instance:
5243 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
5244 *
5245 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
5246 * future versions. Please call $frame->expand() on it anyway so that your code keeps
5247 * working if/when this is changed.
5248 *
5249 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
5250 * expansion.
5251 *
5252 * Please read the documentation in includes/parser/Preprocessor.php for more information
5253 * about the methods available in PPFrame and PPNode.
5254 *
5255 * @throws MWException
5256 * @return string|callable The old callback function for this name, if any
5257 */
5264 # Add to function cache
5268 }
5274 # Case
5277 }
5278 # Add leading hash
5281 }
5282 # Remove trailing colon
5285 }
5287 }
5289 }
5291 /**
5292 * Get all registered function hook identifiers
5293 *
5294 * @return array
5295 */
5298 }
5300 /**
5301 * Create a tag function, e.g. "<test>some stuff</test>".
5302 * Unlike tag hooks, tag functions are parsed at preprocessor level.
5303 * Unlike parser functions, their content is not preprocessed.
5304 * @param string $tag
5305 * @param callable $callback
5306 * @param int $flags
5307 * @throws MWException
5308 * @return null
5309 */
5314 }
5321 }
5324 }
5326 /**
5327 * @todo FIXME: Update documentation. makeLinkObj() is deprecated.
5328 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
5329 * Placeholders created in Skin::makeLinkObj()
5330 *
5331 * @param string $text
5332 * @param int $options
5333 */
5336 }
5338 /**
5339 * Replace "<!--LINK-->" link placeholders with plain text of links
5340 * (not HTML-formatted).
5341 *
5342 * @param string $text
5343 * @return string
5344 */
5347 }
5349 /**
5350 * Renders an image gallery from a text with one line per image.
5351 * text labels may be given by using |-style alternative text. E.g.
5352 * Image:one.jpg|The number "1"
5353 * Image:tree.jpg|A tree
5354 * given as text will return the HTML of a gallery with two images,
5355 * labeled 'The number "1"' and
5356 * 'A tree'.
5357 *
5358 * @param string $text
5359 * @param array $params
5360 * @return string HTML
5361 */
5368 }
5373 // If invalid type set, fallback to default.
5375 }
5388 }
5394 }
5397 }
5400 }
5403 }
5410 # match lines like these:
5411 # Image:someimage.jpg|This is some image
5414 # Skip empty lines
5417 }
5421 }
5424 # Bogus title. Ignore these so we don't bomb out later.
5426 }
5428 # We need to get what handler the file uses, to figure out parameters.
5429 # Note, a hook can overide the file name, and chose an entirely different
5430 # file (which potentially could be of a different type and have different handler).
5435 # Don't register it now, as ImageGallery does that later.
5443 );
5446 // We don't want people to specify per-image widths.
5447 // Additionally the width parameter would need special casing anyhow.
5449 }
5459 // look for an |alt= definition while trying not to break existing
5460 // captions with multiple pipes (|) in it, until a more sensible grammar
5461 // is defined for images in galleries
5463 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5464 // splitting on '|' is a bit odd, and different from makeImage.
5481 //check to see if link matches an absolute url, if not then it must be a wiki link.
5488 }
5489 }
5492 // Must be a handler specific parameter.
5496 // Guess not. Append it to the caption.
5499 }
5500 }
5503 // concatenate all other pipes
5505 }
5506 }
5507 // remove the first pipe
5509 }
5512 }
5517 }
5519 /**
5520 * @param string $handler
5521 * @return array
5522 */
5528 }
5530 # Initialise static lists
5537 );
5545 }
5546 }
5547 }
5549 # Add handler params
5555 }
5556 }
5559 }
5560 return array( $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] );
5561 }
5563 /**
5564 * Parse image options text and use it to make an image
5565 *
5566 * @param Title $title
5567 * @param string $options
5568 * @param LinkHolderArray|bool $holders
5569 * @return string HTML
5570 */
5572 # Check if the options text is of the form "options|alt text"
5573 # Options are:
5574 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5575 # * left no resizing, just left align. label is used for alt= only
5576 # * right same, but right aligned
5577 # * none same, but not aligned
5578 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5579 # * center center the image
5580 # * frame Keep original image size, no magnify-button.
5581 # * framed Same as "frame"
5582 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5583 # * upright reduce width for upright images, rounded to full __0 px
5584 # * border draw a 1px border around the image
5585 # * alt Text for HTML alt attribute (defaults to empty)
5586 # * class Set a class for img node
5587 # * link Set the target of the image link. Can be external, interwiki, or local
5588 # vertical-align values (no % or length right now):
5589 # * baseline
5590 # * sub
5591 # * super
5592 # * top
5593 # * text-top
5594 # * middle
5595 # * bottom
5596 # * text-bottom
5600 # Give extensions a chance to select the file revision for us
5605 # Fetch and register the file (file title may be different via hooks)
5608 # Get parameter map
5615 }
5617 # Process the input parameters
5629 # Special case; width and height come in one variable together
5637 }
5638 }
5644 }
5645 }
5646 # else no validation -- bug 13436
5649 # Validate handler parameter
5652 # Validate internal parameters
5657 # @todo FIXME: Possibly check validity here for
5658 # manualthumb? downstream behavior seems odd with
5659 # missing manual thumbs.
5676 }
5678 }
5686 }
5687 }
5692 // use first appearing option, discard others.
5697 # Most other things appear to be empty or numeric...
5699 }
5700 }
5704 }
5705 }
5706 }
5709 }
5710 }
5712 # Process alignment parameters
5715 }
5718 }
5722 # Will the image be presented in a frame, with the caption below?
5728 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5729 # came to also set the caption, ordinary text after the image -- which
5730 # makes no sense, because that just repeats the text multiple times in
5731 # screen readers. It *also* came to set the title attribute.
5732 #
5733 # Now that we have an alt attribute, we should not set the alt text to
5734 # equal the caption: that's worse than useless, it just repeats the
5735 # text. This is the framed/thumbnail case. If there's no caption, we
5736 # use the unnamed parameter for alt text as well, just for the time be-
5737 # ing, if the unnamed param is set and the alt param is not.
5738 #
5739 # For the future, we need to figure out if we want to tweak this more,
5740 # e.g., introducing a title= parameter for the title; ignoring the un-
5741 # named parameter entirely for images without a caption; adding an ex-
5742 # plicit caption= parameter and preserving the old magic unnamed para-
5743 # meter for BC; ...
5746 # No caption or alt text, add the filename as the alt text so
5747 # that screen readers at least get some description of the image
5749 }
5750 # Do not set $params['frame']['title'] because tooltips don't make sense
5751 # for framed images
5754 # No alt text, use the "caption" for the alt text
5758 # No caption, fall back to using the filename for the
5759 # alt text
5761 }
5762 }
5763 # Use the "caption" for the tooltip text
5765 }
5769 # Linker does the rest
5774 # Give the handler a chance to modify the parser object
5777 }
5780 }
5782 /**
5783 * @param string $caption
5784 * @param LinkHolderArray|bool $holders
5785 * @return mixed|string
5786 */
5788 # Strip bad stuff out of the title (tooltip). We can't just use
5789 # replaceLinkHoldersText() here, because if this function is called
5790 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5795 }
5797 # make sure there are no placeholders in thumbnail attributes
5798 # that are later expanded to html- so expand them now and
5799 # remove the tags
5804 }
5806 /**
5807 * Set a flag in the output object indicating that the content is dynamic and
5808 * shouldn't be cached.
5809 */
5815 }
5818 }
5820 /**
5821 * Callback from the Sanitizer for expanding items found in HTML attribute
5822 * values, so they can be safely tested and escaped.
5823 *
5824 * @param string $text
5825 * @param bool|PPFrame $frame
5826 * @return string
5827 */
5832 }
5834 /**
5835 * Accessor
5836 *
5837 * @return array
5838 */
5844 );
5845 }
5847 /**
5848 * Replace transparent tags in $text with the values given by the callbacks.
5849 *
5850 * Transparent tag hooks are like regular XML-style tag hooks, except they
5851 * operate late in the transformation sequence, on HTML instead of wikitext.
5852 *
5853 * @param string $text
5854 *
5855 * @return string
5856 */
5870 );
5873 }
5875 }
5877 }
5879 /**
5880 * Break wikitext input into sections, and either pull or replace
5881 * some particular section's text.
5882 *
5883 * External callers should use the getSection and replaceSection methods.
5884 *
5885 * @param string $text Page wikitext
5886 * @param string|number $sectionId A section identifier string of the form:
5887 * "<flag1> - <flag2> - ... - <section number>"
5888 *
5889 * Currently the only recognised flag is "T", which means the target section number
5890 * was derived during a template inclusion parse, in other words this is a template
5891 * section edit link. If no flags are given, it was an ordinary section edit link.
5892 * This flag is required to avoid a section numbering mismatch when a section is
5893 * enclosed by "<includeonly>" (bug 6563).
5894 *
5895 * The section number 0 pulls the text before the first heading; other numbers will
5896 * pull the given section along with its lower-level subsections. If the section is
5897 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5898 *
5899 * Section 0 is always considered to exist, even if it only contains the empty
5900 * string. If $text is the empty string and section 0 is replaced, $newText is
5901 * returned.
5902 *
5903 * @param string $mode One of "get" or "replace"
5904 * @param string $newText Replacement text for section data.
5905 * @return string For "get", the extracted section text.
5906 * for "replace", the whole page with the section replaced.
5907 */
5916 # Process section extraction flags
5923 }
5924 }
5926 # Check for empty input
5928 # Only sections 0 and T-0 exist in an empty document
5934 }
5940 }
5941 }
5942 }
5944 # Preprocess the text
5947 # <h> nodes indicate section breaks
5948 # They can only occur at the top level, so we can find them by iterating the root's children
5951 # Find the target section
5953 # Section zero doesn't nest, level=big
5962 }
5963 }
5966 }
5968 }
5969 }
5972 # Not found
5977 }
5978 }
5980 # Find the end of the section, including nested sections
5987 }
5988 }
5991 }
5995 # Write out the remainder (in replace mode only)
5997 # Output the replacement text
5998 # Add two newlines on -- trailing whitespace in $newText is conventionally
5999 # stripped by the editor, so we need both newlines to restore the paragraph gap
6000 # Only add trailing whitespace if there is newText
6003 }
6008 }
6009 }
6012 # Re-insert stripped tags
6014 }
6017 }
6019 /**
6020 * This function returns the text of a section, specified by a number ($section).
6021 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
6022 * the first section before any such heading (section 0).
6023 *
6024 * If a section contains subsections, these are also returned.
6025 *
6026 * @param string $text Text to look in
6027 * @param string|number $sectionId Section identifier as a number or string
6028 * (e.g. 0, 1 or 'T-1').
6029 * @param string $defaultText Default to return if section is not found
6030 *
6031 * @return string Text of the requested section
6032 */
6035 }
6037 /**
6038 * This function returns $oldtext after the content of the section
6039 * specified by $section has been replaced with $text. If the target
6040 * section does not exist, $oldtext is returned unchanged.
6041 *
6042 * @param string $oldText Former text of the article
6043 * @param string|number $sectionId Section identifier as a number or string
6044 * (e.g. 0, 1 or 'T-1').
6045 * @param string $newText Replacing text
6046 *
6047 * @return string Modified text
6048 */
6051 }
6053 /**
6054 * Get the ID of the revision we are parsing
6055 *
6056 * @return int|null
6057 */
6060 }
6062 /**
6063 * Get the revision object for $this->mRevisionId
6064 *
6065 * @return Revision|null Either a Revision object or null
6066 * @since 1.23 (public since 1.23)
6067 */
6071 }
6074 }
6078 }
6080 /**
6081 * Get the timestamp associated with the current revision, adjusted for
6082 * the default server-local timestamp
6083 * @return string
6084 */
6094 # The cryptic '' timezone parameter tells to use the site-default
6095 # timezone offset instead of the user settings.
6096 #
6097 # Since this value will be saved into the parser cache, served
6098 # to other users, and potentially even used inside links and such,
6099 # it needs to be consistent for all visitors.
6103 }
6105 }
6107 /**
6108 * Get the name of the user that edited the last revision
6109 *
6110 * @return string User name
6111 */
6116 # if this template is subst: the revision id will be blank,
6117 # so just use the current user's name
6122 }
6123 }
6125 }
6127 /**
6128 * Get the size of the revision
6129 *
6130 * @return int|null Revision size
6131 */
6136 # if this variable is subst: the revision id will be blank,
6137 # so just use the parser input size, because the own substituation
6138 # will change the size.
6143 }
6144 }
6146 }
6148 /**
6149 * Mutator for $mDefaultSort
6150 *
6151 * @param string $sort New value
6152 */
6156 }
6158 /**
6159 * Accessor for $mDefaultSort
6160 * Will use the empty string if none is set.
6161 *
6162 * This value is treated as a prefix, so the
6163 * empty string is equivalent to sorting by
6164 * page name.
6165 *
6166 * @return string
6167 */
6173 }
6174 }
6176 /**
6177 * Accessor for $mDefaultSort
6178 * Unlike getDefaultSort(), will return false if none is set
6179 *
6180 * @return string|bool
6181 */
6184 }
6186 /**
6187 * Try to guess the section anchor name based on a wikitext fragment
6188 * presumably extracted from a heading, for example "Header" from
6189 * "== Header ==".
6190 *
6191 * @param string $text
6192 *
6193 * @return string
6194 */
6196 # Strip out wikitext links(they break the anchor)
6200 }
6202 /**
6203 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
6204 * instead. For use in redirects, since IE6 interprets Redirect: headers
6205 * as something other than UTF-8 (apparently?), resulting in breakage.
6206 *
6207 * @param string $text The section name
6208 * @return string An anchor
6209 */
6211 # Strip out wikitext links(they break the anchor)
6215 }
6217 /**
6218 * Strips a text string of wikitext for use in a section anchor
6219 *
6220 * Accepts a text string and then removes all wikitext from the
6221 * string and leaves only the resultant text (i.e. the result of
6222 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
6223 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
6224 * to create valid section anchors by mimicing the output of the
6225 * parser when headings are parsed.
6226 *
6227 * @param string $text Text string to be stripped of wikitext
6228 * for use in a Section anchor
6229 * @return string Filtered text string
6230 */
6232 # Strip internal link markup
6236 # Strip external link markup
6237 # @todo FIXME: Not tolerant to blank link text
6238 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
6239 # on how many empty links there are on the page - need to figure that out.
6240 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
6242 # Parse wikitext quotes (italics & bold)
6245 # Strip HTML tags
6248 }
6250 /**
6251 * strip/replaceVariables/unstrip for preprocessor regression testing
6252 *
6253 * @param string $text
6254 * @param Title $title
6255 * @param ParserOptions $options
6256 * @param int $outputType
6257 *
6258 * @return string
6259 */
6260 public function testSrvus( $text, Title $title, ParserOptions $options, $outputType = self::OT_HTML ) {
6268 }
6270 /**
6271 * @param string $text
6272 * @param Title $title
6273 * @param ParserOptions $options
6274 * @return string
6275 */
6278 }
6280 /**
6281 * @param string $text
6282 * @param Title $title
6283 * @param ParserOptions $options
6284 * @return string
6285 */
6288 }
6290 /**
6291 * Call a callback function on all regions of the given text that are not
6292 * inside strip markers, and replace those regions with the return value
6293 * of the callback. For example, with input:
6294 *
6295 * aaa<MARKER>bbb
6296 *
6297 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
6298 * two strings will be replaced with the value returned by the callback in
6299 * each case.
6300 *
6301 * @param string $s
6302 * @param callable $callback
6303 *
6304 * @return string
6305 */
6324 }
6325 }
6326 }
6328 }
6330 /**
6331 * Remove any strip markers found in the given text.
6332 *
6333 * @param string $text Input string
6334 * @return string
6335 */
6338 }
6340 /**
6341 * Save the parser state required to convert the given half-parsed text to
6342 * HTML. "Half-parsed" in this context means the output of
6343 * recursiveTagParse() or internalParse(). This output has strip markers
6344 * from replaceVariables (extensionSubstitution() etc.), and link
6345 * placeholders from replaceLinkHolders().
6346 *
6347 * Returns an array which can be serialized and stored persistently. This
6348 * array can later be loaded into another parser instance with
6349 * unserializeHalfParsedText(). The text can then be safely incorporated into
6350 * the return value of a parser hook.
6351 *
6352 * @param string $text
6353 *
6354 * @return array
6355 */
6363 );
6366 }
6368 /**
6369 * Load the parser state given in the $data array, which is assumed to
6370 * have been generated by serializeHalfParsedText(). The text contents is
6371 * extracted from the array, and its markers are transformed into markers
6372 * appropriate for the current Parser instance. This transformed text is
6373 * returned, and can be safely included in the return value of a parser
6374 * hook.
6375 *
6376 * If the $data array has been stored persistently, the caller should first
6377 * check whether it is still valid, by calling isValidHalfParsedText().
6378 *
6379 * @param array $data Serialized data
6380 * @throws MWException
6381 * @return string
6382 */
6386 }
6388 # First, extract the strip state.
6392 # Now renumber links
6395 # Should be good to go.
6397 }
6399 /**
6400 * Returns true if the given array, presumed to be generated by
6401 * serializeHalfParsedText(), is compatible with the current version of the
6402 * parser.
6403 *
6404 * @param array $data
6405 *
6406 * @return bool
6407 */
6410 }
6412 /**
6413 * Parsed a width param of imagelink like 300px or 200x300px
6414 *
6415 * @param string $value
6416 *
6417 * @return array
6418 * @since 1.20
6419 */
6424 }
6426 # (bug 13500) In both cases (width/height and width only),
6427 # permit trailing "px" for backward compatibility.
6436 }
6438 }
6440 /**
6441 * Lock the current instance of the parser.
6442 *
6443 * This is meant to stop someone from calling the parser
6444 * recursively and messing up all the strip state.
6445 *
6446 * @throws MWException If parser is in a parse
6447 * @return ScopedCallback The lock will be released once the return value goes out of scope.
6448 */
6453 }
6459 } );
6462 }
6464 /**
6465 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6466 *
6467 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6468 * or if there is more than one <p/> tag in the input HTML.
6469 *
6470 * @param string $html
6471 * @return string
6472 * @since 1.24
6473 */
6479 }
6480 }
6483 }
6485 /**
6486 * Return this parser if it is not doing anything, otherwise
6487 * get a fresh parser. You can use this method by doing
6488 * $myParser = $wgParser->getFreshParser(), or more simply
6489 * $wgParser->getFreshParser()->parse( ... );
6490 * if you're unsure if $wgParser is safe to use.
6491 *
6492 * @since 1.24
6493 * @return Parser A parser object that is not parsing anything
6494 */
6501 }
6502 }
6503 }