| blob | 2f443c016450a686a75510416d4ca102dde85da4 |
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 # Regular expression for a non-newline space
96 # State constants for the definition list colon extraction
106 # Flags for preprocessToDom
109 # Allowed values for $this->mOutputType
110 # Parameter to startExternalParse().
117 # Marker Suffix needs to be accessible staticly.
120 # Markers used for wrapping the table of contents
124 # Persistent:
138 # Initialised by initialiseVariables()
140 /**
141 * @var MagicWordArray
142 */
145 /**
146 * @var MagicWordArray
147 */
149 # Initialised in constructor
152 # Cleared with clearState():
153 /**
154 * @var ParserOutput
155 */
159 /**
160 * @var StripState
161 */
165 /**
166 * @var LinkHolderArray
167 */
177 /**
178 * @var User
179 */
182 # Temporary
183 # These are variables reset at least once per parse regardless of $clearState
185 /**
186 * @var ParserOptions
187 */
190 /**
191 * @var Title
192 */
204 /**
205 * @var string
206 */
209 /**
210 * @var array Array with the language name of each language link (i.e. the
211 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
212 * duplicate language links to the ParserOutput.
213 */
216 /**
217 * @var MapCacheLRU|null
218 * @since 1.24
219 *
220 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
221 */
224 /**
225 * @var bool Recursive call protection.
226 * This variable should be treated as if it were private.
227 */
230 /** @var SectionProfiler */
233 /**
234 * @param array $conf
235 */
244 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
247 # PECL extension that conflicts with the core DOM extension (bug 13770)
254 }
256 }
258 /**
259 * Reduce memory usage to reduce the impact of circular references
260 */
264 }
267 }
268 }
270 /**
271 * Allow extensions to clean up when the parser is cloned
272 */
276 // Bug 56226: When you create a reference "to" an object field, that
277 // makes the object field itself be a reference too (until the other
278 // reference goes out of scope). When cloning, any field that's a
279 // reference is copied as a reference in the new object. Both of these
280 // are defined PHP5 behaviors, as inconvenient as it is for us when old
281 // hooks from PHP4 days are passing fields by reference.
283 // Make a non-reference copy of the field, then rebind the field to
284 // reference the new copy.
288 }
291 }
293 /**
294 * Do various kinds of initialisation on the first call of the parser
295 */
299 }
307 }
309 /**
310 * Clear Parser state
311 *
312 * @private
313 */
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 }
373 }
375 /**
376 * Convert wikitext to HTML
377 * Do not call this function recursively.
378 *
379 * @param string $text Text we want to parse
380 * @param Title $title
381 * @param ParserOptions $options
382 * @param bool $linestart
383 * @param bool $clearState
384 * @param int $revid Number to pass in {{REVISIONID}}
385 * @return ParserOutput A ParserOutput
386 */
389 ) {
390 /**
391 * First pass--just handle <nowiki> sections, pass the rest off
392 * to internalParse() which does all the real work.
393 */
400 }
408 }
410 # Remove the strip marker tag prefix from the input, if present.
413 }
426 }
429 # No more strip!
436 /**
437 * A converted title will be provided in the output object if title and
438 * content conversion are enabled, the article text does not contain
439 * a conversion-suppressing double-underscore tag, and no
440 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
441 * automatic link conversion.
442 */
447 ) {
454 }
455 }
461 );
462 }
464 # Information on include size limits, for the benefit of users who try to skirt them
472 );
473 }
478 );
482 );
485 );
488 );
491 );
494 );
497 );
503 }
507 ) ) {
513 }
517 }
518 }
519 }
520 // Since we're not really outputting HTML, decode the entities and
521 // then re-encode the things that need hiding inside HTML comments.
525 // Sanitize for comment. Note '‐' in the replacement is U+2010,
526 // which looks much like the problematic '-'.
530 // Add on template profiling data
534 } );
540 }
546 }
547 }
559 }
561 /**
562 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
563 * can be called from an extension tag hook.
564 *
565 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
566 * instead, which means that lists and links have not been fully parsed yet,
567 * and strip markers are still present.
568 *
569 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
570 *
571 * Use this function if you're a parser tag hook and you want to parse
572 * wikitext before or after applying additional transformations, and you
573 * intend to *return the result as hook output*, which will cause it to go
574 * through the rest of parsing process automatically.
575 *
576 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
577 * $text are not expanded
578 *
579 * @param string $text Text extension wants to have parsed
580 * @param bool|PPFrame $frame The frame to use for expanding any template variables
581 * @return string UNSAFE half-parsed HTML
582 */
588 }
590 /**
591 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
592 * point can be called from an extension tag hook.
593 *
594 * The output of this function is fully-parsed HTML that is safe for output.
595 * If you're a parser tag hook, you might want to use recursiveTagParse()
596 * instead.
597 *
598 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
599 * $text are not expanded
600 *
601 * @since 1.25
602 *
603 * @param string $text Text extension wants to have parsed
604 * @param bool|PPFrame $frame The frame to use for expanding any template variables
605 * @return string Fully parsed HTML
606 */
611 }
613 /**
614 * Expand templates and variables in the text, producing valid, static wikitext.
615 * Also removes comments.
616 * Do not call this function recursively.
617 * @param string $text
618 * @param Title $title
619 * @param ParserOptions $options
620 * @param int|null $revid
621 * @param bool|PPFrame $frame
622 * @return mixed|string
623 */
626 ) {
631 }
637 }
639 /**
640 * Recursive parser entry point that can be called from an extension tag
641 * hook.
642 *
643 * @param string $text Text to be expanded
644 * @param bool|PPFrame $frame The frame to use for expanding any template variables
645 * @return string
646 * @since 1.19
647 */
652 }
654 /**
655 * Process the wikitext for the "?preload=" feature. (bug 5210)
656 *
657 * "<noinclude>", "<includeonly>" etc. are parsed as for template
658 * transclusion, comments, templates, arguments, tags hooks and parser
659 * functions are untouched.
660 *
661 * @param string $text
662 * @param Title $title
663 * @param ParserOptions $options
664 * @param array $params
665 * @return string
666 */
667 public function getPreloadText( $text, Title $title, ParserOptions $options, $params = array() ) {
671 # Parser (re)initialisation
680 }
682 /**
683 * Get a random string
684 *
685 * @return string
686 */
689 }
691 /**
692 * Set the current user.
693 * Should only be used when doing pre-save transform.
694 *
695 * @param User|null $user User object or null (to reset)
696 */
699 }
701 /**
702 * Accessor for mUniqPrefix.
703 *
704 * @return string
705 */
708 # @todo FIXME: This is probably *horribly wrong*
709 # LanguageConverter seems to want $wgParser's uniqPrefix, however
710 # if this is called for a parser cache hit, the parser may not
711 # have ever been initialized in the first place.
712 # Not really sure what the heck is supposed to be going on here.
714 # throw new MWException( "Accessing uninitialized mUniqPrefix" );
715 }
717 }
719 /**
720 * Set the context title
721 *
722 * @param Title $t
723 */
727 }
730 # Strip the fragment to avoid various odd effects
735 }
736 }
738 /**
739 * Accessor for the Title object
740 *
741 * @return Title
742 */
745 }
747 /**
748 * Accessor/mutator for the Title object
749 *
750 * @param Title $x Title object or null to just get the current one
751 * @return Title
752 */
755 }
757 /**
758 * Set the output type
759 *
760 * @param int $ot New value
761 */
764 # Shortcut alias
770 );
771 }
773 /**
774 * Accessor/mutator for the output type
775 *
776 * @param int|null $x New value or null to just get the current one
777 * @return int
778 */
781 }
783 /**
784 * Get the ParserOutput object
785 *
786 * @return ParserOutput
787 */
790 }
792 /**
793 * Get the ParserOptions object
794 *
795 * @return ParserOptions
796 */
799 }
801 /**
802 * Accessor/mutator for the ParserOptions object
803 *
804 * @param ParserOptions $x New value or null to just get the current one
805 * @return ParserOptions Current ParserOptions object
806 */
809 }
811 /**
812 * @return int
813 */
816 }
818 /**
819 * @param int $id
820 */
823 }
825 /**
826 * Get a language object for use in parser functions such as {{FORMATNUM:}}
827 * @return Language
828 */
831 }
833 /**
834 * Get the target language for the content being parsed. This is usually the
835 * language that the content is in.
836 *
837 * @since 1.19
838 *
839 * @throws MWException
840 * @return Language
841 */
851 }
854 }
856 /**
857 * Get the language object for language conversion
858 * @return Language|null
859 */
862 }
864 /**
865 * Get a User object either from $this->mUser, if set, or from the
866 * ParserOptions object otherwise
867 *
868 * @return User
869 */
873 }
875 }
877 /**
878 * Get a preprocessor object
879 *
880 * @return Preprocessor
881 */
886 }
888 }
890 /**
891 * Replaces all occurrences of HTML-style comments and the given tags
892 * in the text with a random marker and returns the next text. The output
893 * parameter $matches will be an associative array filled with data in
894 * the form:
895 *
896 * @code
897 * 'UNIQ-xxxxx' => array(
898 * 'element',
899 * 'tag content',
900 * array( 'param' => 'x' ),
901 * '<element param="x">tag content</element>' ) )
902 * @endcode
903 *
904 * @param array $elements List of element names. Comments are always extracted.
905 * @param string $text Source text string.
906 * @param array $matches Out parameter, Array: extracted tags
907 * @param string $uniq_prefix
908 * @return string Stripped text
909 */
923 }
925 # comment
931 # tag
936 }
942 # Empty element tag, <tag />
951 }
955 # No end tag -- let it run out to the end of the text.
961 }
962 }
968 }
970 }
972 /**
973 * Get a list of strippable XML-like elements
974 *
975 * @return array
976 */
979 }
981 /**
982 * Add an item to the strip state
983 * Returns the unique tag which must be inserted into the stripped text
984 * The tag will be replaced with the original text in unstrip()
985 *
986 * @param string $text
987 *
988 * @return string
989 */
995 }
997 /**
998 * parse the wiki syntax used to render tables
999 *
1000 * @private
1001 * @param string $text
1002 * @return string
1003 */
1021 }
1027 # First check if we are starting a new table
1040 # Don't do any of the following
1044 # We are ending a table
1050 }
1054 }
1058 }
1062 # Now we have a table row
1065 # Whats after the tag is now only attributes
1078 }
1082 }
1091 ) {
1092 # This might be cell elements, td, th or captions
1096 }
1102 }
1104 # Split up multiple cells on the same line.
1105 # FIXME : This can result in improper nesting of tags processed
1106 # by earlier parser steps, but should avoid splitting up eg
1107 # attribute values containing literal "||".
1112 # Loop through each table cell
1119 }
1124 }
1130 }
1140 }
1144 # A cell could contain both parameters and data
1147 # Bug 553: Note that a '|' inside an invalid link should not
1148 # be mistaken as delimiting cell parameters
1157 }
1161 }
1162 }
1164 }
1166 # Closing open td, tr && table
1170 }
1173 }
1176 }
1179 }
1181 # Remove trailing line-ending (b/c)
1184 }
1186 # special case: don't return empty table
1189 }
1192 }
1194 /**
1195 * Helper function for parse() that transforms wiki markup into half-parsed
1196 * HTML. Only called for $mOutputType == self::OT_HTML.
1197 *
1198 * @private
1199 *
1200 * @param string $text
1201 * @param bool $isMain
1202 * @param bool $frame
1203 *
1204 * @return string
1205 */
1210 # Hook to suspend the parser in this state
1211 if ( !Hooks::run( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState ) ) ) {
1213 }
1215 # if $frame is provided, then use $frame for replacing any variables
1217 # use frame depth to infer how include/noinclude tags should be handled
1218 # depth=0 means this is the top-level document; otherwise it's an included document
1223 }
1227 # if $frame is not provided, then use old-style replaceVariables
1229 }
1237 );
1240 # Tables need to come after variable replacement for things to work
1241 # properly; putting them before other transformations should keep
1242 # exciting things like link expansions from showing up in surprising
1243 # places.
1255 # replaceInternalLinks may sometimes leave behind
1256 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1263 }
1265 /**
1266 * Helper function for parse() that transforms half-parsed HTML into fully
1267 * parsed HTML.
1268 *
1269 * @param string $text
1270 * @param bool $isMain
1271 * @param bool $linestart
1272 * @return string
1273 */
1279 # Clean up special characters, only run once, next-to-last before doBlockLevels
1281 # french spaces, last one Guillemet-left
1282 # only if there is something before the space
1284 # french spaces, Guillemet-right
1287 );
1294 /**
1295 * The input doesn't get language converted if
1296 * a) It's disabled
1297 * b) Content isn't converted
1298 * c) It's a conversion table
1299 * d) it is an interface message (which is in the user language)
1300 */
1303 ) {
1305 # The position of the convert() call should not be changed. it
1306 # assumes that the links are all replaced and the only thing left
1307 # is the <nowiki> mark.
1309 }
1310 }
1316 }
1326 # attempt to sanitize at least some nesting problems
1327 # (bug #2702 and quite a few others)
1329 # ''Something [http://www.cool.com cool''] -->
1330 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1333 # fix up an anchor inside another anchor, only
1334 # at least for a single single nested link (bug 3695)
1337 # fix div inside inline elements- doBlockLevels won't wrap a line which
1338 # contains a div, so fix it up here; replace
1339 # div with escaped text
1342 # remove empty italic or bold tag pairs, some
1343 # introduced by rules above
1345 );
1351 }
1355 }
1358 }
1360 /**
1361 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1362 * magic external links.
1363 *
1364 * DML
1365 * @private
1366 *
1367 * @param string $text
1368 *
1369 * @return string
1370 */
1378 '!(?: # Start cases
1387 [0-9Xx] # check digit
1391 }
1393 /**
1394 * @throws MWException
1395 * @param array $m
1396 * @return HTML|string
1397 */
1400 # Skip anchor
1403 # Skip HTML element
1406 # Free external link
1409 # RFC or PMID
1423 }
1427 # ISBN
1435 ));
1442 }
1443 }
1445 /**
1446 * Make a free external link, given a user-supplied URL
1447 *
1448 * @param string $url
1449 *
1450 * @return string HTML
1451 * @private
1452 */
1457 # The characters '<' and '>' (which were escaped by
1458 # removeHTMLtags()) should not be included in
1459 # URLs, per RFC 2396.
1464 }
1466 # Move trailing punctuation to $trail
1468 # If there is no left bracket, then consider right brackets fair game too
1471 }
1475 # Don't break a trailing HTML entity by moving the ; into $trail
1476 # This is in hot code, so use substr_compare to avoid having to
1477 # create a new string object for the comparison
1479 # more optimization: instead of running preg_match with a $
1480 # anchor, which can be slow, do the match on the reversed
1481 # string starting at the desired offset.
1482 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1485 }
1486 }
1490 }
1494 # Is this an external image?
1497 # Not an image, make a link
1502 # Register it in the output object...
1503 # Replace unnecessary URL escape codes with their equivalent characters
1506 }
1508 }
1510 /**
1511 * Parse headers and return html
1512 *
1513 * @private
1514 *
1515 * @param string $text
1516 *
1517 * @return string
1518 */
1523 }
1525 }
1527 /**
1528 * Replace single quotes with HTML markup
1529 * @private
1530 *
1531 * @param string $text
1532 *
1533 * @return string The altered text
1534 */
1540 }
1543 }
1545 /**
1546 * Helper function for doAllQuotes()
1547 *
1548 * @param string $text
1549 *
1550 * @return string
1551 */
1557 }
1559 // First, do some preliminary work. This may shift some apostrophes from
1560 // being mark-up to being text. It also counts the number of occurrences
1561 // of bold and italics mark-ups.
1566 // If there are ever four apostrophes, assume the first is supposed to
1567 // be text, and the remaining three constitute mark-up for bold text.
1568 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1574 // If there are more than 5 apostrophes in a row, assume they're all
1575 // text except for the last 5.
1576 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1580 }
1581 // Count the number of occurrences of bold and italics mark-ups.
1589 }
1590 }
1592 // If there is an odd number of both bold and italics, it is likely
1593 // that one of the bold ones was meant to be an apostrophe followed
1594 // by italics. Which one we cannot know for certain, but it is more
1595 // likely to be one that has a single-letter word before it.
1607 }
1611 // if $firstsingleletterword is set, we don't
1612 // look at the other options, so we can bail early.
1614 }
1618 }
1619 }
1620 }
1621 }
1623 // If there is a single-letter word, use it!
1628 // If not, but there's a multi-letter word, use that one.
1632 // ... otherwise use the first one that has neither.
1633 // (notice that it is possible for all three to be -1 if, for example,
1634 // there is only one pentuple-apostrophe in the line)
1637 }
1638 }
1640 // Now let's actually convert our apostrophic mush to HTML!
1651 }
1670 }
1687 }
1707 }
1708 }
1709 }
1711 }
1712 // Now close all remaining tags. Notice that the order is important.
1715 }
1718 }
1721 }
1722 // There might be lonely ''''', so make sure we have a buffer
1725 }
1727 }
1729 /**
1730 * Replace external links (REL)
1731 *
1732 * Note: this is all very hackish and the order of execution matters a lot.
1733 * Make sure to run tests/parserTests.php if you change this code.
1734 *
1735 * @private
1736 *
1737 * @param string $text
1738 *
1739 * @throws MWException
1740 * @return string
1741 */
1748 }
1758 # The characters '<' and '>' (which were escaped by
1759 # removeHTMLtags()) should not be included in
1760 # URLs, per RFC 2396.
1765 }
1767 # If the link text is an image URL, replace it with an <img> tag
1768 # This happened by accident in the original parser, but some people used it extensively
1772 }
1776 # Set linktype for CSS - if URL==text, link is essentially free
1779 # No link text, e.g. [http://domain.tld/some.link]
1781 # Autonumber
1786 # Have link text, e.g. [http://domain.tld/some.link text]s
1787 # Check for trail
1789 }
1795 # Use the encoded URL
1796 # This means that users can paste URLs directly into the text
1797 # Funny characters like ö aren't valid in URLs anyway
1798 # This was changed in August 2004
1802 # Register link in the output object.
1803 # Replace unnecessary URL escape codes with the referenced character
1804 # This prevents spammers from hiding links from the filters
1807 }
1810 }
1812 /**
1813 * Get the rel attribute for a particular external link.
1814 *
1815 * @since 1.21
1816 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1817 * nofollow if appropriate
1818 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1819 * @return string|null Rel attribute for $url
1820 */
1826 ) {
1828 }
1830 }
1832 /**
1833 * Get an associative array of additional HTML attributes appropriate for a
1834 * particular external link. This currently may include rel => nofollow
1835 * (depending on configuration, namespace, and the URL's domain) and/or a
1836 * target attribute (depending on configuration).
1837 *
1838 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1839 * nofollow if appropriate
1840 * @return array Associative array of HTML attributes
1841 */
1848 }
1850 }
1852 /**
1853 * Replace unusual escape codes in a URL with their equivalent characters
1854 *
1855 * @deprecated since 1.24, use normalizeLinkUrl
1856 * @param string $url
1857 * @return string
1858 */
1862 }
1864 /**
1865 * Replace unusual escape codes in a URL with their equivalent characters
1866 *
1867 * This generally follows the syntax defined in RFC 3986, with special
1868 * consideration for HTTP query strings.
1869 *
1870 * @param string $url
1871 * @return string
1872 */
1874 # First, make sure unsafe characters are encoded
1878 },
1879 $url
1880 );
1885 # Fragment part - 'fragment'
1891 }
1893 # Query part - 'query' minus &=+;
1899 }
1901 # Scheme and path part - 'pchar'
1902 # (we assume no userinfo or encoded colons in the host)
1907 }
1914 # Unescape it
1917 # Leave it escaped, but use uppercase for a-f
1919 }
1920 };
1922 }
1924 /**
1925 * make an image if it's allowed, either through the global
1926 * option, through the exception, or through the on-wiki whitelist
1927 *
1928 * @param string $url
1929 *
1930 * @return string
1931 */
1936 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1943 }
1944 }
1949 }
1953 ) {
1955 # Image found
1957 }
1958 }
1961 ) {
1965 );
1968 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1971 }
1973 # Image matches a whitelist entry
1976 }
1977 }
1978 }
1980 }
1982 /**
1983 * Process [[ ]] wikilinks
1984 *
1985 * @param string $s
1986 *
1987 * @return string Processed text
1988 *
1989 * @private
1990 */
1994 }
1996 /**
1997 * Process [[ ]] wikilinks (RIL)
1998 * @param string $s
1999 * @throws MWException
2000 * @return LinkHolderArray
2001 *
2002 * @private
2003 */
2008 # the % is needed to support urlencoded titles as well
2011 # Match a link having the form [[namespace:link|alternate]]trail
2013 # Match cases where there is no "]]", which might still be images
2015 }
2019 # split the entire text string on occurrences of [[
2021 # get the first element (all text up to first [[), and remove the space we added
2030 # Match the end of a line for a word that's not followed by whitespace,
2031 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2035 }
2039 }
2048 }
2051 }
2055 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2056 # Loop for each link
2058 // @codingStandardsIgnoreStart
2060 # Check for excessive memory usage
2062 # Too big
2063 # Do the existence check, replace the link holders and clear the array
2066 }
2074 }
2075 # first link
2079 }
2080 }
2086 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2087 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2088 # the real problem is with the $e1 regex
2089 # See bug 1300.
2090 #
2091 # Still some problems for cases where the ] is meant to be outside punctuation,
2092 # and no image is in sight. See bug 2095.
2093 #
2097 ) {
2100 }
2101 # fix up urlencoded title texts
2103 # Should anchors '#' also be rejected?
2105 }
2108 # Invalid, but might be an image with a link in its caption
2113 }
2118 }
2122 # Don't allow internal links to pages containing
2123 # PROTO: where PROTO is a valid URL protocol; these
2124 # should be external links.
2128 }
2130 # Make subpage if necessary
2135 }
2139 # Strip off leading ':'
2141 }
2147 }
2156 # look at the next 'line' to see if we can close it there
2161 }
2164 # the first ]] closes the inner link, the second the image
2170 # if there's exactly one ]] that's fine, we'll keep looking
2173 # if $next_line is invalid too, we need look no further
2176 }
2177 }
2179 # we couldn't find the end of this imageLink, so output it raw
2180 # but don't ignore what might be perfectly normal links in the text we've examined
2183 # note: no $trail, because without an end, there *is* no trail
2185 }
2188 # note: no $trail, because without an end, there *is* no trail
2190 }
2191 }
2197 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2198 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2199 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2200 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2202 }
2204 # Link not escaped by : , create the various objects
2206 # Interwikis
2211 )
2212 ) {
2213 # Bug 24502: filter duplicates
2217 }
2222 }
2227 # if no parameters were passed, $text
2228 # becomes something like "File:Foo.png",
2229 # which we don't want to pass on to the
2230 # image generator
2233 # recursively parse links inside the image caption
2234 # actually, this will parse them in any other parameters, too,
2235 # but it might be hard to fix that, and it doesn't matter ATM
2238 }
2239 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2244 }
2246 }
2255 }
2261 /**
2262 * Strip the whitespace Category links produce, see bug 87
2263 */
2267 }
2268 }
2270 # Self-link checking. For some languages, variants of the title are checked in
2271 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2272 # for linking to a different variant.
2276 }
2278 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2279 # @todo FIXME: Should do batch file existence checks, see comment below
2281 # Give extensions a chance to select the file revision for us
2286 # Fetch and register the file (file title may be different via hooks)
2288 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2292 }
2294 # Some titles, such as valid special pages or files in foreign repos, should
2295 # be shown as bluelinks even though they're not included in the page table
2296 #
2297 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2298 # batch file existence checks for NS_FILE and NS_MEDIA
2303 # Links will be added to the output link list after checking
2305 }
2306 }
2308 }
2310 /**
2311 * Render a forced-blue link inline; protect against double expansion of
2312 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2313 * Since this little disaster has to split off the trail text to avoid
2314 * breaking URLs in the following text without breaking trails on the
2315 * wiki links, it's been made into a horrible function.
2316 *
2317 * @param Title $nt
2318 * @param string $text
2319 * @param array|string $query
2320 * @param string $trail
2321 * @param string $prefix
2322 * @return string HTML-wikitext mix oh yuck
2323 */
2324 public function makeKnownLinkHolder( $nt, $text = '', $query = array(), $trail = '', $prefix = '' ) {
2329 }
2332 }
2337 }
2339 /**
2340 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2341 * going to go through further parsing steps before inline URL expansion.
2342 *
2343 * Not needed quite as much as it used to be since free links are a bit
2344 * more sensible these days. But bracketed links are still an issue.
2345 *
2346 * @param string $text More-or-less HTML
2347 * @return string Less-or-more HTML with NOPARSE bits
2348 */
2352 }
2354 /**
2355 * Return true if subpage links should be expanded on this page.
2356 * @return bool
2357 */
2359 # Some namespaces don't allow subpages
2361 }
2363 /**
2364 * Handle link to subpage if necessary
2365 *
2366 * @param string $target The source of the link
2367 * @param string &$text The link text, modified as necessary
2368 * @return string The full name of the link
2369 * @private
2370 */
2373 }
2375 /**#@+
2376 * Used by doBlockLevels()
2377 * @private
2378 *
2379 * @return string
2380 */
2385 }
2389 }
2391 /**
2392 * getCommon() returns the length of the longest common substring
2393 * of both arguments, starting at the beginning of both.
2394 * @private
2395 *
2396 * @param string $st1
2397 * @param string $st2
2398 *
2399 * @return int
2400 */
2406 }
2411 }
2412 }
2414 }
2416 /**
2417 * These next three functions open, continue, and close the list
2418 * element appropriate to the prefix character passed into them.
2419 * @private
2420 *
2421 * @param string $char
2422 *
2423 * @return string
2424 */
2439 }
2442 }
2444 /**
2445 * TODO: document
2446 * @param string $char
2447 * @private
2448 *
2449 * @return string
2450 */
2458 }
2465 }
2466 }
2468 }
2470 /**
2471 * @todo Document
2472 * @param string $char
2473 * @private
2474 *
2475 * @return string
2476 */
2488 }
2491 }
2493 }
2494 /**#@-*/
2496 /**
2497 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2498 *
2499 * @param string $text
2500 * @param bool $linestart Whether or not this is at the start of a line.
2501 * @private
2502 * @return string The lists rendered as HTML
2503 */
2506 # Parsing through the text line by line. The main thing
2507 # happening here is handling of block-level elements p, pre,
2508 # and making lists from lines starting with * # : etc.
2509 #
2519 # Fix up $linestart
2524 }
2525 # * = ul
2526 # # = ol
2527 # ; = dt
2528 # : = dd
2533 # If not in a <pre> element, scan for and figure out what prefixes are there.
2535 # Multiple prefixes may abut each other for nested lists.
2539 # eh?
2540 # ; and : are both from definition-lists, so they're equivalent
2541 # for the purposes of determining whether or not we need to open/close
2542 # elements.
2547 # Don't interpret any other prefixes in preformatted text
2551 }
2553 # List generation
2555 # Same as the last item, so no need to deal with nesting or opening stuff
2560 # The one nasty exception: definition lists work like this:
2561 # ; title : definition text
2562 # So we check for : in the remainder text to split up the
2563 # title and definition, without b0rking links.
2568 }
2569 }
2571 # We need to open or close prefixes, or both.
2573 # Either open or close a level...
2577 # Close all the prefixes which aren't shared.
2581 }
2583 # Continue the current prefix if appropriate.
2586 }
2588 # Open prefixes where appropriate.
2591 }
2597 # @todo FIXME: This is dupe of code above
2601 }
2602 }
2604 }
2607 }
2609 }
2611 # If we have no prefixes, go to paragraph mode.
2613 # No prefix (not in list)--go to paragraph mode
2614 # XXX: use a stack for nestable elements like span, table and div
2616 '/(?:<table|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|'
2618 $t
2619 );
2625 $t
2626 );
2630 # @todo bug 5718: paragraph closed
2634 }
2636 while ( preg_match( '/<(\\/?)blockquote[\s>]/i', $t, $bqMatch, PREG_OFFSET_CAPTURE, $bqOffset ) ) {
2639 }
2645 ) {
2646 # pre
2651 }
2654 # paragraph
2667 }
2668 }
2677 }
2678 }
2679 }
2680 }
2681 }
2682 # somewhere above we forget to get out of pre block (bug 785)
2685 }
2690 }
2691 }
2692 }
2698 }
2699 }
2703 }
2706 }
2708 /**
2709 * Split up a string on ':', ignoring any occurrences inside tags
2710 * to prevent illegal overlapping.
2711 *
2712 * @param string $str The string to split
2713 * @param string &$before Set to everything before the ':'
2714 * @param string &$after Set to everything after the ':'
2715 * @throws MWException
2716 * @return string The position of the ':', or false if none found
2717 */
2722 # Nothing to find!
2724 }
2728 # Easy; no tag nesting to worry about
2732 }
2734 # Ugly state machine to walk through avoiding tags.
2742 # (Using the number is a performance hack for common cases)
2746 # Could be either a <start> tag or an </end> tag
2751 # We found it!
2755 }
2756 # Embedded in a tag; don't break it.
2759 # Skip ahead looking for something interesting
2762 # Nothing else interesting
2764 }
2768 # We found it!
2772 }
2773 }
2775 # Nothing else interesting to find; abort!
2776 # We're nested, but there's no close tags left. Abort!
2778 }
2779 # Skip ahead to next tag start
2782 }
2785 # In a <tag>
2792 # Slash may be followed by >?
2796 # ignore
2797 }
2808 # Illegal early close? This shouldn't happen D:
2813 }
2816 # In a </tag>
2822 }
2824 }
2828 # Yes, a self-closed tag <blah/>
2831 # Probably we're jumping the gun, and this is an attribute
2833 }
2838 }
2845 }
2852 }
2856 }
2857 }
2859 wfDebug( __METHOD__ . ": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2861 }
2863 }
2865 /**
2866 * Return value of a magic variable (like PAGENAME)
2867 *
2868 * @private
2869 *
2870 * @param int $index
2871 * @param bool|PPFrame $frame
2872 *
2873 * @throws MWException
2874 * @return string
2875 */
2881 // If no title set, bad things are going to happen
2882 // later. Title should always be set since this
2883 // should only be called in the middle of a parse
2884 // operation (but the unit-tests do funky stuff)
2887 }
2889 /**
2890 * Some of these require message or data lookups and can be
2891 * expensive to check many times.
2892 */
2896 }
2897 }
2942 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
2976 ) ) );
2986 ) ) );
2994 }
3002 }
3015 # 0 means the page doesn't exist in the database,
3016 # which means the user is previewing a new page.
3017 # The vary-revision flag must be set, because the magic word
3018 # will have a different value once the page is saved.
3021 }
3025 # Let the edit saving system know we should parse the page
3026 # *after* a revision ID has been assigned.
3032 # Let the edit saving system know we should parse the page
3033 # *after* a revision ID has been assigned. This is for null edits.
3039 # Let the edit saving system know we should parse the page
3040 # *after* a revision ID has been assigned. This is for null edits.
3046 # Let the edit saving system know we should parse the page
3047 # *after* a revision ID has been assigned. This is for null edits.
3053 # Let the edit saving system know we should parse the page
3054 # *after* a revision ID has been assigned. This is for null edits.
3060 # Let the edit saving system know we should parse the page
3061 # *after* a revision ID has been assigned. This is for null edits.
3067 # Let the edit saving system know we should parse the page
3068 # *after* a revision ID has been assigned. This is for null edits.
3074 # Let the edit saving system know we should parse the page
3075 # *after* a revision ID has been assigned. This is for null edits.
3081 # Let the edit saving system know we should parse the page
3082 # *after* a revision ID has been assigned. This is for null edits.
3123 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3124 # int to remove the padding
3133 );
3142 false
3143 );
3149 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3150 # int to remove the padding
3211 );
3214 }
3218 }
3221 }
3223 /**
3224 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
3225 *
3226 * @private
3227 */
3234 }
3236 /**
3237 * Preprocess some wikitext and return the document tree.
3238 * This is the ghost of replace_variables().
3239 *
3240 * @param string $text The text to parse
3241 * @param int $flags Bitwise combination of:
3242 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
3243 * included. Default is to assume a direct page view.
3244 *
3245 * The generated DOM tree must depend only on the input text and the flags.
3246 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
3247 *
3248 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
3249 * change in the DOM tree for a given text, must be passed through the section identifier
3250 * in the section edit link and thus back to extractSections().
3251 *
3252 * The output of this function is currently only cached in process memory, but a persistent
3253 * cache may be implemented at a later date which takes further advantage of these strict
3254 * dependency requirements.
3255 *
3256 * @return PPNode
3257 */
3261 }
3263 /**
3264 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3265 *
3266 * @param string $s
3267 *
3268 * @return array
3269 */
3279 }
3281 }
3283 /**
3284 * Replace magic variables, templates, and template arguments
3285 * with the appropriate text. Templates are substituted recursively,
3286 * taking care to avoid infinite loops.
3287 *
3288 * Note that the substitution depends on value of $mOutputType:
3289 * self::OT_WIKI: only {{subst:}} templates
3290 * self::OT_PREPROCESS: templates but not extension tags
3291 * self::OT_HTML: all templates and extension tags
3292 *
3293 * @param string $text The text to transform
3294 * @param bool|PPFrame $frame Object describing the arguments passed to the
3295 * template. Arguments may also be provided as an associative array, as
3296 * was the usual case before MW1.12. Providing arguments this way may be
3297 * useful for extensions wishing to perform variable replacement
3298 * explicitly.
3299 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
3300 * double-brace expansion.
3301 * @return string
3302 */
3304 # Is there any text? Also, Prevent too big inclusions!
3307 }
3315 }
3322 }
3324 /**
3325 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3326 *
3327 * @param array $args
3328 *
3329 * @return array
3330 */
3343 }
3346 }
3347 }
3348 }
3351 }
3353 /**
3354 * Warn the user when a parser limitation is reached
3355 * Will warn at most once the user per limitation type
3356 *
3357 * @param string $limitationType Should be one of:
3358 * 'expensive-parserfunction' (corresponding messages:
3359 * 'expensive-parserfunction-warning',
3360 * 'expensive-parserfunction-category')
3361 * 'post-expand-template-argument' (corresponding messages:
3362 * 'post-expand-template-argument-warning',
3363 * 'post-expand-template-argument-category')
3364 * 'post-expand-template-inclusion' (corresponding messages:
3365 * 'post-expand-template-inclusion-warning',
3366 * 'post-expand-template-inclusion-category')
3367 * 'node-count-exceeded' (corresponding messages:
3368 * 'node-count-exceeded-warning',
3369 * 'node-count-exceeded-category')
3370 * 'expansion-depth-exceeded' (corresponding messages:
3371 * 'expansion-depth-exceeded-warning',
3372 * 'expansion-depth-exceeded-category')
3373 * @param string|int|null $current Current value
3374 * @param string|int|null $max Maximum allowed, when an explicit limit has been
3375 * exceeded, provide the values (optional)
3376 */
3378 # does no harm if $current and $max are present but are unnecessary for the message
3383 }
3385 /**
3386 * Return the text of a template, after recursively
3387 * replacing any variables or templates within the template.
3388 *
3389 * @param array $piece The parts of the template
3390 * $piece['title']: the title, i.e. the part before the |
3391 * $piece['parts']: the parameter array
3392 * $piece['lineStart']: whether the brace was at the start of a line
3393 * @param PPFrame $frame The current frame, contains template arguments
3394 * @throws Exception
3395 * @return string The text of the template
3396 */
3399 // Flags
3401 // $text has been filled
3403 // wiki markup in $text should be escaped
3405 // $text is HTML, armour it against wikitext transformation
3407 // Force interwiki transclusion to be done in raw mode not rendered
3409 // $text is a DOM node needing expansion in a child frame
3411 // $text is a DOM node needing expansion in the current frame
3414 # Title object, where $text came from
3417 # $part1 is the bit before the first |, and must contain only title characters.
3418 # Various prefixes will be stripped from it later.
3423 # Original title text preserved for various purposes
3426 # $args is a list of argument nodes, starting from index 0, not including $part1
3427 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3428 # below won't work b/c this $args isn't an object
3433 # SUBST
3438 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3439 # Decide whether to expand template or keep wikitext as-is.
3445 }
3451 }
3452 }
3457 }
3458 }
3460 # Variables
3467 }
3469 }
3470 }
3472 # MSG, MSGNW and RAW
3474 # Check for MSGNW:
3479 # Remove obsolete MSG:
3482 }
3484 # Check for RAW:
3488 }
3489 }
3491 # Parser functions
3500 }
3505 }
3507 # The interface for parser functions allows for extracting
3508 # flags into the local scope. Extract any forwarded flags
3509 # here.
3511 }
3512 }
3514 # Finish mangling title and then check for loops.
3515 # Set $title to a Title object and $titleText to the PDBK
3518 # Split the title into page and subpage
3524 }
3528 # Check for language variants if the template is not found
3531 }
3532 # Do recursion depth check
3540 }
3541 }
3542 }
3544 # Load from database
3551 ) {
3552 // Pass the template arguments as URL parameters.
3553 // "uselang" will have no effect since the Language object
3554 // is forced to the one defined in ParserOptions.
3563 }
3564 }
3566 // Create a new context to execute the special page
3579 }
3589 }
3590 }
3592 # If the title is valid but undisplayable, make a link to it
3596 }
3598 # Interwiki transclusion
3604 # Preprocess it like a template
3607 }
3609 }
3611 # Do infinite loop check
3612 # This has to be done after redirect resolution to avoid infinite loops via redirects
3619 }
3620 }
3622 # If we haven't found text to substitute by now, we're done
3623 # Recover the source wikitext and return it
3628 }
3630 }
3632 # Expand DOM-style return values in a child frame
3634 # Clean up argument array
3640 # Expansion is eligible for the empty-frame cache
3643 # Uncached expansion
3645 }
3646 }
3650 }
3654 }
3656 # Replace raw HTML by a placeholder
3660 # Escape nowiki-style return values
3665 ) {
3666 # Bug 529: if the template begins with a table or block-level
3667 # element, it should be treated as beginning a new line.
3668 # This behavior is somewhat controversial.
3670 }
3673 # Error, oversize inclusion
3675 # Make a working, properly escaped link if possible (bug 23588)
3678 # This will probably not be a working link, but at least it may
3679 # provide some hint of where the problem is
3682 }
3686 }
3692 }
3695 }
3697 /**
3698 * Call a parser function and return an array with text and flags.
3699 *
3700 * The returned array will always contain a boolean 'found', indicating
3701 * whether the parser function was found or not. It may also contain the
3702 * following:
3703 * text: string|object, resulting wikitext or PP DOM object
3704 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3705 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3706 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3707 * nowiki: bool, wiki markup in $text should be escaped
3708 *
3709 * @since 1.21
3710 * @param PPFrame $frame The current frame, contains template arguments
3711 * @param string $function Function name
3712 * @param array $args Arguments to the function
3713 * @throws MWException
3714 * @return array
3715 */
3720 # Case sensitive functions
3724 # Case insensitive functions
3730 }
3731 }
3735 # Workaround for PHP bug 35229 and similar
3738 }
3742 # Convert arguments to PPNodes and collect for appending to $allArgs
3749 }
3750 }
3752 # Add a frame parameter, and pass the arguments as an array
3756 # Convert arguments to plain text and append to $allArgs
3764 }
3765 }
3766 }
3770 # The interface for function hooks allows them to return a wikitext
3771 # string or an array containing the string and any flags. This mungs
3772 # things around to match what this method should return.
3777 );
3781 }
3785 );
3786 }
3792 }
3795 }
3800 }
3803 }
3805 /**
3806 * Get the semi-parsed DOM representation of a template with a given title,
3807 * and its redirect destination title. Cached.
3808 *
3809 * @param Title $title
3810 *
3811 * @return array
3812 */
3821 }
3824 }
3826 # Cache miss, go to the database
3832 }
3840 }
3843 }
3845 /**
3846 * Fetch the current revision of a given title. Note that the revision
3847 * (and even the title) may not exist in the database, so everything
3848 * contributing to the output of the parser should use this method
3849 * where possible, rather than getting the revisions themselves. This
3850 * method also caches its results, so using it benefits performance.
3851 *
3852 * @since 1.24
3853 * @param Title $title
3854 * @return Revision
3855 */
3860 }
3863 // Defaults to Parser::statelessFetchRevision()
3865 );
3866 }
3868 }
3870 /**
3871 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3872 * without passing them on to it.
3873 *
3874 * @since 1.24
3875 * @param Title $title
3876 * @param Parser|bool $parser
3877 * @return Revision
3878 */
3881 }
3883 /**
3884 * Fetch the unparsed text of a template and register a reference to it.
3885 * @param Title $title
3886 * @return array ( string or false, Title )
3887 */
3889 // Defaults to Parser::statelessFetchTemplate()
3898 // If we transclude ourselves, the final result
3899 // will change based on the new version of the page
3901 }
3902 }
3903 }
3905 }
3907 /**
3908 * Fetch the unparsed text of a template and register a reference to it.
3909 * @param Title $title
3910 * @return string|bool
3911 */
3915 }
3917 /**
3918 * Static function to get a template
3919 * Can be overridden via ParserOptions::setTemplateCallback().
3920 *
3921 * @param Title $title
3922 * @param bool|Parser $parser
3923 *
3924 * @return array
3925 */
3931 # Loop to fetch the article, with up to 1 redirect
3933 # Give extensions a chance to select the revision instead
3944 );
3946 }
3947 # Get the revision
3954 }
3956 # If there is no current revision, there is no page
3960 }
3967 # We fetched a rev from a different title; register it too...
3972 }
3981 }
3988 }
3993 }
3996 }
3997 # Redirect?
4000 }
4005 }
4007 /**
4008 * Fetch a file and its title and register a reference to it.
4009 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4010 * @param Title $title
4011 * @param array $options Array of options to RepoGroup::findFile
4012 * @return File|bool
4013 */
4017 }
4019 /**
4020 * Fetch a file and its title and register a reference to it.
4021 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4022 * @param Title $title
4023 * @param array $options Array of options to RepoGroup::findFile
4024 * @return array ( File or false, Title of file )
4025 */
4031 # Register the file as a dependency...
4034 # Update fetched file title
4037 }
4039 }
4041 /**
4042 * Helper function for fetchFileAndTitle.
4043 *
4044 * Also useful if you need to fetch a file but not use it yet,
4045 * for example to get the file's handler.
4046 *
4047 * @param Title $title
4048 * @param array $options Array of options to RepoGroup::findFile
4049 * @return File|bool
4050 */
4058 }
4060 }
4062 /**
4063 * Transclude an interwiki link.
4064 *
4065 * @param Title $title
4066 * @param string $action
4067 *
4068 * @return string
4069 */
4075 }
4081 }
4083 }
4085 /**
4086 * @param string $url
4087 * @return mixed|string
4088 */
4097 }
4104 // Though we failed to fetch the content, this status is useless.
4109 }
4116 ) );
4118 }
4120 /**
4121 * Triple brace replacement -- used for template arguments
4122 * @private
4123 *
4124 * @param array $piece
4125 * @param PPFrame $frame
4126 *
4127 * @return array
4128 */
4141 )
4142 ) {
4143 # No match in frame, use the supplied default
4145 }
4149 }
4152 # No match anywhere
4154 }
4157 }
4162 }
4165 }
4167 /**
4168 * Return the text to be used for a given extension tag.
4169 * This is the ghost of strip().
4170 *
4171 * @param array $params Associative array of parameters:
4172 * name PPNode for the tag name
4173 * attr PPNode for unparsed text where tag attributes are thought to be
4174 * attributes Optional associative array of parsed attributes
4175 * inner Contents of extension element
4176 * noClose Original text did not have a close tag
4177 * @param PPFrame $frame
4178 *
4179 * @throws MWException
4180 * @return string
4181 */
4195 }
4201 }
4204 # Workaround for PHP bug 35229 and similar
4207 }
4214 }
4220 }
4223 # Extract flags to local scope (to override $markerType)
4228 }
4232 }
4237 }
4238 }
4244 }
4245 }
4255 }
4257 }
4259 /**
4260 * Increment an include size counter
4261 *
4262 * @param string $type The type of expansion
4263 * @param int $size The size of the text
4264 * @return bool False if this inclusion would take it over the maximum, true otherwise
4265 */
4272 }
4273 }
4275 /**
4276 * Increment the expensive function count
4277 *
4278 * @return bool False if the limit has been exceeded
4279 */
4283 }
4285 /**
4286 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
4287 * Fills $this->mDoubleUnderscores, returns the modified text
4288 *
4289 * @param string $text
4290 *
4291 * @return string
4292 */
4295 # The position of __TOC__ needs to be recorded
4301 # Set a placeholder. At the end we'll fill it in with the TOC.
4304 # Only keep the first one.
4306 }
4308 # Now match and remove the rest of them
4314 }
4317 }
4320 ) {
4322 }
4323 # (bug 8068) Allow control over whether robots index a page.
4324 #
4325 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
4326 # is not desirable, the last one on the page should win.
4330 }
4334 }
4336 # Cache all double underscores in the database
4339 }
4342 }
4344 /**
4345 * @see ParserOutput::addTrackingCategory()
4346 * @param string $msg Message key
4347 * @return bool Whether the addition was successful
4348 */
4351 }
4353 /**
4354 * This function accomplishes several tasks:
4355 * 1) Auto-number headings if that option is enabled
4356 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
4357 * 3) Add a Table of contents on the top for users who have enabled the option
4358 * 4) Auto-anchor headings
4359 *
4360 * It loops through all headlines, collects the necessary data, then splits up the
4361 * string and re-inserts the newly formatted headlines.
4362 *
4363 * @param string $text
4364 * @param string $origText Original, untouched wikitext
4365 * @param bool $isMain
4366 * @return mixed|string
4367 * @private
4368 */
4372 # Inhibit editsection links if requested in the page
4378 }
4381 }
4383 # Get all headlines for numbering them and adding funky stuff like [edit]
4384 # links - this is for later, but we need the number of headlines right now
4389 $matches
4390 );
4392 # if there are fewer than 4 headlines in the article, do not show TOC
4393 # unless it's been explicitly enabled.
4397 # Allow user to stipulate that a page should have a "new section"
4398 # link added via __NEWSECTIONLINK__
4401 }
4403 # Allow user to remove the "new section"
4404 # link via __NONEWSECTIONLINK__
4407 }
4409 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4410 # override above conditions and always show TOC above first header
4414 }
4416 # headline counter
4420 # Ugh .. the TOC should have neat indentation levels which can be
4421 # passed to the skin functions. These are determined here
4453 }
4457 }
4461 # Increase TOC level
4468 }
4470 # Decrease TOC level, find level to jump to
4474 # Found last matching level
4478 # Found first matching level below current level
4481 }
4482 }
4485 }
4488 # Unindent only if the previous toc level was shown :p
4493 }
4494 }
4496 # No change in level, end TOC line
4499 }
4500 }
4504 # count number of headlines for each level
4511 }
4514 }
4515 }
4517 # The safe header is a version of the header text safe to use for links
4519 # Remove link placeholders by the link text.
4520 # <!--LINK number-->
4521 # turns into
4522 # link text with suffix
4523 # Do this before unstrip since link text can contain strip markers
4526 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4529 # Strip out HTML (first regex removes any tag not allowed)
4530 # Allowed tags are:
4531 # * <sup> and <sub> (bug 8393)
4532 # * <i> (bug 26375)
4533 # * <b> (r105284)
4534 # * <bdi> (bug 72884)
4535 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4536 #
4537 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4538 # to allow setting directionality in toc items.
4542 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b))(?: .*?)?>#'
4543 ),
4545 $safeHeadline
4546 );
4549 # For the anchor, strip out HTML-y stuff period
4553 # Save headline for section edit hint before it's escaped
4557 # For reverse compatibility, provide an id that's
4558 # HTML4-compatible, like we used to.
4559 #
4560 # It may be worth noting, academically, that it's possible for
4561 # the legacy anchor to conflict with a non-legacy headline
4562 # anchor on the page. In this case likely the "correct" thing
4563 # would be to either drop the legacy anchors or make sure
4564 # they're numbered first. However, this would require people
4565 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4566 # manually, so let's not bother worrying about it.
4572 # No reason to have both (in fact, we can't)
4574 }
4579 }
4581 # HTML names must be case-insensitively unique (bug 10721).
4582 # This does not apply to Unicode characters per
4583 # http://www.w3.org/TR/html5/infrastructure.html#case-sensitivity-and-string-comparison
4584 # @todo FIXME: We may be changing them depending on the current locale.
4590 }
4592 # Create the anchor for linking from the TOC to the section
4601 }
4608 }
4610 # Don't number the heading if it is the only one (looks silly)
4612 # the two are different if the line contains a link
4616 $numbering
4618 }
4623 }
4625 # Add the section to the section tree
4626 # Find the DOM node for this header
4633 }
4634 }
4638 }
4648 );
4650 # give headline the correct <h#> tag
4652 // Output edit section links as markers with styles that can be customized by skins
4654 # Put a T flag in the section identifier, to indicate to extractSections()
4655 # that sections inside <includeonly> should be counted.
4663 }
4664 // We use a bit of pesudo-xml for editsection markers. The
4665 // language converter is run later on. Using a UNIQ style marker
4666 // leads to the converter screwing up the tokens when it
4667 // converts stuff. And trying to insert strip tags fails too. At
4668 // this point all real inputted tags have already been escaped,
4669 // so we don't have to worry about a user trying to input one of
4670 // these markers directly. We use a page and section attribute
4671 // to stop the language converter from converting these
4672 // important bits of data, but put the headline hint inside a
4673 // content block because the language converter is supposed to
4674 // be able to convert that piece of data.
4675 // Gets replaced with html in ParserOutput::getText
4682 }
4685 }
4691 }
4695 # Never ever show TOC if no headers
4698 }
4703 }
4708 }
4712 }
4714 # split up and insert constructed headlines
4718 // build an array of document sections
4721 // $head is zero-based, sections aren't.
4726 }
4728 /**
4729 * Send a hook, one per section.
4730 * The idea here is to be able to make section-level DIVs, but to do so in a
4731 * lower-impact, more correct way than r50769
4732 *
4733 * $this : caller
4734 * $section : the section number
4735 * &$sectionContent : ref to the content of the section
4736 * $showEditLinks : boolean describing whether this section has an edit link
4737 */
4741 }
4744 // append the TOC at the beginning
4745 // Top anchor now in skin
4747 }
4755 }
4756 }
4758 /**
4759 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4760 * conversion, substituting signatures, {{subst:}} templates, etc.
4761 *
4762 * @param string $text The text to transform
4763 * @param Title $title The Title object for the current article
4764 * @param User $user The User object describing the current user
4765 * @param ParserOptions $options Parsing options
4766 * @param bool $clearState Whether to clear the parser state first
4767 * @return string The altered wiki markup
4768 */
4771 ) {
4774 }
4781 );
4785 }
4791 }
4793 /**
4794 * Pre-save transform helper function
4795 *
4796 * @param string $text
4797 * @param User $user
4798 *
4799 * @return string
4800 */
4804 # Note: This is the timestamp saved as hardcoded wikitext to
4805 # the database, we use $wgContLang here in order to give
4806 # everyone the same signature and use the default one rather
4807 # than the one selected in each user's preferences.
4808 # (see also bug 12815)
4814 # Allow translation of timezones through wiki. format() can return
4815 # whatever crap the system uses, localised or not, so we cannot
4816 # ship premade translations.
4821 }
4825 # Variable replacement
4826 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4829 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4830 # which may corrupt this parser instance via its wfMessage()->text() call-
4832 # Signatures
4838 ) );
4840 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4844 // [[ns:page (context)|]]
4846 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4848 // [[ns:page (context), context|]] (using either single or double-width comma)
4850 // [[|page]] (reverse pipe trick: add context from page title)
4853 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4865 # if there's no context, don't bother duplicating the title
4867 }
4869 # Trim trailing whitespace
4873 }
4875 /**
4876 * Fetch the user's signature text, if any, and normalize to
4877 * validated, ready-to-insert wikitext.
4878 * If you have pre-fetched the nickname or the fancySig option, you can
4879 * specify them here to save a database query.
4880 * Do not reuse this parser instance after calling getUserSig(),
4881 * as it may have changed if it's the $wgParser.
4882 *
4883 * @param User $user
4884 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4885 * @param bool|null $fancySig whether the nicknname is the complete signature
4886 * or null to use default value
4887 * @return string
4888 */
4894 # If not given, retrieve from the user object.
4897 }
4901 }
4909 # Sig. might contain markup; validate this
4911 # Validated; clean up (if needed) and return it
4914 # Failed to validate; fall back to the default
4917 }
4918 }
4920 # Make sure nickname doesnt get a sig in a sig
4923 # If we're still here, make it a link to the user page
4930 }
4932 /**
4933 * Check that the user's signature contains no bad XML
4934 *
4935 * @param string $text
4936 * @return string|bool An expanded string, or false if invalid.
4937 */
4940 }
4942 /**
4943 * Clean up signature text
4944 *
4945 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
4946 * 2) Substitute all transclusions
4947 *
4948 * @param string $text
4949 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
4950 * @return string Signature text
4951 */
4957 }
4959 # Option to disable this feature
4962 }
4964 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4965 # => Move this logic to braceSubstitution()
4967 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4978 }
4981 }
4983 /**
4984 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
4985 *
4986 * @param string $text
4987 * @return string Signature text with /~{3,5}/ removed
4988 */
4992 }
4994 /**
4995 * Set up some variables which are usually set up in parse()
4996 * so that an external function can call some class members with confidence
4997 *
4998 * @param Title|null $title
4999 * @param ParserOptions $options
5000 * @param int $outputType
5001 * @param bool $clearState
5002 */
5005 ) {
5007 }
5009 /**
5010 * @param Title|null $title
5011 * @param ParserOptions $options
5012 * @param int $outputType
5013 * @param bool $clearState
5014 */
5017 ) {
5023 }
5024 }
5026 /**
5027 * Wrapper for preprocess()
5028 *
5029 * @param string $text The text to preprocess
5030 * @param ParserOptions $options Options
5031 * @param Title|null $title Title object or null to use $wgTitle
5032 * @return string
5033 */
5037 # Guard against infinite recursion
5040 }
5046 }
5052 }
5054 /**
5055 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
5056 * The callback should have the following form:
5057 * function myParserHook( $text, $params, $parser, $frame ) { ... }
5058 *
5059 * Transform and return $text. Use $parser for any required context, e.g. use
5060 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
5061 *
5062 * Hooks may return extended information by returning an array, of which the
5063 * first numbered element (index 0) must be the return string, and all other
5064 * entries are extracted into local variables within an internal function
5065 * in the Parser class.
5066 *
5067 * This interface (introduced r61913) appears to be undocumented, but
5068 * 'markerName' is used by some core tag hooks to override which strip
5069 * array their results are placed in. **Use great caution if attempting
5070 * this interface, as it is not documented and injudicious use could smash
5071 * private variables.**
5072 *
5073 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5074 * @param callable $callback The callback function (and object) to use for the tag
5075 * @throws MWException
5076 * @return callable|null The old value of the mTagHooks array associated with the hook
5077 */
5082 }
5087 }
5090 }
5092 /**
5093 * As setHook(), but letting the contents be parsed.
5094 *
5095 * Transparent tag hooks are like regular XML-style tag hooks, except they
5096 * operate late in the transformation sequence, on HTML instead of wikitext.
5097 *
5098 * This is probably obsoleted by things dealing with parser frames?
5099 * The only extension currently using it is geoserver.
5100 *
5101 * @since 1.10
5102 * @todo better document or deprecate this
5103 *
5104 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5105 * @param callable $callback The callback function (and object) to use for the tag
5106 * @throws MWException
5107 * @return callable|null The old value of the mTagHooks array associated with the hook
5108 */
5113 }
5114 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
5118 }
5120 /**
5121 * Remove all tag hooks
5122 */
5127 }
5129 /**
5130 * Create a function, e.g. {{sum:1|2|3}}
5131 * The callback function should have the form:
5132 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
5133 *
5134 * Or with Parser::SFH_OBJECT_ARGS:
5135 * function myParserFunction( $parser, $frame, $args ) { ... }
5136 *
5137 * The callback may either return the text result of the function, or an array with the text
5138 * in element 0, and a number of flags in the other elements. The names of the flags are
5139 * specified in the keys. Valid flags are:
5140 * found The text returned is valid, stop processing the template. This
5141 * is on by default.
5142 * nowiki Wiki markup in the return value should be escaped
5143 * isHTML The returned text is HTML, armour it against wikitext transformation
5144 *
5145 * @param string $id The magic word ID
5146 * @param callable $callback The callback function (and object) to use
5147 * @param int $flags A combination of the following flags:
5148 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
5149 *
5150 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
5151 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
5152 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
5153 * the arguments, and to control the way they are expanded.
5154 *
5155 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
5156 * arguments, for instance:
5157 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
5158 *
5159 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
5160 * future versions. Please call $frame->expand() on it anyway so that your code keeps
5161 * working if/when this is changed.
5162 *
5163 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
5164 * expansion.
5165 *
5166 * Please read the documentation in includes/parser/Preprocessor.php for more information
5167 * about the methods available in PPFrame and PPNode.
5168 *
5169 * @throws MWException
5170 * @return string|callable The old callback function for this name, if any
5171 */
5178 # Add to function cache
5182 }
5188 # Case
5191 }
5192 # Add leading hash
5195 }
5196 # Remove trailing colon
5199 }
5201 }
5203 }
5205 /**
5206 * Get all registered function hook identifiers
5207 *
5208 * @return array
5209 */
5212 }
5214 /**
5215 * Create a tag function, e.g. "<test>some stuff</test>".
5216 * Unlike tag hooks, tag functions are parsed at preprocessor level.
5217 * Unlike parser functions, their content is not preprocessed.
5218 * @param string $tag
5219 * @param callable $callback
5220 * @param int $flags
5221 * @throws MWException
5222 * @return null
5223 */
5228 }
5235 }
5238 }
5240 /**
5241 * @todo FIXME: Update documentation. makeLinkObj() is deprecated.
5242 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
5243 * Placeholders created in Skin::makeLinkObj()
5244 *
5245 * @param string $text
5246 * @param int $options
5247 */
5250 }
5252 /**
5253 * Replace "<!--LINK-->" link placeholders with plain text of links
5254 * (not HTML-formatted).
5255 *
5256 * @param string $text
5257 * @return string
5258 */
5261 }
5263 /**
5264 * Renders an image gallery from a text with one line per image.
5265 * text labels may be given by using |-style alternative text. E.g.
5266 * Image:one.jpg|The number "1"
5267 * Image:tree.jpg|A tree
5268 * given as text will return the HTML of a gallery with two images,
5269 * labeled 'The number "1"' and
5270 * 'A tree'.
5271 *
5272 * @param string $text
5273 * @param array $params
5274 * @return string HTML
5275 */
5281 }
5286 // If invalid type set, fallback to default.
5288 }
5301 }
5307 }
5310 }
5313 }
5316 }
5323 # match lines like these:
5324 # Image:someimage.jpg|This is some image
5327 # Skip empty lines
5330 }
5334 }
5337 # Bogus title. Ignore these so we don't bomb out later.
5339 }
5341 # We need to get what handler the file uses, to figure out parameters.
5342 # Note, a hook can overide the file name, and chose an entirely different
5343 # file (which potentially could be of a different type and have different handler).
5348 # Don't register it now, as ImageGallery does that later.
5355 );
5358 // We don't want people to specify per-image widths.
5359 // Additionally the width parameter would need special casing anyhow.
5361 }
5370 // look for an |alt= definition while trying not to break existing
5371 // captions with multiple pipes (|) in it, until a more sensible grammar
5372 // is defined for images in galleries
5374 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5375 // splitting on '|' is a bit odd, and different from makeImage.
5392 //check to see if link matches an absolute url, if not then it must be a wiki link.
5399 }
5400 }
5403 // Must be a handler specific parameter.
5407 // Guess not. Append it to the caption.
5410 }
5411 }
5414 // concatenate all other pipes
5416 }
5417 }
5418 // remove the first pipe
5420 }
5423 }
5427 }
5429 /**
5430 * @param string $handler
5431 * @return array
5432 */
5438 }
5440 # Initialise static lists
5447 );
5455 }
5456 }
5457 }
5459 # Add handler params
5465 }
5466 }
5469 }
5470 return array( $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] );
5471 }
5473 /**
5474 * Parse image options text and use it to make an image
5475 *
5476 * @param Title $title
5477 * @param string $options
5478 * @param LinkHolderArray|bool $holders
5479 * @return string HTML
5480 */
5482 # Check if the options text is of the form "options|alt text"
5483 # Options are:
5484 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5485 # * left no resizing, just left align. label is used for alt= only
5486 # * right same, but right aligned
5487 # * none same, but not aligned
5488 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5489 # * center center the image
5490 # * frame Keep original image size, no magnify-button.
5491 # * framed Same as "frame"
5492 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5493 # * upright reduce width for upright images, rounded to full __0 px
5494 # * border draw a 1px border around the image
5495 # * alt Text for HTML alt attribute (defaults to empty)
5496 # * class Set a class for img node
5497 # * link Set the target of the image link. Can be external, interwiki, or local
5498 # vertical-align values (no % or length right now):
5499 # * baseline
5500 # * sub
5501 # * super
5502 # * top
5503 # * text-top
5504 # * middle
5505 # * bottom
5506 # * text-bottom
5510 # Give extensions a chance to select the file revision for us
5515 # Fetch and register the file (file title may be different via hooks)
5518 # Get parameter map
5525 }
5527 # Process the input parameters
5539 # Special case; width and height come in one variable together
5547 }
5548 }
5554 }
5555 }
5556 # else no validation -- bug 13436
5559 # Validate handler parameter
5562 # Validate internal parameters
5567 # @todo FIXME: Possibly check validity here for
5568 # manualthumb? downstream behavior seems odd with
5569 # missing manual thumbs.
5586 }
5588 }
5596 }
5597 }
5602 // use first appearing option, discard others.
5607 # Most other things appear to be empty or numeric...
5609 }
5610 }
5614 }
5615 }
5616 }
5619 }
5620 }
5622 # Process alignment parameters
5625 }
5628 }
5632 # Will the image be presented in a frame, with the caption below?
5638 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5639 # came to also set the caption, ordinary text after the image -- which
5640 # makes no sense, because that just repeats the text multiple times in
5641 # screen readers. It *also* came to set the title attribute.
5642 #
5643 # Now that we have an alt attribute, we should not set the alt text to
5644 # equal the caption: that's worse than useless, it just repeats the
5645 # text. This is the framed/thumbnail case. If there's no caption, we
5646 # use the unnamed parameter for alt text as well, just for the time be-
5647 # ing, if the unnamed param is set and the alt param is not.
5648 #
5649 # For the future, we need to figure out if we want to tweak this more,
5650 # e.g., introducing a title= parameter for the title; ignoring the un-
5651 # named parameter entirely for images without a caption; adding an ex-
5652 # plicit caption= parameter and preserving the old magic unnamed para-
5653 # meter for BC; ...
5656 # No caption or alt text, add the filename as the alt text so
5657 # that screen readers at least get some description of the image
5659 }
5660 # Do not set $params['frame']['title'] because tooltips don't make sense
5661 # for framed images
5664 # No alt text, use the "caption" for the alt text
5668 # No caption, fall back to using the filename for the
5669 # alt text
5671 }
5672 }
5673 # Use the "caption" for the tooltip text
5675 }
5679 # Linker does the rest
5684 # Give the handler a chance to modify the parser object
5687 }
5690 }
5692 /**
5693 * @param string $caption
5694 * @param LinkHolderArray|bool $holders
5695 * @return mixed|string
5696 */
5698 # Strip bad stuff out of the title (tooltip). We can't just use
5699 # replaceLinkHoldersText() here, because if this function is called
5700 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5705 }
5707 # make sure there are no placeholders in thumbnail attributes
5708 # that are later expanded to html- so expand them now and
5709 # remove the tags
5714 }
5716 /**
5717 * Set a flag in the output object indicating that the content is dynamic and
5718 * shouldn't be cached.
5719 */
5725 }
5728 }
5730 /**
5731 * Callback from the Sanitizer for expanding items found in HTML attribute
5732 * values, so they can be safely tested and escaped.
5733 *
5734 * @param string $text
5735 * @param bool|PPFrame $frame
5736 * @return string
5737 */
5742 }
5744 /**
5745 * Accessor
5746 *
5747 * @return array
5748 */
5754 );
5755 }
5757 /**
5758 * Replace transparent tags in $text with the values given by the callbacks.
5759 *
5760 * Transparent tag hooks are like regular XML-style tag hooks, except they
5761 * operate late in the transformation sequence, on HTML instead of wikitext.
5762 *
5763 * @param string $text
5764 *
5765 * @return string
5766 */
5780 );
5783 }
5785 }
5787 }
5789 /**
5790 * Break wikitext input into sections, and either pull or replace
5791 * some particular section's text.
5792 *
5793 * External callers should use the getSection and replaceSection methods.
5794 *
5795 * @param string $text Page wikitext
5796 * @param string|number $sectionId A section identifier string of the form:
5797 * "<flag1> - <flag2> - ... - <section number>"
5798 *
5799 * Currently the only recognised flag is "T", which means the target section number
5800 * was derived during a template inclusion parse, in other words this is a template
5801 * section edit link. If no flags are given, it was an ordinary section edit link.
5802 * This flag is required to avoid a section numbering mismatch when a section is
5803 * enclosed by "<includeonly>" (bug 6563).
5804 *
5805 * The section number 0 pulls the text before the first heading; other numbers will
5806 * pull the given section along with its lower-level subsections. If the section is
5807 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5808 *
5809 * Section 0 is always considered to exist, even if it only contains the empty
5810 * string. If $text is the empty string and section 0 is replaced, $newText is
5811 * returned.
5812 *
5813 * @param string $mode One of "get" or "replace"
5814 * @param string $newText Replacement text for section data.
5815 * @return string For "get", the extracted section text.
5816 * for "replace", the whole page with the section replaced.
5817 */
5826 # Process section extraction flags
5833 }
5834 }
5836 # Check for empty input
5838 # Only sections 0 and T-0 exist in an empty document
5844 }
5850 }
5851 }
5852 }
5854 # Preprocess the text
5857 # <h> nodes indicate section breaks
5858 # They can only occur at the top level, so we can find them by iterating the root's children
5861 # Find the target section
5863 # Section zero doesn't nest, level=big
5872 }
5873 }
5876 }
5878 }
5879 }
5882 # Not found
5887 }
5888 }
5890 # Find the end of the section, including nested sections
5897 }
5898 }
5901 }
5905 # Write out the remainder (in replace mode only)
5907 # Output the replacement text
5908 # Add two newlines on -- trailing whitespace in $newText is conventionally
5909 # stripped by the editor, so we need both newlines to restore the paragraph gap
5910 # Only add trailing whitespace if there is newText
5913 }
5918 }
5919 }
5922 # Re-insert stripped tags
5924 }
5927 }
5929 /**
5930 * This function returns the text of a section, specified by a number ($section).
5931 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5932 * the first section before any such heading (section 0).
5933 *
5934 * If a section contains subsections, these are also returned.
5935 *
5936 * @param string $text Text to look in
5937 * @param string|number $sectionId Section identifier as a number or string
5938 * (e.g. 0, 1 or 'T-1').
5939 * @param string $defaultText Default to return if section is not found
5940 *
5941 * @return string Text of the requested section
5942 */
5945 }
5947 /**
5948 * This function returns $oldtext after the content of the section
5949 * specified by $section has been replaced with $text. If the target
5950 * section does not exist, $oldtext is returned unchanged.
5951 *
5952 * @param string $oldText Former text of the article
5953 * @param string|number $sectionId Section identifier as a number or string
5954 * (e.g. 0, 1 or 'T-1').
5955 * @param string $newText Replacing text
5956 *
5957 * @return string Modified text
5958 */
5961 }
5963 /**
5964 * Get the ID of the revision we are parsing
5965 *
5966 * @return int|null
5967 */
5970 }
5972 /**
5973 * Get the revision object for $this->mRevisionId
5974 *
5975 * @return Revision|null Either a Revision object or null
5976 * @since 1.23 (public since 1.23)
5977 */
5981 }
5984 }
5988 }
5990 /**
5991 * Get the timestamp associated with the current revision, adjusted for
5992 * the default server-local timestamp
5993 * @return string
5994 */
6003 # The cryptic '' timezone parameter tells to use the site-default
6004 # timezone offset instead of the user settings.
6005 #
6006 # Since this value will be saved into the parser cache, served
6007 # to other users, and potentially even used inside links and such,
6008 # it needs to be consistent for all visitors.
6011 }
6013 }
6015 /**
6016 * Get the name of the user that edited the last revision
6017 *
6018 * @return string User name
6019 */
6024 # if this template is subst: the revision id will be blank,
6025 # so just use the current user's name
6030 }
6031 }
6033 }
6035 /**
6036 * Get the size of the revision
6037 *
6038 * @return int|null Revision size
6039 */
6044 # if this variable is subst: the revision id will be blank,
6045 # so just use the parser input size, because the own substituation
6046 # will change the size.
6051 }
6052 }
6054 }
6056 /**
6057 * Mutator for $mDefaultSort
6058 *
6059 * @param string $sort New value
6060 */
6064 }
6066 /**
6067 * Accessor for $mDefaultSort
6068 * Will use the empty string if none is set.
6069 *
6070 * This value is treated as a prefix, so the
6071 * empty string is equivalent to sorting by
6072 * page name.
6073 *
6074 * @return string
6075 */
6081 }
6082 }
6084 /**
6085 * Accessor for $mDefaultSort
6086 * Unlike getDefaultSort(), will return false if none is set
6087 *
6088 * @return string|bool
6089 */
6092 }
6094 /**
6095 * Try to guess the section anchor name based on a wikitext fragment
6096 * presumably extracted from a heading, for example "Header" from
6097 * "== Header ==".
6098 *
6099 * @param string $text
6100 *
6101 * @return string
6102 */
6104 # Strip out wikitext links(they break the anchor)
6108 }
6110 /**
6111 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
6112 * instead. For use in redirects, since IE6 interprets Redirect: headers
6113 * as something other than UTF-8 (apparently?), resulting in breakage.
6114 *
6115 * @param string $text The section name
6116 * @return string An anchor
6117 */
6119 # Strip out wikitext links(they break the anchor)
6123 }
6125 /**
6126 * Strips a text string of wikitext for use in a section anchor
6127 *
6128 * Accepts a text string and then removes all wikitext from the
6129 * string and leaves only the resultant text (i.e. the result of
6130 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
6131 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
6132 * to create valid section anchors by mimicing the output of the
6133 * parser when headings are parsed.
6134 *
6135 * @param string $text Text string to be stripped of wikitext
6136 * for use in a Section anchor
6137 * @return string Filtered text string
6138 */
6140 # Strip internal link markup
6144 # Strip external link markup
6145 # @todo FIXME: Not tolerant to blank link text
6146 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
6147 # on how many empty links there are on the page - need to figure that out.
6148 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
6150 # Parse wikitext quotes (italics & bold)
6153 # Strip HTML tags
6156 }
6158 /**
6159 * strip/replaceVariables/unstrip for preprocessor regression testing
6160 *
6161 * @param string $text
6162 * @param Title $title
6163 * @param ParserOptions $options
6164 * @param int $outputType
6165 *
6166 * @return string
6167 */
6168 public function testSrvus( $text, Title $title, ParserOptions $options, $outputType = self::OT_HTML ) {
6176 }
6178 /**
6179 * @param string $text
6180 * @param Title $title
6181 * @param ParserOptions $options
6182 * @return string
6183 */
6186 }
6188 /**
6189 * @param string $text
6190 * @param Title $title
6191 * @param ParserOptions $options
6192 * @return string
6193 */
6196 }
6198 /**
6199 * Call a callback function on all regions of the given text that are not
6200 * inside strip markers, and replace those regions with the return value
6201 * of the callback. For example, with input:
6202 *
6203 * aaa<MARKER>bbb
6204 *
6205 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
6206 * two strings will be replaced with the value returned by the callback in
6207 * each case.
6208 *
6209 * @param string $s
6210 * @param callable $callback
6211 *
6212 * @return string
6213 */
6232 }
6233 }
6234 }
6236 }
6238 /**
6239 * Remove any strip markers found in the given text.
6240 *
6241 * @param string $text Input string
6242 * @return string
6243 */
6246 }
6248 /**
6249 * Save the parser state required to convert the given half-parsed text to
6250 * HTML. "Half-parsed" in this context means the output of
6251 * recursiveTagParse() or internalParse(). This output has strip markers
6252 * from replaceVariables (extensionSubstitution() etc.), and link
6253 * placeholders from replaceLinkHolders().
6254 *
6255 * Returns an array which can be serialized and stored persistently. This
6256 * array can later be loaded into another parser instance with
6257 * unserializeHalfParsedText(). The text can then be safely incorporated into
6258 * the return value of a parser hook.
6259 *
6260 * @param string $text
6261 *
6262 * @return array
6263 */
6270 );
6272 }
6274 /**
6275 * Load the parser state given in the $data array, which is assumed to
6276 * have been generated by serializeHalfParsedText(). The text contents is
6277 * extracted from the array, and its markers are transformed into markers
6278 * appropriate for the current Parser instance. This transformed text is
6279 * returned, and can be safely included in the return value of a parser
6280 * hook.
6281 *
6282 * If the $data array has been stored persistently, the caller should first
6283 * check whether it is still valid, by calling isValidHalfParsedText().
6284 *
6285 * @param array $data Serialized data
6286 * @throws MWException
6287 * @return string
6288 */
6292 }
6294 # First, extract the strip state.
6298 # Now renumber links
6301 # Should be good to go.
6303 }
6305 /**
6306 * Returns true if the given array, presumed to be generated by
6307 * serializeHalfParsedText(), is compatible with the current version of the
6308 * parser.
6309 *
6310 * @param array $data
6311 *
6312 * @return bool
6313 */
6316 }
6318 /**
6319 * Parsed a width param of imagelink like 300px or 200x300px
6320 *
6321 * @param string $value
6322 *
6323 * @return array
6324 * @since 1.20
6325 */
6330 }
6332 # (bug 13500) In both cases (width/height and width only),
6333 # permit trailing "px" for backward compatibility.
6342 }
6344 }
6346 /**
6347 * Lock the current instance of the parser.
6348 *
6349 * This is meant to stop someone from calling the parser
6350 * recursively and messing up all the strip state.
6351 *
6352 * @throws MWException If parser is in a parse
6353 * @return ScopedCallback The lock will be released once the return value goes out of scope.
6354 */
6359 }
6365 } );
6368 }
6370 /**
6371 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6372 *
6373 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6374 * or if there is more than one <p/> tag in the input HTML.
6375 *
6376 * @param string $html
6377 * @return string
6378 * @since 1.24
6379 */
6385 }
6386 }
6389 }
6391 /**
6392 * Return this parser if it is not doing anything, otherwise
6393 * get a fresh parser. You can use this method by doing
6394 * $myParser = $wgParser->getFreshParser(), or more simply
6395 * $wgParser->getFreshParser()->parse( ... );
6396 * if you're unsure if $wgParser is safe to use.
6397 *
6398 * @since 1.24
6399 * @return Parser A parser object that is not parsing anything
6400 */
6407 }
6408 }
6409 }