| blob | fe0c81f338b83f4ee0f324f54bfee18ae77b97e4 |
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
82 # Also available as global constants from Defines.php
86 # Constants needed for external link processing
87 # Everything except bracket, space, or control characters
88 # \p{Zs} is unicode 'separator, space' category. It covers the space 0x20
89 # as well as U+3000 is IDEOGRAPHIC SPACE for bug 19052
94 # State constants for the definition list colon extraction
104 # Flags for preprocessToDom
107 # Allowed values for $this->mOutputType
108 # Parameter to startExternalParse().
115 # Marker Suffix needs to be accessible staticly.
118 # Markers used for wrapping the table of contents
122 # Persistent:
136 # Initialised by initialiseVariables()
138 /**
139 * @var MagicWordArray
140 */
143 /**
144 * @var MagicWordArray
145 */
147 public $mConf, $mPreprocessor, $mExtLinkBracketedRegex, $mUrlProtocols; # Initialised in constructor
149 # Cleared with clearState():
150 /**
151 * @var ParserOutput
152 */
156 /**
157 * @var StripState
158 */
162 /**
163 * @var LinkHolderArray
164 */
174 /**
175 * @var User
176 */
179 # Temporary
180 # These are variables reset at least once per parse regardless of $clearState
182 /**
183 * @var ParserOptions
184 */
187 /**
188 * @var Title
189 */
201 /**
202 * @var string
203 */
206 /**
207 * @var array Array with the language name of each language link (i.e. the
208 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
209 * duplicate language links to the ParserOutput.
210 */
213 /**
214 * @var MapCacheLRU|null
215 * @since 1.24
216 *
217 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
218 */
221 /**
222 * @var bool Recursive call protection.
223 * This variable should be treated as if it were private.
224 */
227 /**
228 * @param array $conf
229 */
238 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
241 # PECL extension that conflicts with the core DOM extension (bug 13770)
248 }
250 }
252 /**
253 * Reduce memory usage to reduce the impact of circular references
254 */
258 }
261 }
262 }
264 /**
265 * Allow extensions to clean up when the parser is cloned
266 */
270 // Bug 56226: When you create a reference "to" an object field, that
271 // makes the object field itself be a reference too (until the other
272 // reference goes out of scope). When cloning, any field that's a
273 // reference is copied as a reference in the new object. Both of these
274 // are defined PHP5 behaviors, as inconvenient as it is for us when old
275 // hooks from PHP4 days are passing fields by reference.
277 // Make a non-reference copy of the field, then rebind the field to
278 // reference the new copy.
282 }
285 }
287 /**
288 * Do various kinds of initialisation on the first call of the parser
289 */
293 }
304 }
306 /**
307 * Clear Parser state
308 *
309 * @private
310 */
315 }
333 /**
334 * Prefix for temporary replacement strings for the multipass parser.
335 * \x07 should never appear in input as it's disallowed in XML.
336 * Using it at the front also gives us a little extra robustness
337 * since it shouldn't match when butted up against identifier-like
338 * string constructs.
339 *
340 * Must not consist of all title characters, or else it will change
341 * the behavior of <nowiki> in a link.
342 */
346 # Clear these on every parse, bug 4549
354 );
363 # Fix cloning
366 }
370 }
372 /**
373 * Convert wikitext to HTML
374 * Do not call this function recursively.
375 *
376 * @param string $text Text we want to parse
377 * @param Title $title
378 * @param ParserOptions $options
379 * @param bool $linestart
380 * @param bool $clearState
381 * @param int $revid Number to pass in {{REVISIONID}}
382 * @return ParserOutput A ParserOutput
383 */
386 ) {
387 /**
388 * First pass--just handle <nowiki> sections, pass the rest off
389 * to internalParse() which does all the real work.
390 */
399 }
407 }
409 # Remove the strip marker tag prefix from the input, if present.
412 }
425 }
428 # No more strip!
435 # Clean up special characters, only run once, next-to-last before doBlockLevels
437 # french spaces, last one Guillemet-left
438 # only if there is something before the space
440 # french spaces, Guillemet-right
443 );
450 /**
451 * The input doesn't get language converted if
452 * a) It's disabled
453 * b) Content isn't converted
454 * c) It's a conversion table
455 * d) it is an interface message (which is in the user language)
456 */
459 ) {
461 # The position of the convert() call should not be changed. it
462 # assumes that the links are all replaced and the only thing left
463 # is the <nowiki> mark.
465 }
466 }
468 /**
469 * A converted title will be provided in the output object if title and
470 * content conversion are enabled, the article text does not contain
471 * a conversion-suppressing double-underscore tag, and no
472 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
473 * automatic link conversion.
474 */
479 ) {
486 }
487 }
501 # attempt to sanitize at least some nesting problems
502 # (bug #2702 and quite a few others)
504 # ''Something [http://www.cool.com cool''] -->
505 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
508 # fix up an anchor inside another anchor, only
509 # at least for a single single nested link (bug 3695)
512 # fix div inside inline elements- doBlockLevels won't wrap a line which
513 # contains a div, so fix it up here; replace
514 # div with escaped text
517 # remove empty italic or bold tag pairs, some
518 # introduced by rules above
520 );
526 }
532 );
533 }
537 # Information on include size limits, for the benefit of users who try to skirt them
545 );
546 }
551 );
555 );
558 );
561 );
564 );
567 );
570 );
576 }
580 ) ) {
586 }
590 }
591 }
592 }
593 // Since we're not really outputting HTML, decode the entities and
594 // then re-encode the things that need hiding inside HTML comments.
598 // Sanitize for comment. Note '‐' in the replacement is U+2010,
599 // which looks much like the problematic '-'.
606 }
607 }
621 }
623 /**
624 * Recursive parser entry point that can be called from an extension tag
625 * hook.
626 *
627 * If $frame is not provided, then template variables (e.g., {{{1}}}) within $text are not expanded
628 *
629 * @param string $text Text extension wants to have parsed
630 * @param bool|PPFrame $frame The frame to use for expanding any template variables
631 *
632 * @return string
633 */
641 }
643 /**
644 * Expand templates and variables in the text, producing valid, static wikitext.
645 * Also removes comments.
646 * Do not call this function recursively.
647 * @param string $text
648 * @param Title $title
649 * @param ParserOptions $options
650 * @param int|null $revid
651 * @param bool|PPFrame $frame
652 * @return mixed|string
653 */
654 public function preprocess( $text, Title $title = null, ParserOptions $options, $revid = null, $frame = false ) {
660 }
667 }
669 /**
670 * Recursive parser entry point that can be called from an extension tag
671 * hook.
672 *
673 * @param string $text Text to be expanded
674 * @param bool|PPFrame $frame The frame to use for expanding any template variables
675 * @return string
676 * @since 1.19
677 */
684 }
686 /**
687 * Process the wikitext for the "?preload=" feature. (bug 5210)
688 *
689 * "<noinclude>", "<includeonly>" etc. are parsed as for template
690 * transclusion, comments, templates, arguments, tags hooks and parser
691 * functions are untouched.
692 *
693 * @param string $text
694 * @param Title $title
695 * @param ParserOptions $options
696 * @param array $params
697 * @return string
698 */
699 public function getPreloadText( $text, Title $title, ParserOptions $options, $params = array() ) {
703 # Parser (re)initialisation
712 }
714 /**
715 * Get a random string
716 *
717 * @return string
718 */
721 }
723 /**
724 * Set the current user.
725 * Should only be used when doing pre-save transform.
726 *
727 * @param User|null $user User object or null (to reset)
728 */
731 }
733 /**
734 * Accessor for mUniqPrefix.
735 *
736 * @return string
737 */
740 # @todo FIXME: This is probably *horribly wrong*
741 # LanguageConverter seems to want $wgParser's uniqPrefix, however
742 # if this is called for a parser cache hit, the parser may not
743 # have ever been initialized in the first place.
744 # Not really sure what the heck is supposed to be going on here.
746 # throw new MWException( "Accessing uninitialized mUniqPrefix" );
747 }
749 }
751 /**
752 * Set the context title
753 *
754 * @param Title $t
755 */
759 }
762 # Strip the fragment to avoid various odd effects
767 }
768 }
770 /**
771 * Accessor for the Title object
772 *
773 * @return Title
774 */
777 }
779 /**
780 * Accessor/mutator for the Title object
781 *
782 * @param Title $x Title object or null to just get the current one
783 * @return Title
784 */
787 }
789 /**
790 * Set the output type
791 *
792 * @param int $ot New value
793 */
796 # Shortcut alias
802 );
803 }
805 /**
806 * Accessor/mutator for the output type
807 *
808 * @param int|null $x New value or null to just get the current one
809 * @return int
810 */
813 }
815 /**
816 * Get the ParserOutput object
817 *
818 * @return ParserOutput
819 */
822 }
824 /**
825 * Get the ParserOptions object
826 *
827 * @return ParserOptions
828 */
831 }
833 /**
834 * Accessor/mutator for the ParserOptions object
835 *
836 * @param ParserOptions $x New value or null to just get the current one
837 * @return ParserOptions Current ParserOptions object
838 */
841 }
843 /**
844 * @return int
845 */
848 }
850 /**
851 * @param int $id
852 */
855 }
857 /**
858 * Get a language object for use in parser functions such as {{FORMATNUM:}}
859 * @return Language
860 */
863 }
865 /**
866 * Get the target language for the content being parsed. This is usually the
867 * language that the content is in.
868 *
869 * @since 1.19
870 *
871 * @throws MWException
872 * @return Language
873 */
883 }
886 }
888 /**
889 * Get the language object for language conversion
890 * @return Language|null
891 */
894 }
896 /**
897 * Get a User object either from $this->mUser, if set, or from the
898 * ParserOptions object otherwise
899 *
900 * @return User
901 */
905 }
907 }
909 /**
910 * Get a preprocessor object
911 *
912 * @return Preprocessor
913 */
918 }
920 }
922 /**
923 * Replaces all occurrences of HTML-style comments and the given tags
924 * in the text with a random marker and returns the next text. The output
925 * parameter $matches will be an associative array filled with data in
926 * the form:
927 *
928 * @code
929 * 'UNIQ-xxxxx' => array(
930 * 'element',
931 * 'tag content',
932 * array( 'param' => 'x' ),
933 * '<element param="x">tag content</element>' ) )
934 * @endcode
935 *
936 * @param array $elements List of element names. Comments are always extracted.
937 * @param string $text Source text string.
938 * @param array $matches Out parameter, Array: extracted tags
939 * @param string $uniq_prefix
940 * @return string Stripped text
941 */
955 }
957 # comment
963 # tag
968 }
974 # Empty element tag, <tag />
983 }
987 # No end tag -- let it run out to the end of the text.
993 }
994 }
1000 }
1002 }
1004 /**
1005 * Get a list of strippable XML-like elements
1006 *
1007 * @return array
1008 */
1011 }
1013 /**
1014 * Add an item to the strip state
1015 * Returns the unique tag which must be inserted into the stripped text
1016 * The tag will be replaced with the original text in unstrip()
1017 *
1018 * @param string $text
1019 *
1020 * @return string
1021 */
1027 }
1029 /**
1030 * parse the wiki syntax used to render tables
1031 *
1032 * @private
1033 * @param string $text
1034 * @return string
1035 */
1054 }
1060 # First check if we are starting a new table
1073 # Don't do any of the following
1077 # We are ending a table
1083 }
1087 }
1091 }
1095 # Now we have a table row
1098 # Whats after the tag is now only attributes
1111 }
1115 }
1124 ) {
1125 # This might be cell elements, td, th or captions
1129 }
1135 }
1137 # Split up multiple cells on the same line.
1138 # FIXME : This can result in improper nesting of tags processed
1139 # by earlier parser steps, but should avoid splitting up eg
1140 # attribute values containing literal "||".
1145 # Loop through each table cell
1152 }
1157 }
1163 }
1173 }
1177 # A cell could contain both parameters and data
1180 # Bug 553: Note that a '|' inside an invalid link should not
1181 # be mistaken as delimiting cell parameters
1190 }
1194 }
1195 }
1197 }
1199 # Closing open td, tr && table
1203 }
1206 }
1209 }
1212 }
1214 # Remove trailing line-ending (b/c)
1217 }
1219 # special case: don't return empty table
1222 }
1227 }
1229 /**
1230 * Helper function for parse() that transforms wiki markup into
1231 * HTML. Only called for $mOutputType == self::OT_HTML.
1232 *
1233 * @private
1234 *
1235 * @param string $text
1236 * @param bool $isMain
1237 * @param bool $frame
1238 *
1239 * @return string
1240 */
1246 # Hook to suspend the parser in this state
1247 if ( !wfRunHooks( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState ) ) ) {
1250 }
1252 # if $frame is provided, then use $frame for replacing any variables
1254 # use frame depth to infer how include/noinclude tags should be handled
1255 # depth=0 means this is the top-level document; otherwise it's an included document
1260 }
1264 # if $frame is not provided, then use old-style replaceVariables
1266 }
1274 );
1277 # Tables need to come after variable replacement for things to work
1278 # properly; putting them before other transformations should keep
1279 # exciting things like link expansions from showing up in surprising
1280 # places.
1292 # replaceInternalLinks may sometimes leave behind
1293 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1301 }
1303 /**
1304 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1305 * magic external links.
1306 *
1307 * DML
1308 * @private
1309 *
1310 * @param string $text
1311 *
1312 * @return string
1313 */
1319 '!(?: # Start cases
1323 (?:RFC|PMID)\s+([0-9]+) | # m[4]: RFC or PMID, capture number
1325 (?: 97[89] [\ \-]? )? # optional 13-digit ISBN prefix
1326 (?: [0-9] [\ \-]? ){9} # 9 digits with opt. delimiters
1327 [0-9Xx] # check digit
1332 }
1334 /**
1335 * @throws MWException
1336 * @param array $m
1337 * @return HTML|string
1338 */
1341 # Skip anchor
1344 # Skip HTML element
1347 # Free external link
1350 # RFC or PMID
1364 }
1368 # ISBN
1374 ));
1381 }
1382 }
1384 /**
1385 * Make a free external link, given a user-supplied URL
1386 *
1387 * @param string $url
1388 *
1389 * @return string HTML
1390 * @private
1391 */
1397 # The characters '<' and '>' (which were escaped by
1398 # removeHTMLtags()) should not be included in
1399 # URLs, per RFC 2396.
1404 }
1406 # Move trailing punctuation to $trail
1408 # If there is no left bracket, then consider right brackets fair game too
1411 }
1417 }
1421 # Is this an external image?
1424 # Not an image, make a link
1429 # Register it in the output object...
1430 # Replace unnecessary URL escape codes with their equivalent characters
1433 }
1436 }
1438 /**
1439 * Parse headers and return html
1440 *
1441 * @private
1442 *
1443 * @param string $text
1444 *
1445 * @return string
1446 */
1452 }
1455 }
1457 /**
1458 * Replace single quotes with HTML markup
1459 * @private
1460 *
1461 * @param string $text
1462 *
1463 * @return string The altered text
1464 */
1471 }
1475 }
1477 /**
1478 * Helper function for doAllQuotes()
1479 *
1480 * @param string $text
1481 *
1482 * @return string
1483 */
1489 }
1491 // First, do some preliminary work. This may shift some apostrophes from
1492 // being mark-up to being text. It also counts the number of occurrences
1493 // of bold and italics mark-ups.
1498 // If there are ever four apostrophes, assume the first is supposed to
1499 // be text, and the remaining three constitute mark-up for bold text.
1500 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1506 // If there are more than 5 apostrophes in a row, assume they're all
1507 // text except for the last 5.
1508 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1512 }
1513 // Count the number of occurrences of bold and italics mark-ups.
1521 }
1522 }
1524 // If there is an odd number of both bold and italics, it is likely
1525 // that one of the bold ones was meant to be an apostrophe followed
1526 // by italics. Which one we cannot know for certain, but it is more
1527 // likely to be one that has a single-letter word before it.
1539 }
1543 // if $firstsingleletterword is set, we don't
1544 // look at the other options, so we can bail early.
1546 }
1550 }
1551 }
1552 }
1553 }
1555 // If there is a single-letter word, use it!
1560 // If not, but there's a multi-letter word, use that one.
1564 // ... otherwise use the first one that has neither.
1565 // (notice that it is possible for all three to be -1 if, for example,
1566 // there is only one pentuple-apostrophe in the line)
1569 }
1570 }
1572 // Now let's actually convert our apostrophic mush to HTML!
1583 }
1602 }
1619 }
1639 }
1640 }
1641 }
1643 }
1644 // Now close all remaining tags. Notice that the order is important.
1647 }
1650 }
1653 }
1654 // There might be lonely ''''', so make sure we have a buffer
1657 }
1659 }
1661 /**
1662 * Replace external links (REL)
1663 *
1664 * Note: this is all very hackish and the order of execution matters a lot.
1665 * Make sure to run tests/parserTests.php if you change this code.
1666 *
1667 * @private
1668 *
1669 * @param string $text
1670 *
1671 * @throws MWException
1672 * @return string
1673 */
1682 }
1692 # The characters '<' and '>' (which were escaped by
1693 # removeHTMLtags()) should not be included in
1694 # URLs, per RFC 2396.
1699 }
1701 # If the link text is an image URL, replace it with an <img> tag
1702 # This happened by accident in the original parser, but some people used it extensively
1706 }
1710 # Set linktype for CSS - if URL==text, link is essentially free
1713 # No link text, e.g. [http://domain.tld/some.link]
1715 # Autonumber
1720 # Have link text, e.g. [http://domain.tld/some.link text]s
1721 # Check for trail
1723 }
1729 # Use the encoded URL
1730 # This means that users can paste URLs directly into the text
1731 # Funny characters like ö aren't valid in URLs anyway
1732 # This was changed in August 2004
1736 # Register link in the output object.
1737 # Replace unnecessary URL escape codes with the referenced character
1738 # This prevents spammers from hiding links from the filters
1741 }
1745 }
1747 /**
1748 * Get the rel attribute for a particular external link.
1749 *
1750 * @since 1.21
1751 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1752 * nofollow if appropriate
1753 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1754 * @return string|null Rel attribute for $url
1755 */
1761 ) {
1763 }
1765 }
1767 /**
1768 * Get an associative array of additional HTML attributes appropriate for a
1769 * particular external link. This currently may include rel => nofollow
1770 * (depending on configuration, namespace, and the URL's domain) and/or a
1771 * target attribute (depending on configuration).
1772 *
1773 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1774 * nofollow if appropriate
1775 * @return array Associative array of HTML attributes
1776 */
1783 }
1785 }
1787 /**
1788 * Replace unusual escape codes in a URL with their equivalent characters
1789 *
1790 * @deprecated since 1.24, use normalizeLinkUrl
1791 * @param string $url
1792 * @return string
1793 */
1797 }
1799 /**
1800 * Replace unusual escape codes in a URL with their equivalent characters
1801 *
1802 * This generally follows the syntax defined in RFC 3986, with special
1803 * consideration for HTTP query strings.
1804 *
1805 * @param string $url
1806 * @return string
1807 */
1809 # First, make sure unsafe characters are encoded
1813 },
1814 $url
1815 );
1820 # Fragment part - 'fragment'
1826 }
1828 # Query part - 'query' minus &=+;
1834 }
1836 # Scheme and path part - 'pchar'
1837 # (we assume no userinfo or encoded colons in the host)
1842 }
1849 # Unescape it
1852 # Leave it escaped, but use uppercase for a-f
1854 }
1855 };
1857 }
1859 /**
1860 * make an image if it's allowed, either through the global
1861 * option, through the exception, or through the on-wiki whitelist
1862 *
1863 * @param string $url
1864 *
1865 * @return string
1866 */
1871 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1878 }
1879 }
1884 }
1888 ) {
1890 # Image found
1892 }
1893 }
1896 ) {
1900 );
1903 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1906 }
1908 # Image matches a whitelist entry
1911 }
1912 }
1913 }
1915 }
1917 /**
1918 * Process [[ ]] wikilinks
1919 *
1920 * @param string $s
1921 *
1922 * @return string Processed text
1923 *
1924 * @private
1925 */
1929 }
1931 /**
1932 * Process [[ ]] wikilinks (RIL)
1933 * @param string $s
1934 * @throws MWException
1935 * @return LinkHolderArray
1936 *
1937 * @private
1938 */
1945 # the % is needed to support urlencoded titles as well
1948 # Match a link having the form [[namespace:link|alternate]]trail
1950 # Match cases where there is no "]]", which might still be images
1952 }
1956 # split the entire text string on occurrences of [[
1958 # get the first element (all text up to first [[), and remove the space we added
1967 # Match the end of a line for a word that's not followed by whitespace,
1968 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
1972 }
1978 }
1987 }
1990 }
1995 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
1996 # Loop for each link
1998 // @codingStandardsIgnoreStart
2000 # Check for excessive memory usage
2002 # Too big
2003 # Do the existence check, replace the link holders and clear the array
2006 }
2015 }
2016 # first link
2020 }
2022 }
2029 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2030 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2031 # the real problem is with the $e1 regex
2032 # See bug 1300.
2033 #
2034 # Still some problems for cases where the ] is meant to be outside punctuation,
2035 # and no image is in sight. See bug 2095.
2036 #
2040 ) {
2043 }
2044 # fix up urlencoded title texts
2046 # Should anchors '#' also be rejected?
2048 }
2051 # Invalid, but might be an image with a link in its caption
2056 }
2062 }
2068 # Don't allow internal links to pages containing
2069 # PROTO: where PROTO is a valid URL protocol; these
2070 # should be external links.
2075 }
2077 # Make subpage if necessary
2082 }
2086 # Strip off leading ':'
2088 }
2097 }
2108 # look at the next 'line' to see if we can close it there
2113 }
2116 # the first ]] closes the inner link, the second the image
2122 # if there's exactly one ]] that's fine, we'll keep looking
2125 # if $next_line is invalid too, we need look no further
2128 }
2129 }
2131 # we couldn't find the end of this imageLink, so output it raw
2132 # but don't ignore what might be perfectly normal links in the text we've examined
2135 # note: no $trail, because without an end, there *is* no trail
2138 }
2141 # note: no $trail, because without an end, there *is* no trail
2144 }
2146 }
2152 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2153 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2154 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2155 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2157 }
2159 # Link not escaped by : , create the various objects
2161 # Interwikis
2167 )
2168 ) {
2169 # Bug 24502: filter duplicates
2173 }
2179 }
2186 # if no parameters were passed, $text
2187 # becomes something like "File:Foo.png",
2188 # which we don't want to pass on to the
2189 # image generator
2192 # recursively parse links inside the image caption
2193 # actually, this will parse them in any other parameters, too,
2194 # but it might be hard to fix that, and it doesn't matter ATM
2197 }
2198 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2203 }
2206 }
2216 }
2222 /**
2223 * Strip the whitespace Category links produce, see bug 87
2224 */
2229 }
2230 }
2232 # Self-link checking. For some languages, variants of the title are checked in
2233 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2234 # for linking to a different variant.
2238 }
2240 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2241 # @todo FIXME: Should do batch file existence checks, see comment below
2244 # Give extensions a chance to select the file revision for us
2249 # Fetch and register the file (file title may be different via hooks)
2251 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2256 }
2259 # Some titles, such as valid special pages or files in foreign repos, should
2260 # be shown as bluelinks even though they're not included in the page table
2261 #
2262 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2263 # batch file existence checks for NS_FILE and NS_MEDIA
2268 # Links will be added to the output link list after checking
2270 }
2272 }
2275 }
2277 /**
2278 * Render a forced-blue link inline; protect against double expansion of
2279 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2280 * Since this little disaster has to split off the trail text to avoid
2281 * breaking URLs in the following text without breaking trails on the
2282 * wiki links, it's been made into a horrible function.
2283 *
2284 * @param Title $nt
2285 * @param string $text
2286 * @param array|string $query
2287 * @param string $trail
2288 * @param string $prefix
2289 * @return string HTML-wikitext mix oh yuck
2290 */
2291 public function makeKnownLinkHolder( $nt, $text = '', $query = array(), $trail = '', $prefix = '' ) {
2296 }
2299 }
2304 }
2306 /**
2307 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2308 * going to go through further parsing steps before inline URL expansion.
2309 *
2310 * Not needed quite as much as it used to be since free links are a bit
2311 * more sensible these days. But bracketed links are still an issue.
2312 *
2313 * @param string $text More-or-less HTML
2314 * @return string Less-or-more HTML with NOPARSE bits
2315 */
2319 }
2321 /**
2322 * Return true if subpage links should be expanded on this page.
2323 * @return bool
2324 */
2326 # Some namespaces don't allow subpages
2328 }
2330 /**
2331 * Handle link to subpage if necessary
2332 *
2333 * @param string $target The source of the link
2334 * @param string &$text The link text, modified as necessary
2335 * @return string The full name of the link
2336 * @private
2337 */
2340 }
2342 /**#@+
2343 * Used by doBlockLevels()
2344 * @private
2345 *
2346 * @return string
2347 */
2352 }
2356 }
2358 /**
2359 * getCommon() returns the length of the longest common substring
2360 * of both arguments, starting at the beginning of both.
2361 * @private
2362 *
2363 * @param string $st1
2364 * @param string $st2
2365 *
2366 * @return int
2367 */
2373 }
2378 }
2379 }
2381 }
2383 /**
2384 * These next three functions open, continue, and close the list
2385 * element appropriate to the prefix character passed into them.
2386 * @private
2387 *
2388 * @param string $char
2389 *
2390 * @return string
2391 */
2406 }
2409 }
2411 /**
2412 * TODO: document
2413 * @param string $char
2414 * @private
2415 *
2416 * @return string
2417 */
2425 }
2432 }
2433 }
2435 }
2437 /**
2438 * @todo Document
2439 * @param string $char
2440 * @private
2441 *
2442 * @return string
2443 */
2455 }
2458 }
2460 }
2461 /**#@-*/
2463 /**
2464 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2465 *
2466 * @param string $text
2467 * @param bool $linestart Whether or not this is at the start of a line.
2468 * @private
2469 * @return string The lists rendered as HTML
2470 */
2474 # Parsing through the text line by line. The main thing
2475 # happening here is handling of block-level elements p, pre,
2476 # and making lists from lines starting with * # : etc.
2477 #
2487 # Fix up $linestart
2492 }
2493 # * = ul
2494 # # = ol
2495 # ; = dt
2496 # : = dd
2501 # If not in a <pre> element, scan for and figure out what prefixes are there.
2503 # Multiple prefixes may abut each other for nested lists.
2507 # eh?
2508 # ; and : are both from definition-lists, so they're equivalent
2509 # for the purposes of determining whether or not we need to open/close
2510 # elements.
2515 # Don't interpret any other prefixes in preformatted text
2519 }
2521 # List generation
2523 # Same as the last item, so no need to deal with nesting or opening stuff
2528 # The one nasty exception: definition lists work like this:
2529 # ; title : definition text
2530 # So we check for : in the remainder text to split up the
2531 # title and definition, without b0rking links.
2536 }
2537 }
2539 # We need to open or close prefixes, or both.
2541 # Either open or close a level...
2545 # Close all the prefixes which aren't shared.
2549 }
2551 # Continue the current prefix if appropriate.
2554 }
2556 # Open prefixes where appropriate.
2559 }
2565 # @todo FIXME: This is dupe of code above
2569 }
2570 }
2572 }
2575 }
2577 }
2579 # If we have no prefixes, go to paragraph mode.
2582 # No prefix (not in list)--go to paragraph mode
2583 # XXX: use a stack for nestable elements like span, table and div
2585 '/(?:<table|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|'
2587 $t
2588 );
2594 $t
2595 );
2599 # @todo bug 5718: paragraph closed
2603 }
2605 while ( preg_match( '/<(\\/?)blockquote[\s>]/i', $t, $bqMatch, PREG_OFFSET_CAPTURE, $bqOffset ) ) {
2608 }
2614 ) {
2615 # pre
2620 }
2623 # paragraph
2636 }
2637 }
2646 }
2647 }
2648 }
2649 }
2651 }
2652 # somewhere above we forget to get out of pre block (bug 785)
2655 }
2660 }
2661 }
2662 }
2668 }
2669 }
2673 }
2677 }
2679 /**
2680 * Split up a string on ':', ignoring any occurrences inside tags
2681 * to prevent illegal overlapping.
2682 *
2683 * @param string $str The string to split
2684 * @param string &$before Set to everything before the ':'
2685 * @param string &$after Set to everything after the ':'
2686 * @throws MWException
2687 * @return string The position of the ':', or false if none found
2688 */
2694 # Nothing to find!
2697 }
2701 # Easy; no tag nesting to worry about
2706 }
2708 # Ugly state machine to walk through avoiding tags.
2716 # (Using the number is a performance hack for common cases)
2720 # Could be either a <start> tag or an </end> tag
2725 # We found it!
2730 }
2731 # Embedded in a tag; don't break it.
2734 # Skip ahead looking for something interesting
2737 # Nothing else interesting
2740 }
2744 # We found it!
2749 }
2750 }
2752 # Nothing else interesting to find; abort!
2753 # We're nested, but there's no close tags left. Abort!
2755 }
2756 # Skip ahead to next tag start
2759 }
2762 # In a <tag>
2769 # Slash may be followed by >?
2773 # ignore
2774 }
2785 # Illegal early close? This shouldn't happen D:
2790 }
2793 # In a </tag>
2800 }
2802 }
2806 # Yes, a self-closed tag <blah/>
2809 # Probably we're jumping the gun, and this is an attribute
2811 }
2816 }
2823 }
2830 }
2835 }
2836 }
2838 wfDebug( __METHOD__ . ": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2841 }
2844 }
2846 /**
2847 * Return value of a magic variable (like PAGENAME)
2848 *
2849 * @private
2850 *
2851 * @param int $index
2852 * @param bool|PPFrame $frame
2853 *
2854 * @throws MWException
2855 * @return string
2856 */
2862 // If no title set, bad things are going to happen
2863 // later. Title should always be set since this
2864 // should only be called in the middle of a parse
2865 // operation (but the unit-tests do funky stuff)
2868 }
2870 /**
2871 * Some of these require message or data lookups and can be
2872 * expensive to check many times.
2873 */
2877 }
2878 }
2923 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
2957 ) ) );
2967 ) ) );
2975 }
2983 }
2996 # 0 means the page doesn't exist in the database,
2997 # which means the user is previewing a new page.
2998 # The vary-revision flag must be set, because the magic word
2999 # will have a different value once the page is saved.
3002 }
3006 # Let the edit saving system know we should parse the page
3007 # *after* a revision ID has been assigned.
3013 # Let the edit saving system know we should parse the page
3014 # *after* a revision ID has been assigned. This is for null edits.
3020 # Let the edit saving system know we should parse the page
3021 # *after* a revision ID has been assigned. This is for null edits.
3027 # Let the edit saving system know we should parse the page
3028 # *after* a revision ID has been assigned. This is for null edits.
3034 # Let the edit saving system know we should parse the page
3035 # *after* a revision ID has been assigned. This is for null edits.
3041 # Let the edit saving system know we should parse the page
3042 # *after* a revision ID has been assigned. This is for null edits.
3048 # Let the edit saving system know we should parse the page
3049 # *after* a revision ID has been assigned. This is for null edits.
3055 # Let the edit saving system know we should parse the page
3056 # *after* a revision ID has been assigned. This is for null edits.
3062 # Let the edit saving system know we should parse the page
3063 # *after* a revision ID has been assigned. This is for null edits.
3104 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3105 # int to remove the padding
3114 );
3123 false
3124 );
3130 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3131 # int to remove the padding
3192 );
3195 }
3199 }
3202 }
3204 /**
3205 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
3206 *
3207 * @private
3208 */
3217 }
3219 /**
3220 * Preprocess some wikitext and return the document tree.
3221 * This is the ghost of replace_variables().
3222 *
3223 * @param string $text The text to parse
3224 * @param int $flags Bitwise combination of:
3225 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
3226 * included. Default is to assume a direct page view.
3227 *
3228 * The generated DOM tree must depend only on the input text and the flags.
3229 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
3230 *
3231 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
3232 * change in the DOM tree for a given text, must be passed through the section identifier
3233 * in the section edit link and thus back to extractSections().
3234 *
3235 * The output of this function is currently only cached in process memory, but a persistent
3236 * cache may be implemented at a later date which takes further advantage of these strict
3237 * dependency requirements.
3238 *
3239 * @return PPNode
3240 */
3244 }
3246 /**
3247 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3248 *
3249 * @param string $s
3250 *
3251 * @return array
3252 */
3262 }
3264 }
3266 /**
3267 * Replace magic variables, templates, and template arguments
3268 * with the appropriate text. Templates are substituted recursively,
3269 * taking care to avoid infinite loops.
3270 *
3271 * Note that the substitution depends on value of $mOutputType:
3272 * self::OT_WIKI: only {{subst:}} templates
3273 * self::OT_PREPROCESS: templates but not extension tags
3274 * self::OT_HTML: all templates and extension tags
3275 *
3276 * @param string $text The text to transform
3277 * @param bool|PPFrame $frame Object describing the arguments passed to the
3278 * template. Arguments may also be provided as an associative array, as
3279 * was the usual case before MW1.12. Providing arguments this way may be
3280 * useful for extensions wishing to perform variable replacement
3281 * explicitly.
3282 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
3283 * double-brace expansion.
3284 * @return string
3285 */
3287 # Is there any text? Also, Prevent too big inclusions!
3290 }
3299 }
3307 }
3309 /**
3310 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3311 *
3312 * @param array $args
3313 *
3314 * @return array
3315 */
3328 }
3331 }
3332 }
3333 }
3336 }
3338 /**
3339 * Warn the user when a parser limitation is reached
3340 * Will warn at most once the user per limitation type
3341 *
3342 * @param string $limitationType Should be one of:
3343 * 'expensive-parserfunction' (corresponding messages:
3344 * 'expensive-parserfunction-warning',
3345 * 'expensive-parserfunction-category')
3346 * 'post-expand-template-argument' (corresponding messages:
3347 * 'post-expand-template-argument-warning',
3348 * 'post-expand-template-argument-category')
3349 * 'post-expand-template-inclusion' (corresponding messages:
3350 * 'post-expand-template-inclusion-warning',
3351 * 'post-expand-template-inclusion-category')
3352 * 'node-count-exceeded' (corresponding messages:
3353 * 'node-count-exceeded-warning',
3354 * 'node-count-exceeded-category')
3355 * 'expansion-depth-exceeded' (corresponding messages:
3356 * 'expansion-depth-exceeded-warning',
3357 * 'expansion-depth-exceeded-category')
3358 * @param string|int|null $current Current value
3359 * @param string|int|null $max Maximum allowed, when an explicit limit has been
3360 * exceeded, provide the values (optional)
3361 */
3363 # does no harm if $current and $max are present but are unnecessary for the message
3368 }
3370 /**
3371 * Return the text of a template, after recursively
3372 * replacing any variables or templates within the template.
3373 *
3374 * @param array $piece The parts of the template
3375 * $piece['title']: the title, i.e. the part before the |
3376 * $piece['parts']: the parameter array
3377 * $piece['lineStart']: whether the brace was at the start of a line
3378 * @param PPFrame $frame The current frame, contains template arguments
3379 * @throws Exception
3380 * @return string The text of the template
3381 */
3386 // Flags
3388 // $text has been filled
3390 // wiki markup in $text should be escaped
3392 // $text is HTML, armour it against wikitext transformation
3394 // Force interwiki transclusion to be done in raw mode not rendered
3396 // $text is a DOM node needing expansion in a child frame
3398 // $text is a DOM node needing expansion in the current frame
3401 # Title object, where $text came from
3404 # $part1 is the bit before the first |, and must contain only title characters.
3405 # Various prefixes will be stripped from it later.
3410 # Original title text preserved for various purposes
3413 # $args is a list of argument nodes, starting from index 0, not including $part1
3414 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3415 # below won't work b/c this $args isn't an object
3421 # SUBST
3427 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3428 # Decide whether to expand template or keep wikitext as-is.
3434 }
3440 }
3441 }
3446 }
3447 }
3449 # Variables
3456 }
3458 }
3459 }
3461 # MSG, MSGNW and RAW
3463 # Check for MSGNW:
3468 # Remove obsolete MSG:
3471 }
3473 # Check for RAW:
3477 }
3478 }
3481 # Parser functions
3491 }
3498 }
3500 # The interface for parser functions allows for extracting
3501 # flags into the local scope. Extract any forwarded flags
3502 # here.
3504 }
3506 }
3508 # Finish mangling title and then check for loops.
3509 # Set $title to a Title object and $titleText to the PDBK
3512 # Split the title into page and subpage
3518 }
3522 # Check for language variants if the template is not found
3525 }
3526 # Do recursion depth check
3534 }
3535 }
3536 }
3538 # Load from database
3541 # Too many unique items can kill profiling DBs/collectors
3544 }
3550 ) {
3551 // Pass the template arguments as URL parameters.
3552 // "uselang" will have no effect since the Language object
3553 // is forced to the one defined in ParserOptions.
3562 }
3563 }
3565 // Create a new context to execute the special page
3578 }
3588 }
3589 }
3591 # If the title is valid but undisplayable, make a link to it
3595 }
3597 # Interwiki transclusion
3603 # Preprocess it like a template
3606 }
3608 }
3610 # Do infinite loop check
3611 # This has to be done after redirect resolution to avoid infinite loops via redirects
3618 }
3620 }
3622 # If we haven't found text to substitute by now, we're done
3623 # Recover the source wikitext and return it
3628 }
3631 }
3633 # Expand DOM-style return values in a child frame
3635 # Clean up argument array
3641 # Expansion is eligible for the empty-frame cache
3644 # Uncached expansion
3646 }
3647 }
3651 }
3655 }
3657 # Replace raw HTML by a placeholder
3661 # Escape nowiki-style return values
3666 ) {
3667 # Bug 529: if the template begins with a table or block-level
3668 # element, it should be treated as beginning a new line.
3669 # This behavior is somewhat controversial.
3671 }
3674 # Error, oversize inclusion
3676 # Make a working, properly escaped link if possible (bug 23588)
3679 # This will probably not be a working link, but at least it may
3680 # provide some hint of where the problem is
3683 }
3687 }
3693 }
3697 }
3699 /**
3700 * Call a parser function and return an array with text and flags.
3701 *
3702 * The returned array will always contain a boolean 'found', indicating
3703 * whether the parser function was found or not. It may also contain the
3704 * following:
3705 * text: string|object, resulting wikitext or PP DOM object
3706 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3707 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3708 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3709 * nowiki: bool, wiki markup in $text should be escaped
3710 *
3711 * @since 1.21
3712 * @param PPFrame $frame The current frame, contains template arguments
3713 * @param string $function Function name
3714 * @param array $args Arguments to the function
3715 * @throws MWException
3716 * @return array
3717 */
3723 # Case sensitive functions
3727 # Case insensitive functions
3734 }
3735 }
3740 # Workaround for PHP bug 35229 and similar
3745 }
3749 # Convert arguments to PPNodes and collect for appending to $allArgs
3756 }
3757 }
3759 # Add a frame parameter, and pass the arguments as an array
3763 # Convert arguments to plain text and append to $allArgs
3771 }
3772 }
3773 }
3777 # The interface for function hooks allows them to return a wikitext
3778 # string or an array containing the string and any flags. This mungs
3779 # things around to match what this method should return.
3784 );
3788 }
3792 );
3793 }
3799 }
3802 }
3807 }
3812 }
3814 /**
3815 * Get the semi-parsed DOM representation of a template with a given title,
3816 * and its redirect destination title. Cached.
3817 *
3818 * @param Title $title
3819 *
3820 * @return array
3821 */
3830 }
3833 }
3835 # Cache miss, go to the database
3841 }
3849 }
3852 }
3854 /**
3855 * Fetch the current revision of a given title. Note that the revision
3856 * (and even the title) may not exist in the database, so everything
3857 * contributing to the output of the parser should use this method
3858 * where possible, rather than getting the revisions themselves. This
3859 * method also caches its results, so using it benefits performance.
3860 *
3861 * @since 1.24
3862 * @param Title $title
3863 * @return Revision
3864 */
3869 }
3872 // Defaults to Parser::statelessFetchRevision()
3874 );
3875 }
3877 }
3879 /**
3880 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3881 * without passing them on to it.
3882 *
3883 * @since 1.24
3884 * @param Title $title
3885 * @param Parser|bool $parser
3886 * @return Revision
3887 */
3890 }
3892 /**
3893 * Fetch the unparsed text of a template and register a reference to it.
3894 * @param Title $title
3895 * @return array ( string or false, Title )
3896 */
3898 // Defaults to Parser::statelessFetchTemplate()
3907 // If we transclude ourselves, the final result
3908 // will change based on the new version of the page
3910 }
3911 }
3912 }
3914 }
3916 /**
3917 * Fetch the unparsed text of a template and register a reference to it.
3918 * @param Title $title
3919 * @return string|bool
3920 */
3924 }
3926 /**
3927 * Static function to get a template
3928 * Can be overridden via ParserOptions::setTemplateCallback().
3929 *
3930 * @param Title $title
3931 * @param bool|Parser $parser
3932 *
3933 * @return array
3934 */
3940 # Loop to fetch the article, with up to 1 redirect
3942 # Give extensions a chance to select the revision instead
3953 );
3955 }
3956 # Get the revision
3963 }
3965 # If there is no current revision, there is no page
3969 }
3976 # We fetched a rev from a different title; register it too...
3981 }
3990 }
3997 }
4002 }
4005 }
4006 # Redirect?
4009 }
4014 }
4016 /**
4017 * Fetch a file and its title and register a reference to it.
4018 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4019 * @param Title $title
4020 * @param array $options Array of options to RepoGroup::findFile
4021 * @return File|bool
4022 */
4026 }
4028 /**
4029 * Fetch a file and its title and register a reference to it.
4030 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4031 * @param Title $title
4032 * @param array $options Array of options to RepoGroup::findFile
4033 * @return array ( File or false, Title of file )
4034 */
4040 # Register the file as a dependency...
4043 # Update fetched file title
4046 }
4048 }
4050 /**
4051 * Helper function for fetchFileAndTitle.
4052 *
4053 * Also useful if you need to fetch a file but not use it yet,
4054 * for example to get the file's handler.
4055 *
4056 * @param Title $title
4057 * @param array $options Array of options to RepoGroup::findFile
4058 * @return File|bool
4059 */
4067 }
4069 }
4071 /**
4072 * Transclude an interwiki link.
4073 *
4074 * @param Title $title
4075 * @param string $action
4076 *
4077 * @return string
4078 */
4084 }
4090 }
4092 }
4094 /**
4095 * @param string $url
4096 * @return mixed|string
4097 */
4106 }
4113 // Though we failed to fetch the content, this status is useless.
4118 }
4125 ) );
4127 }
4129 /**
4130 * Triple brace replacement -- used for template arguments
4131 * @private
4132 *
4133 * @param array $piece
4134 * @param PPFrame $frame
4135 *
4136 * @return array
4137 */
4151 )
4152 ) {
4153 # No match in frame, use the supplied default
4155 }
4159 }
4162 # No match anywhere
4164 }
4167 }
4172 }
4176 }
4178 /**
4179 * Return the text to be used for a given extension tag.
4180 * This is the ghost of strip().
4181 *
4182 * @param array $params Associative array of parameters:
4183 * name PPNode for the tag name
4184 * attr PPNode for unparsed text where tag attributes are thought to be
4185 * attributes Optional associative array of parsed attributes
4186 * inner Contents of extension element
4187 * noClose Original text did not have a close tag
4188 * @param PPFrame $frame
4189 *
4190 * @throws MWException
4191 * @return string
4192 */
4206 }
4212 }
4215 # Workaround for PHP bug 35229 and similar
4218 }
4225 }
4231 }
4234 # Extract flags to local scope (to override $markerType)
4239 }
4243 }
4248 }
4249 }
4255 }
4256 }
4266 }
4268 }
4270 /**
4271 * Increment an include size counter
4272 *
4273 * @param string $type The type of expansion
4274 * @param int $size The size of the text
4275 * @return bool False if this inclusion would take it over the maximum, true otherwise
4276 */
4283 }
4284 }
4286 /**
4287 * Increment the expensive function count
4288 *
4289 * @return bool False if the limit has been exceeded
4290 */
4294 }
4296 /**
4297 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
4298 * Fills $this->mDoubleUnderscores, returns the modified text
4299 *
4300 * @param string $text
4301 *
4302 * @return string
4303 */
4307 # The position of __TOC__ needs to be recorded
4313 # Set a placeholder. At the end we'll fill it in with the TOC.
4316 # Only keep the first one.
4318 }
4320 # Now match and remove the rest of them
4326 }
4329 }
4332 ) {
4334 }
4335 # (bug 8068) Allow control over whether robots index a page.
4336 #
4337 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
4338 # is not desirable, the last one on the page should win.
4342 }
4346 }
4348 # Cache all double underscores in the database
4351 }
4355 }
4357 /**
4358 * @see ParserOutput::addTrackingCategory()
4359 * @param string $msg Message key
4360 * @return bool Whether the addition was successful
4361 */
4364 }
4366 /**
4367 * This function accomplishes several tasks:
4368 * 1) Auto-number headings if that option is enabled
4369 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
4370 * 3) Add a Table of contents on the top for users who have enabled the option
4371 * 4) Auto-anchor headings
4372 *
4373 * It loops through all headlines, collects the necessary data, then splits up the
4374 * string and re-inserts the newly formatted headlines.
4375 *
4376 * @param string $text
4377 * @param string $origText Original, untouched wikitext
4378 * @param bool $isMain
4379 * @return mixed|string
4380 * @private
4381 */
4385 # Inhibit editsection links if requested in the page
4391 }
4394 }
4396 # Get all headlines for numbering them and adding funky stuff like [edit]
4397 # links - this is for later, but we need the number of headlines right now
4402 $matches
4403 );
4405 # if there are fewer than 4 headlines in the article, do not show TOC
4406 # unless it's been explicitly enabled.
4410 # Allow user to stipulate that a page should have a "new section"
4411 # link added via __NEWSECTIONLINK__
4414 }
4416 # Allow user to remove the "new section"
4417 # link via __NONEWSECTIONLINK__
4420 }
4422 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4423 # override above conditions and always show TOC above first header
4427 }
4429 # headline counter
4433 # Ugh .. the TOC should have neat indentation levels which can be
4434 # passed to the skin functions. These are determined here
4466 }
4470 }
4474 # Increase TOC level
4481 }
4483 # Decrease TOC level, find level to jump to
4487 # Found last matching level
4491 # Found first matching level below current level
4494 }
4495 }
4498 }
4501 # Unindent only if the previous toc level was shown :p
4506 }
4507 }
4509 # No change in level, end TOC line
4512 }
4513 }
4517 # count number of headlines for each level
4524 }
4527 }
4528 }
4530 # The safe header is a version of the header text safe to use for links
4532 # Remove link placeholders by the link text.
4533 # <!--LINK number-->
4534 # turns into
4535 # link text with suffix
4536 # Do this before unstrip since link text can contain strip markers
4539 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4542 # Strip out HTML (first regex removes any tag not allowed)
4543 # Allowed tags are:
4544 # * <sup> and <sub> (bug 8393)
4545 # * <i> (bug 26375)
4546 # * <b> (r105284)
4547 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4548 #
4549 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4550 # to allow setting directionality in toc items.
4555 ),
4557 $safeHeadline
4558 );
4561 # For the anchor, strip out HTML-y stuff period
4565 # Save headline for section edit hint before it's escaped
4569 # For reverse compatibility, provide an id that's
4570 # HTML4-compatible, like we used to.
4571 #
4572 # It may be worth noting, academically, that it's possible for
4573 # the legacy anchor to conflict with a non-legacy headline
4574 # anchor on the page. In this case likely the "correct" thing
4575 # would be to either drop the legacy anchors or make sure
4576 # they're numbered first. However, this would require people
4577 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4578 # manually, so let's not bother worrying about it.
4584 # No reason to have both (in fact, we can't)
4586 }
4591 }
4593 # HTML names must be case-insensitively unique (bug 10721).
4594 # This does not apply to Unicode characters per
4595 # http://dev.w3.org/html5/spec/infrastructure.html#case-sensitivity-and-string-comparison
4596 # @todo FIXME: We may be changing them depending on the current locale.
4602 }
4604 # count how many in assoc. array so we can track dupes in anchors
4609 }
4614 }
4616 # Don't number the heading if it is the only one (looks silly)
4618 # the two are different if the line contains a link
4622 $numbering
4624 }
4626 # Create the anchor for linking from the TOC to the section
4631 }
4634 }
4638 }
4640 # Add the section to the section tree
4641 # Find the DOM node for this header
4648 }
4649 }
4653 }
4663 );
4665 # give headline the correct <h#> tag
4667 // Output edit section links as markers with styles that can be customized by skins
4669 # Put a T flag in the section identifier, to indicate to extractSections()
4670 # that sections inside <includeonly> should be counted.
4678 }
4679 // We use a bit of pesudo-xml for editsection markers. The
4680 // language converter is run later on. Using a UNIQ style marker
4681 // leads to the converter screwing up the tokens when it
4682 // converts stuff. And trying to insert strip tags fails too. At
4683 // this point all real inputted tags have already been escaped,
4684 // so we don't have to worry about a user trying to input one of
4685 // these markers directly. We use a page and section attribute
4686 // to stop the language converter from converting these
4687 // important bits of data, but put the headline hint inside a
4688 // content block because the language converter is supposed to
4689 // be able to convert that piece of data.
4690 // Gets replaced with html in ParserOutput::getText
4697 }
4700 }
4706 }
4710 # Never ever show TOC if no headers
4713 }
4718 }
4723 }
4727 }
4729 # split up and insert constructed headlines
4733 // build an array of document sections
4736 // $head is zero-based, sections aren't.
4741 }
4743 /**
4744 * Send a hook, one per section.
4745 * The idea here is to be able to make section-level DIVs, but to do so in a
4746 * lower-impact, more correct way than r50769
4747 *
4748 * $this : caller
4749 * $section : the section number
4750 * &$sectionContent : ref to the content of the section
4751 * $showEditLinks : boolean describing whether this section has an edit link
4752 */
4756 }
4759 // append the TOC at the beginning
4760 // Top anchor now in skin
4762 }
4770 }
4771 }
4773 /**
4774 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4775 * conversion, substitting signatures, {{subst:}} templates, etc.
4776 *
4777 * @param string $text The text to transform
4778 * @param Title $title The Title object for the current article
4779 * @param User $user The User object describing the current user
4780 * @param ParserOptions $options Parsing options
4781 * @param bool $clearState Whether to clear the parser state first
4782 * @return string The altered wiki markup
4783 */
4786 ) {
4789 }
4795 );
4799 }
4805 }
4807 /**
4808 * Pre-save transform helper function
4809 *
4810 * @param string $text
4811 * @param User $user
4812 *
4813 * @return string
4814 */
4818 # Note: This is the timestamp saved as hardcoded wikitext to
4819 # the database, we use $wgContLang here in order to give
4820 # everyone the same signature and use the default one rather
4821 # than the one selected in each user's preferences.
4822 # (see also bug 12815)
4828 # Allow translation of timezones through wiki. format() can return
4829 # whatever crap the system uses, localised or not, so we cannot
4830 # ship premade translations.
4835 }
4839 # Variable replacement
4840 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4843 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4844 # which may corrupt this parser instance via its wfMessage()->text() call-
4846 # Signatures
4852 ) );
4854 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4858 // [[ns:page (context)|]]
4860 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4862 // [[ns:page (context), context|]] (using either single or double-width comma)
4864 // [[|page]] (reverse pipe trick: add context from page title)
4867 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4879 # if there's no context, don't bother duplicating the title
4881 }
4883 # Trim trailing whitespace
4887 }
4889 /**
4890 * Fetch the user's signature text, if any, and normalize to
4891 * validated, ready-to-insert wikitext.
4892 * If you have pre-fetched the nickname or the fancySig option, you can
4893 * specify them here to save a database query.
4894 * Do not reuse this parser instance after calling getUserSig(),
4895 * as it may have changed if it's the $wgParser.
4896 *
4897 * @param User $user
4898 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4899 * @param bool|null $fancySig whether the nicknname is the complete signature
4900 * or null to use default value
4901 * @return string
4902 */
4908 # If not given, retrieve from the user object.
4911 }
4915 }
4923 # Sig. might contain markup; validate this
4925 # Validated; clean up (if needed) and return it
4928 # Failed to validate; fall back to the default
4931 }
4932 }
4934 # Make sure nickname doesnt get a sig in a sig
4937 # If we're still here, make it a link to the user page
4944 }
4946 /**
4947 * Check that the user's signature contains no bad XML
4948 *
4949 * @param string $text
4950 * @return string|bool An expanded string, or false if invalid.
4951 */
4954 }
4956 /**
4957 * Clean up signature text
4958 *
4959 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
4960 * 2) Substitute all transclusions
4961 *
4962 * @param string $text
4963 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
4964 * @return string Signature text
4965 */
4971 }
4973 # Option to disable this feature
4976 }
4978 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4979 # => Move this logic to braceSubstitution()
4981 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4992 }
4995 }
4997 /**
4998 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
4999 *
5000 * @param string $text
5001 * @return string Signature text with /~{3,5}/ removed
5002 */
5006 }
5008 /**
5009 * Set up some variables which are usually set up in parse()
5010 * so that an external function can call some class members with confidence
5011 *
5012 * @param Title|null $title
5013 * @param ParserOptions $options
5014 * @param int $outputType
5015 * @param bool $clearState
5016 */
5019 ) {
5021 }
5023 /**
5024 * @param Title|null $title
5025 * @param ParserOptions $options
5026 * @param int $outputType
5027 * @param bool $clearState
5028 */
5031 ) {
5037 }
5038 }
5040 /**
5041 * Wrapper for preprocess()
5042 *
5043 * @param string $text The text to preprocess
5044 * @param ParserOptions $options Options
5045 * @param Title|null $title Title object or null to use $wgTitle
5046 * @return string
5047 */
5051 # Guard against infinite recursion
5054 }
5061 }
5068 }
5070 /**
5071 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
5072 * The callback should have the following form:
5073 * function myParserHook( $text, $params, $parser, $frame ) { ... }
5074 *
5075 * Transform and return $text. Use $parser for any required context, e.g. use
5076 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
5077 *
5078 * Hooks may return extended information by returning an array, of which the
5079 * first numbered element (index 0) must be the return string, and all other
5080 * entries are extracted into local variables within an internal function
5081 * in the Parser class.
5082 *
5083 * This interface (introduced r61913) appears to be undocumented, but
5084 * 'markerName' is used by some core tag hooks to override which strip
5085 * array their results are placed in. **Use great caution if attempting
5086 * this interface, as it is not documented and injudicious use could smash
5087 * private variables.**
5088 *
5089 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5090 * @param callable $callback The callback function (and object) to use for the tag
5091 * @throws MWException
5092 * @return callable|null The old value of the mTagHooks array associated with the hook
5093 */
5098 }
5103 }
5106 }
5108 /**
5109 * As setHook(), but letting the contents be parsed.
5110 *
5111 * Transparent tag hooks are like regular XML-style tag hooks, except they
5112 * operate late in the transformation sequence, on HTML instead of wikitext.
5113 *
5114 * This is probably obsoleted by things dealing with parser frames?
5115 * The only extension currently using it is geoserver.
5116 *
5117 * @since 1.10
5118 * @todo better document or deprecate this
5119 *
5120 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5121 * @param callable $callback The callback function (and object) to use for the tag
5122 * @throws MWException
5123 * @return callable|null The old value of the mTagHooks array associated with the hook
5124 */
5129 }
5130 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
5134 }
5136 /**
5137 * Remove all tag hooks
5138 */
5143 }
5145 /**
5146 * Create a function, e.g. {{sum:1|2|3}}
5147 * The callback function should have the form:
5148 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
5149 *
5150 * Or with SFH_OBJECT_ARGS:
5151 * function myParserFunction( $parser, $frame, $args ) { ... }
5152 *
5153 * The callback may either return the text result of the function, or an array with the text
5154 * in element 0, and a number of flags in the other elements. The names of the flags are
5155 * specified in the keys. Valid flags are:
5156 * found The text returned is valid, stop processing the template. This
5157 * is on by default.
5158 * nowiki Wiki markup in the return value should be escaped
5159 * isHTML The returned text is HTML, armour it against wikitext transformation
5160 *
5161 * @param string $id The magic word ID
5162 * @param callable $callback The callback function (and object) to use
5163 * @param int $flags A combination of the following flags:
5164 * SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
5165 *
5166 * SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text. This
5167 * allows for conditional expansion of the parse tree, allowing you to eliminate dead
5168 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
5169 * the arguments, and to control the way they are expanded.
5170 *
5171 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
5172 * arguments, for instance:
5173 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
5174 *
5175 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
5176 * future versions. Please call $frame->expand() on it anyway so that your code keeps
5177 * working if/when this is changed.
5178 *
5179 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
5180 * expansion.
5181 *
5182 * Please read the documentation in includes/parser/Preprocessor.php for more information
5183 * about the methods available in PPFrame and PPNode.
5184 *
5185 * @throws MWException
5186 * @return string|callable The old callback function for this name, if any
5187 */
5194 # Add to function cache
5198 }
5204 # Case
5207 }
5208 # Add leading hash
5211 }
5212 # Remove trailing colon
5215 }
5217 }
5219 }
5221 /**
5222 * Get all registered function hook identifiers
5223 *
5224 * @return array
5225 */
5228 }
5230 /**
5231 * Create a tag function, e.g. "<test>some stuff</test>".
5232 * Unlike tag hooks, tag functions are parsed at preprocessor level.
5233 * Unlike parser functions, their content is not preprocessed.
5234 * @param string $tag
5235 * @param callable $callback
5236 * @param int $flags
5237 * @throws MWException
5238 * @return null
5239 */
5244 }
5251 }
5254 }
5256 /**
5257 * @todo FIXME: Update documentation. makeLinkObj() is deprecated.
5258 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
5259 * Placeholders created in Skin::makeLinkObj()
5260 *
5261 * @param string $text
5262 * @param int $options
5263 *
5264 * @return array Array of link CSS classes, indexed by PDBK.
5265 */
5268 }
5270 /**
5271 * Replace "<!--LINK-->" link placeholders with plain text of links
5272 * (not HTML-formatted).
5273 *
5274 * @param string $text
5275 * @return string
5276 */
5279 }
5281 /**
5282 * Renders an image gallery from a text with one line per image.
5283 * text labels may be given by using |-style alternative text. E.g.
5284 * Image:one.jpg|The number "1"
5285 * Image:tree.jpg|A tree
5286 * given as text will return the HTML of a gallery with two images,
5287 * labeled 'The number "1"' and
5288 * 'A tree'.
5289 *
5290 * @param string $text
5291 * @param array $params
5292 * @return string HTML
5293 */
5300 }
5305 // If invalid type set, fallback to default.
5307 }
5320 }
5326 }
5329 }
5332 }
5335 }
5342 # match lines like these:
5343 # Image:someimage.jpg|This is some image
5346 # Skip empty lines
5349 }
5353 }
5356 # Bogus title. Ignore these so we don't bomb out later.
5358 }
5360 # We need to get what handler the file uses, to figure out parameters.
5361 # Note, a hook can overide the file name, and chose an entirely different
5362 # file (which potentially could be of a different type and have different handler).
5367 # Don't register it now, as ImageGallery does that later.
5375 );
5378 // We don't want people to specify per-image widths.
5379 // Additionally the width parameter would need special casing anyhow.
5381 }
5391 // look for an |alt= definition while trying not to break existing
5392 // captions with multiple pipes (|) in it, until a more sensible grammar
5393 // is defined for images in galleries
5395 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5396 // splitting on '|' is a bit odd, and different from makeImage.
5413 //check to see if link matches an absolute url, if not then it must be a wiki link.
5420 }
5421 }
5424 // Must be a handler specific parameter.
5428 // Guess not. Append it to the caption.
5431 }
5432 }
5435 // concatenate all other pipes
5437 }
5438 }
5439 // remove the first pipe
5441 }
5444 }
5449 }
5451 /**
5452 * @param string $handler
5453 * @return array
5454 */
5460 }
5462 # Initialise static lists
5469 );
5477 }
5478 }
5479 }
5481 # Add handler params
5487 }
5488 }
5491 }
5492 return array( $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] );
5493 }
5495 /**
5496 * Parse image options text and use it to make an image
5497 *
5498 * @param Title $title
5499 * @param string $options
5500 * @param LinkHolderArray|bool $holders
5501 * @return string HTML
5502 */
5504 # Check if the options text is of the form "options|alt text"
5505 # Options are:
5506 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5507 # * left no resizing, just left align. label is used for alt= only
5508 # * right same, but right aligned
5509 # * none same, but not aligned
5510 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5511 # * center center the image
5512 # * frame Keep original image size, no magnify-button.
5513 # * framed Same as "frame"
5514 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5515 # * upright reduce width for upright images, rounded to full __0 px
5516 # * border draw a 1px border around the image
5517 # * alt Text for HTML alt attribute (defaults to empty)
5518 # * class Set a class for img node
5519 # * link Set the target of the image link. Can be external, interwiki, or local
5520 # vertical-align values (no % or length right now):
5521 # * baseline
5522 # * sub
5523 # * super
5524 # * top
5525 # * text-top
5526 # * middle
5527 # * bottom
5528 # * text-bottom
5532 # Give extensions a chance to select the file revision for us
5537 # Fetch and register the file (file title may be different via hooks)
5540 # Get parameter map
5547 }
5549 # Process the input parameters
5561 # Special case; width and height come in one variable together
5569 }
5570 }
5576 }
5577 }
5578 # else no validation -- bug 13436
5581 # Validate handler parameter
5584 # Validate internal parameters
5589 # @todo FIXME: Possibly check validity here for
5590 # manualthumb? downstream behavior seems odd with
5591 # missing manual thumbs.
5608 }
5610 }
5618 }
5619 }
5624 // use first appearing option, discard others.
5629 # Most other things appear to be empty or numeric...
5631 }
5632 }
5636 }
5637 }
5638 }
5641 }
5642 }
5644 # Process alignment parameters
5647 }
5650 }
5654 # Will the image be presented in a frame, with the caption below?
5660 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5661 # came to also set the caption, ordinary text after the image -- which
5662 # makes no sense, because that just repeats the text multiple times in
5663 # screen readers. It *also* came to set the title attribute.
5664 #
5665 # Now that we have an alt attribute, we should not set the alt text to
5666 # equal the caption: that's worse than useless, it just repeats the
5667 # text. This is the framed/thumbnail case. If there's no caption, we
5668 # use the unnamed parameter for alt text as well, just for the time be-
5669 # ing, if the unnamed param is set and the alt param is not.
5670 #
5671 # For the future, we need to figure out if we want to tweak this more,
5672 # e.g., introducing a title= parameter for the title; ignoring the un-
5673 # named parameter entirely for images without a caption; adding an ex-
5674 # plicit caption= parameter and preserving the old magic unnamed para-
5675 # meter for BC; ...
5678 # No caption or alt text, add the filename as the alt text so
5679 # that screen readers at least get some description of the image
5681 }
5682 # Do not set $params['frame']['title'] because tooltips don't make sense
5683 # for framed images
5686 # No alt text, use the "caption" for the alt text
5690 # No caption, fall back to using the filename for the
5691 # alt text
5693 }
5694 }
5695 # Use the "caption" for the tooltip text
5697 }
5701 # Linker does the rest
5706 # Give the handler a chance to modify the parser object
5709 }
5712 }
5714 /**
5715 * @param string $caption
5716 * @param LinkHolderArray|bool $holders
5717 * @return mixed|string
5718 */
5720 # Strip bad stuff out of the title (tooltip). We can't just use
5721 # replaceLinkHoldersText() here, because if this function is called
5722 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5727 }
5729 # make sure there are no placeholders in thumbnail attributes
5730 # that are later expanded to html- so expand them now and
5731 # remove the tags
5736 }
5738 /**
5739 * Set a flag in the output object indicating that the content is dynamic and
5740 * shouldn't be cached.
5741 */
5747 }
5750 }
5752 /**
5753 * Callback from the Sanitizer for expanding items found in HTML attribute
5754 * values, so they can be safely tested and escaped.
5755 *
5756 * @param string $text
5757 * @param bool|PPFrame $frame
5758 * @return string
5759 */
5764 }
5766 /**
5767 * Accessor
5768 *
5769 * @return array
5770 */
5776 );
5777 }
5779 /**
5780 * Replace transparent tags in $text with the values given by the callbacks.
5781 *
5782 * Transparent tag hooks are like regular XML-style tag hooks, except they
5783 * operate late in the transformation sequence, on HTML instead of wikitext.
5784 *
5785 * @param string $text
5786 *
5787 * @return string
5788 */
5802 );
5805 }
5807 }
5809 }
5811 /**
5812 * Break wikitext input into sections, and either pull or replace
5813 * some particular section's text.
5814 *
5815 * External callers should use the getSection and replaceSection methods.
5816 *
5817 * @param string $text Page wikitext
5818 * @param string|number $sectionId A section identifier string of the form:
5819 * "<flag1> - <flag2> - ... - <section number>"
5820 *
5821 * Currently the only recognised flag is "T", which means the target section number
5822 * was derived during a template inclusion parse, in other words this is a template
5823 * section edit link. If no flags are given, it was an ordinary section edit link.
5824 * This flag is required to avoid a section numbering mismatch when a section is
5825 * enclosed by "<includeonly>" (bug 6563).
5826 *
5827 * The section number 0 pulls the text before the first heading; other numbers will
5828 * pull the given section along with its lower-level subsections. If the section is
5829 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5830 *
5831 * Section 0 is always considered to exist, even if it only contains the empty
5832 * string. If $text is the empty string and section 0 is replaced, $newText is
5833 * returned.
5834 *
5835 * @param string $mode One of "get" or "replace"
5836 * @param string $newText Replacement text for section data.
5837 * @return string For "get", the extracted section text.
5838 * for "replace", the whole page with the section replaced.
5839 */
5848 # Process section extraction flags
5855 }
5856 }
5858 # Check for empty input
5860 # Only sections 0 and T-0 exist in an empty document
5866 }
5872 }
5873 }
5874 }
5876 # Preprocess the text
5879 # <h> nodes indicate section breaks
5880 # They can only occur at the top level, so we can find them by iterating the root's children
5883 # Find the target section
5885 # Section zero doesn't nest, level=big
5894 }
5895 }
5898 }
5900 }
5901 }
5904 # Not found
5909 }
5910 }
5912 # Find the end of the section, including nested sections
5919 }
5920 }
5923 }
5927 # Write out the remainder (in replace mode only)
5929 # Output the replacement text
5930 # Add two newlines on -- trailing whitespace in $newText is conventionally
5931 # stripped by the editor, so we need both newlines to restore the paragraph gap
5932 # Only add trailing whitespace if there is newText
5935 }
5940 }
5941 }
5944 # Re-insert stripped tags
5946 }
5949 }
5951 /**
5952 * This function returns the text of a section, specified by a number ($section).
5953 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5954 * the first section before any such heading (section 0).
5955 *
5956 * If a section contains subsections, these are also returned.
5957 *
5958 * @param string $text Text to look in
5959 * @param string|number $sectionId Section identifier as a number or string
5960 * (e.g. 0, 1 or 'T-1').
5961 * @param string $defaultText Default to return if section is not found
5962 *
5963 * @return string Text of the requested section
5964 */
5967 }
5969 /**
5970 * This function returns $oldtext after the content of the section
5971 * specified by $section has been replaced with $text. If the target
5972 * section does not exist, $oldtext is returned unchanged.
5973 *
5974 * @param string $oldText Former text of the article
5975 * @param string|number $sectionId Section identifier as a number or string
5976 * (e.g. 0, 1 or 'T-1').
5977 * @param string $newText Replacing text
5978 *
5979 * @return string Modified text
5980 */
5983 }
5985 /**
5986 * Get the ID of the revision we are parsing
5987 *
5988 * @return int|null
5989 */
5992 }
5994 /**
5995 * Get the revision object for $this->mRevisionId
5996 *
5997 * @return Revision|null Either a Revision object or null
5998 * @since 1.23 (public since 1.23)
5999 */
6003 }
6006 }
6010 }
6012 /**
6013 * Get the timestamp associated with the current revision, adjusted for
6014 * the default server-local timestamp
6015 * @return string
6016 */
6026 # The cryptic '' timezone parameter tells to use the site-default
6027 # timezone offset instead of the user settings.
6028 #
6029 # Since this value will be saved into the parser cache, served
6030 # to other users, and potentially even used inside links and such,
6031 # it needs to be consistent for all visitors.
6035 }
6037 }
6039 /**
6040 * Get the name of the user that edited the last revision
6041 *
6042 * @return string User name
6043 */
6048 # if this template is subst: the revision id will be blank,
6049 # so just use the current user's name
6054 }
6055 }
6057 }
6059 /**
6060 * Get the size of the revision
6061 *
6062 * @return int|null Revision size
6063 */
6068 # if this variable is subst: the revision id will be blank,
6069 # so just use the parser input size, because the own substituation
6070 # will change the size.
6075 }
6076 }
6078 }
6080 /**
6081 * Mutator for $mDefaultSort
6082 *
6083 * @param string $sort New value
6084 */
6088 }
6090 /**
6091 * Accessor for $mDefaultSort
6092 * Will use the empty string if none is set.
6093 *
6094 * This value is treated as a prefix, so the
6095 * empty string is equivalent to sorting by
6096 * page name.
6097 *
6098 * @return string
6099 */
6105 }
6106 }
6108 /**
6109 * Accessor for $mDefaultSort
6110 * Unlike getDefaultSort(), will return false if none is set
6111 *
6112 * @return string|bool
6113 */
6116 }
6118 /**
6119 * Try to guess the section anchor name based on a wikitext fragment
6120 * presumably extracted from a heading, for example "Header" from
6121 * "== Header ==".
6122 *
6123 * @param string $text
6124 *
6125 * @return string
6126 */
6128 # Strip out wikitext links(they break the anchor)
6132 }
6134 /**
6135 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
6136 * instead. For use in redirects, since IE6 interprets Redirect: headers
6137 * as something other than UTF-8 (apparently?), resulting in breakage.
6138 *
6139 * @param string $text The section name
6140 * @return string An anchor
6141 */
6143 # Strip out wikitext links(they break the anchor)
6147 }
6149 /**
6150 * Strips a text string of wikitext for use in a section anchor
6151 *
6152 * Accepts a text string and then removes all wikitext from the
6153 * string and leaves only the resultant text (i.e. the result of
6154 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
6155 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
6156 * to create valid section anchors by mimicing the output of the
6157 * parser when headings are parsed.
6158 *
6159 * @param string $text Text string to be stripped of wikitext
6160 * for use in a Section anchor
6161 * @return string Filtered text string
6162 */
6164 # Strip internal link markup
6168 # Strip external link markup
6169 # @todo FIXME: Not tolerant to blank link text
6170 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
6171 # on how many empty links there are on the page - need to figure that out.
6172 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
6174 # Parse wikitext quotes (italics & bold)
6177 # Strip HTML tags
6180 }
6182 /**
6183 * strip/replaceVariables/unstrip for preprocessor regression testing
6184 *
6185 * @param string $text
6186 * @param Title $title
6187 * @param ParserOptions $options
6188 * @param int $outputType
6189 *
6190 * @return string
6191 */
6192 public function testSrvus( $text, Title $title, ParserOptions $options, $outputType = self::OT_HTML ) {
6200 }
6202 /**
6203 * @param string $text
6204 * @param Title $title
6205 * @param ParserOptions $options
6206 * @return string
6207 */
6210 }
6212 /**
6213 * @param string $text
6214 * @param Title $title
6215 * @param ParserOptions $options
6216 * @return string
6217 */
6220 }
6222 /**
6223 * Call a callback function on all regions of the given text that are not
6224 * inside strip markers, and replace those regions with the return value
6225 * of the callback. For example, with input:
6226 *
6227 * aaa<MARKER>bbb
6228 *
6229 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
6230 * two strings will be replaced with the value returned by the callback in
6231 * each case.
6232 *
6233 * @param string $s
6234 * @param callable $callback
6235 *
6236 * @return string
6237 */
6256 }
6257 }
6258 }
6260 }
6262 /**
6263 * Remove any strip markers found in the given text.
6264 *
6265 * @param string $text Input string
6266 * @return string
6267 */
6270 }
6272 /**
6273 * Save the parser state required to convert the given half-parsed text to
6274 * HTML. "Half-parsed" in this context means the output of
6275 * recursiveTagParse() or internalParse(). This output has strip markers
6276 * from replaceVariables (extensionSubstitution() etc.), and link
6277 * placeholders from replaceLinkHolders().
6278 *
6279 * Returns an array which can be serialized and stored persistently. This
6280 * array can later be loaded into another parser instance with
6281 * unserializeHalfParsedText(). The text can then be safely incorporated into
6282 * the return value of a parser hook.
6283 *
6284 * @param string $text
6285 *
6286 * @return array
6287 */
6295 );
6298 }
6300 /**
6301 * Load the parser state given in the $data array, which is assumed to
6302 * have been generated by serializeHalfParsedText(). The text contents is
6303 * extracted from the array, and its markers are transformed into markers
6304 * appropriate for the current Parser instance. This transformed text is
6305 * returned, and can be safely included in the return value of a parser
6306 * hook.
6307 *
6308 * If the $data array has been stored persistently, the caller should first
6309 * check whether it is still valid, by calling isValidHalfParsedText().
6310 *
6311 * @param array $data Serialized data
6312 * @throws MWException
6313 * @return string
6314 */
6318 }
6320 # First, extract the strip state.
6324 # Now renumber links
6327 # Should be good to go.
6329 }
6331 /**
6332 * Returns true if the given array, presumed to be generated by
6333 * serializeHalfParsedText(), is compatible with the current version of the
6334 * parser.
6335 *
6336 * @param array $data
6337 *
6338 * @return bool
6339 */
6342 }
6344 /**
6345 * Parsed a width param of imagelink like 300px or 200x300px
6346 *
6347 * @param string $value
6348 *
6349 * @return array
6350 * @since 1.20
6351 */
6356 }
6358 # (bug 13500) In both cases (width/height and width only),
6359 # permit trailing "px" for backward compatibility.
6368 }
6370 }
6372 /**
6373 * Lock the current instance of the parser.
6374 *
6375 * This is meant to stop someone from calling the parser
6376 * recursively and messing up all the strip state.
6377 *
6378 * @throws MWException If parser is in a parse
6379 * @return ScopedCallback The lock will be released once the return value goes out of scope.
6380 */
6385 }
6391 } );
6394 }
6396 /**
6397 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6398 *
6399 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6400 * or if there is more than one <p/> tag in the input HTML.
6401 *
6402 * @param string $html
6403 * @return string
6404 * @since 1.24
6405 */
6411 }
6412 }
6415 }
6417 /**
6418 * Return this parser if it is not doing anything, otherwise
6419 * get a fresh parser. You can use this method by doing
6420 * $myParser = $wgParser->getFreshParser(), or more simply
6421 * $wgParser->getFreshParser()->parse( ... );
6422 * if you're unsure if $wgParser is safe to use.
6423 *
6424 * @since 1.24
6425 * @return Parser A parser object that is not parsing anything
6426 */
6433 }
6434 }
6435 }