| blob | 3027ab6efeb8051142e58ecb8a309f17f10488dc |
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 }
308 }
310 /**
311 * Clear Parser state
312 *
313 * @private
314 */
318 }
336 /**
337 * Prefix for temporary replacement strings for the multipass parser.
338 * \x07 should never appear in input as it's disallowed in XML.
339 * Using it at the front also gives us a little extra robustness
340 * since it shouldn't match when butted up against identifier-like
341 * string constructs.
342 *
343 * Must not consist of all title characters, or else it will change
344 * the behavior of <nowiki> in a link.
345 */
349 # Clear these on every parse, bug 4549
357 );
366 # Fix cloning
369 }
374 }
376 /**
377 * Convert wikitext to HTML
378 * Do not call this function recursively.
379 *
380 * @param string $text Text we want to parse
381 * @param Title $title
382 * @param ParserOptions $options
383 * @param bool $linestart
384 * @param bool $clearState
385 * @param int $revid Number to pass in {{REVISIONID}}
386 * @return ParserOutput A ParserOutput
387 */
390 ) {
391 /**
392 * First pass--just handle <nowiki> sections, pass the rest off
393 * to internalParse() which does all the real work.
394 */
402 }
410 }
412 # Remove the strip marker tag prefix from the input, if present.
415 }
428 }
431 # No more strip!
438 /**
439 * A converted title will be provided in the output object if title and
440 * content conversion are enabled, the article text does not contain
441 * a conversion-suppressing double-underscore tag, and no
442 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
443 * automatic link conversion.
444 */
449 ) {
456 }
457 }
463 );
464 }
466 # Information on include size limits, for the benefit of users who try to skirt them
474 );
475 }
480 );
484 );
487 );
490 );
493 );
496 );
499 );
505 }
509 ) ) {
515 }
519 }
520 }
521 }
522 // Since we're not really outputting HTML, decode the entities and
523 // then re-encode the things that need hiding inside HTML comments.
527 // Sanitize for comment. Note '‐' in the replacement is U+2010,
528 // which looks much like the problematic '-'.
532 // Add on template profiling data
536 } );
542 }
548 }
549 }
562 }
564 /**
565 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
566 * can be called from an extension tag hook.
567 *
568 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
569 * instead, which means that lists and links have not been fully parsed yet,
570 * and strip markers are still present.
571 *
572 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
573 *
574 * Use this function if you're a parser tag hook and you want to parse
575 * wikitext before or after applying additional transformations, and you
576 * intend to *return the result as hook output*, which will cause it to go
577 * through the rest of parsing process automatically.
578 *
579 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
580 * $text are not expanded
581 *
582 * @param string $text Text extension wants to have parsed
583 * @param bool|PPFrame $frame The frame to use for expanding any template variables
584 * @return string UNSAFE half-parsed HTML
585 */
591 }
593 /**
594 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
595 * point can be called from an extension tag hook.
596 *
597 * The output of this function is fully-parsed HTML that is safe for output.
598 * If you're a parser tag hook, you might want to use recursiveTagParse()
599 * instead.
600 *
601 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
602 * $text are not expanded
603 *
604 * @since 1.25
605 *
606 * @param string $text Text extension wants to have parsed
607 * @param bool|PPFrame $frame The frame to use for expanding any template variables
608 * @return string Fully parsed HTML
609 */
614 }
616 /**
617 * Expand templates and variables in the text, producing valid, static wikitext.
618 * Also removes comments.
619 * Do not call this function recursively.
620 * @param string $text
621 * @param Title $title
622 * @param ParserOptions $options
623 * @param int|null $revid
624 * @param bool|PPFrame $frame
625 * @return mixed|string
626 */
629 ) {
634 }
640 }
642 /**
643 * Recursive parser entry point that can be called from an extension tag
644 * hook.
645 *
646 * @param string $text Text to be expanded
647 * @param bool|PPFrame $frame The frame to use for expanding any template variables
648 * @return string
649 * @since 1.19
650 */
655 }
657 /**
658 * Process the wikitext for the "?preload=" feature. (bug 5210)
659 *
660 * "<noinclude>", "<includeonly>" etc. are parsed as for template
661 * transclusion, comments, templates, arguments, tags hooks and parser
662 * functions are untouched.
663 *
664 * @param string $text
665 * @param Title $title
666 * @param ParserOptions $options
667 * @param array $params
668 * @return string
669 */
670 public function getPreloadText( $text, Title $title, ParserOptions $options, $params = array() ) {
674 # Parser (re)initialisation
683 }
685 /**
686 * Get a random string
687 *
688 * @return string
689 */
692 }
694 /**
695 * Set the current user.
696 * Should only be used when doing pre-save transform.
697 *
698 * @param User|null $user User object or null (to reset)
699 */
702 }
704 /**
705 * Accessor for mUniqPrefix.
706 *
707 * @return string
708 */
711 # @todo FIXME: This is probably *horribly wrong*
712 # LanguageConverter seems to want $wgParser's uniqPrefix, however
713 # if this is called for a parser cache hit, the parser may not
714 # have ever been initialized in the first place.
715 # Not really sure what the heck is supposed to be going on here.
717 # throw new MWException( "Accessing uninitialized mUniqPrefix" );
718 }
720 }
722 /**
723 * Set the context title
724 *
725 * @param Title $t
726 */
730 }
733 # Strip the fragment to avoid various odd effects
738 }
739 }
741 /**
742 * Accessor for the Title object
743 *
744 * @return Title
745 */
748 }
750 /**
751 * Accessor/mutator for the Title object
752 *
753 * @param Title $x Title object or null to just get the current one
754 * @return Title
755 */
758 }
760 /**
761 * Set the output type
762 *
763 * @param int $ot New value
764 */
767 # Shortcut alias
773 );
774 }
776 /**
777 * Accessor/mutator for the output type
778 *
779 * @param int|null $x New value or null to just get the current one
780 * @return int
781 */
784 }
786 /**
787 * Get the ParserOutput object
788 *
789 * @return ParserOutput
790 */
793 }
795 /**
796 * Get the ParserOptions object
797 *
798 * @return ParserOptions
799 */
802 }
804 /**
805 * Accessor/mutator for the ParserOptions object
806 *
807 * @param ParserOptions $x New value or null to just get the current one
808 * @return ParserOptions Current ParserOptions object
809 */
812 }
814 /**
815 * @return int
816 */
819 }
821 /**
822 * @param int $id
823 */
826 }
828 /**
829 * Get a language object for use in parser functions such as {{FORMATNUM:}}
830 * @return Language
831 */
834 }
836 /**
837 * Get the target language for the content being parsed. This is usually the
838 * language that the content is in.
839 *
840 * @since 1.19
841 *
842 * @throws MWException
843 * @return Language
844 */
854 }
857 }
859 /**
860 * Get the language object for language conversion
861 * @return Language|null
862 */
865 }
867 /**
868 * Get a User object either from $this->mUser, if set, or from the
869 * ParserOptions object otherwise
870 *
871 * @return User
872 */
876 }
878 }
880 /**
881 * Get a preprocessor object
882 *
883 * @return Preprocessor
884 */
889 }
891 }
893 /**
894 * Replaces all occurrences of HTML-style comments and the given tags
895 * in the text with a random marker and returns the next text. The output
896 * parameter $matches will be an associative array filled with data in
897 * the form:
898 *
899 * @code
900 * 'UNIQ-xxxxx' => array(
901 * 'element',
902 * 'tag content',
903 * array( 'param' => 'x' ),
904 * '<element param="x">tag content</element>' ) )
905 * @endcode
906 *
907 * @param array $elements List of element names. Comments are always extracted.
908 * @param string $text Source text string.
909 * @param array $matches Out parameter, Array: extracted tags
910 * @param string $uniq_prefix
911 * @return string Stripped text
912 */
926 }
928 # comment
934 # tag
939 }
945 # Empty element tag, <tag />
954 }
958 # No end tag -- let it run out to the end of the text.
964 }
965 }
971 }
973 }
975 /**
976 * Get a list of strippable XML-like elements
977 *
978 * @return array
979 */
982 }
984 /**
985 * Add an item to the strip state
986 * Returns the unique tag which must be inserted into the stripped text
987 * The tag will be replaced with the original text in unstrip()
988 *
989 * @param string $text
990 *
991 * @return string
992 */
998 }
1000 /**
1001 * parse the wiki syntax used to render tables
1002 *
1003 * @private
1004 * @param string $text
1005 * @return string
1006 */
1024 }
1030 # First check if we are starting a new table
1043 # Don't do any of the following
1047 # We are ending a table
1053 }
1057 }
1061 }
1065 # Now we have a table row
1068 # Whats after the tag is now only attributes
1081 }
1085 }
1094 ) {
1095 # This might be cell elements, td, th or captions
1099 }
1105 }
1107 # Split up multiple cells on the same line.
1108 # FIXME : This can result in improper nesting of tags processed
1109 # by earlier parser steps, but should avoid splitting up eg
1110 # attribute values containing literal "||".
1115 # Loop through each table cell
1122 }
1127 }
1133 }
1143 }
1147 # A cell could contain both parameters and data
1150 # Bug 553: Note that a '|' inside an invalid link should not
1151 # be mistaken as delimiting cell parameters
1160 }
1164 }
1165 }
1167 }
1169 # Closing open td, tr && table
1173 }
1176 }
1179 }
1182 }
1184 # Remove trailing line-ending (b/c)
1187 }
1189 # special case: don't return empty table
1192 }
1196 }
1198 /**
1199 * Helper function for parse() that transforms wiki markup into half-parsed
1200 * HTML. Only called for $mOutputType == self::OT_HTML.
1201 *
1202 * @private
1203 *
1204 * @param string $text
1205 * @param bool $isMain
1206 * @param bool $frame
1207 *
1208 * @return string
1209 */
1214 # Hook to suspend the parser in this state
1215 if ( !Hooks::run( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState ) ) ) {
1217 }
1219 # if $frame is provided, then use $frame for replacing any variables
1221 # use frame depth to infer how include/noinclude tags should be handled
1222 # depth=0 means this is the top-level document; otherwise it's an included document
1227 }
1231 # if $frame is not provided, then use old-style replaceVariables
1233 }
1241 );
1244 # Tables need to come after variable replacement for things to work
1245 # properly; putting them before other transformations should keep
1246 # exciting things like link expansions from showing up in surprising
1247 # places.
1259 # replaceInternalLinks may sometimes leave behind
1260 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1267 }
1269 /**
1270 * Helper function for parse() that transforms half-parsed HTML into fully
1271 * parsed HTML.
1272 *
1273 * @param string $text
1274 * @param bool $isMain
1275 * @param bool $linestart
1276 * @return string
1277 */
1283 # Clean up special characters, only run once, next-to-last before doBlockLevels
1285 # french spaces, last one Guillemet-left
1286 # only if there is something before the space
1288 # french spaces, Guillemet-right
1291 );
1298 /**
1299 * The input doesn't get language converted if
1300 * a) It's disabled
1301 * b) Content isn't converted
1302 * c) It's a conversion table
1303 * d) it is an interface message (which is in the user language)
1304 */
1307 ) {
1309 # The position of the convert() call should not be changed. it
1310 # assumes that the links are all replaced and the only thing left
1311 # is the <nowiki> mark.
1313 }
1314 }
1320 }
1330 # attempt to sanitize at least some nesting problems
1331 # (bug #2702 and quite a few others)
1333 # ''Something [http://www.cool.com cool''] -->
1334 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1337 # fix up an anchor inside another anchor, only
1338 # at least for a single single nested link (bug 3695)
1341 # fix div inside inline elements- doBlockLevels won't wrap a line which
1342 # contains a div, so fix it up here; replace
1343 # div with escaped text
1346 # remove empty italic or bold tag pairs, some
1347 # introduced by rules above
1349 );
1355 }
1359 }
1362 }
1364 /**
1365 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1366 * magic external links.
1367 *
1368 * DML
1369 * @private
1370 *
1371 * @param string $text
1372 *
1373 * @return string
1374 */
1382 '!(?: # Start cases
1391 [0-9Xx] # check digit
1395 }
1397 /**
1398 * @throws MWException
1399 * @param array $m
1400 * @return HTML|string
1401 */
1404 # Skip anchor
1407 # Skip HTML element
1410 # Free external link
1413 # RFC or PMID
1427 }
1431 # ISBN
1439 ));
1446 }
1447 }
1449 /**
1450 * Make a free external link, given a user-supplied URL
1451 *
1452 * @param string $url
1453 *
1454 * @return string HTML
1455 * @private
1456 */
1461 # The characters '<' and '>' (which were escaped by
1462 # removeHTMLtags()) should not be included in
1463 # URLs, per RFC 2396.
1468 }
1470 # Move trailing punctuation to $trail
1472 # If there is no left bracket, then consider right brackets fair game too
1475 }
1479 # Don't break a trailing HTML entity by moving the ; into $trail
1480 # This is in hot code, so use substr_compare to avoid having to
1481 # create a new string object for the comparison
1483 # more optimization: instead of running preg_match with a $
1484 # anchor, which can be slow, do the match on the reversed
1485 # string starting at the desired offset.
1486 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1489 }
1490 }
1494 }
1498 # Is this an external image?
1501 # Not an image, make a link
1506 # Register it in the output object...
1507 # Replace unnecessary URL escape codes with their equivalent characters
1510 }
1512 }
1514 /**
1515 * Parse headers and return html
1516 *
1517 * @private
1518 *
1519 * @param string $text
1520 *
1521 * @return string
1522 */
1527 }
1529 }
1531 /**
1532 * Replace single quotes with HTML markup
1533 * @private
1534 *
1535 * @param string $text
1536 *
1537 * @return string The altered text
1538 */
1544 }
1547 }
1549 /**
1550 * Helper function for doAllQuotes()
1551 *
1552 * @param string $text
1553 *
1554 * @return string
1555 */
1561 }
1563 // First, do some preliminary work. This may shift some apostrophes from
1564 // being mark-up to being text. It also counts the number of occurrences
1565 // of bold and italics mark-ups.
1570 // If there are ever four apostrophes, assume the first is supposed to
1571 // be text, and the remaining three constitute mark-up for bold text.
1572 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1578 // If there are more than 5 apostrophes in a row, assume they're all
1579 // text except for the last 5.
1580 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1584 }
1585 // Count the number of occurrences of bold and italics mark-ups.
1593 }
1594 }
1596 // If there is an odd number of both bold and italics, it is likely
1597 // that one of the bold ones was meant to be an apostrophe followed
1598 // by italics. Which one we cannot know for certain, but it is more
1599 // likely to be one that has a single-letter word before it.
1611 }
1615 // if $firstsingleletterword is set, we don't
1616 // look at the other options, so we can bail early.
1618 }
1622 }
1623 }
1624 }
1625 }
1627 // If there is a single-letter word, use it!
1632 // If not, but there's a multi-letter word, use that one.
1636 // ... otherwise use the first one that has neither.
1637 // (notice that it is possible for all three to be -1 if, for example,
1638 // there is only one pentuple-apostrophe in the line)
1641 }
1642 }
1644 // Now let's actually convert our apostrophic mush to HTML!
1655 }
1674 }
1691 }
1711 }
1712 }
1713 }
1715 }
1716 // Now close all remaining tags. Notice that the order is important.
1719 }
1722 }
1725 }
1726 // There might be lonely ''''', so make sure we have a buffer
1729 }
1731 }
1733 /**
1734 * Replace external links (REL)
1735 *
1736 * Note: this is all very hackish and the order of execution matters a lot.
1737 * Make sure to run tests/parserTests.php if you change this code.
1738 *
1739 * @private
1740 *
1741 * @param string $text
1742 *
1743 * @throws MWException
1744 * @return string
1745 */
1752 }
1762 # The characters '<' and '>' (which were escaped by
1763 # removeHTMLtags()) should not be included in
1764 # URLs, per RFC 2396.
1769 }
1771 # If the link text is an image URL, replace it with an <img> tag
1772 # This happened by accident in the original parser, but some people used it extensively
1776 }
1780 # Set linktype for CSS - if URL==text, link is essentially free
1783 # No link text, e.g. [http://domain.tld/some.link]
1785 # Autonumber
1790 # Have link text, e.g. [http://domain.tld/some.link text]s
1791 # Check for trail
1793 }
1799 # Use the encoded URL
1800 # This means that users can paste URLs directly into the text
1801 # Funny characters like ö aren't valid in URLs anyway
1802 # This was changed in August 2004
1806 # Register link in the output object.
1807 # Replace unnecessary URL escape codes with the referenced character
1808 # This prevents spammers from hiding links from the filters
1811 }
1814 }
1816 /**
1817 * Get the rel attribute for a particular external link.
1818 *
1819 * @since 1.21
1820 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1821 * nofollow if appropriate
1822 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1823 * @return string|null Rel attribute for $url
1824 */
1830 ) {
1832 }
1834 }
1836 /**
1837 * Get an associative array of additional HTML attributes appropriate for a
1838 * particular external link. This currently may include rel => nofollow
1839 * (depending on configuration, namespace, and the URL's domain) and/or a
1840 * target attribute (depending on configuration).
1841 *
1842 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1843 * nofollow if appropriate
1844 * @return array Associative array of HTML attributes
1845 */
1852 }
1854 }
1856 /**
1857 * Replace unusual escape codes in a URL with their equivalent characters
1858 *
1859 * @deprecated since 1.24, use normalizeLinkUrl
1860 * @param string $url
1861 * @return string
1862 */
1866 }
1868 /**
1869 * Replace unusual escape codes in a URL with their equivalent characters
1870 *
1871 * This generally follows the syntax defined in RFC 3986, with special
1872 * consideration for HTTP query strings.
1873 *
1874 * @param string $url
1875 * @return string
1876 */
1878 # First, make sure unsafe characters are encoded
1882 },
1883 $url
1884 );
1889 # Fragment part - 'fragment'
1895 }
1897 # Query part - 'query' minus &=+;
1903 }
1905 # Scheme and path part - 'pchar'
1906 # (we assume no userinfo or encoded colons in the host)
1911 }
1918 # Unescape it
1921 # Leave it escaped, but use uppercase for a-f
1923 }
1924 };
1926 }
1928 /**
1929 * make an image if it's allowed, either through the global
1930 * option, through the exception, or through the on-wiki whitelist
1931 *
1932 * @param string $url
1933 *
1934 * @return string
1935 */
1940 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1947 }
1948 }
1953 }
1957 ) {
1959 # Image found
1961 }
1962 }
1965 ) {
1969 );
1972 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
1975 }
1977 # Image matches a whitelist entry
1980 }
1981 }
1982 }
1984 }
1986 /**
1987 * Process [[ ]] wikilinks
1988 *
1989 * @param string $s
1990 *
1991 * @return string Processed text
1992 *
1993 * @private
1994 */
1998 }
2000 /**
2001 * Process [[ ]] wikilinks (RIL)
2002 * @param string $s
2003 * @throws MWException
2004 * @return LinkHolderArray
2005 *
2006 * @private
2007 */
2013 # the % is needed to support urlencoded titles as well
2016 # Match a link having the form [[namespace:link|alternate]]trail
2018 # Match cases where there is no "]]", which might still be images
2020 }
2024 # split the entire text string on occurrences of [[
2026 # get the first element (all text up to first [[), and remove the space we added
2035 # Match the end of a line for a word that's not followed by whitespace,
2036 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2040 }
2045 }
2054 }
2057 }
2062 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2063 # Loop for each link
2065 // @codingStandardsIgnoreStart
2067 # Check for excessive memory usage
2069 # Too big
2070 # Do the existence check, replace the link holders and clear the array
2073 }
2082 }
2083 # first link
2087 }
2089 }
2096 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2097 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2098 # the real problem is with the $e1 regex
2099 # See bug 1300.
2100 #
2101 # Still some problems for cases where the ] is meant to be outside punctuation,
2102 # and no image is in sight. See bug 2095.
2103 #
2107 ) {
2110 }
2111 # fix up urlencoded title texts
2113 # Should anchors '#' also be rejected?
2115 }
2118 # Invalid, but might be an image with a link in its caption
2123 }
2129 }
2135 # Don't allow internal links to pages containing
2136 # PROTO: where PROTO is a valid URL protocol; these
2137 # should be external links.
2142 }
2144 # Make subpage if necessary
2149 }
2153 # Strip off leading ':'
2155 }
2164 }
2175 # look at the next 'line' to see if we can close it there
2180 }
2183 # the first ]] closes the inner link, the second the image
2189 # if there's exactly one ]] that's fine, we'll keep looking
2192 # if $next_line is invalid too, we need look no further
2195 }
2196 }
2198 # we couldn't find the end of this imageLink, so output it raw
2199 # but don't ignore what might be perfectly normal links in the text we've examined
2202 # note: no $trail, because without an end, there *is* no trail
2205 }
2208 # note: no $trail, because without an end, there *is* no trail
2211 }
2213 }
2219 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2220 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2221 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2222 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2224 }
2226 # Link not escaped by : , create the various objects
2228 # Interwikis
2234 )
2235 ) {
2236 # Bug 24502: filter duplicates
2240 }
2246 }
2253 # if no parameters were passed, $text
2254 # becomes something like "File:Foo.png",
2255 # which we don't want to pass on to the
2256 # image generator
2259 # recursively parse links inside the image caption
2260 # actually, this will parse them in any other parameters, too,
2261 # but it might be hard to fix that, and it doesn't matter ATM
2264 }
2265 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2270 }
2273 }
2283 }
2289 /**
2290 * Strip the whitespace Category links produce, see bug 87
2291 */
2296 }
2297 }
2299 # Self-link checking. For some languages, variants of the title are checked in
2300 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2301 # for linking to a different variant.
2305 }
2307 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2308 # @todo FIXME: Should do batch file existence checks, see comment below
2311 # Give extensions a chance to select the file revision for us
2316 # Fetch and register the file (file title may be different via hooks)
2318 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2323 }
2326 # Some titles, such as valid special pages or files in foreign repos, should
2327 # be shown as bluelinks even though they're not included in the page table
2328 #
2329 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2330 # batch file existence checks for NS_FILE and NS_MEDIA
2335 # Links will be added to the output link list after checking
2337 }
2339 }
2341 }
2343 /**
2344 * Render a forced-blue link inline; protect against double expansion of
2345 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2346 * Since this little disaster has to split off the trail text to avoid
2347 * breaking URLs in the following text without breaking trails on the
2348 * wiki links, it's been made into a horrible function.
2349 *
2350 * @param Title $nt
2351 * @param string $text
2352 * @param array|string $query
2353 * @param string $trail
2354 * @param string $prefix
2355 * @return string HTML-wikitext mix oh yuck
2356 */
2357 public function makeKnownLinkHolder( $nt, $text = '', $query = array(), $trail = '', $prefix = '' ) {
2362 }
2365 }
2370 }
2372 /**
2373 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2374 * going to go through further parsing steps before inline URL expansion.
2375 *
2376 * Not needed quite as much as it used to be since free links are a bit
2377 * more sensible these days. But bracketed links are still an issue.
2378 *
2379 * @param string $text More-or-less HTML
2380 * @return string Less-or-more HTML with NOPARSE bits
2381 */
2385 }
2387 /**
2388 * Return true if subpage links should be expanded on this page.
2389 * @return bool
2390 */
2392 # Some namespaces don't allow subpages
2394 }
2396 /**
2397 * Handle link to subpage if necessary
2398 *
2399 * @param string $target The source of the link
2400 * @param string &$text The link text, modified as necessary
2401 * @return string The full name of the link
2402 * @private
2403 */
2406 }
2408 /**#@+
2409 * Used by doBlockLevels()
2410 * @private
2411 *
2412 * @return string
2413 */
2418 }
2422 }
2424 /**
2425 * getCommon() returns the length of the longest common substring
2426 * of both arguments, starting at the beginning of both.
2427 * @private
2428 *
2429 * @param string $st1
2430 * @param string $st2
2431 *
2432 * @return int
2433 */
2439 }
2444 }
2445 }
2447 }
2449 /**
2450 * These next three functions open, continue, and close the list
2451 * element appropriate to the prefix character passed into them.
2452 * @private
2453 *
2454 * @param string $char
2455 *
2456 * @return string
2457 */
2472 }
2475 }
2477 /**
2478 * TODO: document
2479 * @param string $char
2480 * @private
2481 *
2482 * @return string
2483 */
2491 }
2498 }
2499 }
2501 }
2503 /**
2504 * @todo Document
2505 * @param string $char
2506 * @private
2507 *
2508 * @return string
2509 */
2521 }
2524 }
2526 }
2527 /**#@-*/
2529 /**
2530 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2531 *
2532 * @param string $text
2533 * @param bool $linestart Whether or not this is at the start of a line.
2534 * @private
2535 * @return string The lists rendered as HTML
2536 */
2539 # Parsing through the text line by line. The main thing
2540 # happening here is handling of block-level elements p, pre,
2541 # and making lists from lines starting with * # : etc.
2542 #
2552 # Fix up $linestart
2557 }
2558 # * = ul
2559 # # = ol
2560 # ; = dt
2561 # : = dd
2566 # If not in a <pre> element, scan for and figure out what prefixes are there.
2568 # Multiple prefixes may abut each other for nested lists.
2572 # eh?
2573 # ; and : are both from definition-lists, so they're equivalent
2574 # for the purposes of determining whether or not we need to open/close
2575 # elements.
2580 # Don't interpret any other prefixes in preformatted text
2584 }
2586 # List generation
2588 # Same as the last item, so no need to deal with nesting or opening stuff
2593 # The one nasty exception: definition lists work like this:
2594 # ; title : definition text
2595 # So we check for : in the remainder text to split up the
2596 # title and definition, without b0rking links.
2601 }
2602 }
2604 # We need to open or close prefixes, or both.
2606 # Either open or close a level...
2610 # Close all the prefixes which aren't shared.
2614 }
2616 # Continue the current prefix if appropriate.
2619 }
2621 # Open prefixes where appropriate.
2624 }
2630 # @todo FIXME: This is dupe of code above
2634 }
2635 }
2637 }
2640 }
2642 }
2644 # If we have no prefixes, go to paragraph mode.
2647 # No prefix (not in list)--go to paragraph mode
2648 # XXX: use a stack for nestable elements like span, table and div
2650 '/(?:<table|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|'
2652 $t
2653 );
2659 $t
2660 );
2664 # @todo bug 5718: paragraph closed
2668 }
2670 while ( preg_match( '/<(\\/?)blockquote[\s>]/i', $t, $bqMatch, PREG_OFFSET_CAPTURE, $bqOffset ) ) {
2673 }
2679 ) {
2680 # pre
2685 }
2688 # paragraph
2701 }
2702 }
2711 }
2712 }
2713 }
2714 }
2716 }
2717 # somewhere above we forget to get out of pre block (bug 785)
2720 }
2725 }
2726 }
2727 }
2733 }
2734 }
2738 }
2741 }
2743 /**
2744 * Split up a string on ':', ignoring any occurrences inside tags
2745 * to prevent illegal overlapping.
2746 *
2747 * @param string $str The string to split
2748 * @param string &$before Set to everything before the ':'
2749 * @param string &$after Set to everything after the ':'
2750 * @throws MWException
2751 * @return string The position of the ':', or false if none found
2752 */
2757 # Nothing to find!
2759 }
2763 # Easy; no tag nesting to worry about
2767 }
2769 # Ugly state machine to walk through avoiding tags.
2777 # (Using the number is a performance hack for common cases)
2781 # Could be either a <start> tag or an </end> tag
2786 # We found it!
2790 }
2791 # Embedded in a tag; don't break it.
2794 # Skip ahead looking for something interesting
2797 # Nothing else interesting
2799 }
2803 # We found it!
2807 }
2808 }
2810 # Nothing else interesting to find; abort!
2811 # We're nested, but there's no close tags left. Abort!
2813 }
2814 # Skip ahead to next tag start
2817 }
2820 # In a <tag>
2827 # Slash may be followed by >?
2831 # ignore
2832 }
2843 # Illegal early close? This shouldn't happen D:
2848 }
2851 # In a </tag>
2857 }
2859 }
2863 # Yes, a self-closed tag <blah/>
2866 # Probably we're jumping the gun, and this is an attribute
2868 }
2873 }
2880 }
2887 }
2891 }
2892 }
2894 wfDebug( __METHOD__ . ": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2896 }
2898 }
2900 /**
2901 * Return value of a magic variable (like PAGENAME)
2902 *
2903 * @private
2904 *
2905 * @param int $index
2906 * @param bool|PPFrame $frame
2907 *
2908 * @throws MWException
2909 * @return string
2910 */
2916 // If no title set, bad things are going to happen
2917 // later. Title should always be set since this
2918 // should only be called in the middle of a parse
2919 // operation (but the unit-tests do funky stuff)
2922 }
2924 /**
2925 * Some of these require message or data lookups and can be
2926 * expensive to check many times.
2927 */
2931 }
2932 }
2977 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
3011 ) ) );
3021 ) ) );
3029 }
3037 }
3050 # 0 means the page doesn't exist in the database,
3051 # which means the user is previewing a new page.
3052 # The vary-revision flag must be set, because the magic word
3053 # will have a different value once the page is saved.
3056 }
3060 # Let the edit saving system know we should parse the page
3061 # *after* a revision ID has been assigned.
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.
3088 # Let the edit saving system know we should parse the page
3089 # *after* a revision ID has been assigned. This is for null edits.
3095 # Let the edit saving system know we should parse the page
3096 # *after* a revision ID has been assigned. This is for null edits.
3102 # Let the edit saving system know we should parse the page
3103 # *after* a revision ID has been assigned. This is for null edits.
3109 # Let the edit saving system know we should parse the page
3110 # *after* a revision ID has been assigned. This is for null edits.
3116 # Let the edit saving system know we should parse the page
3117 # *after* a revision ID has been assigned. This is for null edits.
3158 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3159 # int to remove the padding
3168 );
3177 false
3178 );
3184 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3185 # int to remove the padding
3246 );
3249 }
3253 }
3256 }
3258 /**
3259 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
3260 *
3261 * @private
3262 */
3269 }
3271 /**
3272 * Preprocess some wikitext and return the document tree.
3273 * This is the ghost of replace_variables().
3274 *
3275 * @param string $text The text to parse
3276 * @param int $flags Bitwise combination of:
3277 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
3278 * included. Default is to assume a direct page view.
3279 *
3280 * The generated DOM tree must depend only on the input text and the flags.
3281 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
3282 *
3283 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
3284 * change in the DOM tree for a given text, must be passed through the section identifier
3285 * in the section edit link and thus back to extractSections().
3286 *
3287 * The output of this function is currently only cached in process memory, but a persistent
3288 * cache may be implemented at a later date which takes further advantage of these strict
3289 * dependency requirements.
3290 *
3291 * @return PPNode
3292 */
3296 }
3298 /**
3299 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3300 *
3301 * @param string $s
3302 *
3303 * @return array
3304 */
3314 }
3316 }
3318 /**
3319 * Replace magic variables, templates, and template arguments
3320 * with the appropriate text. Templates are substituted recursively,
3321 * taking care to avoid infinite loops.
3322 *
3323 * Note that the substitution depends on value of $mOutputType:
3324 * self::OT_WIKI: only {{subst:}} templates
3325 * self::OT_PREPROCESS: templates but not extension tags
3326 * self::OT_HTML: all templates and extension tags
3327 *
3328 * @param string $text The text to transform
3329 * @param bool|PPFrame $frame Object describing the arguments passed to the
3330 * template. Arguments may also be provided as an associative array, as
3331 * was the usual case before MW1.12. Providing arguments this way may be
3332 * useful for extensions wishing to perform variable replacement
3333 * explicitly.
3334 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
3335 * double-brace expansion.
3336 * @return string
3337 */
3339 # Is there any text? Also, Prevent too big inclusions!
3342 }
3350 }
3357 }
3359 /**
3360 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3361 *
3362 * @param array $args
3363 *
3364 * @return array
3365 */
3378 }
3381 }
3382 }
3383 }
3386 }
3388 /**
3389 * Warn the user when a parser limitation is reached
3390 * Will warn at most once the user per limitation type
3391 *
3392 * @param string $limitationType Should be one of:
3393 * 'expensive-parserfunction' (corresponding messages:
3394 * 'expensive-parserfunction-warning',
3395 * 'expensive-parserfunction-category')
3396 * 'post-expand-template-argument' (corresponding messages:
3397 * 'post-expand-template-argument-warning',
3398 * 'post-expand-template-argument-category')
3399 * 'post-expand-template-inclusion' (corresponding messages:
3400 * 'post-expand-template-inclusion-warning',
3401 * 'post-expand-template-inclusion-category')
3402 * 'node-count-exceeded' (corresponding messages:
3403 * 'node-count-exceeded-warning',
3404 * 'node-count-exceeded-category')
3405 * 'expansion-depth-exceeded' (corresponding messages:
3406 * 'expansion-depth-exceeded-warning',
3407 * 'expansion-depth-exceeded-category')
3408 * @param string|int|null $current Current value
3409 * @param string|int|null $max Maximum allowed, when an explicit limit has been
3410 * exceeded, provide the values (optional)
3411 */
3413 # does no harm if $current and $max are present but are unnecessary for the message
3418 }
3420 /**
3421 * Return the text of a template, after recursively
3422 * replacing any variables or templates within the template.
3423 *
3424 * @param array $piece The parts of the template
3425 * $piece['title']: the title, i.e. the part before the |
3426 * $piece['parts']: the parameter array
3427 * $piece['lineStart']: whether the brace was at the start of a line
3428 * @param PPFrame $frame The current frame, contains template arguments
3429 * @throws Exception
3430 * @return string The text of the template
3431 */
3435 // Flags
3437 // $text has been filled
3439 // wiki markup in $text should be escaped
3441 // $text is HTML, armour it against wikitext transformation
3443 // Force interwiki transclusion to be done in raw mode not rendered
3445 // $text is a DOM node needing expansion in a child frame
3447 // $text is a DOM node needing expansion in the current frame
3450 # Title object, where $text came from
3453 # $part1 is the bit before the first |, and must contain only title characters.
3454 # Various prefixes will be stripped from it later.
3459 # Original title text preserved for various purposes
3462 # $args is a list of argument nodes, starting from index 0, not including $part1
3463 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3464 # below won't work b/c this $args isn't an object
3470 # SUBST
3476 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3477 # Decide whether to expand template or keep wikitext as-is.
3483 }
3489 }
3490 }
3495 }
3496 }
3498 # Variables
3505 }
3507 }
3508 }
3510 # MSG, MSGNW and RAW
3512 # Check for MSGNW:
3517 # Remove obsolete MSG:
3520 }
3522 # Check for RAW:
3526 }
3527 }
3530 # Parser functions
3540 }
3546 }
3548 # The interface for parser functions allows for extracting
3549 # flags into the local scope. Extract any forwarded flags
3550 # here.
3552 }
3554 }
3556 # Finish mangling title and then check for loops.
3557 # Set $title to a Title object and $titleText to the PDBK
3560 # Split the title into page and subpage
3566 }
3570 # Check for language variants if the template is not found
3573 }
3574 # Do recursion depth check
3582 }
3583 }
3584 }
3586 # Load from database
3594 ) {
3595 // Pass the template arguments as URL parameters.
3596 // "uselang" will have no effect since the Language object
3597 // is forced to the one defined in ParserOptions.
3606 }
3607 }
3609 // Create a new context to execute the special page
3622 }
3632 }
3633 }
3635 # If the title is valid but undisplayable, make a link to it
3639 }
3641 # Interwiki transclusion
3647 # Preprocess it like a template
3650 }
3652 }
3654 # Do infinite loop check
3655 # This has to be done after redirect resolution to avoid infinite loops via redirects
3662 }
3664 }
3666 # If we haven't found text to substitute by now, we're done
3667 # Recover the source wikitext and return it
3672 }
3674 }
3676 # Expand DOM-style return values in a child frame
3678 # Clean up argument array
3684 # Expansion is eligible for the empty-frame cache
3687 # Uncached expansion
3689 }
3690 }
3694 }
3698 }
3700 # Replace raw HTML by a placeholder
3704 # Escape nowiki-style return values
3709 ) {
3710 # Bug 529: if the template begins with a table or block-level
3711 # element, it should be treated as beginning a new line.
3712 # This behavior is somewhat controversial.
3714 }
3717 # Error, oversize inclusion
3719 # Make a working, properly escaped link if possible (bug 23588)
3722 # This will probably not be a working link, but at least it may
3723 # provide some hint of where the problem is
3726 }
3730 }
3736 }
3739 }
3741 /**
3742 * Call a parser function and return an array with text and flags.
3743 *
3744 * The returned array will always contain a boolean 'found', indicating
3745 * whether the parser function was found or not. It may also contain the
3746 * following:
3747 * text: string|object, resulting wikitext or PP DOM object
3748 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3749 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3750 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3751 * nowiki: bool, wiki markup in $text should be escaped
3752 *
3753 * @since 1.21
3754 * @param PPFrame $frame The current frame, contains template arguments
3755 * @param string $function Function name
3756 * @param array $args Arguments to the function
3757 * @throws MWException
3758 * @return array
3759 */
3764 # Case sensitive functions
3768 # Case insensitive functions
3774 }
3775 }
3780 # Workaround for PHP bug 35229 and similar
3784 }
3788 # Convert arguments to PPNodes and collect for appending to $allArgs
3795 }
3796 }
3798 # Add a frame parameter, and pass the arguments as an array
3802 # Convert arguments to plain text and append to $allArgs
3810 }
3811 }
3812 }
3816 # The interface for function hooks allows them to return a wikitext
3817 # string or an array containing the string and any flags. This mungs
3818 # things around to match what this method should return.
3823 );
3827 }
3831 );
3832 }
3838 }
3841 }
3846 }
3850 }
3852 /**
3853 * Get the semi-parsed DOM representation of a template with a given title,
3854 * and its redirect destination title. Cached.
3855 *
3856 * @param Title $title
3857 *
3858 * @return array
3859 */
3868 }
3871 }
3873 # Cache miss, go to the database
3879 }
3887 }
3890 }
3892 /**
3893 * Fetch the current revision of a given title. Note that the revision
3894 * (and even the title) may not exist in the database, so everything
3895 * contributing to the output of the parser should use this method
3896 * where possible, rather than getting the revisions themselves. This
3897 * method also caches its results, so using it benefits performance.
3898 *
3899 * @since 1.24
3900 * @param Title $title
3901 * @return Revision
3902 */
3907 }
3910 // Defaults to Parser::statelessFetchRevision()
3912 );
3913 }
3915 }
3917 /**
3918 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3919 * without passing them on to it.
3920 *
3921 * @since 1.24
3922 * @param Title $title
3923 * @param Parser|bool $parser
3924 * @return Revision
3925 */
3928 }
3930 /**
3931 * Fetch the unparsed text of a template and register a reference to it.
3932 * @param Title $title
3933 * @return array ( string or false, Title )
3934 */
3936 // Defaults to Parser::statelessFetchTemplate()
3945 // If we transclude ourselves, the final result
3946 // will change based on the new version of the page
3948 }
3949 }
3950 }
3952 }
3954 /**
3955 * Fetch the unparsed text of a template and register a reference to it.
3956 * @param Title $title
3957 * @return string|bool
3958 */
3962 }
3964 /**
3965 * Static function to get a template
3966 * Can be overridden via ParserOptions::setTemplateCallback().
3967 *
3968 * @param Title $title
3969 * @param bool|Parser $parser
3970 *
3971 * @return array
3972 */
3978 # Loop to fetch the article, with up to 1 redirect
3980 # Give extensions a chance to select the revision instead
3991 );
3993 }
3994 # Get the revision
4001 }
4003 # If there is no current revision, there is no page
4007 }
4014 # We fetched a rev from a different title; register it too...
4019 }
4028 }
4035 }
4040 }
4043 }
4044 # Redirect?
4047 }
4052 }
4054 /**
4055 * Fetch a file and its title and register a reference to it.
4056 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4057 * @param Title $title
4058 * @param array $options Array of options to RepoGroup::findFile
4059 * @return File|bool
4060 */
4064 }
4066 /**
4067 * Fetch a file and its title and register a reference to it.
4068 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4069 * @param Title $title
4070 * @param array $options Array of options to RepoGroup::findFile
4071 * @return array ( File or false, Title of file )
4072 */
4078 # Register the file as a dependency...
4081 # Update fetched file title
4084 }
4086 }
4088 /**
4089 * Helper function for fetchFileAndTitle.
4090 *
4091 * Also useful if you need to fetch a file but not use it yet,
4092 * for example to get the file's handler.
4093 *
4094 * @param Title $title
4095 * @param array $options Array of options to RepoGroup::findFile
4096 * @return File|bool
4097 */
4105 }
4107 }
4109 /**
4110 * Transclude an interwiki link.
4111 *
4112 * @param Title $title
4113 * @param string $action
4114 *
4115 * @return string
4116 */
4122 }
4128 }
4130 }
4132 /**
4133 * @param string $url
4134 * @return mixed|string
4135 */
4144 }
4151 // Though we failed to fetch the content, this status is useless.
4156 }
4163 ) );
4165 }
4167 /**
4168 * Triple brace replacement -- used for template arguments
4169 * @private
4170 *
4171 * @param array $piece
4172 * @param PPFrame $frame
4173 *
4174 * @return array
4175 */
4188 )
4189 ) {
4190 # No match in frame, use the supplied default
4192 }
4196 }
4199 # No match anywhere
4201 }
4204 }
4209 }
4212 }
4214 /**
4215 * Return the text to be used for a given extension tag.
4216 * This is the ghost of strip().
4217 *
4218 * @param array $params Associative array of parameters:
4219 * name PPNode for the tag name
4220 * attr PPNode for unparsed text where tag attributes are thought to be
4221 * attributes Optional associative array of parsed attributes
4222 * inner Contents of extension element
4223 * noClose Original text did not have a close tag
4224 * @param PPFrame $frame
4225 *
4226 * @throws MWException
4227 * @return string
4228 */
4242 }
4248 }
4251 # Workaround for PHP bug 35229 and similar
4254 }
4261 }
4267 }
4270 # Extract flags to local scope (to override $markerType)
4275 }
4279 }
4284 }
4285 }
4291 }
4292 }
4302 }
4304 }
4306 /**
4307 * Increment an include size counter
4308 *
4309 * @param string $type The type of expansion
4310 * @param int $size The size of the text
4311 * @return bool False if this inclusion would take it over the maximum, true otherwise
4312 */
4319 }
4320 }
4322 /**
4323 * Increment the expensive function count
4324 *
4325 * @return bool False if the limit has been exceeded
4326 */
4330 }
4332 /**
4333 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
4334 * Fills $this->mDoubleUnderscores, returns the modified text
4335 *
4336 * @param string $text
4337 *
4338 * @return string
4339 */
4342 # The position of __TOC__ needs to be recorded
4348 # Set a placeholder. At the end we'll fill it in with the TOC.
4351 # Only keep the first one.
4353 }
4355 # Now match and remove the rest of them
4361 }
4364 }
4367 ) {
4369 }
4370 # (bug 8068) Allow control over whether robots index a page.
4371 #
4372 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
4373 # is not desirable, the last one on the page should win.
4377 }
4381 }
4383 # Cache all double underscores in the database
4386 }
4389 }
4391 /**
4392 * @see ParserOutput::addTrackingCategory()
4393 * @param string $msg Message key
4394 * @return bool Whether the addition was successful
4395 */
4398 }
4400 /**
4401 * This function accomplishes several tasks:
4402 * 1) Auto-number headings if that option is enabled
4403 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
4404 * 3) Add a Table of contents on the top for users who have enabled the option
4405 * 4) Auto-anchor headings
4406 *
4407 * It loops through all headlines, collects the necessary data, then splits up the
4408 * string and re-inserts the newly formatted headlines.
4409 *
4410 * @param string $text
4411 * @param string $origText Original, untouched wikitext
4412 * @param bool $isMain
4413 * @return mixed|string
4414 * @private
4415 */
4419 # Inhibit editsection links if requested in the page
4425 }
4428 }
4430 # Get all headlines for numbering them and adding funky stuff like [edit]
4431 # links - this is for later, but we need the number of headlines right now
4436 $matches
4437 );
4439 # if there are fewer than 4 headlines in the article, do not show TOC
4440 # unless it's been explicitly enabled.
4444 # Allow user to stipulate that a page should have a "new section"
4445 # link added via __NEWSECTIONLINK__
4448 }
4450 # Allow user to remove the "new section"
4451 # link via __NONEWSECTIONLINK__
4454 }
4456 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4457 # override above conditions and always show TOC above first header
4461 }
4463 # headline counter
4467 # Ugh .. the TOC should have neat indentation levels which can be
4468 # passed to the skin functions. These are determined here
4500 }
4504 }
4508 # Increase TOC level
4515 }
4517 # Decrease TOC level, find level to jump to
4521 # Found last matching level
4525 # Found first matching level below current level
4528 }
4529 }
4532 }
4535 # Unindent only if the previous toc level was shown :p
4540 }
4541 }
4543 # No change in level, end TOC line
4546 }
4547 }
4551 # count number of headlines for each level
4558 }
4561 }
4562 }
4564 # The safe header is a version of the header text safe to use for links
4566 # Remove link placeholders by the link text.
4567 # <!--LINK number-->
4568 # turns into
4569 # link text with suffix
4570 # Do this before unstrip since link text can contain strip markers
4573 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4576 # Strip out HTML (first regex removes any tag not allowed)
4577 # Allowed tags are:
4578 # * <sup> and <sub> (bug 8393)
4579 # * <i> (bug 26375)
4580 # * <b> (r105284)
4581 # * <bdi> (bug 72884)
4582 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4583 #
4584 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4585 # to allow setting directionality in toc items.
4590 ),
4592 $safeHeadline
4593 );
4596 # For the anchor, strip out HTML-y stuff period
4600 # Save headline for section edit hint before it's escaped
4604 # For reverse compatibility, provide an id that's
4605 # HTML4-compatible, like we used to.
4606 #
4607 # It may be worth noting, academically, that it's possible for
4608 # the legacy anchor to conflict with a non-legacy headline
4609 # anchor on the page. In this case likely the "correct" thing
4610 # would be to either drop the legacy anchors or make sure
4611 # they're numbered first. However, this would require people
4612 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4613 # manually, so let's not bother worrying about it.
4619 # No reason to have both (in fact, we can't)
4621 }
4626 }
4628 # HTML names must be case-insensitively unique (bug 10721).
4629 # This does not apply to Unicode characters per
4630 # http://dev.w3.org/html5/spec/infrastructure.html#case-sensitivity-and-string-comparison
4631 # @todo FIXME: We may be changing them depending on the current locale.
4637 }
4639 # count how many in assoc. array so we can track dupes in anchors
4644 }
4649 }
4651 # Don't number the heading if it is the only one (looks silly)
4653 # the two are different if the line contains a link
4657 $numbering
4659 }
4661 # Create the anchor for linking from the TOC to the section
4666 }
4669 }
4673 }
4675 # Add the section to the section tree
4676 # Find the DOM node for this header
4683 }
4684 }
4688 }
4698 );
4700 # give headline the correct <h#> tag
4702 // Output edit section links as markers with styles that can be customized by skins
4704 # Put a T flag in the section identifier, to indicate to extractSections()
4705 # that sections inside <includeonly> should be counted.
4713 }
4714 // We use a bit of pesudo-xml for editsection markers. The
4715 // language converter is run later on. Using a UNIQ style marker
4716 // leads to the converter screwing up the tokens when it
4717 // converts stuff. And trying to insert strip tags fails too. At
4718 // this point all real inputted tags have already been escaped,
4719 // so we don't have to worry about a user trying to input one of
4720 // these markers directly. We use a page and section attribute
4721 // to stop the language converter from converting these
4722 // important bits of data, but put the headline hint inside a
4723 // content block because the language converter is supposed to
4724 // be able to convert that piece of data.
4725 // Gets replaced with html in ParserOutput::getText
4732 }
4735 }
4741 }
4745 # Never ever show TOC if no headers
4748 }
4753 }
4758 }
4762 }
4764 # split up and insert constructed headlines
4768 // build an array of document sections
4771 // $head is zero-based, sections aren't.
4776 }
4778 /**
4779 * Send a hook, one per section.
4780 * The idea here is to be able to make section-level DIVs, but to do so in a
4781 * lower-impact, more correct way than r50769
4782 *
4783 * $this : caller
4784 * $section : the section number
4785 * &$sectionContent : ref to the content of the section
4786 * $showEditLinks : boolean describing whether this section has an edit link
4787 */
4791 }
4794 // append the TOC at the beginning
4795 // Top anchor now in skin
4797 }
4805 }
4806 }
4808 /**
4809 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4810 * conversion, substituting signatures, {{subst:}} templates, etc.
4811 *
4812 * @param string $text The text to transform
4813 * @param Title $title The Title object for the current article
4814 * @param User $user The User object describing the current user
4815 * @param ParserOptions $options Parsing options
4816 * @param bool $clearState Whether to clear the parser state first
4817 * @return string The altered wiki markup
4818 */
4821 ) {
4824 }
4831 );
4835 }
4841 }
4843 /**
4844 * Pre-save transform helper function
4845 *
4846 * @param string $text
4847 * @param User $user
4848 *
4849 * @return string
4850 */
4854 # Note: This is the timestamp saved as hardcoded wikitext to
4855 # the database, we use $wgContLang here in order to give
4856 # everyone the same signature and use the default one rather
4857 # than the one selected in each user's preferences.
4858 # (see also bug 12815)
4864 # Allow translation of timezones through wiki. format() can return
4865 # whatever crap the system uses, localised or not, so we cannot
4866 # ship premade translations.
4871 }
4875 # Variable replacement
4876 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4879 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4880 # which may corrupt this parser instance via its wfMessage()->text() call-
4882 # Signatures
4888 ) );
4890 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4894 // [[ns:page (context)|]]
4896 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4898 // [[ns:page (context), context|]] (using either single or double-width comma)
4900 // [[|page]] (reverse pipe trick: add context from page title)
4903 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4915 # if there's no context, don't bother duplicating the title
4917 }
4919 # Trim trailing whitespace
4923 }
4925 /**
4926 * Fetch the user's signature text, if any, and normalize to
4927 * validated, ready-to-insert wikitext.
4928 * If you have pre-fetched the nickname or the fancySig option, you can
4929 * specify them here to save a database query.
4930 * Do not reuse this parser instance after calling getUserSig(),
4931 * as it may have changed if it's the $wgParser.
4932 *
4933 * @param User $user
4934 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4935 * @param bool|null $fancySig whether the nicknname is the complete signature
4936 * or null to use default value
4937 * @return string
4938 */
4944 # If not given, retrieve from the user object.
4947 }
4951 }
4959 # Sig. might contain markup; validate this
4961 # Validated; clean up (if needed) and return it
4964 # Failed to validate; fall back to the default
4967 }
4968 }
4970 # Make sure nickname doesnt get a sig in a sig
4973 # If we're still here, make it a link to the user page
4980 }
4982 /**
4983 * Check that the user's signature contains no bad XML
4984 *
4985 * @param string $text
4986 * @return string|bool An expanded string, or false if invalid.
4987 */
4990 }
4992 /**
4993 * Clean up signature text
4994 *
4995 * 1) Strip ~~~, ~~~~ and ~~~~~ out of signatures @see cleanSigInSig
4996 * 2) Substitute all transclusions
4997 *
4998 * @param string $text
4999 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
5000 * @return string Signature text
5001 */
5007 }
5009 # Option to disable this feature
5012 }
5014 # @todo FIXME: Regex doesn't respect extension tags or nowiki
5015 # => Move this logic to braceSubstitution()
5017 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
5028 }
5031 }
5033 /**
5034 * Strip ~~~, ~~~~ and ~~~~~ out of signatures
5035 *
5036 * @param string $text
5037 * @return string Signature text with /~{3,5}/ removed
5038 */
5042 }
5044 /**
5045 * Set up some variables which are usually set up in parse()
5046 * so that an external function can call some class members with confidence
5047 *
5048 * @param Title|null $title
5049 * @param ParserOptions $options
5050 * @param int $outputType
5051 * @param bool $clearState
5052 */
5055 ) {
5057 }
5059 /**
5060 * @param Title|null $title
5061 * @param ParserOptions $options
5062 * @param int $outputType
5063 * @param bool $clearState
5064 */
5067 ) {
5073 }
5074 }
5076 /**
5077 * Wrapper for preprocess()
5078 *
5079 * @param string $text The text to preprocess
5080 * @param ParserOptions $options Options
5081 * @param Title|null $title Title object or null to use $wgTitle
5082 * @return string
5083 */
5087 # Guard against infinite recursion
5090 }
5096 }
5102 }
5104 /**
5105 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
5106 * The callback should have the following form:
5107 * function myParserHook( $text, $params, $parser, $frame ) { ... }
5108 *
5109 * Transform and return $text. Use $parser for any required context, e.g. use
5110 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
5111 *
5112 * Hooks may return extended information by returning an array, of which the
5113 * first numbered element (index 0) must be the return string, and all other
5114 * entries are extracted into local variables within an internal function
5115 * in the Parser class.
5116 *
5117 * This interface (introduced r61913) appears to be undocumented, but
5118 * 'markerName' is used by some core tag hooks to override which strip
5119 * array their results are placed in. **Use great caution if attempting
5120 * this interface, as it is not documented and injudicious use could smash
5121 * private variables.**
5122 *
5123 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5124 * @param callable $callback The callback function (and object) to use for the tag
5125 * @throws MWException
5126 * @return callable|null The old value of the mTagHooks array associated with the hook
5127 */
5132 }
5137 }
5140 }
5142 /**
5143 * As setHook(), but letting the contents be parsed.
5144 *
5145 * Transparent tag hooks are like regular XML-style tag hooks, except they
5146 * operate late in the transformation sequence, on HTML instead of wikitext.
5147 *
5148 * This is probably obsoleted by things dealing with parser frames?
5149 * The only extension currently using it is geoserver.
5150 *
5151 * @since 1.10
5152 * @todo better document or deprecate this
5153 *
5154 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5155 * @param callable $callback The callback function (and object) to use for the tag
5156 * @throws MWException
5157 * @return callable|null The old value of the mTagHooks array associated with the hook
5158 */
5163 }
5164 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
5168 }
5170 /**
5171 * Remove all tag hooks
5172 */
5177 }
5179 /**
5180 * Create a function, e.g. {{sum:1|2|3}}
5181 * The callback function should have the form:
5182 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
5183 *
5184 * Or with Parser::SFH_OBJECT_ARGS:
5185 * function myParserFunction( $parser, $frame, $args ) { ... }
5186 *
5187 * The callback may either return the text result of the function, or an array with the text
5188 * in element 0, and a number of flags in the other elements. The names of the flags are
5189 * specified in the keys. Valid flags are:
5190 * found The text returned is valid, stop processing the template. This
5191 * is on by default.
5192 * nowiki Wiki markup in the return value should be escaped
5193 * isHTML The returned text is HTML, armour it against wikitext transformation
5194 *
5195 * @param string $id The magic word ID
5196 * @param callable $callback The callback function (and object) to use
5197 * @param int $flags A combination of the following flags:
5198 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
5199 *
5200 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
5201 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
5202 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
5203 * the arguments, and to control the way they are expanded.
5204 *
5205 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
5206 * arguments, for instance:
5207 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
5208 *
5209 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
5210 * future versions. Please call $frame->expand() on it anyway so that your code keeps
5211 * working if/when this is changed.
5212 *
5213 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
5214 * expansion.
5215 *
5216 * Please read the documentation in includes/parser/Preprocessor.php for more information
5217 * about the methods available in PPFrame and PPNode.
5218 *
5219 * @throws MWException
5220 * @return string|callable The old callback function for this name, if any
5221 */
5228 # Add to function cache
5232 }
5238 # Case
5241 }
5242 # Add leading hash
5245 }
5246 # Remove trailing colon
5249 }
5251 }
5253 }
5255 /**
5256 * Get all registered function hook identifiers
5257 *
5258 * @return array
5259 */
5262 }
5264 /**
5265 * Create a tag function, e.g. "<test>some stuff</test>".
5266 * Unlike tag hooks, tag functions are parsed at preprocessor level.
5267 * Unlike parser functions, their content is not preprocessed.
5268 * @param string $tag
5269 * @param callable $callback
5270 * @param int $flags
5271 * @throws MWException
5272 * @return null
5273 */
5278 }
5285 }
5288 }
5290 /**
5291 * @todo FIXME: Update documentation. makeLinkObj() is deprecated.
5292 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
5293 * Placeholders created in Skin::makeLinkObj()
5294 *
5295 * @param string $text
5296 * @param int $options
5297 */
5300 }
5302 /**
5303 * Replace "<!--LINK-->" link placeholders with plain text of links
5304 * (not HTML-formatted).
5305 *
5306 * @param string $text
5307 * @return string
5308 */
5311 }
5313 /**
5314 * Renders an image gallery from a text with one line per image.
5315 * text labels may be given by using |-style alternative text. E.g.
5316 * Image:one.jpg|The number "1"
5317 * Image:tree.jpg|A tree
5318 * given as text will return the HTML of a gallery with two images,
5319 * labeled 'The number "1"' and
5320 * 'A tree'.
5321 *
5322 * @param string $text
5323 * @param array $params
5324 * @return string HTML
5325 */
5331 }
5336 // If invalid type set, fallback to default.
5338 }
5351 }
5357 }
5360 }
5363 }
5366 }
5373 # match lines like these:
5374 # Image:someimage.jpg|This is some image
5377 # Skip empty lines
5380 }
5384 }
5387 # Bogus title. Ignore these so we don't bomb out later.
5389 }
5391 # We need to get what handler the file uses, to figure out parameters.
5392 # Note, a hook can overide the file name, and chose an entirely different
5393 # file (which potentially could be of a different type and have different handler).
5398 # Don't register it now, as ImageGallery does that later.
5406 );
5409 // We don't want people to specify per-image widths.
5410 // Additionally the width parameter would need special casing anyhow.
5412 }
5422 // look for an |alt= definition while trying not to break existing
5423 // captions with multiple pipes (|) in it, until a more sensible grammar
5424 // is defined for images in galleries
5426 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5427 // splitting on '|' is a bit odd, and different from makeImage.
5444 //check to see if link matches an absolute url, if not then it must be a wiki link.
5451 }
5452 }
5455 // Must be a handler specific parameter.
5459 // Guess not. Append it to the caption.
5462 }
5463 }
5466 // concatenate all other pipes
5468 }
5469 }
5470 // remove the first pipe
5472 }
5475 }
5479 }
5481 /**
5482 * @param string $handler
5483 * @return array
5484 */
5490 }
5492 # Initialise static lists
5499 );
5507 }
5508 }
5509 }
5511 # Add handler params
5517 }
5518 }
5521 }
5522 return array( $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] );
5523 }
5525 /**
5526 * Parse image options text and use it to make an image
5527 *
5528 * @param Title $title
5529 * @param string $options
5530 * @param LinkHolderArray|bool $holders
5531 * @return string HTML
5532 */
5534 # Check if the options text is of the form "options|alt text"
5535 # Options are:
5536 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5537 # * left no resizing, just left align. label is used for alt= only
5538 # * right same, but right aligned
5539 # * none same, but not aligned
5540 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5541 # * center center the image
5542 # * frame Keep original image size, no magnify-button.
5543 # * framed Same as "frame"
5544 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5545 # * upright reduce width for upright images, rounded to full __0 px
5546 # * border draw a 1px border around the image
5547 # * alt Text for HTML alt attribute (defaults to empty)
5548 # * class Set a class for img node
5549 # * link Set the target of the image link. Can be external, interwiki, or local
5550 # vertical-align values (no % or length right now):
5551 # * baseline
5552 # * sub
5553 # * super
5554 # * top
5555 # * text-top
5556 # * middle
5557 # * bottom
5558 # * text-bottom
5562 # Give extensions a chance to select the file revision for us
5567 # Fetch and register the file (file title may be different via hooks)
5570 # Get parameter map
5577 }
5579 # Process the input parameters
5591 # Special case; width and height come in one variable together
5599 }
5600 }
5606 }
5607 }
5608 # else no validation -- bug 13436
5611 # Validate handler parameter
5614 # Validate internal parameters
5619 # @todo FIXME: Possibly check validity here for
5620 # manualthumb? downstream behavior seems odd with
5621 # missing manual thumbs.
5638 }
5640 }
5648 }
5649 }
5654 // use first appearing option, discard others.
5659 # Most other things appear to be empty or numeric...
5661 }
5662 }
5666 }
5667 }
5668 }
5671 }
5672 }
5674 # Process alignment parameters
5677 }
5680 }
5684 # Will the image be presented in a frame, with the caption below?
5690 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5691 # came to also set the caption, ordinary text after the image -- which
5692 # makes no sense, because that just repeats the text multiple times in
5693 # screen readers. It *also* came to set the title attribute.
5694 #
5695 # Now that we have an alt attribute, we should not set the alt text to
5696 # equal the caption: that's worse than useless, it just repeats the
5697 # text. This is the framed/thumbnail case. If there's no caption, we
5698 # use the unnamed parameter for alt text as well, just for the time be-
5699 # ing, if the unnamed param is set and the alt param is not.
5700 #
5701 # For the future, we need to figure out if we want to tweak this more,
5702 # e.g., introducing a title= parameter for the title; ignoring the un-
5703 # named parameter entirely for images without a caption; adding an ex-
5704 # plicit caption= parameter and preserving the old magic unnamed para-
5705 # meter for BC; ...
5708 # No caption or alt text, add the filename as the alt text so
5709 # that screen readers at least get some description of the image
5711 }
5712 # Do not set $params['frame']['title'] because tooltips don't make sense
5713 # for framed images
5716 # No alt text, use the "caption" for the alt text
5720 # No caption, fall back to using the filename for the
5721 # alt text
5723 }
5724 }
5725 # Use the "caption" for the tooltip text
5727 }
5731 # Linker does the rest
5736 # Give the handler a chance to modify the parser object
5739 }
5742 }
5744 /**
5745 * @param string $caption
5746 * @param LinkHolderArray|bool $holders
5747 * @return mixed|string
5748 */
5750 # Strip bad stuff out of the title (tooltip). We can't just use
5751 # replaceLinkHoldersText() here, because if this function is called
5752 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5757 }
5759 # make sure there are no placeholders in thumbnail attributes
5760 # that are later expanded to html- so expand them now and
5761 # remove the tags
5766 }
5768 /**
5769 * Set a flag in the output object indicating that the content is dynamic and
5770 * shouldn't be cached.
5771 */
5777 }
5780 }
5782 /**
5783 * Callback from the Sanitizer for expanding items found in HTML attribute
5784 * values, so they can be safely tested and escaped.
5785 *
5786 * @param string $text
5787 * @param bool|PPFrame $frame
5788 * @return string
5789 */
5794 }
5796 /**
5797 * Accessor
5798 *
5799 * @return array
5800 */
5806 );
5807 }
5809 /**
5810 * Replace transparent tags in $text with the values given by the callbacks.
5811 *
5812 * Transparent tag hooks are like regular XML-style tag hooks, except they
5813 * operate late in the transformation sequence, on HTML instead of wikitext.
5814 *
5815 * @param string $text
5816 *
5817 * @return string
5818 */
5832 );
5835 }
5837 }
5839 }
5841 /**
5842 * Break wikitext input into sections, and either pull or replace
5843 * some particular section's text.
5844 *
5845 * External callers should use the getSection and replaceSection methods.
5846 *
5847 * @param string $text Page wikitext
5848 * @param string|number $sectionId A section identifier string of the form:
5849 * "<flag1> - <flag2> - ... - <section number>"
5850 *
5851 * Currently the only recognised flag is "T", which means the target section number
5852 * was derived during a template inclusion parse, in other words this is a template
5853 * section edit link. If no flags are given, it was an ordinary section edit link.
5854 * This flag is required to avoid a section numbering mismatch when a section is
5855 * enclosed by "<includeonly>" (bug 6563).
5856 *
5857 * The section number 0 pulls the text before the first heading; other numbers will
5858 * pull the given section along with its lower-level subsections. If the section is
5859 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5860 *
5861 * Section 0 is always considered to exist, even if it only contains the empty
5862 * string. If $text is the empty string and section 0 is replaced, $newText is
5863 * returned.
5864 *
5865 * @param string $mode One of "get" or "replace"
5866 * @param string $newText Replacement text for section data.
5867 * @return string For "get", the extracted section text.
5868 * for "replace", the whole page with the section replaced.
5869 */
5878 # Process section extraction flags
5885 }
5886 }
5888 # Check for empty input
5890 # Only sections 0 and T-0 exist in an empty document
5896 }
5902 }
5903 }
5904 }
5906 # Preprocess the text
5909 # <h> nodes indicate section breaks
5910 # They can only occur at the top level, so we can find them by iterating the root's children
5913 # Find the target section
5915 # Section zero doesn't nest, level=big
5924 }
5925 }
5928 }
5930 }
5931 }
5934 # Not found
5939 }
5940 }
5942 # Find the end of the section, including nested sections
5949 }
5950 }
5953 }
5957 # Write out the remainder (in replace mode only)
5959 # Output the replacement text
5960 # Add two newlines on -- trailing whitespace in $newText is conventionally
5961 # stripped by the editor, so we need both newlines to restore the paragraph gap
5962 # Only add trailing whitespace if there is newText
5965 }
5970 }
5971 }
5974 # Re-insert stripped tags
5976 }
5979 }
5981 /**
5982 * This function returns the text of a section, specified by a number ($section).
5983 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5984 * the first section before any such heading (section 0).
5985 *
5986 * If a section contains subsections, these are also returned.
5987 *
5988 * @param string $text Text to look in
5989 * @param string|number $sectionId Section identifier as a number or string
5990 * (e.g. 0, 1 or 'T-1').
5991 * @param string $defaultText Default to return if section is not found
5992 *
5993 * @return string Text of the requested section
5994 */
5997 }
5999 /**
6000 * This function returns $oldtext after the content of the section
6001 * specified by $section has been replaced with $text. If the target
6002 * section does not exist, $oldtext is returned unchanged.
6003 *
6004 * @param string $oldText Former text of the article
6005 * @param string|number $sectionId Section identifier as a number or string
6006 * (e.g. 0, 1 or 'T-1').
6007 * @param string $newText Replacing text
6008 *
6009 * @return string Modified text
6010 */
6013 }
6015 /**
6016 * Get the ID of the revision we are parsing
6017 *
6018 * @return int|null
6019 */
6022 }
6024 /**
6025 * Get the revision object for $this->mRevisionId
6026 *
6027 * @return Revision|null Either a Revision object or null
6028 * @since 1.23 (public since 1.23)
6029 */
6033 }
6036 }
6040 }
6042 /**
6043 * Get the timestamp associated with the current revision, adjusted for
6044 * the default server-local timestamp
6045 * @return string
6046 */
6055 # The cryptic '' timezone parameter tells to use the site-default
6056 # timezone offset instead of the user settings.
6057 #
6058 # Since this value will be saved into the parser cache, served
6059 # to other users, and potentially even used inside links and such,
6060 # it needs to be consistent for all visitors.
6063 }
6065 }
6067 /**
6068 * Get the name of the user that edited the last revision
6069 *
6070 * @return string User name
6071 */
6076 # if this template is subst: the revision id will be blank,
6077 # so just use the current user's name
6082 }
6083 }
6085 }
6087 /**
6088 * Get the size of the revision
6089 *
6090 * @return int|null Revision size
6091 */
6096 # if this variable is subst: the revision id will be blank,
6097 # so just use the parser input size, because the own substituation
6098 # will change the size.
6103 }
6104 }
6106 }
6108 /**
6109 * Mutator for $mDefaultSort
6110 *
6111 * @param string $sort New value
6112 */
6116 }
6118 /**
6119 * Accessor for $mDefaultSort
6120 * Will use the empty string if none is set.
6121 *
6122 * This value is treated as a prefix, so the
6123 * empty string is equivalent to sorting by
6124 * page name.
6125 *
6126 * @return string
6127 */
6133 }
6134 }
6136 /**
6137 * Accessor for $mDefaultSort
6138 * Unlike getDefaultSort(), will return false if none is set
6139 *
6140 * @return string|bool
6141 */
6144 }
6146 /**
6147 * Try to guess the section anchor name based on a wikitext fragment
6148 * presumably extracted from a heading, for example "Header" from
6149 * "== Header ==".
6150 *
6151 * @param string $text
6152 *
6153 * @return string
6154 */
6156 # Strip out wikitext links(they break the anchor)
6160 }
6162 /**
6163 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
6164 * instead. For use in redirects, since IE6 interprets Redirect: headers
6165 * as something other than UTF-8 (apparently?), resulting in breakage.
6166 *
6167 * @param string $text The section name
6168 * @return string An anchor
6169 */
6171 # Strip out wikitext links(they break the anchor)
6175 }
6177 /**
6178 * Strips a text string of wikitext for use in a section anchor
6179 *
6180 * Accepts a text string and then removes all wikitext from the
6181 * string and leaves only the resultant text (i.e. the result of
6182 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
6183 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
6184 * to create valid section anchors by mimicing the output of the
6185 * parser when headings are parsed.
6186 *
6187 * @param string $text Text string to be stripped of wikitext
6188 * for use in a Section anchor
6189 * @return string Filtered text string
6190 */
6192 # Strip internal link markup
6196 # Strip external link markup
6197 # @todo FIXME: Not tolerant to blank link text
6198 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
6199 # on how many empty links there are on the page - need to figure that out.
6200 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
6202 # Parse wikitext quotes (italics & bold)
6205 # Strip HTML tags
6208 }
6210 /**
6211 * strip/replaceVariables/unstrip for preprocessor regression testing
6212 *
6213 * @param string $text
6214 * @param Title $title
6215 * @param ParserOptions $options
6216 * @param int $outputType
6217 *
6218 * @return string
6219 */
6220 public function testSrvus( $text, Title $title, ParserOptions $options, $outputType = self::OT_HTML ) {
6228 }
6230 /**
6231 * @param string $text
6232 * @param Title $title
6233 * @param ParserOptions $options
6234 * @return string
6235 */
6238 }
6240 /**
6241 * @param string $text
6242 * @param Title $title
6243 * @param ParserOptions $options
6244 * @return string
6245 */
6248 }
6250 /**
6251 * Call a callback function on all regions of the given text that are not
6252 * inside strip markers, and replace those regions with the return value
6253 * of the callback. For example, with input:
6254 *
6255 * aaa<MARKER>bbb
6256 *
6257 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
6258 * two strings will be replaced with the value returned by the callback in
6259 * each case.
6260 *
6261 * @param string $s
6262 * @param callable $callback
6263 *
6264 * @return string
6265 */
6284 }
6285 }
6286 }
6288 }
6290 /**
6291 * Remove any strip markers found in the given text.
6292 *
6293 * @param string $text Input string
6294 * @return string
6295 */
6298 }
6300 /**
6301 * Save the parser state required to convert the given half-parsed text to
6302 * HTML. "Half-parsed" in this context means the output of
6303 * recursiveTagParse() or internalParse(). This output has strip markers
6304 * from replaceVariables (extensionSubstitution() etc.), and link
6305 * placeholders from replaceLinkHolders().
6306 *
6307 * Returns an array which can be serialized and stored persistently. This
6308 * array can later be loaded into another parser instance with
6309 * unserializeHalfParsedText(). The text can then be safely incorporated into
6310 * the return value of a parser hook.
6311 *
6312 * @param string $text
6313 *
6314 * @return array
6315 */
6322 );
6324 }
6326 /**
6327 * Load the parser state given in the $data array, which is assumed to
6328 * have been generated by serializeHalfParsedText(). The text contents is
6329 * extracted from the array, and its markers are transformed into markers
6330 * appropriate for the current Parser instance. This transformed text is
6331 * returned, and can be safely included in the return value of a parser
6332 * hook.
6333 *
6334 * If the $data array has been stored persistently, the caller should first
6335 * check whether it is still valid, by calling isValidHalfParsedText().
6336 *
6337 * @param array $data Serialized data
6338 * @throws MWException
6339 * @return string
6340 */
6344 }
6346 # First, extract the strip state.
6350 # Now renumber links
6353 # Should be good to go.
6355 }
6357 /**
6358 * Returns true if the given array, presumed to be generated by
6359 * serializeHalfParsedText(), is compatible with the current version of the
6360 * parser.
6361 *
6362 * @param array $data
6363 *
6364 * @return bool
6365 */
6368 }
6370 /**
6371 * Parsed a width param of imagelink like 300px or 200x300px
6372 *
6373 * @param string $value
6374 *
6375 * @return array
6376 * @since 1.20
6377 */
6382 }
6384 # (bug 13500) In both cases (width/height and width only),
6385 # permit trailing "px" for backward compatibility.
6394 }
6396 }
6398 /**
6399 * Lock the current instance of the parser.
6400 *
6401 * This is meant to stop someone from calling the parser
6402 * recursively and messing up all the strip state.
6403 *
6404 * @throws MWException If parser is in a parse
6405 * @return ScopedCallback The lock will be released once the return value goes out of scope.
6406 */
6411 }
6417 } );
6420 }
6422 /**
6423 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6424 *
6425 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6426 * or if there is more than one <p/> tag in the input HTML.
6427 *
6428 * @param string $html
6429 * @return string
6430 * @since 1.24
6431 */
6437 }
6438 }
6441 }
6443 /**
6444 * Return this parser if it is not doing anything, otherwise
6445 * get a fresh parser. You can use this method by doing
6446 * $myParser = $wgParser->getFreshParser(), or more simply
6447 * $wgParser->getFreshParser()->parse( ... );
6448 * if you're unsure if $wgParser is safe to use.
6449 *
6450 * @since 1.24
6451 * @return Parser A parser object that is not parsing anything
6452 */
6459 }
6460 }
6461 }