| blob | bd2f13181e22abbbfd9f403aa07748283ec0683c |
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 */
26 /**
27 * @defgroup Parser Parser
28 */
30 /**
31 * PHP Parser - Processes wiki markup (which uses a more user-friendly
32 * syntax, such as "[[link]]" for making links), and provides a one-way
33 * transformation of that wiki markup it into (X)HTML output / markup
34 * (which in turn the browser understands, and can display).
35 *
36 * There are seven main entry points into the Parser class:
37 *
38 * - Parser::parse()
39 * produces HTML output
40 * - Parser::preSaveTransform()
41 * produces altered wiki markup
42 * - Parser::preprocess()
43 * removes HTML comments and expands templates
44 * - Parser::cleanSig() and Parser::cleanSigInSig()
45 * cleans a signature before saving it to preferences
46 * - Parser::getSection()
47 * return the content of a section from an article for section editing
48 * - Parser::replaceSection()
49 * replaces a section by number inside an article
50 * - Parser::getPreloadText()
51 * removes <noinclude> sections and <includeonly> tags
52 *
53 * Globals used:
54 * object: $wgContLang
55 *
56 * @warning $wgUser or $wgTitle or $wgRequest or $wgLang. Keep them away!
57 *
58 * @par Settings:
59 * $wgNamespacesWithSubpages
60 *
61 * @par Settings only within ParserOptions:
62 * $wgAllowExternalImages
63 * $wgAllowSpecialInclusion
64 * $wgInterwikiMagic
65 * $wgMaxArticleSize
66 *
67 * @ingroup Parser
68 */
70 /**
71 * Update this version number when the ParserOutput format
72 * changes in an incompatible way, so the parser cache
73 * can automatically discard old data.
74 */
77 /**
78 * Update this version number when the output of serialiseHalfParsedText()
79 * changes in an incompatible way
80 */
83 # Flags for Parser::setFunctionHook
87 # Constants needed for external link processing
88 # Everything except bracket, space, or control characters
89 # \p{Zs} is unicode 'separator, space' category. It covers the space 0x20
90 # as well as U+3000 is IDEOGRAPHIC SPACE for bug 19052
92 # Simplified expression to match an IPv4 or IPv6 address, or
93 # at least one character of a host name (embeds EXT_LINK_URL_CLASS)
95 # RegExp to make image URLs (embeds IPv6 part of EXT_LINK_ADDR)
96 // @codingStandardsIgnoreStart Generic.Files.LineLength
97 const EXT_IMAGE_REGEX = '/^(http:\/\/|https:\/\/)((?:\\[(?i:[0-9a-f:.]+)\\])?[^][<>"\\x00-\\x20\\x7F\p{Zs}]+)
99 // @codingStandardsIgnoreEnd
101 # Regular expression for a non-newline space
104 # Flags for preprocessToDom
107 # Allowed values for $this->mOutputType
108 # Parameter to startExternalParse().
115 /**
116 * @var string Prefix and suffix for temporary replacement strings
117 * for the multipass parser.
118 *
119 * \x7f should never appear in input as it's disallowed in XML.
120 * Using it at the front also gives us a little extra robustness
121 * since it shouldn't match when butted up against identifier-like
122 * string constructs.
123 *
124 * Must not consist of all title characters, or else it will change
125 * the behavior of <nowiki> in a link.
126 *
127 * Must have a character that needs escaping in attributes, otherwise
128 * someone could put a strip marker in an attribute, to get around
129 * escaping quote marks, and break out of the attribute. Thus we add
130 * `'".
131 */
135 # Markers used for wrapping the table of contents
139 # Persistent:
153 # Initialised by initialiseVariables()
155 /**
156 * @var MagicWordArray
157 */
160 /**
161 * @var MagicWordArray
162 */
164 # Initialised in constructor
167 # Initialized in getPreprocessor()
168 /** @var Preprocessor */
171 # Cleared with clearState():
172 /**
173 * @var ParserOutput
174 */
178 /**
179 * @var StripState
180 */
184 /**
185 * @var LinkHolderArray
186 */
196 /**
197 * @var User
198 */
201 # Temporary
202 # These are variables reset at least once per parse regardless of $clearState
204 /**
205 * @var ParserOptions
206 */
209 /**
210 * @var Title
211 */
223 /**
224 * @var string Deprecated accessor for the strip marker prefix.
225 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
226 **/
229 /**
230 * @var array Array with the language name of each language link (i.e. the
231 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
232 * duplicate language links to the ParserOutput.
233 */
236 /**
237 * @var MapCacheLRU|null
238 * @since 1.24
239 *
240 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
241 */
244 /**
245 * @var bool Recursive call protection.
246 * This variable should be treated as if it were private.
247 */
250 /** @var SectionProfiler */
253 /**
254 * @var LinkRenderer
255 */
258 /**
259 * @param array $conf
260 */
270 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
273 # PECL extension that conflicts with the core DOM extension (bug 13770)
280 }
282 }
284 /**
285 * Reduce memory usage to reduce the impact of circular references
286 */
290 }
293 }
294 }
296 /**
297 * Allow extensions to clean up when the parser is cloned
298 */
302 // Bug 56226: When you create a reference "to" an object field, that
303 // makes the object field itself be a reference too (until the other
304 // reference goes out of scope). When cloning, any field that's a
305 // reference is copied as a reference in the new object. Both of these
306 // are defined PHP5 behaviors, as inconvenient as it is for us when old
307 // hooks from PHP4 days are passing fields by reference.
309 // Make a non-reference copy of the field, then rebind the field to
310 // reference the new copy.
314 }
317 }
319 /**
320 * Do various kinds of initialisation on the first call of the parser
321 */
325 }
333 }
335 /**
336 * Clear Parser state
337 *
338 * @private
339 */
343 }
359 # Clear these on every parse, bug 4549
367 ];
376 # Fix cloning
379 }
384 }
386 /**
387 * Convert wikitext to HTML
388 * Do not call this function recursively.
389 *
390 * @param string $text Text we want to parse
391 * @param Title $title
392 * @param ParserOptions $options
393 * @param bool $linestart
394 * @param bool $clearState
395 * @param int $revid Number to pass in {{REVISIONID}}
396 * @return ParserOutput A ParserOutput
397 */
400 ) {
401 /**
402 * First pass--just handle <nowiki> sections, pass the rest off
403 * to internalParse() which does all the real work.
404 */
409 // We use U+007F DELETE to construct strip markers, so we have to make
410 // sure that this character does not occur in the input text.
413 }
421 }
434 }
437 # No more strip!
444 /**
445 * A converted title will be provided in the output object if title and
446 * content conversion are enabled, the article text does not contain
447 * a conversion-suppressing double-underscore tag, and no
448 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
449 * automatic link conversion.
450 */
455 ) {
462 }
463 }
469 );
470 }
472 # Information on include size limits, for the benefit of users who try to skirt them
480 );
481 }
486 );
490 );
493 );
496 );
499 );
502 );
505 );
511 }
521 ) ) {
527 }
531 }
532 }
533 }
534 // Since we're not really outputting HTML, decode the entities and
535 // then re-encode the things that need hiding inside HTML comments.
539 // Sanitize for comment. Note '‐' in the replacement is U+2010,
540 // which looks much like the problematic '-'.
544 // Add on template profiling data
548 } );
554 }
560 }
561 }
573 }
575 /**
576 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
577 * can be called from an extension tag hook.
578 *
579 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
580 * instead, which means that lists and links have not been fully parsed yet,
581 * and strip markers are still present.
582 *
583 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
584 *
585 * Use this function if you're a parser tag hook and you want to parse
586 * wikitext before or after applying additional transformations, and you
587 * intend to *return the result as hook output*, which will cause it to go
588 * through the rest of parsing process automatically.
589 *
590 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
591 * $text are not expanded
592 *
593 * @param string $text Text extension wants to have parsed
594 * @param bool|PPFrame $frame The frame to use for expanding any template variables
595 * @return string UNSAFE half-parsed HTML
596 */
602 }
604 /**
605 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
606 * point can be called from an extension tag hook.
607 *
608 * The output of this function is fully-parsed HTML that is safe for output.
609 * If you're a parser tag hook, you might want to use recursiveTagParse()
610 * instead.
611 *
612 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
613 * $text are not expanded
614 *
615 * @since 1.25
616 *
617 * @param string $text Text extension wants to have parsed
618 * @param bool|PPFrame $frame The frame to use for expanding any template variables
619 * @return string Fully parsed HTML
620 */
625 }
627 /**
628 * Expand templates and variables in the text, producing valid, static wikitext.
629 * Also removes comments.
630 * Do not call this function recursively.
631 * @param string $text
632 * @param Title $title
633 * @param ParserOptions $options
634 * @param int|null $revid
635 * @param bool|PPFrame $frame
636 * @return mixed|string
637 */
640 ) {
645 }
651 }
653 /**
654 * Recursive parser entry point that can be called from an extension tag
655 * hook.
656 *
657 * @param string $text Text to be expanded
658 * @param bool|PPFrame $frame The frame to use for expanding any template variables
659 * @return string
660 * @since 1.19
661 */
666 }
668 /**
669 * Process the wikitext for the "?preload=" feature. (bug 5210)
670 *
671 * "<noinclude>", "<includeonly>" etc. are parsed as for template
672 * transclusion, comments, templates, arguments, tags hooks and parser
673 * functions are untouched.
674 *
675 * @param string $text
676 * @param Title $title
677 * @param ParserOptions $options
678 * @param array $params
679 * @return string
680 */
685 # Parser (re)initialisation
694 }
696 /**
697 * Get a random string
698 *
699 * @return string
700 * @deprecated since 1.26; use wfRandomString() instead.
701 */
705 }
707 /**
708 * Set the current user.
709 * Should only be used when doing pre-save transform.
710 *
711 * @param User|null $user User object or null (to reset)
712 */
715 }
717 /**
718 * Accessor for mUniqPrefix.
719 *
720 * @return string
721 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
722 */
726 }
728 /**
729 * Set the context title
730 *
731 * @param Title $t
732 */
736 }
739 # Strip the fragment to avoid various odd effects
743 }
744 }
746 /**
747 * Accessor for the Title object
748 *
749 * @return Title
750 */
753 }
755 /**
756 * Accessor/mutator for the Title object
757 *
758 * @param Title $x Title object or null to just get the current one
759 * @return Title
760 */
763 }
765 /**
766 * Set the output type
767 *
768 * @param int $ot New value
769 */
772 # Shortcut alias
778 ];
779 }
781 /**
782 * Accessor/mutator for the output type
783 *
784 * @param int|null $x New value or null to just get the current one
785 * @return int
786 */
789 }
791 /**
792 * Get the ParserOutput object
793 *
794 * @return ParserOutput
795 */
798 }
800 /**
801 * Get the ParserOptions object
802 *
803 * @return ParserOptions
804 */
807 }
809 /**
810 * Accessor/mutator for the ParserOptions object
811 *
812 * @param ParserOptions $x New value or null to just get the current one
813 * @return ParserOptions Current ParserOptions object
814 */
817 }
819 /**
820 * @return int
821 */
824 }
826 /**
827 * @param int $id
828 */
831 }
833 /**
834 * Get a language object for use in parser functions such as {{FORMATNUM:}}
835 * @return Language
836 */
839 }
841 /**
842 * Get the target language for the content being parsed. This is usually the
843 * language that the content is in.
844 *
845 * @since 1.19
846 *
847 * @throws MWException
848 * @return Language
849 */
859 }
862 }
864 /**
865 * Get the language object for language conversion
866 * @return Language|null
867 */
870 }
872 /**
873 * Get a User object either from $this->mUser, if set, or from the
874 * ParserOptions object otherwise
875 *
876 * @return User
877 */
881 }
883 }
885 /**
886 * Get a preprocessor object
887 *
888 * @return Preprocessor
889 */
894 }
896 }
898 /**
899 * Get a LinkRenderer instance to make links with
900 *
901 * @since 1.28
902 * @return LinkRenderer
903 */
910 );
911 }
914 }
916 /**
917 * Replaces all occurrences of HTML-style comments and the given tags
918 * in the text with a random marker and returns the next text. The output
919 * parameter $matches will be an associative array filled with data in
920 * the form:
921 *
922 * @code
923 * 'UNIQ-xxxxx' => array(
924 * 'element',
925 * 'tag content',
926 * array( 'param' => 'x' ),
927 * '<element param="x">tag content</element>' ) )
928 * @endcode
929 *
930 * @param array $elements List of element names. Comments are always extracted.
931 * @param string $text Source text string.
932 * @param array $matches Out parameter, Array: extracted tags
933 * @param string|null $uniq_prefix
934 * @return string Stripped text
935 * @since 1.26 The uniq_prefix argument is deprecated.
936 */
937 public static function extractTagsAndParams( $elements, $text, &$matches, $uniq_prefix = null ) {
940 }
953 }
955 # comment
961 # tag
966 }
972 # Empty element tag, <tag />
981 }
985 # No end tag -- let it run out to the end of the text.
991 }
992 }
998 }
1000 }
1002 /**
1003 * Get a list of strippable XML-like elements
1004 *
1005 * @return array
1006 */
1009 }
1011 /**
1012 * Add an item to the strip state
1013 * Returns the unique tag which must be inserted into the stripped text
1014 * The tag will be replaced with the original text in unstrip()
1015 *
1016 * @param string $text
1017 *
1018 * @return string
1019 */
1025 }
1027 /**
1028 * parse the wiki syntax used to render tables
1029 *
1030 * @private
1031 * @param string $text
1032 * @return string
1033 */
1051 }
1058 # First check if we are starting a new table
1071 # Don't do any of the following
1075 # We are ending a table
1081 }
1085 }
1089 }
1093 # Now we have a table row
1096 # Whats after the tag is now only attributes
1109 }
1113 }
1122 ) {
1123 # This might be cell elements, td, th or captions
1129 }
1131 // Implies both are valid for table headings.
1134 }
1136 # Split up multiple cells on the same line.
1137 # FIXME : This can result in improper nesting of tags processed
1138 # by earlier parser steps.
1143 # Loop through each table cell
1150 }
1155 }
1161 }
1171 }
1175 # A cell could contain both parameters and data
1178 # Bug 553: Note that a '|' inside an invalid link should not
1179 # be mistaken as delimiting cell parameters
1188 }
1192 }
1193 }
1195 }
1197 # Closing open td, tr && table
1201 }
1204 }
1207 }
1210 }
1212 # Remove trailing line-ending (b/c)
1215 }
1217 # special case: don't return empty table
1220 }
1223 }
1225 /**
1226 * Helper function for parse() that transforms wiki markup into half-parsed
1227 * HTML. Only called for $mOutputType == self::OT_HTML.
1228 *
1229 * @private
1230 *
1231 * @param string $text The text to parse
1232 * @param bool $isMain Whether this is being called from the main parse() function
1233 * @param PPFrame|bool $frame A pre-processor frame
1234 *
1235 * @return string
1236 */
1241 # Hook to suspend the parser in this state
1244 }
1246 # if $frame is provided, then use $frame for replacing any variables
1248 # use frame depth to infer how include/noinclude tags should be handled
1249 # depth=0 means this is the top-level document; otherwise it's an included document
1254 }
1258 # if $frame is not provided, then use old-style replaceVariables
1260 }
1268 );
1271 # Tables need to come after variable replacement for things to work
1272 # properly; putting them before other transformations should keep
1273 # exciting things like link expansions from showing up in surprising
1274 # places.
1286 # replaceInternalLinks may sometimes leave behind
1287 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1294 }
1296 /**
1297 * Helper function for parse() that transforms half-parsed HTML into fully
1298 * parsed HTML.
1299 *
1300 * @param string $text
1301 * @param bool $isMain
1302 * @param bool $linestart
1303 * @return string
1304 */
1310 }
1312 # Clean up special characters, only run once, next-to-last before doBlockLevels
1314 # french spaces, last one Guillemet-left
1315 # only if there is something before the space
1317 # french spaces, Guillemet-right
1320 ];
1327 /**
1328 * The input doesn't get language converted if
1329 * a) It's disabled
1330 * b) Content isn't converted
1331 * c) It's a conversion table
1332 * d) it is an interface message (which is in the user language)
1333 */
1336 ) {
1338 # The position of the convert() call should not be changed. it
1339 # assumes that the links are all replaced and the only thing left
1340 # is the <nowiki> mark.
1342 }
1343 }
1349 }
1360 # attempt to sanitize at least some nesting problems
1361 # (bug #2702 and quite a few others)
1363 # ''Something [http://www.cool.com cool''] -->
1364 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1367 # fix up an anchor inside another anchor, only
1368 # at least for a single single nested link (bug 3695)
1371 # fix div inside inline elements- doBlockLevels won't wrap a line which
1372 # contains a div, so fix it up here; replace
1373 # div with escaped text
1376 # remove empty italic or bold tag pairs, some
1377 # introduced by rules above
1379 ];
1385 }
1389 }
1392 }
1394 /**
1395 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1396 * magic external links.
1397 *
1398 * DML
1399 * @private
1400 *
1401 * @param string $text
1402 *
1403 * @return string
1404 */
1413 '!(?: # Start cases
1415 (<.*?>) | # m[2]: Skip stuff inside
1418 # m[4]: Post-protocol path
1424 [0-9Xx] # check digit
1428 }
1430 /**
1431 * @throws MWException
1432 * @param array $m
1433 * @return HTML|string
1434 */
1437 # Skip anchor
1440 # Skip HTML element
1443 # Free external link
1446 # RFC or PMID
1460 }
1462 return Linker::makeExternalLink( $url, "{$keyword} {$id}", true, $cssClass, [], $this->mTitle );
1464 # ISBN
1472 ] );
1476 [
1479 ]
1480 );
1483 }
1484 }
1486 /**
1487 * Make a free external link, given a user-supplied URL
1488 *
1489 * @param string $url
1490 * @param int $numPostProto
1491 * The number of characters after the protocol.
1492 * @return string HTML
1493 * @private
1494 */
1498 # The characters '<' and '>' (which were escaped by
1499 # removeHTMLtags()) should not be included in
1500 # URLs, per RFC 2396.
1501 # Make terminate a URL as well (bug T84937)
1507 PREG_OFFSET_CAPTURE
1508 ) ) {
1511 }
1513 # Move trailing punctuation to $trail
1515 # If there is no left bracket, then consider right brackets fair game too
1518 }
1522 # Don't break a trailing HTML entity by moving the ; into $trail
1523 # This is in hot code, so use substr_compare to avoid having to
1524 # create a new string object for the comparison
1526 # more optimization: instead of running preg_match with a $
1527 # anchor, which can be slow, do the match on the reversed
1528 # string starting at the desired offset.
1529 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1532 }
1533 }
1537 }
1539 # Verify that we still have a real URL after trail removal, and
1540 # not just lone protocol
1543 }
1547 # Is this an external image?
1550 # Not an image, make a link
1555 # Register it in the output object...
1556 # Replace unnecessary URL escape codes with their equivalent characters
1559 }
1561 }
1563 /**
1564 * Parse headers and return html
1565 *
1566 * @private
1567 *
1568 * @param string $text
1569 *
1570 * @return string
1571 */
1576 }
1578 }
1580 /**
1581 * Replace single quotes with HTML markup
1582 * @private
1583 *
1584 * @param string $text
1585 *
1586 * @return string The altered text
1587 */
1593 }
1596 }
1598 /**
1599 * Helper function for doAllQuotes()
1600 *
1601 * @param string $text
1602 *
1603 * @return string
1604 */
1610 }
1612 // First, do some preliminary work. This may shift some apostrophes from
1613 // being mark-up to being text. It also counts the number of occurrences
1614 // of bold and italics mark-ups.
1619 // If there are ever four apostrophes, assume the first is supposed to
1620 // be text, and the remaining three constitute mark-up for bold text.
1621 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1627 // If there are more than 5 apostrophes in a row, assume they're all
1628 // text except for the last 5.
1629 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1633 }
1634 // Count the number of occurrences of bold and italics mark-ups.
1642 }
1643 }
1645 // If there is an odd number of both bold and italics, it is likely
1646 // that one of the bold ones was meant to be an apostrophe followed
1647 // by italics. Which one we cannot know for certain, but it is more
1648 // likely to be one that has a single-letter word before it.
1660 }
1663 // if $firstsingleletterword is set, we don't
1664 // look at the other options, so we can bail early.
1669 }
1670 }
1671 }
1672 }
1674 // If there is a single-letter word, use it!
1679 // If not, but there's a multi-letter word, use that one.
1683 // ... otherwise use the first one that has neither.
1684 // (notice that it is possible for all three to be -1 if, for example,
1685 // there is only one pentuple-apostrophe in the line)
1688 }
1689 }
1691 // Now let's actually convert our apostrophic mush to HTML!
1702 }
1721 }
1738 }
1758 }
1759 }
1760 }
1762 }
1763 // Now close all remaining tags. Notice that the order is important.
1766 }
1769 }
1772 }
1773 // There might be lonely ''''', so make sure we have a buffer
1776 }
1778 }
1780 /**
1781 * Replace external links (REL)
1782 *
1783 * Note: this is all very hackish and the order of execution matters a lot.
1784 * Make sure to run tests/parserTests.php if you change this code.
1785 *
1786 * @private
1787 *
1788 * @param string $text
1789 *
1790 * @throws MWException
1791 * @return string
1792 */
1799 }
1809 # The characters '<' and '>' (which were escaped by
1810 # removeHTMLtags()) should not be included in
1811 # URLs, per RFC 2396.
1816 }
1818 # If the link text is an image URL, replace it with an <img> tag
1819 # This happened by accident in the original parser, but some people used it extensively
1823 }
1827 # Set linktype for CSS - if URL==text, link is essentially free
1830 # No link text, e.g. [http://domain.tld/some.link]
1832 # Autonumber
1837 # Have link text, e.g. [http://domain.tld/some.link text]s
1838 # Check for trail
1840 }
1846 # Use the encoded URL
1847 # This means that users can paste URLs directly into the text
1848 # Funny characters like ö aren't valid in URLs anyway
1849 # This was changed in August 2004
1853 # Register link in the output object.
1854 # Replace unnecessary URL escape codes with the referenced character
1855 # This prevents spammers from hiding links from the filters
1858 }
1861 }
1863 /**
1864 * Get the rel attribute for a particular external link.
1865 *
1866 * @since 1.21
1867 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1868 * nofollow if appropriate
1869 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1870 * @return string|null Rel attribute for $url
1871 */
1877 ) {
1879 }
1881 }
1883 /**
1884 * Get an associative array of additional HTML attributes appropriate for a
1885 * particular external link. This currently may include rel => nofollow
1886 * (depending on configuration, namespace, and the URL's domain) and/or a
1887 * target attribute (depending on configuration).
1888 *
1889 * @param string $url URL to extract the domain from for rel =>
1890 * nofollow if appropriate
1891 * @return array Associative array of HTML attributes
1892 */
1901 // T133507. New windows can navigate parent cross-origin.
1902 // Including noreferrer due to lacking browser
1903 // support of noopener. Eventually noreferrer should be removed.
1906 }
1908 }
1909 }
1912 }
1914 /**
1915 * Replace unusual escape codes in a URL with their equivalent characters
1916 *
1917 * @deprecated since 1.24, use normalizeLinkUrl
1918 * @param string $url
1919 * @return string
1920 */
1924 }
1926 /**
1927 * Replace unusual escape codes in a URL with their equivalent characters
1928 *
1929 * This generally follows the syntax defined in RFC 3986, with special
1930 * consideration for HTTP query strings.
1931 *
1932 * @param string $url
1933 * @return string
1934 */
1936 # First, make sure unsafe characters are encoded
1940 },
1941 $url
1942 );
1947 # Fragment part - 'fragment'
1953 }
1955 # Query part - 'query' minus &=+;
1961 }
1963 # Scheme and path part - 'pchar'
1964 # (we assume no userinfo or encoded colons in the host)
1969 }
1976 # Unescape it
1979 # Leave it escaped, but use uppercase for a-f
1981 }
1982 };
1984 }
1986 /**
1987 * make an image if it's allowed, either through the global
1988 * option, through the exception, or through the on-wiki whitelist
1989 *
1990 * @param string $url
1991 *
1992 * @return string
1993 */
1998 # $imagesfrom could be either a single string or an array of strings, parse out the latter
2005 }
2006 }
2011 }
2015 ) {
2017 # Image found
2019 }
2020 }
2023 ) {
2027 );
2030 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
2033 }
2035 # Image matches a whitelist entry
2038 }
2039 }
2040 }
2042 }
2044 /**
2045 * Process [[ ]] wikilinks
2046 *
2047 * @param string $s
2048 *
2049 * @return string Processed text
2050 *
2051 * @private
2052 */
2056 }
2058 /**
2059 * Process [[ ]] wikilinks (RIL)
2060 * @param string $s
2061 * @throws MWException
2062 * @return LinkHolderArray
2063 *
2064 * @private
2065 */
2070 # the % is needed to support urlencoded titles as well
2073 # Match a link having the form [[namespace:link|alternate]]trail
2075 # Match cases where there is no "]]", which might still be images
2077 }
2081 # split the entire text string on occurrences of [[
2083 # get the first element (all text up to first [[), and remove the space we added
2092 # Match the end of a line for a word that's not followed by whitespace,
2093 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2097 }
2101 }
2110 }
2113 }
2117 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2118 # Loop for each link
2120 // @codingStandardsIgnoreEnd
2122 # Check for excessive memory usage
2124 # Too big
2125 # Do the existence check, replace the link holders and clear the array
2128 }
2136 }
2137 # first link
2141 }
2142 }
2148 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2149 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2150 # the real problem is with the $e1 regex
2151 # See bug 1300.
2152 # Still some problems for cases where the ] is meant to be outside punctuation,
2153 # and no image is in sight. See bug 2095.
2157 ) {
2160 }
2161 # fix up urlencoded title texts
2163 # Should anchors '#' also be rejected?
2165 }
2168 # Invalid, but might be an image with a link in its caption
2173 }
2178 }
2182 # Don't allow internal links to pages containing
2183 # PROTO: where PROTO is a valid URL protocol; these
2184 # should be external links.
2188 }
2190 # Make subpage if necessary
2195 }
2199 # Strip off leading ':'
2201 }
2208 }
2217 # look at the next 'line' to see if we can close it there
2222 }
2225 # the first ]] closes the inner link, the second the image
2231 # if there's exactly one ]] that's fine, we'll keep looking
2234 # if $next_line is invalid too, we need look no further
2237 }
2238 }
2240 # we couldn't find the end of this imageLink, so output it raw
2241 # but don't ignore what might be perfectly normal links in the text we've examined
2244 # note: no $trail, because without an end, there *is* no trail
2246 }
2249 # note: no $trail, because without an end, there *is* no trail
2251 }
2252 }
2258 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2259 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2260 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2261 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2263 }
2265 # Link not escaped by : , create the various objects
2267 # Interwikis
2272 )
2273 ) {
2274 # Bug 24502: filter duplicates
2278 }
2283 }
2288 # if no parameters were passed, $text
2289 # becomes something like "File:Foo.png",
2290 # which we don't want to pass on to the
2291 # image generator
2294 # recursively parse links inside the image caption
2295 # actually, this will parse them in any other parameters, too,
2296 # but it might be hard to fix that, and it doesn't matter ATM
2299 }
2300 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2304 }
2312 }
2318 /**
2319 * Strip the whitespace Category links produce, see bug 87
2320 */
2324 }
2325 }
2327 # Self-link checking. For some languages, variants of the title are checked in
2328 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2329 # for linking to a different variant.
2333 }
2335 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2336 # @todo FIXME: Should do batch file existence checks, see comment below
2338 # Give extensions a chance to select the file revision for us
2343 # Fetch and register the file (file title may be different via hooks)
2345 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2349 }
2351 # Some titles, such as valid special pages or files in foreign repos, should
2352 # be shown as bluelinks even though they're not included in the page table
2353 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2354 # batch file existence checks for NS_FILE and NS_MEDIA
2359 # Links will be added to the output link list after checking
2361 }
2362 }
2364 }
2366 /**
2367 * Render a forced-blue link inline; protect against double expansion of
2368 * URLs if we're in a mode that prepends full URL prefixes to internal links.
2369 * Since this little disaster has to split off the trail text to avoid
2370 * breaking URLs in the following text without breaking trails on the
2371 * wiki links, it's been made into a horrible function.
2372 *
2373 * @param Title $nt
2374 * @param string $text
2375 * @param string $trail
2376 * @param string $prefix
2377 * @return string HTML-wikitext mix oh yuck
2378 */
2384 }
2388 );
2391 }
2393 /**
2394 * Insert a NOPARSE hacky thing into any inline links in a chunk that's
2395 * going to go through further parsing steps before inline URL expansion.
2396 *
2397 * Not needed quite as much as it used to be since free links are a bit
2398 * more sensible these days. But bracketed links are still an issue.
2399 *
2400 * @param string $text More-or-less HTML
2401 * @return string Less-or-more HTML with NOPARSE bits
2402 */
2406 }
2408 /**
2409 * Return true if subpage links should be expanded on this page.
2410 * @return bool
2411 */
2413 # Some namespaces don't allow subpages
2415 }
2417 /**
2418 * Handle link to subpage if necessary
2419 *
2420 * @param string $target The source of the link
2421 * @param string &$text The link text, modified as necessary
2422 * @return string The full name of the link
2423 * @private
2424 */
2427 }
2429 /**
2430 * Make lists from lines starting with ':', '*', '#', etc. (DBL)
2431 *
2432 * @param string $text
2433 * @param bool $linestart Whether or not this is at the start of a line.
2434 * @private
2435 * @return string The lists rendered as HTML
2436 */
2439 }
2441 /**
2442 * Return value of a magic variable (like PAGENAME)
2443 *
2444 * @private
2445 *
2446 * @param int $index
2447 * @param bool|PPFrame $frame
2448 *
2449 * @throws MWException
2450 * @return string
2451 */
2457 // If no title set, bad things are going to happen
2458 // later. Title should always be set since this
2459 // should only be called in the middle of a parse
2460 // operation (but the unit-tests do funky stuff)
2463 }
2465 /**
2466 * Some of these require message or data lookups and can be
2467 * expensive to check many times.
2468 */
2472 }
2473 }
2518 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
2552 ) ) );
2562 ) ) );
2570 }
2578 }
2591 # 0 means the page doesn't exist in the database,
2592 # which means the user is previewing a new page.
2593 # The vary-revision flag must be set, because the magic word
2594 # will have a different value once the page is saved.
2597 }
2601 # Let the edit saving system know we should parse the page
2602 # *after* a revision ID has been assigned.
2609 }
2612 # Let the edit saving system know we should parse the page
2613 # *after* a revision ID has been assigned. This is for null edits.
2619 # Let the edit saving system know we should parse the page
2620 # *after* a revision ID has been assigned. This is for null edits.
2626 # Let the edit saving system know we should parse the page
2627 # *after* a revision ID has been assigned. This is for null edits.
2633 # Let the edit saving system know we should parse the page
2634 # *after* a revision ID has been assigned. This is for null edits.
2640 # Let the edit saving system know we should parse the page
2641 # *after* a revision ID has been assigned. This is for null edits.
2647 # Let the edit saving system know we should parse the page
2648 # *after* a revision ID has been assigned. This is for null edits.
2654 # Let the edit saving system know we should parse the page
2655 # *after* a revision ID has been assigned for null edits.
2699 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2700 # int to remove the padding
2709 );
2718 false
2719 );
2725 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
2726 # int to remove the padding
2787 );
2790 }
2794 }
2797 }
2799 /**
2800 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
2801 *
2802 * @private
2803 */
2810 }
2812 /**
2813 * Preprocess some wikitext and return the document tree.
2814 * This is the ghost of replace_variables().
2815 *
2816 * @param string $text The text to parse
2817 * @param int $flags Bitwise combination of:
2818 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
2819 * included. Default is to assume a direct page view.
2820 *
2821 * The generated DOM tree must depend only on the input text and the flags.
2822 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
2823 *
2824 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
2825 * change in the DOM tree for a given text, must be passed through the section identifier
2826 * in the section edit link and thus back to extractSections().
2827 *
2828 * The output of this function is currently only cached in process memory, but a persistent
2829 * cache may be implemented at a later date which takes further advantage of these strict
2830 * dependency requirements.
2831 *
2832 * @return PPNode
2833 */
2837 }
2839 /**
2840 * Return a three-element array: leading whitespace, string contents, trailing whitespace
2841 *
2842 * @param string $s
2843 *
2844 * @return array
2845 */
2855 }
2857 }
2859 /**
2860 * Replace magic variables, templates, and template arguments
2861 * with the appropriate text. Templates are substituted recursively,
2862 * taking care to avoid infinite loops.
2863 *
2864 * Note that the substitution depends on value of $mOutputType:
2865 * self::OT_WIKI: only {{subst:}} templates
2866 * self::OT_PREPROCESS: templates but not extension tags
2867 * self::OT_HTML: all templates and extension tags
2868 *
2869 * @param string $text The text to transform
2870 * @param bool|PPFrame $frame Object describing the arguments passed to the
2871 * template. Arguments may also be provided as an associative array, as
2872 * was the usual case before MW1.12. Providing arguments this way may be
2873 * useful for extensions wishing to perform variable replacement
2874 * explicitly.
2875 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
2876 * double-brace expansion.
2877 * @return string
2878 */
2880 # Is there any text? Also, Prevent too big inclusions!
2884 }
2892 }
2899 }
2901 /**
2902 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
2903 *
2904 * @param array $args
2905 *
2906 * @return array
2907 */
2920 }
2923 }
2924 }
2925 }
2928 }
2930 /**
2931 * Warn the user when a parser limitation is reached
2932 * Will warn at most once the user per limitation type
2933 *
2934 * The results are shown during preview and run through the Parser (See EditPage.php)
2935 *
2936 * @param string $limitationType Should be one of:
2937 * 'expensive-parserfunction' (corresponding messages:
2938 * 'expensive-parserfunction-warning',
2939 * 'expensive-parserfunction-category')
2940 * 'post-expand-template-argument' (corresponding messages:
2941 * 'post-expand-template-argument-warning',
2942 * 'post-expand-template-argument-category')
2943 * 'post-expand-template-inclusion' (corresponding messages:
2944 * 'post-expand-template-inclusion-warning',
2945 * 'post-expand-template-inclusion-category')
2946 * 'node-count-exceeded' (corresponding messages:
2947 * 'node-count-exceeded-warning',
2948 * 'node-count-exceeded-category')
2949 * 'expansion-depth-exceeded' (corresponding messages:
2950 * 'expansion-depth-exceeded-warning',
2951 * 'expansion-depth-exceeded-category')
2952 * @param string|int|null $current Current value
2953 * @param string|int|null $max Maximum allowed, when an explicit limit has been
2954 * exceeded, provide the values (optional)
2955 */
2957 # does no harm if $current and $max are present but are unnecessary for the message
2958 # Not doing ->inLanguage( $this->mOptions->getUserLangObj() ), since this is shown
2959 # only during preview, and that would split the parser cache unnecessarily.
2964 }
2966 /**
2967 * Return the text of a template, after recursively
2968 * replacing any variables or templates within the template.
2969 *
2970 * @param array $piece The parts of the template
2971 * $piece['title']: the title, i.e. the part before the |
2972 * $piece['parts']: the parameter array
2973 * $piece['lineStart']: whether the brace was at the start of a line
2974 * @param PPFrame $frame The current frame, contains template arguments
2975 * @throws Exception
2976 * @return string The text of the template
2977 */
2980 // Flags
2982 // $text has been filled
2984 // wiki markup in $text should be escaped
2986 // $text is HTML, armour it against wikitext transformation
2988 // Force interwiki transclusion to be done in raw mode not rendered
2990 // $text is a DOM node needing expansion in a child frame
2992 // $text is a DOM node needing expansion in the current frame
2995 # Title object, where $text came from
2998 # $part1 is the bit before the first |, and must contain only title characters.
2999 # Various prefixes will be stripped from it later.
3004 # Original title text preserved for various purposes
3007 # $args is a list of argument nodes, starting from index 0, not including $part1
3008 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3009 # below won't work b/c this $args isn't an object
3014 # SUBST
3018 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3019 # Decide whether to expand template or keep wikitext as-is.
3025 }
3031 }
3032 }
3037 }
3038 }
3040 # Variables
3047 }
3049 }
3050 }
3052 # MSG, MSGNW and RAW
3054 # Check for MSGNW:
3059 # Remove obsolete MSG:
3062 }
3064 # Check for RAW:
3068 }
3069 }
3071 # Parser functions
3080 }
3085 }
3087 # The interface for parser functions allows for extracting
3088 # flags into the local scope. Extract any forwarded flags
3089 # here.
3091 }
3092 }
3094 # Finish mangling title and then check for loops.
3095 # Set $title to a Title object and $titleText to the PDBK
3098 # Split the title into page and subpage
3104 }
3108 # Check for language variants if the template is not found
3111 }
3112 # Do recursion depth check
3120 }
3121 }
3122 }
3124 # Load from database
3131 ) {
3133 // Pass the template arguments as URL parameters.
3134 // "uselang" will have no effect since the Language object
3135 // is forced to the one defined in ParserOptions.
3144 }
3145 }
3147 // Create a new context to execute the special page
3154 // If this page is cached, then we better not be per user.
3156 }
3166 }
3167 }
3177 }
3178 }
3180 # If the title is valid but undisplayable, make a link to it
3184 }
3186 # Interwiki transclusion
3192 # Preprocess it like a template
3195 }
3197 }
3199 # Do infinite loop check
3200 # This has to be done after redirect resolution to avoid infinite loops via redirects
3207 }
3208 }
3210 # If we haven't found text to substitute by now, we're done
3211 # Recover the source wikitext and return it
3216 }
3218 }
3220 # Expand DOM-style return values in a child frame
3222 # Clean up argument array
3228 # Expansion is eligible for the empty-frame cache
3231 # Uncached expansion
3233 }
3234 }
3238 }
3242 }
3244 # Replace raw HTML by a placeholder
3248 # Escape nowiki-style return values
3253 ) {
3254 # Bug 529: if the template begins with a table or block-level
3255 # element, it should be treated as beginning a new line.
3256 # This behavior is somewhat controversial.
3258 }
3261 # Error, oversize inclusion
3263 # Make a working, properly escaped link if possible (bug 23588)
3266 # This will probably not be a working link, but at least it may
3267 # provide some hint of where the problem is
3270 }
3274 }
3280 }
3283 }
3285 /**
3286 * Call a parser function and return an array with text and flags.
3287 *
3288 * The returned array will always contain a boolean 'found', indicating
3289 * whether the parser function was found or not. It may also contain the
3290 * following:
3291 * text: string|object, resulting wikitext or PP DOM object
3292 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3293 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3294 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3295 * nowiki: bool, wiki markup in $text should be escaped
3296 *
3297 * @since 1.21
3298 * @param PPFrame $frame The current frame, contains template arguments
3299 * @param string $function Function name
3300 * @param array $args Arguments to the function
3301 * @throws MWException
3302 * @return array
3303 */
3307 # Case sensitive functions
3311 # Case insensitive functions
3317 }
3318 }
3322 # Workaround for PHP bug 35229 and similar
3325 }
3329 # Convert arguments to PPNodes and collect for appending to $allArgs
3336 }
3337 }
3339 # Add a frame parameter, and pass the arguments as an array
3343 # Convert arguments to plain text and append to $allArgs
3351 }
3352 }
3353 }
3357 # The interface for function hooks allows them to return a wikitext
3358 # string or an array containing the string and any flags. This mungs
3359 # things around to match what this method should return.
3364 ];
3368 }
3372 ];
3373 }
3379 }
3382 }
3387 }
3390 }
3392 /**
3393 * Get the semi-parsed DOM representation of a template with a given title,
3394 * and its redirect destination title. Cached.
3395 *
3396 * @param Title $title
3397 *
3398 * @return array
3399 */
3408 }
3411 }
3413 # Cache miss, go to the database
3419 }
3427 }
3430 }
3432 /**
3433 * Fetch the current revision of a given title. Note that the revision
3434 * (and even the title) may not exist in the database, so everything
3435 * contributing to the output of the parser should use this method
3436 * where possible, rather than getting the revisions themselves. This
3437 * method also caches its results, so using it benefits performance.
3438 *
3439 * @since 1.24
3440 * @param Title $title
3441 * @return Revision
3442 */
3447 }
3450 // Defaults to Parser::statelessFetchRevision()
3452 );
3453 }
3455 }
3457 /**
3458 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3459 * without passing them on to it.
3460 *
3461 * @since 1.24
3462 * @param Title $title
3463 * @param Parser|bool $parser
3464 * @return Revision
3465 */
3468 }
3470 /**
3471 * Fetch the unparsed text of a template and register a reference to it.
3472 * @param Title $title
3473 * @return array ( string or false, Title )
3474 */
3476 // Defaults to Parser::statelessFetchTemplate()
3479 // We use U+007F DELETE to distinguish strip markers from regular text.
3483 }
3489 // If we transclude ourselves, the final result
3490 // will change based on the new version of the page
3492 }
3493 }
3494 }
3496 }
3498 /**
3499 * Fetch the unparsed text of a template and register a reference to it.
3500 * @param Title $title
3501 * @return string|bool
3502 */
3505 }
3507 /**
3508 * Static function to get a template
3509 * Can be overridden via ParserOptions::setTemplateCallback().
3510 *
3511 * @param Title $title
3512 * @param bool|Parser $parser
3513 *
3514 * @return array
3515 */
3521 # Loop to fetch the article, with up to 1 redirect
3522 // @codingStandardsIgnoreStart Generic.CodeAnalysis.ForLoopWithTestFunctionCall.NotAllowed
3524 // @codingStandardsIgnoreEnd
3525 # Give extensions a chance to select the revision instead
3536 ];
3538 }
3539 # Get the revision
3546 }
3548 # If there is no current revision, there is no page
3552 }
3559 # We fetched a rev from a different title; register it too...
3564 }
3573 }
3580 }
3585 }
3588 }
3589 # Redirect?
3592 }
3597 }
3599 /**
3600 * Fetch a file and its title and register a reference to it.
3601 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
3602 * @param Title $title
3603 * @param array $options Array of options to RepoGroup::findFile
3604 * @return File|bool
3605 */
3608 }
3610 /**
3611 * Fetch a file and its title and register a reference to it.
3612 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
3613 * @param Title $title
3614 * @param array $options Array of options to RepoGroup::findFile
3615 * @return array ( File or false, Title of file )
3616 */
3622 # Register the file as a dependency...
3625 # Update fetched file title
3628 }
3630 }
3632 /**
3633 * Helper function for fetchFileAndTitle.
3634 *
3635 * Also useful if you need to fetch a file but not use it yet,
3636 * for example to get the file's handler.
3637 *
3638 * @param Title $title
3639 * @param array $options Array of options to RepoGroup::findFile
3640 * @return File|bool
3641 */
3649 }
3651 }
3653 /**
3654 * Transclude an interwiki link.
3655 *
3656 * @param Title $title
3657 * @param string $action
3658 *
3659 * @return string
3660 */
3666 }
3672 }
3674 }
3676 /**
3677 * @param string $url
3678 * @return mixed|string
3679 */
3688 }
3695 // Though we failed to fetch the content, this status is useless.
3700 }
3707 ] );
3709 }
3711 /**
3712 * Triple brace replacement -- used for template arguments
3713 * @private
3714 *
3715 * @param array $piece
3716 * @param PPFrame $frame
3717 *
3718 * @return array
3719 */
3732 )
3733 ) {
3734 # No match in frame, use the supplied default
3736 }
3740 }
3743 # No match anywhere
3745 }
3748 }
3753 }
3756 }
3758 /**
3759 * Return the text to be used for a given extension tag.
3760 * This is the ghost of strip().
3761 *
3762 * @param array $params Associative array of parameters:
3763 * name PPNode for the tag name
3764 * attr PPNode for unparsed text where tag attributes are thought to be
3765 * attributes Optional associative array of parsed attributes
3766 * inner Contents of extension element
3767 * noClose Original text did not have a close tag
3768 * @param PPFrame $frame
3769 *
3770 * @throws MWException
3771 * @return string
3772 */
3786 }
3792 }
3795 # Workaround for PHP bug 35229 and similar
3798 }
3805 }
3811 }
3814 # Extract flags to local scope (to override $markerType)
3819 }
3823 }
3828 }
3829 }
3835 }
3836 }
3846 }
3848 }
3850 /**
3851 * Increment an include size counter
3852 *
3853 * @param string $type The type of expansion
3854 * @param int $size The size of the text
3855 * @return bool False if this inclusion would take it over the maximum, true otherwise
3856 */
3863 }
3864 }
3866 /**
3867 * Increment the expensive function count
3868 *
3869 * @return bool False if the limit has been exceeded
3870 */
3874 }
3876 /**
3877 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
3878 * Fills $this->mDoubleUnderscores, returns the modified text
3879 *
3880 * @param string $text
3881 *
3882 * @return string
3883 */
3886 # The position of __TOC__ needs to be recorded
3892 # Set a placeholder. At the end we'll fill it in with the TOC.
3895 # Only keep the first one.
3897 }
3899 # Now match and remove the rest of them
3905 }
3908 }
3911 ) {
3913 }
3914 # (bug 8068) Allow control over whether robots index a page.
3915 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
3916 # is not desirable, the last one on the page should win.
3920 }
3924 }
3926 # Cache all double underscores in the database
3929 }
3932 }
3934 /**
3935 * @see ParserOutput::addTrackingCategory()
3936 * @param string $msg Message key
3937 * @return bool Whether the addition was successful
3938 */
3941 }
3943 /**
3944 * This function accomplishes several tasks:
3945 * 1) Auto-number headings if that option is enabled
3946 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
3947 * 3) Add a Table of contents on the top for users who have enabled the option
3948 * 4) Auto-anchor headings
3949 *
3950 * It loops through all headlines, collects the necessary data, then splits up the
3951 * string and re-inserts the newly formatted headlines.
3952 *
3953 * @param string $text
3954 * @param string $origText Original, untouched wikitext
3955 * @param bool $isMain
3956 * @return mixed|string
3957 * @private
3958 */
3962 # Inhibit editsection links if requested in the page
3968 }
3971 }
3973 # Get all headlines for numbering them and adding funky stuff like [edit]
3974 # links - this is for later, but we need the number of headlines right now
3979 $matches
3980 );
3982 # if there are fewer than 4 headlines in the article, do not show TOC
3983 # unless it's been explicitly enabled.
3987 # Allow user to stipulate that a page should have a "new section"
3988 # link added via __NEWSECTIONLINK__
3991 }
3993 # Allow user to remove the "new section"
3994 # link via __NONEWSECTIONLINK__
3997 }
3999 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4000 # override above conditions and always show TOC above first header
4004 }
4006 # headline counter
4010 # Ugh .. the TOC should have neat indentation levels which can be
4011 # passed to the skin functions. These are determined here
4045 }
4049 }
4053 # Increase TOC level
4060 }
4062 # Decrease TOC level, find level to jump to
4066 # Found last matching level
4070 # Found first matching level below current level
4073 }
4074 }
4077 }
4080 # Unindent only if the previous toc level was shown :p
4085 }
4086 }
4088 # No change in level, end TOC line
4091 }
4092 }
4096 # count number of headlines for each level
4103 }
4106 }
4107 }
4109 # The safe header is a version of the header text safe to use for links
4111 # Remove link placeholders by the link text.
4112 # <!--LINK number-->
4113 # turns into
4114 # link text with suffix
4115 # Do this before unstrip since link text can contain strip markers
4118 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4121 # Strip out HTML (first regex removes any tag not allowed)
4122 # Allowed tags are:
4123 # * <sup> and <sub> (bug 8393)
4124 # * <i> (bug 26375)
4125 # * <b> (r105284)
4126 # * <bdi> (bug 72884)
4127 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4128 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4129 # to allow setting directionality in toc items.
4131 [
4133 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b))(?: .*?)?>#'
4134 ],
4136 $safeHeadline
4137 );
4139 # Strip '<span></span>', which is the result from the above if
4140 # <span id="foo"></span> is used to produce an additional anchor
4141 # for a section.
4146 # For the anchor, strip out HTML-y stuff period
4150 # Save headline for section edit hint before it's escaped
4154 # For reverse compatibility, provide an id that's
4155 # HTML4-compatible, like we used to.
4156 # It may be worth noting, academically, that it's possible for
4157 # the legacy anchor to conflict with a non-legacy headline
4158 # anchor on the page. In this case likely the "correct" thing
4159 # would be to either drop the legacy anchors or make sure
4160 # they're numbered first. However, this would require people
4161 # to type in section names like "abc_.D7.93.D7.90.D7.A4"
4162 # manually, so let's not bother worrying about it.
4168 # No reason to have both (in fact, we can't)
4170 }
4175 }
4177 # HTML names must be case-insensitively unique (bug 10721).
4178 # This does not apply to Unicode characters per
4179 # http://www.w3.org/TR/html5/infrastructure.html#case-sensitivity-and-string-comparison
4180 # @todo FIXME: We may be changing them depending on the current locale.
4186 }
4188 # Create the anchor for linking from the TOC to the section
4192 // @codingStandardsIgnoreStart
4194 // @codingStandardsIgnoreEnd
4199 }
4201 // @codingStandardsIgnoreStart
4203 // @codingStandardsIgnoreEnd
4208 }
4210 # Don't number the heading if it is the only one (looks silly)
4212 # the two are different if the line contains a link
4216 $numbering
4218 }
4223 }
4225 # Add the section to the section tree
4226 # Find the DOM node for this header
4233 }
4234 }
4238 }
4248 ];
4250 # give headline the correct <h#> tag
4252 // Output edit section links as markers with styles that can be customized by skins
4254 # Put a T flag in the section identifier, to indicate to extractSections()
4255 # that sections inside <includeonly> should be counted.
4263 }
4264 // We use a bit of pesudo-xml for editsection markers. The
4265 // language converter is run later on. Using a UNIQ style marker
4266 // leads to the converter screwing up the tokens when it
4267 // converts stuff. And trying to insert strip tags fails too. At
4268 // this point all real inputted tags have already been escaped,
4269 // so we don't have to worry about a user trying to input one of
4270 // these markers directly. We use a page and section attribute
4271 // to stop the language converter from converting these
4272 // important bits of data, but put the headline hint inside a
4273 // content block because the language converter is supposed to
4274 // be able to convert that piece of data.
4275 // Gets replaced with html in ParserOutput::getText
4282 }
4285 }
4291 }
4295 # Never ever show TOC if no headers
4298 }
4303 }
4308 }
4312 }
4314 # split up and insert constructed headlines
4318 // build an array of document sections
4321 // $head is zero-based, sections aren't.
4326 }
4328 /**
4329 * Send a hook, one per section.
4330 * The idea here is to be able to make section-level DIVs, but to do so in a
4331 * lower-impact, more correct way than r50769
4332 *
4333 * $this : caller
4334 * $section : the section number
4335 * &$sectionContent : ref to the content of the section
4336 * $showEditLinks : boolean describing whether this section has an edit link
4337 */
4341 }
4344 // append the TOC at the beginning
4345 // Top anchor now in skin
4347 }
4355 }
4356 }
4358 /**
4359 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4360 * conversion, substituting signatures, {{subst:}} templates, etc.
4361 *
4362 * @param string $text The text to transform
4363 * @param Title $title The Title object for the current article
4364 * @param User $user The User object describing the current user
4365 * @param ParserOptions $options Parsing options
4366 * @param bool $clearState Whether to clear the parser state first
4367 * @return string The altered wiki markup
4368 */
4371 ) {
4374 }
4381 ];
4385 }
4391 }
4393 /**
4394 * Pre-save transform helper function
4395 *
4396 * @param string $text
4397 * @param User $user
4398 *
4399 * @return string
4400 */
4404 # Note: This is the timestamp saved as hardcoded wikitext to
4405 # the database, we use $wgContLang here in order to give
4406 # everyone the same signature and use the default one rather
4407 # than the one selected in each user's preferences.
4408 # (see also bug 12815)
4416 # Variable replacement
4417 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4420 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4421 # which may corrupt this parser instance via its wfMessage()->text() call-
4423 # Signatures
4429 ] );
4431 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4435 // [[ns:page (context)|]]
4437 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4439 // [[ns:page (context), context|]] (using either single or double-width comma)
4441 // [[|page]] (reverse pipe trick: add context from page title)
4444 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4456 # if there's no context, don't bother duplicating the title
4458 }
4460 # Trim trailing whitespace
4464 }
4466 /**
4467 * Fetch the user's signature text, if any, and normalize to
4468 * validated, ready-to-insert wikitext.
4469 * If you have pre-fetched the nickname or the fancySig option, you can
4470 * specify them here to save a database query.
4471 * Do not reuse this parser instance after calling getUserSig(),
4472 * as it may have changed if it's the $wgParser.
4473 *
4474 * @param User $user
4475 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4476 * @param bool|null $fancySig whether the nicknname is the complete signature
4477 * or null to use default value
4478 * @return string
4479 */
4485 # If not given, retrieve from the user object.
4488 }
4492 }
4500 # Sig. might contain markup; validate this
4502 # Validated; clean up (if needed) and return it
4505 # Failed to validate; fall back to the default
4508 }
4509 }
4511 # Make sure nickname doesnt get a sig in a sig
4514 # If we're still here, make it a link to the user page
4521 }
4523 /**
4524 * Check that the user's signature contains no bad XML
4525 *
4526 * @param string $text
4527 * @return string|bool An expanded string, or false if invalid.
4528 */
4531 }
4533 /**
4534 * Clean up signature text
4535 *
4536 * 1) Strip 3, 4 or 5 tildes out of signatures @see cleanSigInSig
4537 * 2) Substitute all transclusions
4538 *
4539 * @param string $text
4540 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
4541 * @return string Signature text
4542 */
4548 }
4550 # Option to disable this feature
4553 }
4555 # @todo FIXME: Regex doesn't respect extension tags or nowiki
4556 # => Move this logic to braceSubstitution()
4558 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
4569 }
4572 }
4574 /**
4575 * Strip 3, 4 or 5 tildes out of signatures.
4576 *
4577 * @param string $text
4578 * @return string Signature text with /~{3,5}/ removed
4579 */
4583 }
4585 /**
4586 * Set up some variables which are usually set up in parse()
4587 * so that an external function can call some class members with confidence
4588 *
4589 * @param Title|null $title
4590 * @param ParserOptions $options
4591 * @param int $outputType
4592 * @param bool $clearState
4593 */
4596 ) {
4598 }
4600 /**
4601 * @param Title|null $title
4602 * @param ParserOptions $options
4603 * @param int $outputType
4604 * @param bool $clearState
4605 */
4608 ) {
4614 }
4615 }
4617 /**
4618 * Wrapper for preprocess()
4619 *
4620 * @param string $text The text to preprocess
4621 * @param ParserOptions $options Options
4622 * @param Title|null $title Title object or null to use $wgTitle
4623 * @return string
4624 */
4628 # Guard against infinite recursion
4631 }
4637 }
4643 }
4645 /**
4646 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
4647 * The callback should have the following form:
4648 * function myParserHook( $text, $params, $parser, $frame ) { ... }
4649 *
4650 * Transform and return $text. Use $parser for any required context, e.g. use
4651 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
4652 *
4653 * Hooks may return extended information by returning an array, of which the
4654 * first numbered element (index 0) must be the return string, and all other
4655 * entries are extracted into local variables within an internal function
4656 * in the Parser class.
4657 *
4658 * This interface (introduced r61913) appears to be undocumented, but
4659 * 'markerType' is used by some core tag hooks to override which strip
4660 * array their results are placed in. **Use great caution if attempting
4661 * this interface, as it is not documented and injudicious use could smash
4662 * private variables.**
4663 *
4664 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
4665 * @param callable $callback The callback function (and object) to use for the tag
4666 * @throws MWException
4667 * @return callable|null The old value of the mTagHooks array associated with the hook
4668 */
4673 }
4678 }
4681 }
4683 /**
4684 * As setHook(), but letting the contents be parsed.
4685 *
4686 * Transparent tag hooks are like regular XML-style tag hooks, except they
4687 * operate late in the transformation sequence, on HTML instead of wikitext.
4688 *
4689 * This is probably obsoleted by things dealing with parser frames?
4690 * The only extension currently using it is geoserver.
4691 *
4692 * @since 1.10
4693 * @todo better document or deprecate this
4694 *
4695 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
4696 * @param callable $callback The callback function (and object) to use for the tag
4697 * @throws MWException
4698 * @return callable|null The old value of the mTagHooks array associated with the hook
4699 */
4704 }
4705 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
4709 }
4711 /**
4712 * Remove all tag hooks
4713 */
4718 }
4720 /**
4721 * Create a function, e.g. {{sum:1|2|3}}
4722 * The callback function should have the form:
4723 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
4724 *
4725 * Or with Parser::SFH_OBJECT_ARGS:
4726 * function myParserFunction( $parser, $frame, $args ) { ... }
4727 *
4728 * The callback may either return the text result of the function, or an array with the text
4729 * in element 0, and a number of flags in the other elements. The names of the flags are
4730 * specified in the keys. Valid flags are:
4731 * found The text returned is valid, stop processing the template. This
4732 * is on by default.
4733 * nowiki Wiki markup in the return value should be escaped
4734 * isHTML The returned text is HTML, armour it against wikitext transformation
4735 *
4736 * @param string $id The magic word ID
4737 * @param callable $callback The callback function (and object) to use
4738 * @param int $flags A combination of the following flags:
4739 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
4740 *
4741 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
4742 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
4743 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
4744 * the arguments, and to control the way they are expanded.
4745 *
4746 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
4747 * arguments, for instance:
4748 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
4749 *
4750 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
4751 * future versions. Please call $frame->expand() on it anyway so that your code keeps
4752 * working if/when this is changed.
4753 *
4754 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
4755 * expansion.
4756 *
4757 * Please read the documentation in includes/parser/Preprocessor.php for more information
4758 * about the methods available in PPFrame and PPNode.
4759 *
4760 * @throws MWException
4761 * @return string|callable The old callback function for this name, if any
4762 */
4769 # Add to function cache
4773 }
4779 # Case
4782 }
4783 # Add leading hash
4786 }
4787 # Remove trailing colon
4790 }
4792 }
4794 }
4796 /**
4797 * Get all registered function hook identifiers
4798 *
4799 * @return array
4800 */
4803 }
4805 /**
4806 * Create a tag function, e.g. "<test>some stuff</test>".
4807 * Unlike tag hooks, tag functions are parsed at preprocessor level.
4808 * Unlike parser functions, their content is not preprocessed.
4809 * @param string $tag
4810 * @param callable $callback
4811 * @param int $flags
4812 * @throws MWException
4813 * @return null
4814 */
4819 }
4826 }
4829 }
4831 /**
4832 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
4833 * Placeholders created in Linker::link()
4834 *
4835 * @param string $text
4836 * @param int $options
4837 */
4840 }
4842 /**
4843 * Replace "<!--LINK-->" link placeholders with plain text of links
4844 * (not HTML-formatted).
4845 *
4846 * @param string $text
4847 * @return string
4848 */
4851 }
4853 /**
4854 * Renders an image gallery from a text with one line per image.
4855 * text labels may be given by using |-style alternative text. E.g.
4856 * Image:one.jpg|The number "1"
4857 * Image:tree.jpg|A tree
4858 * given as text will return the HTML of a gallery with two images,
4859 * labeled 'The number "1"' and
4860 * 'A tree'.
4861 *
4862 * @param string $text
4863 * @param array $params
4864 * @return string HTML
4865 */
4871 }
4876 // If invalid type set, fallback to default.
4878 }
4891 }
4897 }
4900 }
4903 }
4906 }
4913 # match lines like these:
4914 # Image:someimage.jpg|This is some image
4917 # Skip empty lines
4920 }
4924 }
4927 # Bogus title. Ignore these so we don't bomb out later.
4929 }
4931 # We need to get what handler the file uses, to figure out parameters.
4932 # Note, a hook can overide the file name, and chose an entirely different
4933 # file (which potentially could be of a different type and have different handler).
4938 # Don't register it now, as ImageGallery does that later.
4945 ];
4948 // We don't want people to specify per-image widths.
4949 // Additionally the width parameter would need special casing anyhow.
4951 }
4960 // look for an |alt= definition while trying not to break existing
4961 // captions with multiple pipes (|) in it, until a more sensible grammar
4962 // is defined for images in galleries
4964 // FIXME: Doing recursiveTagParse at this stage, and the trim before
4965 // splitting on '|' is a bit odd, and different from makeImage.
4983 // check to see if link matches an absolute url, if not then it must be a wiki link.
4990 }
4991 }
4994 // Must be a handler specific parameter.
4998 // Guess not, consider it as caption.
5001 }
5002 }
5005 // Last pipe wins.
5007 }
5008 }
5009 // Remove the pipe.
5011 }
5014 }
5018 }
5020 /**
5021 * @param MediaHandler $handler
5022 * @return array
5023 */
5029 }
5031 # Initialise static lists
5038 ];
5046 }
5047 }
5048 }
5050 # Add handler params
5056 }
5057 }
5060 }
5062 }
5064 /**
5065 * Parse image options text and use it to make an image
5066 *
5067 * @param Title $title
5068 * @param string $options
5069 * @param LinkHolderArray|bool $holders
5070 * @return string HTML
5071 */
5073 # Check if the options text is of the form "options|alt text"
5074 # Options are:
5075 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5076 # * left no resizing, just left align. label is used for alt= only
5077 # * right same, but right aligned
5078 # * none same, but not aligned
5079 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5080 # * center center the image
5081 # * frame Keep original image size, no magnify-button.
5082 # * framed Same as "frame"
5083 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5084 # * upright reduce width for upright images, rounded to full __0 px
5085 # * border draw a 1px border around the image
5086 # * alt Text for HTML alt attribute (defaults to empty)
5087 # * class Set a class for img node
5088 # * link Set the target of the image link. Can be external, interwiki, or local
5089 # vertical-align values (no % or length right now):
5090 # * baseline
5091 # * sub
5092 # * super
5093 # * top
5094 # * text-top
5095 # * middle
5096 # * bottom
5097 # * text-bottom
5101 # Give extensions a chance to select the file revision for us
5106 # Fetch and register the file (file title may be different via hooks)
5109 # Get parameter map
5116 }
5118 # Process the input parameters
5130 # Special case; width and height come in one variable together
5138 }
5139 }
5145 }
5146 }
5147 # else no validation -- bug 13436
5150 # Validate handler parameter
5153 # Validate internal parameters
5158 # @todo FIXME: Possibly check validity here for
5159 # manualthumb? downstream behavior seems odd with
5160 # missing manual thumbs.
5178 }
5180 }
5188 }
5189 }
5194 // use first appearing option, discard others.
5199 # Most other things appear to be empty or numeric...
5201 }
5202 }
5206 }
5207 }
5208 }
5211 }
5212 }
5214 # Process alignment parameters
5217 }
5220 }
5224 # Will the image be presented in a frame, with the caption below?
5230 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5231 # came to also set the caption, ordinary text after the image -- which
5232 # makes no sense, because that just repeats the text multiple times in
5233 # screen readers. It *also* came to set the title attribute.
5234 # Now that we have an alt attribute, we should not set the alt text to
5235 # equal the caption: that's worse than useless, it just repeats the
5236 # text. This is the framed/thumbnail case. If there's no caption, we
5237 # use the unnamed parameter for alt text as well, just for the time be-
5238 # ing, if the unnamed param is set and the alt param is not.
5239 # For the future, we need to figure out if we want to tweak this more,
5240 # e.g., introducing a title= parameter for the title; ignoring the un-
5241 # named parameter entirely for images without a caption; adding an ex-
5242 # plicit caption= parameter and preserving the old magic unnamed para-
5243 # meter for BC; ...
5246 # No caption or alt text, add the filename as the alt text so
5247 # that screen readers at least get some description of the image
5249 }
5250 # Do not set $params['frame']['title'] because tooltips don't make sense
5251 # for framed images
5254 # No alt text, use the "caption" for the alt text
5258 # No caption, fall back to using the filename for the
5259 # alt text
5261 }
5262 }
5263 # Use the "caption" for the tooltip text
5265 }
5269 # Linker does the rest
5274 # Give the handler a chance to modify the parser object
5277 }
5280 }
5282 /**
5283 * @param string $caption
5284 * @param LinkHolderArray|bool $holders
5285 * @return mixed|string
5286 */
5288 # Strip bad stuff out of the title (tooltip). We can't just use
5289 # replaceLinkHoldersText() here, because if this function is called
5290 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5295 }
5297 # make sure there are no placeholders in thumbnail attributes
5298 # that are later expanded to html- so expand them now and
5299 # remove the tags
5304 }
5306 /**
5307 * Set a flag in the output object indicating that the content is dynamic and
5308 * shouldn't be cached.
5309 * @deprecated since 1.28; use getOutput()->updateCacheExpiry()
5310 */
5316 }
5318 }
5320 /**
5321 * Callback from the Sanitizer for expanding items found in HTML attribute
5322 * values, so they can be safely tested and escaped.
5323 *
5324 * @param string $text
5325 * @param bool|PPFrame $frame
5326 * @return string
5327 */
5332 }
5334 /**
5335 * Accessor
5336 *
5337 * @return array
5338 */
5344 );
5345 }
5347 /**
5348 * Replace transparent tags in $text with the values given by the callbacks.
5349 *
5350 * Transparent tag hooks are like regular XML-style tag hooks, except they
5351 * operate late in the transformation sequence, on HTML instead of wikitext.
5352 *
5353 * @param string $text
5354 *
5355 * @return string
5356 */
5370 );
5373 }
5375 }
5377 }
5379 /**
5380 * Break wikitext input into sections, and either pull or replace
5381 * some particular section's text.
5382 *
5383 * External callers should use the getSection and replaceSection methods.
5384 *
5385 * @param string $text Page wikitext
5386 * @param string|number $sectionId A section identifier string of the form:
5387 * "<flag1> - <flag2> - ... - <section number>"
5388 *
5389 * Currently the only recognised flag is "T", which means the target section number
5390 * was derived during a template inclusion parse, in other words this is a template
5391 * section edit link. If no flags are given, it was an ordinary section edit link.
5392 * This flag is required to avoid a section numbering mismatch when a section is
5393 * enclosed by "<includeonly>" (bug 6563).
5394 *
5395 * The section number 0 pulls the text before the first heading; other numbers will
5396 * pull the given section along with its lower-level subsections. If the section is
5397 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5398 *
5399 * Section 0 is always considered to exist, even if it only contains the empty
5400 * string. If $text is the empty string and section 0 is replaced, $newText is
5401 * returned.
5402 *
5403 * @param string $mode One of "get" or "replace"
5404 * @param string $newText Replacement text for section data.
5405 * @return string For "get", the extracted section text.
5406 * for "replace", the whole page with the section replaced.
5407 */
5416 # Process section extraction flags
5423 }
5424 }
5426 # Check for empty input
5428 # Only sections 0 and T-0 exist in an empty document
5434 }
5440 }
5441 }
5442 }
5444 # Preprocess the text
5447 # <h> nodes indicate section breaks
5448 # They can only occur at the top level, so we can find them by iterating the root's children
5451 # Find the target section
5453 # Section zero doesn't nest, level=big
5462 }
5463 }
5466 }
5468 }
5469 }
5472 # Not found
5477 }
5478 }
5480 # Find the end of the section, including nested sections
5487 }
5488 }
5491 }
5495 # Write out the remainder (in replace mode only)
5497 # Output the replacement text
5498 # Add two newlines on -- trailing whitespace in $newText is conventionally
5499 # stripped by the editor, so we need both newlines to restore the paragraph gap
5500 # Only add trailing whitespace if there is newText
5503 }
5508 }
5509 }
5512 # Re-insert stripped tags
5514 }
5517 }
5519 /**
5520 * This function returns the text of a section, specified by a number ($section).
5521 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5522 * the first section before any such heading (section 0).
5523 *
5524 * If a section contains subsections, these are also returned.
5525 *
5526 * @param string $text Text to look in
5527 * @param string|number $sectionId Section identifier as a number or string
5528 * (e.g. 0, 1 or 'T-1').
5529 * @param string $defaultText Default to return if section is not found
5530 *
5531 * @return string Text of the requested section
5532 */
5535 }
5537 /**
5538 * This function returns $oldtext after the content of the section
5539 * specified by $section has been replaced with $text. If the target
5540 * section does not exist, $oldtext is returned unchanged.
5541 *
5542 * @param string $oldText Former text of the article
5543 * @param string|number $sectionId Section identifier as a number or string
5544 * (e.g. 0, 1 or 'T-1').
5545 * @param string $newText Replacing text
5546 *
5547 * @return string Modified text
5548 */
5551 }
5553 /**
5554 * Get the ID of the revision we are parsing
5555 *
5556 * @return int|null
5557 */
5560 }
5562 /**
5563 * Get the revision object for $this->mRevisionId
5564 *
5565 * @return Revision|null Either a Revision object or null
5566 * @since 1.23 (public since 1.23)
5567 */
5571 }
5574 }
5578 );
5580 # If the parse is for a new revision, then the callback should have
5581 # already been set to force the object and should match mRevisionId.
5582 # If not, try to fetch by mRevisionId for sanity.
5585 }
5590 }
5592 /**
5593 * Get the timestamp associated with the current revision, adjusted for
5594 * the default server-local timestamp
5595 * @return string
5596 */
5604 # The cryptic '' timezone parameter tells to use the site-default
5605 # timezone offset instead of the user settings.
5606 # Since this value will be saved into the parser cache, served
5607 # to other users, and potentially even used inside links and such,
5608 # it needs to be consistent for all visitors.
5611 }
5613 }
5615 /**
5616 * Get the name of the user that edited the last revision
5617 *
5618 * @return string User name
5619 */
5624 # if this template is subst: the revision id will be blank,
5625 # so just use the current user's name
5630 }
5631 }
5633 }
5635 /**
5636 * Get the size of the revision
5637 *
5638 * @return int|null Revision size
5639 */
5644 # if this variable is subst: the revision id will be blank,
5645 # so just use the parser input size, because the own substituation
5646 # will change the size.
5651 }
5652 }
5654 }
5656 /**
5657 * Mutator for $mDefaultSort
5658 *
5659 * @param string $sort New value
5660 */
5664 }
5666 /**
5667 * Accessor for $mDefaultSort
5668 * Will use the empty string if none is set.
5669 *
5670 * This value is treated as a prefix, so the
5671 * empty string is equivalent to sorting by
5672 * page name.
5673 *
5674 * @return string
5675 */
5681 }
5682 }
5684 /**
5685 * Accessor for $mDefaultSort
5686 * Unlike getDefaultSort(), will return false if none is set
5687 *
5688 * @return string|bool
5689 */
5692 }
5694 /**
5695 * Try to guess the section anchor name based on a wikitext fragment
5696 * presumably extracted from a heading, for example "Header" from
5697 * "== Header ==".
5698 *
5699 * @param string $text
5700 *
5701 * @return string
5702 */
5704 # Strip out wikitext links(they break the anchor)
5708 }
5710 /**
5711 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
5712 * instead. For use in redirects, since IE6 interprets Redirect: headers
5713 * as something other than UTF-8 (apparently?), resulting in breakage.
5714 *
5715 * @param string $text The section name
5716 * @return string An anchor
5717 */
5719 # Strip out wikitext links(they break the anchor)
5723 }
5725 /**
5726 * Strips a text string of wikitext for use in a section anchor
5727 *
5728 * Accepts a text string and then removes all wikitext from the
5729 * string and leaves only the resultant text (i.e. the result of
5730 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
5731 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
5732 * to create valid section anchors by mimicing the output of the
5733 * parser when headings are parsed.
5734 *
5735 * @param string $text Text string to be stripped of wikitext
5736 * for use in a Section anchor
5737 * @return string Filtered text string
5738 */
5740 # Strip internal link markup
5744 # Strip external link markup
5745 # @todo FIXME: Not tolerant to blank link text
5746 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
5747 # on how many empty links there are on the page - need to figure that out.
5748 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
5750 # Parse wikitext quotes (italics & bold)
5753 # Strip HTML tags
5756 }
5758 /**
5759 * strip/replaceVariables/unstrip for preprocessor regression testing
5760 *
5761 * @param string $text
5762 * @param Title $title
5763 * @param ParserOptions $options
5764 * @param int $outputType
5765 *
5766 * @return string
5767 */
5770 ) {
5778 }
5780 /**
5781 * @param string $text
5782 * @param Title $title
5783 * @param ParserOptions $options
5784 * @return string
5785 */
5788 }
5790 /**
5791 * @param string $text
5792 * @param Title $title
5793 * @param ParserOptions $options
5794 * @return string
5795 */
5798 }
5800 /**
5801 * Call a callback function on all regions of the given text that are not
5802 * inside strip markers, and replace those regions with the return value
5803 * of the callback. For example, with input:
5804 *
5805 * aaa<MARKER>bbb
5806 *
5807 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
5808 * two strings will be replaced with the value returned by the callback in
5809 * each case.
5810 *
5811 * @param string $s
5812 * @param callable $callback
5813 *
5814 * @return string
5815 */
5834 }
5835 }
5836 }
5838 }
5840 /**
5841 * Remove any strip markers found in the given text.
5842 *
5843 * @param string $text Input string
5844 * @return string
5845 */
5848 }
5850 /**
5851 * Save the parser state required to convert the given half-parsed text to
5852 * HTML. "Half-parsed" in this context means the output of
5853 * recursiveTagParse() or internalParse(). This output has strip markers
5854 * from replaceVariables (extensionSubstitution() etc.), and link
5855 * placeholders from replaceLinkHolders().
5856 *
5857 * Returns an array which can be serialized and stored persistently. This
5858 * array can later be loaded into another parser instance with
5859 * unserializeHalfParsedText(). The text can then be safely incorporated into
5860 * the return value of a parser hook.
5861 *
5862 * @param string $text
5863 *
5864 * @return array
5865 */
5872 ];
5874 }
5876 /**
5877 * Load the parser state given in the $data array, which is assumed to
5878 * have been generated by serializeHalfParsedText(). The text contents is
5879 * extracted from the array, and its markers are transformed into markers
5880 * appropriate for the current Parser instance. This transformed text is
5881 * returned, and can be safely included in the return value of a parser
5882 * hook.
5883 *
5884 * If the $data array has been stored persistently, the caller should first
5885 * check whether it is still valid, by calling isValidHalfParsedText().
5886 *
5887 * @param array $data Serialized data
5888 * @throws MWException
5889 * @return string
5890 */
5894 }
5896 # First, extract the strip state.
5900 # Now renumber links
5903 # Should be good to go.
5905 }
5907 /**
5908 * Returns true if the given array, presumed to be generated by
5909 * serializeHalfParsedText(), is compatible with the current version of the
5910 * parser.
5911 *
5912 * @param array $data
5913 *
5914 * @return bool
5915 */
5918 }
5920 /**
5921 * Parsed a width param of imagelink like 300px or 200x300px
5922 *
5923 * @param string $value
5924 *
5925 * @return array
5926 * @since 1.20
5927 */
5932 }
5934 # (bug 13500) In both cases (width/height and width only),
5935 # permit trailing "px" for backward compatibility.
5944 }
5946 }
5948 /**
5949 * Lock the current instance of the parser.
5950 *
5951 * This is meant to stop someone from calling the parser
5952 * recursively and messing up all the strip state.
5953 *
5954 * @throws MWException If parser is in a parse
5955 * @return ScopedCallback The lock will be released once the return value goes out of scope.
5956 */
5961 }
5966 } );
5969 }
5971 /**
5972 * Strip outer <p></p> tag from the HTML source of a single paragraph.
5973 *
5974 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
5975 * or if there is more than one <p/> tag in the input HTML.
5976 *
5977 * @param string $html
5978 * @return string
5979 * @since 1.24
5980 */
5986 }
5987 }
5990 }
5992 /**
5993 * Return this parser if it is not doing anything, otherwise
5994 * get a fresh parser. You can use this method by doing
5995 * $myParser = $wgParser->getFreshParser(), or more simply
5996 * $wgParser->getFreshParser()->parse( ... );
5997 * if you're unsure if $wgParser is safe to use.
5998 *
5999 * @since 1.24
6000 * @return Parser A parser object that is not parsing anything
6001 */
6008 }
6009 }
6011 /**
6012 * Set's up the PHP implementation of OOUI for use in this request
6013 * and instructs OutputPage to enable OOUI for itself.
6014 *
6015 * @since 1.26
6016 */
6020 }
6021 }