| blob | 7b4a65074ba6e4d43579a32164036167cbb243f0 |
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
90 # Simplified expression to match an IPv4 or IPv6 address, or
91 # at least one character of a host name (embeds EXT_LINK_URL_CLASS)
93 # RegExp to make image URLs (embeds IPv6 part of EXT_LINK_ADDR)
94 // @codingStandardsIgnoreStart Generic.Files.LineLength
95 const EXT_IMAGE_REGEX = '/^(http:\/\/|https:\/\/)((?:\\[(?i:[0-9a-f:.]+)\\])?[^][<>"\\x00-\\x20\\x7F\p{Zs}]+)
97 // @codingStandardsIgnoreEnd
99 # Regular expression for a non-newline space
102 # State constants for the definition list colon extraction
112 # Flags for preprocessToDom
115 # Allowed values for $this->mOutputType
116 # Parameter to startExternalParse().
123 /**
124 * @var string Prefix and suffix for temporary replacement strings
125 * for the multipass parser.
126 *
127 * \x7f should never appear in input as it's disallowed in XML.
128 * Using it at the front also gives us a little extra robustness
129 * since it shouldn't match when butted up against identifier-like
130 * string constructs.
131 *
132 * Must not consist of all title characters, or else it will change
133 * the behavior of <nowiki> in a link.
134 */
138 # Markers used for wrapping the table of contents
142 # Persistent:
156 # Initialised by initialiseVariables()
158 /**
159 * @var MagicWordArray
160 */
163 /**
164 * @var MagicWordArray
165 */
167 # Initialised in constructor
170 # Initialized in getPreprocessor()
171 /** @var Preprocessor */
174 # Cleared with clearState():
175 /**
176 * @var ParserOutput
177 */
181 /**
182 * @var StripState
183 */
187 /**
188 * @var LinkHolderArray
189 */
199 /**
200 * @var User
201 */
204 # Temporary
205 # These are variables reset at least once per parse regardless of $clearState
207 /**
208 * @var ParserOptions
209 */
212 /**
213 * @var Title
214 */
226 /**
227 * @var string Deprecated accessor for the strip marker prefix.
228 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
229 **/
232 /**
233 * @var array Array with the language name of each language link (i.e. the
234 * interwiki prefix) in the key, value arbitrary. Used to avoid sending
235 * duplicate language links to the ParserOutput.
236 */
239 /**
240 * @var MapCacheLRU|null
241 * @since 1.24
242 *
243 * A cache of the current revisions of titles. Keys are $title->getPrefixedDbKey()
244 */
247 /**
248 * @var bool Recursive call protection.
249 * This variable should be treated as if it were private.
250 */
253 /** @var SectionProfiler */
256 /**
257 * @param array $conf
258 */
268 # Preprocessor_Hash is much faster than Preprocessor_DOM under HipHop
271 # PECL extension that conflicts with the core DOM extension (bug 13770)
278 }
280 }
282 /**
283 * Reduce memory usage to reduce the impact of circular references
284 */
288 }
291 }
292 }
294 /**
295 * Allow extensions to clean up when the parser is cloned
296 */
300 // Bug 56226: When you create a reference "to" an object field, that
301 // makes the object field itself be a reference too (until the other
302 // reference goes out of scope). When cloning, any field that's a
303 // reference is copied as a reference in the new object. Both of these
304 // are defined PHP5 behaviors, as inconvenient as it is for us when old
305 // hooks from PHP4 days are passing fields by reference.
307 // Make a non-reference copy of the field, then rebind the field to
308 // reference the new copy.
312 }
315 }
317 /**
318 * Do various kinds of initialisation on the first call of the parser
319 */
323 }
331 }
333 /**
334 * Clear Parser state
335 *
336 * @private
337 */
341 }
361 # Clear these on every parse, bug 4549
369 );
378 # Fix cloning
381 }
386 }
388 /**
389 * Convert wikitext to HTML
390 * Do not call this function recursively.
391 *
392 * @param string $text Text we want to parse
393 * @param Title $title
394 * @param ParserOptions $options
395 * @param bool $linestart
396 * @param bool $clearState
397 * @param int $revid Number to pass in {{REVISIONID}}
398 * @return ParserOutput A ParserOutput
399 */
402 ) {
403 /**
404 * First pass--just handle <nowiki> sections, pass the rest off
405 * to internalParse() which does all the real work.
406 */
411 // We use U+007F DELETE to construct strip markers, so we have to make
412 // sure that this character does not occur in the input text.
415 }
423 }
436 }
439 # No more strip!
446 /**
447 * A converted title will be provided in the output object if title and
448 * content conversion are enabled, the article text does not contain
449 * a conversion-suppressing double-underscore tag, and no
450 * {{DISPLAYTITLE:...}} is present. DISPLAYTITLE takes precedence over
451 * automatic link conversion.
452 */
457 ) {
464 }
465 }
471 );
472 }
474 # Information on include size limits, for the benefit of users who try to skirt them
482 );
483 }
488 );
492 );
495 );
498 );
501 );
504 );
507 );
513 }
523 ) ) {
529 }
533 }
534 }
535 }
536 // Since we're not really outputting HTML, decode the entities and
537 // then re-encode the things that need hiding inside HTML comments.
541 // Sanitize for comment. Note '‐' in the replacement is U+2010,
542 // which looks much like the problematic '-'.
546 // Add on template profiling data
550 } );
556 }
562 }
563 }
575 }
577 /**
578 * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
579 * can be called from an extension tag hook.
580 *
581 * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
582 * instead, which means that lists and links have not been fully parsed yet,
583 * and strip markers are still present.
584 *
585 * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
586 *
587 * Use this function if you're a parser tag hook and you want to parse
588 * wikitext before or after applying additional transformations, and you
589 * intend to *return the result as hook output*, which will cause it to go
590 * through the rest of parsing process automatically.
591 *
592 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
593 * $text are not expanded
594 *
595 * @param string $text Text extension wants to have parsed
596 * @param bool|PPFrame $frame The frame to use for expanding any template variables
597 * @return string UNSAFE half-parsed HTML
598 */
604 }
606 /**
607 * Fully parse wikitext to fully parsed HTML. This recursive parser entry
608 * point can be called from an extension tag hook.
609 *
610 * The output of this function is fully-parsed HTML that is safe for output.
611 * If you're a parser tag hook, you might want to use recursiveTagParse()
612 * instead.
613 *
614 * If $frame is not provided, then template variables (e.g., {{{1}}}) within
615 * $text are not expanded
616 *
617 * @since 1.25
618 *
619 * @param string $text Text extension wants to have parsed
620 * @param bool|PPFrame $frame The frame to use for expanding any template variables
621 * @return string Fully parsed HTML
622 */
627 }
629 /**
630 * Expand templates and variables in the text, producing valid, static wikitext.
631 * Also removes comments.
632 * Do not call this function recursively.
633 * @param string $text
634 * @param Title $title
635 * @param ParserOptions $options
636 * @param int|null $revid
637 * @param bool|PPFrame $frame
638 * @return mixed|string
639 */
642 ) {
647 }
653 }
655 /**
656 * Recursive parser entry point that can be called from an extension tag
657 * hook.
658 *
659 * @param string $text Text to be expanded
660 * @param bool|PPFrame $frame The frame to use for expanding any template variables
661 * @return string
662 * @since 1.19
663 */
668 }
670 /**
671 * Process the wikitext for the "?preload=" feature. (bug 5210)
672 *
673 * "<noinclude>", "<includeonly>" etc. are parsed as for template
674 * transclusion, comments, templates, arguments, tags hooks and parser
675 * functions are untouched.
676 *
677 * @param string $text
678 * @param Title $title
679 * @param ParserOptions $options
680 * @param array $params
681 * @return string
682 */
683 public function getPreloadText( $text, Title $title, ParserOptions $options, $params = array() ) {
687 # Parser (re)initialisation
696 }
698 /**
699 * Get a random string
700 *
701 * @return string
702 * @deprecated since 1.26; use wfRandomString() instead.
703 */
707 }
709 /**
710 * Set the current user.
711 * Should only be used when doing pre-save transform.
712 *
713 * @param User|null $user User object or null (to reset)
714 */
717 }
719 /**
720 * Accessor for mUniqPrefix.
721 *
722 * @return string
723 * @deprecated since 1.26; use Parser::MARKER_PREFIX instead.
724 */
728 }
730 /**
731 * Set the context title
732 *
733 * @param Title $t
734 */
738 }
741 # Strip the fragment to avoid various odd effects
746 }
747 }
749 /**
750 * Accessor for the Title object
751 *
752 * @return Title
753 */
756 }
758 /**
759 * Accessor/mutator for the Title object
760 *
761 * @param Title $x Title object or null to just get the current one
762 * @return Title
763 */
766 }
768 /**
769 * Set the output type
770 *
771 * @param int $ot New value
772 */
775 # Shortcut alias
781 );
782 }
784 /**
785 * Accessor/mutator for the output type
786 *
787 * @param int|null $x New value or null to just get the current one
788 * @return int
789 */
792 }
794 /**
795 * Get the ParserOutput object
796 *
797 * @return ParserOutput
798 */
801 }
803 /**
804 * Get the ParserOptions object
805 *
806 * @return ParserOptions
807 */
810 }
812 /**
813 * Accessor/mutator for the ParserOptions object
814 *
815 * @param ParserOptions $x New value or null to just get the current one
816 * @return ParserOptions Current ParserOptions object
817 */
820 }
822 /**
823 * @return int
824 */
827 }
829 /**
830 * @param int $id
831 */
834 }
836 /**
837 * Get a language object for use in parser functions such as {{FORMATNUM:}}
838 * @return Language
839 */
842 }
844 /**
845 * Get the target language for the content being parsed. This is usually the
846 * language that the content is in.
847 *
848 * @since 1.19
849 *
850 * @throws MWException
851 * @return Language
852 */
862 }
865 }
867 /**
868 * Get the language object for language conversion
869 * @return Language|null
870 */
873 }
875 /**
876 * Get a User object either from $this->mUser, if set, or from the
877 * ParserOptions object otherwise
878 *
879 * @return User
880 */
884 }
886 }
888 /**
889 * Get a preprocessor object
890 *
891 * @return Preprocessor
892 */
897 }
899 }
901 /**
902 * Replaces all occurrences of HTML-style comments and the given tags
903 * in the text with a random marker and returns the next text. The output
904 * parameter $matches will be an associative array filled with data in
905 * the form:
906 *
907 * @code
908 * 'UNIQ-xxxxx' => array(
909 * 'element',
910 * 'tag content',
911 * array( 'param' => 'x' ),
912 * '<element param="x">tag content</element>' ) )
913 * @endcode
914 *
915 * @param array $elements List of element names. Comments are always extracted.
916 * @param string $text Source text string.
917 * @param array $matches Out parameter, Array: extracted tags
918 * @param string|null $uniq_prefix
919 * @return string Stripped text
920 * @since 1.26 The uniq_prefix argument is deprecated.
921 */
922 public static function extractTagsAndParams( $elements, $text, &$matches, $uniq_prefix = null ) {
925 }
938 }
940 # comment
946 # tag
951 }
957 # Empty element tag, <tag />
966 }
970 # No end tag -- let it run out to the end of the text.
976 }
977 }
983 }
985 }
987 /**
988 * Get a list of strippable XML-like elements
989 *
990 * @return array
991 */
994 }
996 /**
997 * Add an item to the strip state
998 * Returns the unique tag which must be inserted into the stripped text
999 * The tag will be replaced with the original text in unstrip()
1000 *
1001 * @param string $text
1002 *
1003 * @return string
1004 */
1010 }
1012 /**
1013 * parse the wiki syntax used to render tables
1014 *
1015 * @private
1016 * @param string $text
1017 * @return string
1018 */
1036 }
1043 # First check if we are starting a new table
1056 # Don't do any of the following
1060 # We are ending a table
1066 }
1070 }
1074 }
1078 # Now we have a table row
1081 # Whats after the tag is now only attributes
1094 }
1098 }
1107 ) {
1108 # This might be cell elements, td, th or captions
1114 }
1118 }
1120 # Split up multiple cells on the same line.
1121 # FIXME : This can result in improper nesting of tags processed
1122 # by earlier parser steps, but should avoid splitting up eg
1123 # attribute values containing literal "||".
1128 # Loop through each table cell
1135 }
1140 }
1146 }
1156 }
1160 # A cell could contain both parameters and data
1163 # Bug 553: Note that a '|' inside an invalid link should not
1164 # be mistaken as delimiting cell parameters
1173 }
1177 }
1178 }
1180 }
1182 # Closing open td, tr && table
1186 }
1189 }
1192 }
1195 }
1197 # Remove trailing line-ending (b/c)
1200 }
1202 # special case: don't return empty table
1205 }
1208 }
1210 /**
1211 * Helper function for parse() that transforms wiki markup into half-parsed
1212 * HTML. Only called for $mOutputType == self::OT_HTML.
1213 *
1214 * @private
1215 *
1216 * @param string $text
1217 * @param bool $isMain
1218 * @param PPFrame|bool $frame
1219 *
1220 * @return string
1221 */
1226 # Hook to suspend the parser in this state
1227 if ( !Hooks::run( 'ParserBeforeInternalParse', array( &$this, &$text, &$this->mStripState ) ) ) {
1229 }
1231 # if $frame is provided, then use $frame for replacing any variables
1233 # use frame depth to infer how include/noinclude tags should be handled
1234 # depth=0 means this is the top-level document; otherwise it's an included document
1239 }
1243 # if $frame is not provided, then use old-style replaceVariables
1245 }
1253 );
1256 # Tables need to come after variable replacement for things to work
1257 # properly; putting them before other transformations should keep
1258 # exciting things like link expansions from showing up in surprising
1259 # places.
1271 # replaceInternalLinks may sometimes leave behind
1272 # absolute URLs, which have to be masked to hide them from replaceExternalLinks
1279 }
1281 /**
1282 * Helper function for parse() that transforms half-parsed HTML into fully
1283 * parsed HTML.
1284 *
1285 * @param string $text
1286 * @param bool $isMain
1287 * @param bool $linestart
1288 * @return string
1289 */
1295 }
1297 # Clean up special characters, only run once, next-to-last before doBlockLevels
1299 # french spaces, last one Guillemet-left
1300 # only if there is something before the space
1302 # french spaces, Guillemet-right
1305 );
1312 /**
1313 * The input doesn't get language converted if
1314 * a) It's disabled
1315 * b) Content isn't converted
1316 * c) It's a conversion table
1317 * d) it is an interface message (which is in the user language)
1318 */
1321 ) {
1323 # The position of the convert() call should not be changed. it
1324 # assumes that the links are all replaced and the only thing left
1325 # is the <nowiki> mark.
1327 }
1328 }
1334 }
1345 # attempt to sanitize at least some nesting problems
1346 # (bug #2702 and quite a few others)
1348 # ''Something [http://www.cool.com cool''] -->
1349 # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
1352 # fix up an anchor inside another anchor, only
1353 # at least for a single single nested link (bug 3695)
1356 # fix div inside inline elements- doBlockLevels won't wrap a line which
1357 # contains a div, so fix it up here; replace
1358 # div with escaped text
1361 # remove empty italic or bold tag pairs, some
1362 # introduced by rules above
1364 );
1370 }
1374 }
1377 }
1379 /**
1380 * Replace special strings like "ISBN xxx" and "RFC xxx" with
1381 * magic external links.
1382 *
1383 * DML
1384 * @private
1385 *
1386 * @param string $text
1387 *
1388 * @return string
1389 */
1398 '!(?: # Start cases
1400 (<.*?>) | # m[2]: Skip stuff inside
1403 # m[4]: Post-protocol path
1409 [0-9Xx] # check digit
1413 }
1415 /**
1416 * @throws MWException
1417 * @param array $m
1418 * @return HTML|string
1419 */
1422 # Skip anchor
1425 # Skip HTML element
1428 # Free external link
1431 # RFC or PMID
1445 }
1449 # ISBN
1457 ) );
1464 }
1465 }
1467 /**
1468 * Make a free external link, given a user-supplied URL
1469 *
1470 * @param string $url
1471 * @param int $numPostProto
1472 * The number of characters after the protocol.
1473 * @return string HTML
1474 * @private
1475 */
1479 # The characters '<' and '>' (which were escaped by
1480 # removeHTMLtags()) should not be included in
1481 # URLs, per RFC 2396.
1482 # Make terminate a URL as well (bug T84937)
1488 PREG_OFFSET_CAPTURE
1489 ) ) {
1492 }
1494 # Move trailing punctuation to $trail
1496 # If there is no left bracket, then consider right brackets fair game too
1499 }
1503 # Don't break a trailing HTML entity by moving the ; into $trail
1504 # This is in hot code, so use substr_compare to avoid having to
1505 # create a new string object for the comparison
1507 # more optimization: instead of running preg_match with a $
1508 # anchor, which can be slow, do the match on the reversed
1509 # string starting at the desired offset.
1510 # un-reversed regexp is: /&([a-z]+|#x[\da-f]+|#\d+)$/i
1513 }
1514 }
1518 }
1520 # Verify that we still have a real URL after trail removal, and
1521 # not just lone protocol
1524 }
1528 # Is this an external image?
1531 # Not an image, make a link
1536 # Register it in the output object...
1537 # Replace unnecessary URL escape codes with their equivalent characters
1540 }
1542 }
1544 /**
1545 * Parse headers and return html
1546 *
1547 * @private
1548 *
1549 * @param string $text
1550 *
1551 * @return string
1552 */
1557 }
1559 }
1561 /**
1562 * Replace single quotes with HTML markup
1563 * @private
1564 *
1565 * @param string $text
1566 *
1567 * @return string The altered text
1568 */
1574 }
1577 }
1579 /**
1580 * Helper function for doAllQuotes()
1581 *
1582 * @param string $text
1583 *
1584 * @return string
1585 */
1591 }
1593 // First, do some preliminary work. This may shift some apostrophes from
1594 // being mark-up to being text. It also counts the number of occurrences
1595 // of bold and italics mark-ups.
1600 // If there are ever four apostrophes, assume the first is supposed to
1601 // be text, and the remaining three constitute mark-up for bold text.
1602 // (bug 13227: ''''foo'''' turns into ' ''' foo ' ''')
1608 // If there are more than 5 apostrophes in a row, assume they're all
1609 // text except for the last 5.
1610 // (bug 13227: ''''''foo'''''' turns into ' ''''' foo ' ''''')
1614 }
1615 // Count the number of occurrences of bold and italics mark-ups.
1623 }
1624 }
1626 // If there is an odd number of both bold and italics, it is likely
1627 // that one of the bold ones was meant to be an apostrophe followed
1628 // by italics. Which one we cannot know for certain, but it is more
1629 // likely to be one that has a single-letter word before it.
1641 }
1644 // if $firstsingleletterword is set, we don't
1645 // look at the other options, so we can bail early.
1650 }
1651 }
1652 }
1653 }
1655 // If there is a single-letter word, use it!
1660 // If not, but there's a multi-letter word, use that one.
1664 // ... otherwise use the first one that has neither.
1665 // (notice that it is possible for all three to be -1 if, for example,
1666 // there is only one pentuple-apostrophe in the line)
1669 }
1670 }
1672 // Now let's actually convert our apostrophic mush to HTML!
1683 }
1702 }
1719 }
1739 }
1740 }
1741 }
1743 }
1744 // Now close all remaining tags. Notice that the order is important.
1747 }
1750 }
1753 }
1754 // There might be lonely ''''', so make sure we have a buffer
1757 }
1759 }
1761 /**
1762 * Replace external links (REL)
1763 *
1764 * Note: this is all very hackish and the order of execution matters a lot.
1765 * Make sure to run tests/parserTests.php if you change this code.
1766 *
1767 * @private
1768 *
1769 * @param string $text
1770 *
1771 * @throws MWException
1772 * @return string
1773 */
1780 }
1790 # The characters '<' and '>' (which were escaped by
1791 # removeHTMLtags()) should not be included in
1792 # URLs, per RFC 2396.
1797 }
1799 # If the link text is an image URL, replace it with an <img> tag
1800 # This happened by accident in the original parser, but some people used it extensively
1804 }
1808 # Set linktype for CSS - if URL==text, link is essentially free
1811 # No link text, e.g. [http://domain.tld/some.link]
1813 # Autonumber
1818 # Have link text, e.g. [http://domain.tld/some.link text]s
1819 # Check for trail
1821 }
1827 # Use the encoded URL
1828 # This means that users can paste URLs directly into the text
1829 # Funny characters like ö aren't valid in URLs anyway
1830 # This was changed in August 2004
1834 # Register link in the output object.
1835 # Replace unnecessary URL escape codes with the referenced character
1836 # This prevents spammers from hiding links from the filters
1839 }
1842 }
1844 /**
1845 * Get the rel attribute for a particular external link.
1846 *
1847 * @since 1.21
1848 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1849 * nofollow if appropriate
1850 * @param Title $title Optional Title, for wgNoFollowNsExceptions lookups
1851 * @return string|null Rel attribute for $url
1852 */
1858 ) {
1860 }
1862 }
1864 /**
1865 * Get an associative array of additional HTML attributes appropriate for a
1866 * particular external link. This currently may include rel => nofollow
1867 * (depending on configuration, namespace, and the URL's domain) and/or a
1868 * target attribute (depending on configuration).
1869 *
1870 * @param string|bool $url Optional URL, to extract the domain from for rel =>
1871 * nofollow if appropriate
1872 * @return array Associative array of HTML attributes
1873 */
1880 }
1882 }
1884 /**
1885 * Replace unusual escape codes in a URL with their equivalent characters
1886 *
1887 * @deprecated since 1.24, use normalizeLinkUrl
1888 * @param string $url
1889 * @return string
1890 */
1894 }
1896 /**
1897 * Replace unusual escape codes in a URL with their equivalent characters
1898 *
1899 * This generally follows the syntax defined in RFC 3986, with special
1900 * consideration for HTTP query strings.
1901 *
1902 * @param string $url
1903 * @return string
1904 */
1906 # First, make sure unsafe characters are encoded
1910 },
1911 $url
1912 );
1917 # Fragment part - 'fragment'
1923 }
1925 # Query part - 'query' minus &=+;
1931 }
1933 # Scheme and path part - 'pchar'
1934 # (we assume no userinfo or encoded colons in the host)
1939 }
1946 # Unescape it
1949 # Leave it escaped, but use uppercase for a-f
1951 }
1952 };
1954 }
1956 /**
1957 * make an image if it's allowed, either through the global
1958 * option, through the exception, or through the on-wiki whitelist
1959 *
1960 * @param string $url
1961 *
1962 * @return string
1963 */
1968 # $imagesfrom could be either a single string or an array of strings, parse out the latter
1975 }
1976 }
1981 }
1985 ) {
1987 # Image found
1989 }
1990 }
1993 ) {
1997 );
2000 # Sanitize the regex fragment, make it case-insensitive, ignore blank entries/comments
2003 }
2005 # Image matches a whitelist entry
2008 }
2009 }
2010 }
2012 }
2014 /**
2015 * Process [[ ]] wikilinks
2016 *
2017 * @param string $s
2018 *
2019 * @return string Processed text
2020 *
2021 * @private
2022 */
2026 }
2028 /**
2029 * Process [[ ]] wikilinks (RIL)
2030 * @param string $s
2031 * @throws MWException
2032 * @return LinkHolderArray
2033 *
2034 * @private
2035 */
2040 # the % is needed to support urlencoded titles as well
2043 # Match a link having the form [[namespace:link|alternate]]trail
2045 # Match cases where there is no "]]", which might still be images
2047 }
2051 # split the entire text string on occurrences of [[
2053 # get the first element (all text up to first [[), and remove the space we added
2062 # Match the end of a line for a word that's not followed by whitespace,
2063 # e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
2067 }
2071 }
2080 }
2083 }
2087 // @codingStandardsIgnoreStart Squiz.WhiteSpace.SemicolonSpacing.Incorrect
2088 # Loop for each link
2090 // @codingStandardsIgnoreStart
2092 # Check for excessive memory usage
2094 # Too big
2095 # Do the existence check, replace the link holders and clear the array
2098 }
2106 }
2107 # first link
2111 }
2112 }
2118 # If we get a ] at the beginning of $m[3] that means we have a link that's something like:
2119 # [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
2120 # the real problem is with the $e1 regex
2121 # See bug 1300.
2122 #
2123 # Still some problems for cases where the ] is meant to be outside punctuation,
2124 # and no image is in sight. See bug 2095.
2125 #
2129 ) {
2132 }
2133 # fix up urlencoded title texts
2135 # Should anchors '#' also be rejected?
2137 }
2140 # Invalid, but might be an image with a link in its caption
2145 }
2150 }
2154 # Don't allow internal links to pages containing
2155 # PROTO: where PROTO is a valid URL protocol; these
2156 # should be external links.
2160 }
2162 # Make subpage if necessary
2167 }
2171 # Strip off leading ':'
2173 }
2180 }
2189 # look at the next 'line' to see if we can close it there
2194 }
2197 # the first ]] closes the inner link, the second the image
2203 # if there's exactly one ]] that's fine, we'll keep looking
2206 # if $next_line is invalid too, we need look no further
2209 }
2210 }
2212 # we couldn't find the end of this imageLink, so output it raw
2213 # but don't ignore what might be perfectly normal links in the text we've examined
2216 # note: no $trail, because without an end, there *is* no trail
2218 }
2221 # note: no $trail, because without an end, there *is* no trail
2223 }
2224 }
2230 # Bug 4598 madness. Handle the quotes only if they come from the alternate part
2231 # [[Lista d''e paise d''o munno]] -> <a href="...">Lista d''e paise d''o munno</a>
2232 # [[Criticism of Harry Potter|Criticism of ''Harry Potter'']]
2233 # -> <a href="Criticism of Harry Potter">Criticism of <i>Harry Potter</i></a>
2235 }
2237 # Link not escaped by : , create the various objects
2239 # Interwikis
2244 )
2245 ) {
2246 # Bug 24502: filter duplicates
2250 }
2255 }
2260 # if no parameters were passed, $text
2261 # becomes something like "File:Foo.png",
2262 # which we don't want to pass on to the
2263 # image generator
2266 # recursively parse links inside the image caption
2267 # actually, this will parse them in any other parameters, too,
2268 # but it might be hard to fix that, and it doesn't matter ATM
2271 }
2272 # cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
2277 }
2279 }
2288 }
2294 /**
2295 * Strip the whitespace Category links produce, see bug 87
2296 */
2300 }
2301 }
2303 # Self-link checking. For some languages, variants of the title are checked in
2304 # LinkHolderArray::doVariants() to allow batching the existence checks necessary
2305 # for linking to a different variant.
2309 }
2311 # NS_MEDIA is a pseudo-namespace for linking directly to a file
2312 # @todo FIXME: Should do batch file existence checks, see comment below
2314 # Give extensions a chance to select the file revision for us
2319 # Fetch and register the file (file title may be different via hooks)
2321 # Cloak with NOPARSE to avoid replacement in replaceExternalLinks
2325 }
2327 # Some titles, such as valid special pages or files in foreign repos, should
2328 # be shown as bluelinks even though they're not included in the page table
2329 #
2330 # @todo FIXME: isAlwaysKnown() can be expensive for file links; we should really do
2331 # batch file existence checks for NS_FILE and NS_MEDIA
2336 # Links will be added to the output link list after checking
2338 }
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.
2646 # No prefix (not in list)--go to paragraph mode
2647 # XXX: use a stack for nestable elements like span, table and div
2649 '/(?:<table|<h1|<h2|<h3|<h4|<h5|<h6|<pre|<tr|'
2651 $t
2652 );
2658 $t
2659 );
2663 # @todo bug 5718: paragraph closed
2667 }
2669 while ( preg_match( '/<(\\/?)blockquote[\s>]/i', $t, $bqMatch, PREG_OFFSET_CAPTURE, $bqOffset ) ) {
2672 }
2678 ) {
2679 # pre
2684 }
2687 # paragraph
2700 }
2701 }
2710 }
2711 }
2712 }
2713 }
2714 }
2715 # somewhere above we forget to get out of pre block (bug 785)
2718 }
2723 }
2724 }
2725 }
2731 }
2732 }
2736 }
2739 }
2741 /**
2742 * Split up a string on ':', ignoring any occurrences inside tags
2743 * to prevent illegal overlapping.
2744 *
2745 * @param string $str The string to split
2746 * @param string &$before Set to everything before the ':'
2747 * @param string &$after Set to everything after the ':'
2748 * @throws MWException
2749 * @return string The position of the ':', or false if none found
2750 */
2755 # Nothing to find!
2757 }
2761 # Easy; no tag nesting to worry about
2765 }
2767 # Ugly state machine to walk through avoiding tags.
2775 # (Using the number is a performance hack for common cases)
2779 # Could be either a <start> tag or an </end> tag
2784 # We found it!
2788 }
2789 # Embedded in a tag; don't break it.
2792 # Skip ahead looking for something interesting
2795 # Nothing else interesting
2797 }
2801 # We found it!
2805 }
2806 }
2808 # Nothing else interesting to find; abort!
2809 # We're nested, but there's no close tags left. Abort!
2811 }
2812 # Skip ahead to next tag start
2815 }
2818 # In a <tag>
2825 # Slash may be followed by >?
2829 # ignore
2830 }
2841 # Illegal early close? This shouldn't happen D:
2846 }
2849 # In a </tag>
2855 }
2857 }
2861 # Yes, a self-closed tag <blah/>
2864 # Probably we're jumping the gun, and this is an attribute
2866 }
2871 }
2878 }
2885 }
2889 }
2890 }
2892 wfDebug( __METHOD__ . ": Invalid input; not enough close tags (stack $stack, state $state)\n" );
2894 }
2896 }
2898 /**
2899 * Return value of a magic variable (like PAGENAME)
2900 *
2901 * @private
2902 *
2903 * @param int $index
2904 * @param bool|PPFrame $frame
2905 *
2906 * @throws MWException
2907 * @return string
2908 */
2914 // If no title set, bad things are going to happen
2915 // later. Title should always be set since this
2916 // should only be called in the middle of a parse
2917 // operation (but the unit-tests do funky stuff)
2920 }
2922 /**
2923 * Some of these require message or data lookups and can be
2924 * expensive to check many times.
2925 */
2929 }
2930 }
2975 $value = $pageLang->getMonthAbbreviation( MWTimestamp::getLocalInstance( $ts )->format( 'n' ) );
3009 ) ) );
3019 ) ) );
3027 }
3035 }
3048 # 0 means the page doesn't exist in the database,
3049 # which means the user is previewing a new page.
3050 # The vary-revision flag must be set, because the magic word
3051 # will have a different value once the page is saved.
3054 }
3058 # Let the edit saving system know we should parse the page
3059 # *after* a revision ID has been assigned.
3065 # Let the edit saving system know we should parse the page
3066 # *after* a revision ID has been assigned. This is for null edits.
3072 # Let the edit saving system know we should parse the page
3073 # *after* a revision ID has been assigned. This is for null edits.
3079 # Let the edit saving system know we should parse the page
3080 # *after* a revision ID has been assigned. This is for null edits.
3086 # Let the edit saving system know we should parse the page
3087 # *after* a revision ID has been assigned. This is for null edits.
3093 # Let the edit saving system know we should parse the page
3094 # *after* a revision ID has been assigned. This is for null edits.
3100 # Let the edit saving system know we should parse the page
3101 # *after* a revision ID has been assigned. This is for null edits.
3107 # Let the edit saving system know we should parse the page
3108 # *after* a revision ID has been assigned. This is for null edits.
3114 # Let the edit saving system know we should parse the page
3115 # *after* a revision ID has been assigned. This is for null edits.
3156 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3157 # int to remove the padding
3166 );
3175 false
3176 );
3182 # @bug 4594 PHP5 has it zero padded, PHP4 does not, cast to
3183 # int to remove the padding
3244 );
3247 }
3251 }
3254 }
3256 /**
3257 * initialise the magic variables (like CURRENTMONTHNAME) and substitution modifiers
3258 *
3259 * @private
3260 */
3267 }
3269 /**
3270 * Preprocess some wikitext and return the document tree.
3271 * This is the ghost of replace_variables().
3272 *
3273 * @param string $text The text to parse
3274 * @param int $flags Bitwise combination of:
3275 * - self::PTD_FOR_INCLUSION: Handle "<noinclude>" and "<includeonly>" as if the text is being
3276 * included. Default is to assume a direct page view.
3277 *
3278 * The generated DOM tree must depend only on the input text and the flags.
3279 * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
3280 *
3281 * Any flag added to the $flags parameter here, or any other parameter liable to cause a
3282 * change in the DOM tree for a given text, must be passed through the section identifier
3283 * in the section edit link and thus back to extractSections().
3284 *
3285 * The output of this function is currently only cached in process memory, but a persistent
3286 * cache may be implemented at a later date which takes further advantage of these strict
3287 * dependency requirements.
3288 *
3289 * @return PPNode
3290 */
3294 }
3296 /**
3297 * Return a three-element array: leading whitespace, string contents, trailing whitespace
3298 *
3299 * @param string $s
3300 *
3301 * @return array
3302 */
3312 }
3314 }
3316 /**
3317 * Replace magic variables, templates, and template arguments
3318 * with the appropriate text. Templates are substituted recursively,
3319 * taking care to avoid infinite loops.
3320 *
3321 * Note that the substitution depends on value of $mOutputType:
3322 * self::OT_WIKI: only {{subst:}} templates
3323 * self::OT_PREPROCESS: templates but not extension tags
3324 * self::OT_HTML: all templates and extension tags
3325 *
3326 * @param string $text The text to transform
3327 * @param bool|PPFrame $frame Object describing the arguments passed to the
3328 * template. Arguments may also be provided as an associative array, as
3329 * was the usual case before MW1.12. Providing arguments this way may be
3330 * useful for extensions wishing to perform variable replacement
3331 * explicitly.
3332 * @param bool $argsOnly Only do argument (triple-brace) expansion, not
3333 * double-brace expansion.
3334 * @return string
3335 */
3337 # Is there any text? Also, Prevent too big inclusions!
3341 }
3349 }
3356 }
3358 /**
3359 * Clean up argument array - refactored in 1.9 so parserfunctions can use it, too.
3360 *
3361 * @param array $args
3362 *
3363 * @return array
3364 */
3377 }
3380 }
3381 }
3382 }
3385 }
3387 /**
3388 * Warn the user when a parser limitation is reached
3389 * Will warn at most once the user per limitation type
3390 *
3391 * The results are shown during preview and run through the Parser (See EditPage.php)
3392 *
3393 * @param string $limitationType Should be one of:
3394 * 'expensive-parserfunction' (corresponding messages:
3395 * 'expensive-parserfunction-warning',
3396 * 'expensive-parserfunction-category')
3397 * 'post-expand-template-argument' (corresponding messages:
3398 * 'post-expand-template-argument-warning',
3399 * 'post-expand-template-argument-category')
3400 * 'post-expand-template-inclusion' (corresponding messages:
3401 * 'post-expand-template-inclusion-warning',
3402 * 'post-expand-template-inclusion-category')
3403 * 'node-count-exceeded' (corresponding messages:
3404 * 'node-count-exceeded-warning',
3405 * 'node-count-exceeded-category')
3406 * 'expansion-depth-exceeded' (corresponding messages:
3407 * 'expansion-depth-exceeded-warning',
3408 * 'expansion-depth-exceeded-category')
3409 * @param string|int|null $current Current value
3410 * @param string|int|null $max Maximum allowed, when an explicit limit has been
3411 * exceeded, provide the values (optional)
3412 */
3414 # does no harm if $current and $max are present but are unnecessary for the message
3415 # Not doing ->inLanguage( $this->mOptions->getUserLangObj() ), since this is shown
3416 # only during preview, and that would split the parser cache unnecessarily.
3421 }
3423 /**
3424 * Return the text of a template, after recursively
3425 * replacing any variables or templates within the template.
3426 *
3427 * @param array $piece The parts of the template
3428 * $piece['title']: the title, i.e. the part before the |
3429 * $piece['parts']: the parameter array
3430 * $piece['lineStart']: whether the brace was at the start of a line
3431 * @param PPFrame $frame The current frame, contains template arguments
3432 * @throws Exception
3433 * @return string The text of the template
3434 */
3437 // Flags
3439 // $text has been filled
3441 // wiki markup in $text should be escaped
3443 // $text is HTML, armour it against wikitext transformation
3445 // Force interwiki transclusion to be done in raw mode not rendered
3447 // $text is a DOM node needing expansion in a child frame
3449 // $text is a DOM node needing expansion in the current frame
3452 # Title object, where $text came from
3455 # $part1 is the bit before the first |, and must contain only title characters.
3456 # Various prefixes will be stripped from it later.
3461 # Original title text preserved for various purposes
3464 # $args is a list of argument nodes, starting from index 0, not including $part1
3465 # @todo FIXME: If piece['parts'] is null then the call to getLength()
3466 # below won't work b/c this $args isn't an object
3471 # SUBST
3475 # Possibilities for substMatch: "subst", "safesubst" or FALSE
3476 # Decide whether to expand template or keep wikitext as-is.
3482 }
3488 }
3489 }
3494 }
3495 }
3497 # Variables
3504 }
3506 }
3507 }
3509 # MSG, MSGNW and RAW
3511 # Check for MSGNW:
3516 # Remove obsolete MSG:
3519 }
3521 # Check for RAW:
3525 }
3526 }
3528 # Parser functions
3536 }
3541 }
3543 # The interface for parser functions allows for extracting
3544 # flags into the local scope. Extract any forwarded flags
3545 # here.
3547 }
3548 }
3550 # Finish mangling title and then check for loops.
3551 # Set $title to a Title object and $titleText to the PDBK
3554 # Split the title into page and subpage
3560 }
3564 # Check for language variants if the template is not found
3567 }
3568 # Do recursion depth check
3576 }
3577 }
3578 }
3580 # Load from database
3587 ) {
3588 // Pass the template arguments as URL parameters.
3589 // "uselang" will have no effect since the Language object
3590 // is forced to the one defined in ParserOptions.
3599 }
3600 }
3602 // Create a new context to execute the special page
3615 }
3625 }
3626 }
3628 # If the title is valid but undisplayable, make a link to it
3632 }
3634 # Interwiki transclusion
3640 # Preprocess it like a template
3643 }
3645 }
3647 # Do infinite loop check
3648 # This has to be done after redirect resolution to avoid infinite loops via redirects
3655 }
3656 }
3658 # If we haven't found text to substitute by now, we're done
3659 # Recover the source wikitext and return it
3664 }
3666 }
3668 # Expand DOM-style return values in a child frame
3670 # Clean up argument array
3676 # Expansion is eligible for the empty-frame cache
3679 # Uncached expansion
3681 }
3682 }
3686 }
3690 }
3692 # Replace raw HTML by a placeholder
3696 # Escape nowiki-style return values
3701 ) {
3702 # Bug 529: if the template begins with a table or block-level
3703 # element, it should be treated as beginning a new line.
3704 # This behavior is somewhat controversial.
3706 }
3709 # Error, oversize inclusion
3711 # Make a working, properly escaped link if possible (bug 23588)
3714 # This will probably not be a working link, but at least it may
3715 # provide some hint of where the problem is
3718 }
3722 }
3728 }
3731 }
3733 /**
3734 * Call a parser function and return an array with text and flags.
3735 *
3736 * The returned array will always contain a boolean 'found', indicating
3737 * whether the parser function was found or not. It may also contain the
3738 * following:
3739 * text: string|object, resulting wikitext or PP DOM object
3740 * isHTML: bool, $text is HTML, armour it against wikitext transformation
3741 * isChildObj: bool, $text is a DOM node needing expansion in a child frame
3742 * isLocalObj: bool, $text is a DOM node needing expansion in the current frame
3743 * nowiki: bool, wiki markup in $text should be escaped
3744 *
3745 * @since 1.21
3746 * @param PPFrame $frame The current frame, contains template arguments
3747 * @param string $function Function name
3748 * @param array $args Arguments to the function
3749 * @throws MWException
3750 * @return array
3751 */
3755 # Case sensitive functions
3759 # Case insensitive functions
3765 }
3766 }
3770 # Workaround for PHP bug 35229 and similar
3773 }
3777 # Convert arguments to PPNodes and collect for appending to $allArgs
3784 }
3785 }
3787 # Add a frame parameter, and pass the arguments as an array
3791 # Convert arguments to plain text and append to $allArgs
3799 }
3800 }
3801 }
3805 # The interface for function hooks allows them to return a wikitext
3806 # string or an array containing the string and any flags. This mungs
3807 # things around to match what this method should return.
3812 );
3816 }
3820 );
3821 }
3827 }
3830 }
3835 }
3838 }
3840 /**
3841 * Get the semi-parsed DOM representation of a template with a given title,
3842 * and its redirect destination title. Cached.
3843 *
3844 * @param Title $title
3845 *
3846 * @return array
3847 */
3856 }
3859 }
3861 # Cache miss, go to the database
3867 }
3875 }
3878 }
3880 /**
3881 * Fetch the current revision of a given title. Note that the revision
3882 * (and even the title) may not exist in the database, so everything
3883 * contributing to the output of the parser should use this method
3884 * where possible, rather than getting the revisions themselves. This
3885 * method also caches its results, so using it benefits performance.
3886 *
3887 * @since 1.24
3888 * @param Title $title
3889 * @return Revision
3890 */
3895 }
3898 // Defaults to Parser::statelessFetchRevision()
3900 );
3901 }
3903 }
3905 /**
3906 * Wrapper around Revision::newFromTitle to allow passing additional parameters
3907 * without passing them on to it.
3908 *
3909 * @since 1.24
3910 * @param Title $title
3911 * @param Parser|bool $parser
3912 * @return Revision
3913 */
3916 }
3918 /**
3919 * Fetch the unparsed text of a template and register a reference to it.
3920 * @param Title $title
3921 * @return array ( string or false, Title )
3922 */
3924 // Defaults to Parser::statelessFetchTemplate()
3927 // We use U+007F DELETE to distinguish strip markers from regular text.
3931 }
3937 // If we transclude ourselves, the final result
3938 // will change based on the new version of the page
3940 }
3941 }
3942 }
3944 }
3946 /**
3947 * Fetch the unparsed text of a template and register a reference to it.
3948 * @param Title $title
3949 * @return string|bool
3950 */
3954 }
3956 /**
3957 * Static function to get a template
3958 * Can be overridden via ParserOptions::setTemplateCallback().
3959 *
3960 * @param Title $title
3961 * @param bool|Parser $parser
3962 *
3963 * @return array
3964 */
3970 # Loop to fetch the article, with up to 1 redirect
3972 # Give extensions a chance to select the revision instead
3983 );
3985 }
3986 # Get the revision
3993 }
3995 # If there is no current revision, there is no page
3999 }
4006 # We fetched a rev from a different title; register it too...
4011 }
4020 }
4027 }
4032 }
4035 }
4036 # Redirect?
4039 }
4044 }
4046 /**
4047 * Fetch a file and its title and register a reference to it.
4048 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4049 * @param Title $title
4050 * @param array $options Array of options to RepoGroup::findFile
4051 * @return File|bool
4052 */
4056 }
4058 /**
4059 * Fetch a file and its title and register a reference to it.
4060 * If 'broken' is a key in $options then the file will appear as a broken thumbnail.
4061 * @param Title $title
4062 * @param array $options Array of options to RepoGroup::findFile
4063 * @return array ( File or false, Title of file )
4064 */
4070 # Register the file as a dependency...
4073 # Update fetched file title
4076 }
4078 }
4080 /**
4081 * Helper function for fetchFileAndTitle.
4082 *
4083 * Also useful if you need to fetch a file but not use it yet,
4084 * for example to get the file's handler.
4085 *
4086 * @param Title $title
4087 * @param array $options Array of options to RepoGroup::findFile
4088 * @return File|bool
4089 */
4097 }
4099 }
4101 /**
4102 * Transclude an interwiki link.
4103 *
4104 * @param Title $title
4105 * @param string $action
4106 *
4107 * @return string
4108 */
4114 }
4120 }
4122 }
4124 /**
4125 * @param string $url
4126 * @return mixed|string
4127 */
4136 }
4143 // Though we failed to fetch the content, this status is useless.
4148 }
4155 ) );
4157 }
4159 /**
4160 * Triple brace replacement -- used for template arguments
4161 * @private
4162 *
4163 * @param array $piece
4164 * @param PPFrame $frame
4165 *
4166 * @return array
4167 */
4180 )
4181 ) {
4182 # No match in frame, use the supplied default
4184 }
4188 }
4191 # No match anywhere
4193 }
4196 }
4201 }
4204 }
4206 /**
4207 * Return the text to be used for a given extension tag.
4208 * This is the ghost of strip().
4209 *
4210 * @param array $params Associative array of parameters:
4211 * name PPNode for the tag name
4212 * attr PPNode for unparsed text where tag attributes are thought to be
4213 * attributes Optional associative array of parsed attributes
4214 * inner Contents of extension element
4215 * noClose Original text did not have a close tag
4216 * @param PPFrame $frame
4217 *
4218 * @throws MWException
4219 * @return string
4220 */
4234 }
4240 }
4243 # Workaround for PHP bug 35229 and similar
4246 }
4253 }
4259 }
4262 # Extract flags to local scope (to override $markerType)
4267 }
4271 }
4276 }
4277 }
4283 }
4284 }
4294 }
4296 }
4298 /**
4299 * Increment an include size counter
4300 *
4301 * @param string $type The type of expansion
4302 * @param int $size The size of the text
4303 * @return bool False if this inclusion would take it over the maximum, true otherwise
4304 */
4311 }
4312 }
4314 /**
4315 * Increment the expensive function count
4316 *
4317 * @return bool False if the limit has been exceeded
4318 */
4322 }
4324 /**
4325 * Strip double-underscore items like __NOGALLERY__ and __NOTOC__
4326 * Fills $this->mDoubleUnderscores, returns the modified text
4327 *
4328 * @param string $text
4329 *
4330 * @return string
4331 */
4334 # The position of __TOC__ needs to be recorded
4340 # Set a placeholder. At the end we'll fill it in with the TOC.
4343 # Only keep the first one.
4345 }
4347 # Now match and remove the rest of them
4353 }
4356 }
4359 ) {
4361 }
4362 # (bug 8068) Allow control over whether robots index a page.
4363 #
4364 # @todo FIXME: Bug 14899: __INDEX__ always overrides __NOINDEX__ here! This
4365 # is not desirable, the last one on the page should win.
4369 }
4373 }
4375 # Cache all double underscores in the database
4378 }
4381 }
4383 /**
4384 * @see ParserOutput::addTrackingCategory()
4385 * @param string $msg Message key
4386 * @return bool Whether the addition was successful
4387 */
4390 }
4392 /**
4393 * This function accomplishes several tasks:
4394 * 1) Auto-number headings if that option is enabled
4395 * 2) Add an [edit] link to sections for users who have enabled the option and can edit the page
4396 * 3) Add a Table of contents on the top for users who have enabled the option
4397 * 4) Auto-anchor headings
4398 *
4399 * It loops through all headlines, collects the necessary data, then splits up the
4400 * string and re-inserts the newly formatted headlines.
4401 *
4402 * @param string $text
4403 * @param string $origText Original, untouched wikitext
4404 * @param bool $isMain
4405 * @return mixed|string
4406 * @private
4407 */
4411 # Inhibit editsection links if requested in the page
4417 }
4420 }
4422 # Get all headlines for numbering them and adding funky stuff like [edit]
4423 # links - this is for later, but we need the number of headlines right now
4428 $matches
4429 );
4431 # if there are fewer than 4 headlines in the article, do not show TOC
4432 # unless it's been explicitly enabled.
4436 # Allow user to stipulate that a page should have a "new section"
4437 # link added via __NEWSECTIONLINK__
4440 }
4442 # Allow user to remove the "new section"
4443 # link via __NONEWSECTIONLINK__
4446 }
4448 # if the string __FORCETOC__ (not case-sensitive) occurs in the HTML,
4449 # override above conditions and always show TOC above first header
4453 }
4455 # headline counter
4459 # Ugh .. the TOC should have neat indentation levels which can be
4460 # passed to the skin functions. These are determined here
4494 }
4498 }
4502 # Increase TOC level
4509 }
4511 # Decrease TOC level, find level to jump to
4515 # Found last matching level
4519 # Found first matching level below current level
4522 }
4523 }
4526 }
4529 # Unindent only if the previous toc level was shown :p
4534 }
4535 }
4537 # No change in level, end TOC line
4540 }
4541 }
4545 # count number of headlines for each level
4552 }
4555 }
4556 }
4558 # The safe header is a version of the header text safe to use for links
4560 # Remove link placeholders by the link text.
4561 # <!--LINK number-->
4562 # turns into
4563 # link text with suffix
4564 # Do this before unstrip since link text can contain strip markers
4567 # Avoid insertion of weird stuff like <math> by expanding the relevant sections
4570 # Strip out HTML (first regex removes any tag not allowed)
4571 # Allowed tags are:
4572 # * <sup> and <sub> (bug 8393)
4573 # * <i> (bug 26375)
4574 # * <b> (r105284)
4575 # * <bdi> (bug 72884)
4576 # * <span dir="rtl"> and <span dir="ltr"> (bug 35167)
4577 #
4578 # We strip any parameter from accepted tags (second regex), except dir="rtl|ltr" from <span>,
4579 # to allow setting directionality in toc items.
4583 '#<(/?(?:span(?: dir="(?:rtl|ltr)")?|sup|sub|bdi|i|b))(?: .*?)?>#'
4584 ),
4586 $safeHeadline
4587 );
4589 # Strip '<span></span>', which is the result from the above if
4590 # <span id="foo"></span> is used to produce an additional anchor
4591 # for a section.
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://www.w3.org/TR/html5/infrastructure.html#case-sensitivity-and-string-comparison
4631 # @todo FIXME: We may be changing them depending on the current locale.
4637 }
4639 # Create the anchor for linking from the TOC to the section
4648 }
4655 }
4657 # Don't number the heading if it is the only one (looks silly)
4659 # the two are different if the line contains a link
4663 $numbering
4665 }
4670 }
4672 # Add the section to the section tree
4673 # Find the DOM node for this header
4680 }
4681 }
4685 }
4695 );
4697 # give headline the correct <h#> tag
4699 // Output edit section links as markers with styles that can be customized by skins
4701 # Put a T flag in the section identifier, to indicate to extractSections()
4702 # that sections inside <includeonly> should be counted.
4710 }
4711 // We use a bit of pesudo-xml for editsection markers. The
4712 // language converter is run later on. Using a UNIQ style marker
4713 // leads to the converter screwing up the tokens when it
4714 // converts stuff. And trying to insert strip tags fails too. At
4715 // this point all real inputted tags have already been escaped,
4716 // so we don't have to worry about a user trying to input one of
4717 // these markers directly. We use a page and section attribute
4718 // to stop the language converter from converting these
4719 // important bits of data, but put the headline hint inside a
4720 // content block because the language converter is supposed to
4721 // be able to convert that piece of data.
4722 // Gets replaced with html in ParserOutput::getText
4729 }
4732 }
4738 }
4742 # Never ever show TOC if no headers
4745 }
4750 }
4755 }
4759 }
4761 # split up and insert constructed headlines
4765 // build an array of document sections
4768 // $head is zero-based, sections aren't.
4773 }
4775 /**
4776 * Send a hook, one per section.
4777 * The idea here is to be able to make section-level DIVs, but to do so in a
4778 * lower-impact, more correct way than r50769
4779 *
4780 * $this : caller
4781 * $section : the section number
4782 * &$sectionContent : ref to the content of the section
4783 * $showEditLinks : boolean describing whether this section has an edit link
4784 */
4788 }
4791 // append the TOC at the beginning
4792 // Top anchor now in skin
4794 }
4802 }
4803 }
4805 /**
4806 * Transform wiki markup when saving a page by doing "\r\n" -> "\n"
4807 * conversion, substituting signatures, {{subst:}} templates, etc.
4808 *
4809 * @param string $text The text to transform
4810 * @param Title $title The Title object for the current article
4811 * @param User $user The User object describing the current user
4812 * @param ParserOptions $options Parsing options
4813 * @param bool $clearState Whether to clear the parser state first
4814 * @return string The altered wiki markup
4815 */
4818 ) {
4821 }
4828 );
4832 }
4838 }
4840 /**
4841 * Pre-save transform helper function
4842 *
4843 * @param string $text
4844 * @param User $user
4845 *
4846 * @return string
4847 */
4851 # Note: This is the timestamp saved as hardcoded wikitext to
4852 # the database, we use $wgContLang here in order to give
4853 # everyone the same signature and use the default one rather
4854 # than the one selected in each user's preferences.
4855 # (see also bug 12815)
4863 # Variable replacement
4864 # Because mOutputType is OT_WIKI, this will only process {{subst:xxx}} type tags
4867 # This works almost by chance, as the replaceVariables are done before the getUserSig(),
4868 # which may corrupt this parser instance via its wfMessage()->text() call-
4870 # Signatures
4876 ) );
4878 # Context links ("pipe tricks"): [[|name]] and [[name (context)|]]
4882 // [[ns:page (context)|]]
4884 // [[ns:page（context）|]] (double-width brackets, added in r40257)
4886 // [[ns:page (context), context|]] (using either single or double-width comma)
4888 // [[|page]] (reverse pipe trick: add context from page title)
4891 # try $p1 first, to turn "[[A, B (C)|]]" into "[[A, B (C)|A, B]]"
4903 # if there's no context, don't bother duplicating the title
4905 }
4907 # Trim trailing whitespace
4911 }
4913 /**
4914 * Fetch the user's signature text, if any, and normalize to
4915 * validated, ready-to-insert wikitext.
4916 * If you have pre-fetched the nickname or the fancySig option, you can
4917 * specify them here to save a database query.
4918 * Do not reuse this parser instance after calling getUserSig(),
4919 * as it may have changed if it's the $wgParser.
4920 *
4921 * @param User $user
4922 * @param string|bool $nickname Nickname to use or false to use user's default nickname
4923 * @param bool|null $fancySig whether the nicknname is the complete signature
4924 * or null to use default value
4925 * @return string
4926 */
4932 # If not given, retrieve from the user object.
4935 }
4939 }
4947 # Sig. might contain markup; validate this
4949 # Validated; clean up (if needed) and return it
4952 # Failed to validate; fall back to the default
4955 }
4956 }
4958 # Make sure nickname doesnt get a sig in a sig
4961 # If we're still here, make it a link to the user page
4968 }
4970 /**
4971 * Check that the user's signature contains no bad XML
4972 *
4973 * @param string $text
4974 * @return string|bool An expanded string, or false if invalid.
4975 */
4978 }
4980 /**
4981 * Clean up signature text
4982 *
4983 * 1) Strip 3, 4 or 5 tildes out of signatures @see cleanSigInSig
4984 * 2) Substitute all transclusions
4985 *
4986 * @param string $text
4987 * @param bool $parsing Whether we're cleaning (preferences save) or parsing
4988 * @return string Signature text
4989 */
4995 }
4997 # Option to disable this feature
5000 }
5002 # @todo FIXME: Regex doesn't respect extension tags or nowiki
5003 # => Move this logic to braceSubstitution()
5005 $substRegex = '/\{\{(?!(?:' . $substWord->getBaseRegex() . '))/x' . $substWord->getRegexCase();
5016 }
5019 }
5021 /**
5022 * Strip 3, 4 or 5 tildes out of signatures.
5023 *
5024 * @param string $text
5025 * @return string Signature text with /~{3,5}/ removed
5026 */
5030 }
5032 /**
5033 * Set up some variables which are usually set up in parse()
5034 * so that an external function can call some class members with confidence
5035 *
5036 * @param Title|null $title
5037 * @param ParserOptions $options
5038 * @param int $outputType
5039 * @param bool $clearState
5040 */
5043 ) {
5045 }
5047 /**
5048 * @param Title|null $title
5049 * @param ParserOptions $options
5050 * @param int $outputType
5051 * @param bool $clearState
5052 */
5055 ) {
5061 }
5062 }
5064 /**
5065 * Wrapper for preprocess()
5066 *
5067 * @param string $text The text to preprocess
5068 * @param ParserOptions $options Options
5069 * @param Title|null $title Title object or null to use $wgTitle
5070 * @return string
5071 */
5075 # Guard against infinite recursion
5078 }
5084 }
5090 }
5092 /**
5093 * Create an HTML-style tag, e.g. "<yourtag>special text</yourtag>"
5094 * The callback should have the following form:
5095 * function myParserHook( $text, $params, $parser, $frame ) { ... }
5096 *
5097 * Transform and return $text. Use $parser for any required context, e.g. use
5098 * $parser->getTitle() and $parser->getOptions() not $wgTitle or $wgOut->mParserOptions
5099 *
5100 * Hooks may return extended information by returning an array, of which the
5101 * first numbered element (index 0) must be the return string, and all other
5102 * entries are extracted into local variables within an internal function
5103 * in the Parser class.
5104 *
5105 * This interface (introduced r61913) appears to be undocumented, but
5106 * 'markerType' is used by some core tag hooks to override which strip
5107 * array their results are placed in. **Use great caution if attempting
5108 * this interface, as it is not documented and injudicious use could smash
5109 * private variables.**
5110 *
5111 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5112 * @param callable $callback The callback function (and object) to use for the tag
5113 * @throws MWException
5114 * @return callable|null The old value of the mTagHooks array associated with the hook
5115 */
5120 }
5125 }
5128 }
5130 /**
5131 * As setHook(), but letting the contents be parsed.
5132 *
5133 * Transparent tag hooks are like regular XML-style tag hooks, except they
5134 * operate late in the transformation sequence, on HTML instead of wikitext.
5135 *
5136 * This is probably obsoleted by things dealing with parser frames?
5137 * The only extension currently using it is geoserver.
5138 *
5139 * @since 1.10
5140 * @todo better document or deprecate this
5141 *
5142 * @param string $tag The tag to use, e.g. 'hook' for "<hook>"
5143 * @param callable $callback The callback function (and object) to use for the tag
5144 * @throws MWException
5145 * @return callable|null The old value of the mTagHooks array associated with the hook
5146 */
5151 }
5152 $oldVal = isset( $this->mTransparentTagHooks[$tag] ) ? $this->mTransparentTagHooks[$tag] : null;
5156 }
5158 /**
5159 * Remove all tag hooks
5160 */
5165 }
5167 /**
5168 * Create a function, e.g. {{sum:1|2|3}}
5169 * The callback function should have the form:
5170 * function myParserFunction( &$parser, $arg1, $arg2, $arg3 ) { ... }
5171 *
5172 * Or with Parser::SFH_OBJECT_ARGS:
5173 * function myParserFunction( $parser, $frame, $args ) { ... }
5174 *
5175 * The callback may either return the text result of the function, or an array with the text
5176 * in element 0, and a number of flags in the other elements. The names of the flags are
5177 * specified in the keys. Valid flags are:
5178 * found The text returned is valid, stop processing the template. This
5179 * is on by default.
5180 * nowiki Wiki markup in the return value should be escaped
5181 * isHTML The returned text is HTML, armour it against wikitext transformation
5182 *
5183 * @param string $id The magic word ID
5184 * @param callable $callback The callback function (and object) to use
5185 * @param int $flags A combination of the following flags:
5186 * Parser::SFH_NO_HASH No leading hash, i.e. {{plural:...}} instead of {{#if:...}}
5187 *
5188 * Parser::SFH_OBJECT_ARGS Pass the template arguments as PPNode objects instead of text.
5189 * This allows for conditional expansion of the parse tree, allowing you to eliminate dead
5190 * branches and thus speed up parsing. It is also possible to analyse the parse tree of
5191 * the arguments, and to control the way they are expanded.
5192 *
5193 * The $frame parameter is a PPFrame. This can be used to produce expanded text from the
5194 * arguments, for instance:
5195 * $text = isset( $args[0] ) ? $frame->expand( $args[0] ) : '';
5196 *
5197 * For technical reasons, $args[0] is pre-expanded and will be a string. This may change in
5198 * future versions. Please call $frame->expand() on it anyway so that your code keeps
5199 * working if/when this is changed.
5200 *
5201 * If you want whitespace to be trimmed from $args, you need to do it yourself, post-
5202 * expansion.
5203 *
5204 * Please read the documentation in includes/parser/Preprocessor.php for more information
5205 * about the methods available in PPFrame and PPNode.
5206 *
5207 * @throws MWException
5208 * @return string|callable The old callback function for this name, if any
5209 */
5216 # Add to function cache
5220 }
5226 # Case
5229 }
5230 # Add leading hash
5233 }
5234 # Remove trailing colon
5237 }
5239 }
5241 }
5243 /**
5244 * Get all registered function hook identifiers
5245 *
5246 * @return array
5247 */
5250 }
5252 /**
5253 * Create a tag function, e.g. "<test>some stuff</test>".
5254 * Unlike tag hooks, tag functions are parsed at preprocessor level.
5255 * Unlike parser functions, their content is not preprocessed.
5256 * @param string $tag
5257 * @param callable $callback
5258 * @param int $flags
5259 * @throws MWException
5260 * @return null
5261 */
5266 }
5273 }
5276 }
5278 /**
5279 * Replace "<!--LINK-->" link placeholders with actual links, in the buffer
5280 * Placeholders created in Linker::link()
5281 *
5282 * @param string $text
5283 * @param int $options
5284 */
5287 }
5289 /**
5290 * Replace "<!--LINK-->" link placeholders with plain text of links
5291 * (not HTML-formatted).
5292 *
5293 * @param string $text
5294 * @return string
5295 */
5298 }
5300 /**
5301 * Renders an image gallery from a text with one line per image.
5302 * text labels may be given by using |-style alternative text. E.g.
5303 * Image:one.jpg|The number "1"
5304 * Image:tree.jpg|A tree
5305 * given as text will return the HTML of a gallery with two images,
5306 * labeled 'The number "1"' and
5307 * 'A tree'.
5308 *
5309 * @param string $text
5310 * @param array $params
5311 * @return string HTML
5312 */
5318 }
5323 // If invalid type set, fallback to default.
5325 }
5338 }
5344 }
5347 }
5350 }
5353 }
5360 # match lines like these:
5361 # Image:someimage.jpg|This is some image
5364 # Skip empty lines
5367 }
5371 }
5374 # Bogus title. Ignore these so we don't bomb out later.
5376 }
5378 # We need to get what handler the file uses, to figure out parameters.
5379 # Note, a hook can overide the file name, and chose an entirely different
5380 # file (which potentially could be of a different type and have different handler).
5385 # Don't register it now, as ImageGallery does that later.
5392 );
5395 // We don't want people to specify per-image widths.
5396 // Additionally the width parameter would need special casing anyhow.
5398 }
5407 // look for an |alt= definition while trying not to break existing
5408 // captions with multiple pipes (|) in it, until a more sensible grammar
5409 // is defined for images in galleries
5411 // FIXME: Doing recursiveTagParse at this stage, and the trim before
5412 // splitting on '|' is a bit odd, and different from makeImage.
5430 //check to see if link matches an absolute url, if not then it must be a wiki link.
5437 }
5438 }
5441 // Must be a handler specific parameter.
5445 // Guess not. Append it to the caption.
5448 }
5449 }
5452 // concatenate all other pipes
5454 }
5455 }
5456 // remove the first pipe
5458 }
5461 }
5465 }
5467 /**
5468 * @param MediaHandler $handler
5469 * @return array
5470 */
5476 }
5478 # Initialise static lists
5485 );
5493 }
5494 }
5495 }
5497 # Add handler params
5503 }
5504 }
5507 }
5508 return array( $this->mImageParams[$handlerClass], $this->mImageParamsMagicArray[$handlerClass] );
5509 }
5511 /**
5512 * Parse image options text and use it to make an image
5513 *
5514 * @param Title $title
5515 * @param string $options
5516 * @param LinkHolderArray|bool $holders
5517 * @return string HTML
5518 */
5520 # Check if the options text is of the form "options|alt text"
5521 # Options are:
5522 # * thumbnail make a thumbnail with enlarge-icon and caption, alignment depends on lang
5523 # * left no resizing, just left align. label is used for alt= only
5524 # * right same, but right aligned
5525 # * none same, but not aligned
5526 # * ___px scale to ___ pixels width, no aligning. e.g. use in taxobox
5527 # * center center the image
5528 # * frame Keep original image size, no magnify-button.
5529 # * framed Same as "frame"
5530 # * frameless like 'thumb' but without a frame. Keeps user preferences for width
5531 # * upright reduce width for upright images, rounded to full __0 px
5532 # * border draw a 1px border around the image
5533 # * alt Text for HTML alt attribute (defaults to empty)
5534 # * class Set a class for img node
5535 # * link Set the target of the image link. Can be external, interwiki, or local
5536 # vertical-align values (no % or length right now):
5537 # * baseline
5538 # * sub
5539 # * super
5540 # * top
5541 # * text-top
5542 # * middle
5543 # * bottom
5544 # * text-bottom
5548 # Give extensions a chance to select the file revision for us
5553 # Fetch and register the file (file title may be different via hooks)
5556 # Get parameter map
5563 }
5565 # Process the input parameters
5577 # Special case; width and height come in one variable together
5585 }
5586 }
5592 }
5593 }
5594 # else no validation -- bug 13436
5597 # Validate handler parameter
5600 # Validate internal parameters
5605 # @todo FIXME: Possibly check validity here for
5606 # manualthumb? downstream behavior seems odd with
5607 # missing manual thumbs.
5625 }
5627 }
5635 }
5636 }
5641 // use first appearing option, discard others.
5646 # Most other things appear to be empty or numeric...
5648 }
5649 }
5653 }
5654 }
5655 }
5658 }
5659 }
5661 # Process alignment parameters
5664 }
5667 }
5671 # Will the image be presented in a frame, with the caption below?
5677 # In the old days, [[Image:Foo|text...]] would set alt text. Later it
5678 # came to also set the caption, ordinary text after the image -- which
5679 # makes no sense, because that just repeats the text multiple times in
5680 # screen readers. It *also* came to set the title attribute.
5681 #
5682 # Now that we have an alt attribute, we should not set the alt text to
5683 # equal the caption: that's worse than useless, it just repeats the
5684 # text. This is the framed/thumbnail case. If there's no caption, we
5685 # use the unnamed parameter for alt text as well, just for the time be-
5686 # ing, if the unnamed param is set and the alt param is not.
5687 #
5688 # For the future, we need to figure out if we want to tweak this more,
5689 # e.g., introducing a title= parameter for the title; ignoring the un-
5690 # named parameter entirely for images without a caption; adding an ex-
5691 # plicit caption= parameter and preserving the old magic unnamed para-
5692 # meter for BC; ...
5695 # No caption or alt text, add the filename as the alt text so
5696 # that screen readers at least get some description of the image
5698 }
5699 # Do not set $params['frame']['title'] because tooltips don't make sense
5700 # for framed images
5703 # No alt text, use the "caption" for the alt text
5707 # No caption, fall back to using the filename for the
5708 # alt text
5710 }
5711 }
5712 # Use the "caption" for the tooltip text
5714 }
5718 # Linker does the rest
5723 # Give the handler a chance to modify the parser object
5726 }
5729 }
5731 /**
5732 * @param string $caption
5733 * @param LinkHolderArray|bool $holders
5734 * @return mixed|string
5735 */
5737 # Strip bad stuff out of the title (tooltip). We can't just use
5738 # replaceLinkHoldersText() here, because if this function is called
5739 # from replaceInternalLinks2(), mLinkHolders won't be up-to-date.
5744 }
5746 # make sure there are no placeholders in thumbnail attributes
5747 # that are later expanded to html- so expand them now and
5748 # remove the tags
5753 }
5755 /**
5756 * Set a flag in the output object indicating that the content is dynamic and
5757 * shouldn't be cached.
5758 */
5764 }
5766 }
5768 /**
5769 * Callback from the Sanitizer for expanding items found in HTML attribute
5770 * values, so they can be safely tested and escaped.
5771 *
5772 * @param string $text
5773 * @param bool|PPFrame $frame
5774 * @return string
5775 */
5780 }
5782 /**
5783 * Accessor
5784 *
5785 * @return array
5786 */
5792 );
5793 }
5795 /**
5796 * Replace transparent tags in $text with the values given by the callbacks.
5797 *
5798 * Transparent tag hooks are like regular XML-style tag hooks, except they
5799 * operate late in the transformation sequence, on HTML instead of wikitext.
5800 *
5801 * @param string $text
5802 *
5803 * @return string
5804 */
5818 );
5821 }
5823 }
5825 }
5827 /**
5828 * Break wikitext input into sections, and either pull or replace
5829 * some particular section's text.
5830 *
5831 * External callers should use the getSection and replaceSection methods.
5832 *
5833 * @param string $text Page wikitext
5834 * @param string|number $sectionId A section identifier string of the form:
5835 * "<flag1> - <flag2> - ... - <section number>"
5836 *
5837 * Currently the only recognised flag is "T", which means the target section number
5838 * was derived during a template inclusion parse, in other words this is a template
5839 * section edit link. If no flags are given, it was an ordinary section edit link.
5840 * This flag is required to avoid a section numbering mismatch when a section is
5841 * enclosed by "<includeonly>" (bug 6563).
5842 *
5843 * The section number 0 pulls the text before the first heading; other numbers will
5844 * pull the given section along with its lower-level subsections. If the section is
5845 * not found, $mode=get will return $newtext, and $mode=replace will return $text.
5846 *
5847 * Section 0 is always considered to exist, even if it only contains the empty
5848 * string. If $text is the empty string and section 0 is replaced, $newText is
5849 * returned.
5850 *
5851 * @param string $mode One of "get" or "replace"
5852 * @param string $newText Replacement text for section data.
5853 * @return string For "get", the extracted section text.
5854 * for "replace", the whole page with the section replaced.
5855 */
5864 # Process section extraction flags
5871 }
5872 }
5874 # Check for empty input
5876 # Only sections 0 and T-0 exist in an empty document
5882 }
5888 }
5889 }
5890 }
5892 # Preprocess the text
5895 # <h> nodes indicate section breaks
5896 # They can only occur at the top level, so we can find them by iterating the root's children
5899 # Find the target section
5901 # Section zero doesn't nest, level=big
5910 }
5911 }
5914 }
5916 }
5917 }
5920 # Not found
5925 }
5926 }
5928 # Find the end of the section, including nested sections
5935 }
5936 }
5939 }
5943 # Write out the remainder (in replace mode only)
5945 # Output the replacement text
5946 # Add two newlines on -- trailing whitespace in $newText is conventionally
5947 # stripped by the editor, so we need both newlines to restore the paragraph gap
5948 # Only add trailing whitespace if there is newText
5951 }
5956 }
5957 }
5960 # Re-insert stripped tags
5962 }
5965 }
5967 /**
5968 * This function returns the text of a section, specified by a number ($section).
5969 * A section is text under a heading like == Heading == or \<h1\>Heading\</h1\>, or
5970 * the first section before any such heading (section 0).
5971 *
5972 * If a section contains subsections, these are also returned.
5973 *
5974 * @param string $text Text to look in
5975 * @param string|number $sectionId Section identifier as a number or string
5976 * (e.g. 0, 1 or 'T-1').
5977 * @param string $defaultText Default to return if section is not found
5978 *
5979 * @return string Text of the requested section
5980 */
5983 }
5985 /**
5986 * This function returns $oldtext after the content of the section
5987 * specified by $section has been replaced with $text. If the target
5988 * section does not exist, $oldtext is returned unchanged.
5989 *
5990 * @param string $oldText Former text of the article
5991 * @param string|number $sectionId Section identifier as a number or string
5992 * (e.g. 0, 1 or 'T-1').
5993 * @param string $newText Replacing text
5994 *
5995 * @return string Modified text
5996 */
5999 }
6001 /**
6002 * Get the ID of the revision we are parsing
6003 *
6004 * @return int|null
6005 */
6008 }
6010 /**
6011 * Get the revision object for $this->mRevisionId
6012 *
6013 * @return Revision|null Either a Revision object or null
6014 * @since 1.23 (public since 1.23)
6015 */
6019 }
6022 }
6026 );
6028 # If the parse is for a new revision, then the callback should have
6029 # already been set to force the object and should match mRevisionId.
6030 # If not, try to fetch by mRevisionId for sanity.
6033 }
6038 }
6040 /**
6041 * Get the timestamp associated with the current revision, adjusted for
6042 * the default server-local timestamp
6043 * @return string
6044 */
6052 # The cryptic '' timezone parameter tells to use the site-default
6053 # timezone offset instead of the user settings.
6054 #
6055 # Since this value will be saved into the parser cache, served
6056 # to other users, and potentially even used inside links and such,
6057 # it needs to be consistent for all visitors.
6060 }
6062 }
6064 /**
6065 * Get the name of the user that edited the last revision
6066 *
6067 * @return string User name
6068 */
6073 # if this template is subst: the revision id will be blank,
6074 # so just use the current user's name
6079 }
6080 }
6082 }
6084 /**
6085 * Get the size of the revision
6086 *
6087 * @return int|null Revision size
6088 */
6093 # if this variable is subst: the revision id will be blank,
6094 # so just use the parser input size, because the own substituation
6095 # will change the size.
6100 }
6101 }
6103 }
6105 /**
6106 * Mutator for $mDefaultSort
6107 *
6108 * @param string $sort New value
6109 */
6113 }
6115 /**
6116 * Accessor for $mDefaultSort
6117 * Will use the empty string if none is set.
6118 *
6119 * This value is treated as a prefix, so the
6120 * empty string is equivalent to sorting by
6121 * page name.
6122 *
6123 * @return string
6124 */
6130 }
6131 }
6133 /**
6134 * Accessor for $mDefaultSort
6135 * Unlike getDefaultSort(), will return false if none is set
6136 *
6137 * @return string|bool
6138 */
6141 }
6143 /**
6144 * Try to guess the section anchor name based on a wikitext fragment
6145 * presumably extracted from a heading, for example "Header" from
6146 * "== Header ==".
6147 *
6148 * @param string $text
6149 *
6150 * @return string
6151 */
6153 # Strip out wikitext links(they break the anchor)
6157 }
6159 /**
6160 * Same as guessSectionNameFromWikiText(), but produces legacy anchors
6161 * instead. For use in redirects, since IE6 interprets Redirect: headers
6162 * as something other than UTF-8 (apparently?), resulting in breakage.
6163 *
6164 * @param string $text The section name
6165 * @return string An anchor
6166 */
6168 # Strip out wikitext links(they break the anchor)
6172 }
6174 /**
6175 * Strips a text string of wikitext for use in a section anchor
6176 *
6177 * Accepts a text string and then removes all wikitext from the
6178 * string and leaves only the resultant text (i.e. the result of
6179 * [[User:WikiSysop|Sysop]] would be "Sysop" and the result of
6180 * [[User:WikiSysop]] would be "User:WikiSysop") - this is intended
6181 * to create valid section anchors by mimicing the output of the
6182 * parser when headings are parsed.
6183 *
6184 * @param string $text Text string to be stripped of wikitext
6185 * for use in a Section anchor
6186 * @return string Filtered text string
6187 */
6189 # Strip internal link markup
6193 # Strip external link markup
6194 # @todo FIXME: Not tolerant to blank link text
6195 # I.E. [https://www.mediawiki.org] will render as [1] or something depending
6196 # on how many empty links there are on the page - need to figure that out.
6197 $text = preg_replace( '/\[(?i:' . $this->mUrlProtocols . ')([^ ]+?) ([^[]+)\]/', '$2', $text );
6199 # Parse wikitext quotes (italics & bold)
6202 # Strip HTML tags
6205 }
6207 /**
6208 * strip/replaceVariables/unstrip for preprocessor regression testing
6209 *
6210 * @param string $text
6211 * @param Title $title
6212 * @param ParserOptions $options
6213 * @param int $outputType
6214 *
6215 * @return string
6216 */
6217 public function testSrvus( $text, Title $title, ParserOptions $options, $outputType = self::OT_HTML ) {
6225 }
6227 /**
6228 * @param string $text
6229 * @param Title $title
6230 * @param ParserOptions $options
6231 * @return string
6232 */
6235 }
6237 /**
6238 * @param string $text
6239 * @param Title $title
6240 * @param ParserOptions $options
6241 * @return string
6242 */
6245 }
6247 /**
6248 * Call a callback function on all regions of the given text that are not
6249 * inside strip markers, and replace those regions with the return value
6250 * of the callback. For example, with input:
6251 *
6252 * aaa<MARKER>bbb
6253 *
6254 * This will call the callback function twice, with 'aaa' and 'bbb'. Those
6255 * two strings will be replaced with the value returned by the callback in
6256 * each case.
6257 *
6258 * @param string $s
6259 * @param callable $callback
6260 *
6261 * @return string
6262 */
6281 }
6282 }
6283 }
6285 }
6287 /**
6288 * Remove any strip markers found in the given text.
6289 *
6290 * @param string $text Input string
6291 * @return string
6292 */
6295 }
6297 /**
6298 * Save the parser state required to convert the given half-parsed text to
6299 * HTML. "Half-parsed" in this context means the output of
6300 * recursiveTagParse() or internalParse(). This output has strip markers
6301 * from replaceVariables (extensionSubstitution() etc.), and link
6302 * placeholders from replaceLinkHolders().
6303 *
6304 * Returns an array which can be serialized and stored persistently. This
6305 * array can later be loaded into another parser instance with
6306 * unserializeHalfParsedText(). The text can then be safely incorporated into
6307 * the return value of a parser hook.
6308 *
6309 * @param string $text
6310 *
6311 * @return array
6312 */
6319 );
6321 }
6323 /**
6324 * Load the parser state given in the $data array, which is assumed to
6325 * have been generated by serializeHalfParsedText(). The text contents is
6326 * extracted from the array, and its markers are transformed into markers
6327 * appropriate for the current Parser instance. This transformed text is
6328 * returned, and can be safely included in the return value of a parser
6329 * hook.
6330 *
6331 * If the $data array has been stored persistently, the caller should first
6332 * check whether it is still valid, by calling isValidHalfParsedText().
6333 *
6334 * @param array $data Serialized data
6335 * @throws MWException
6336 * @return string
6337 */
6341 }
6343 # First, extract the strip state.
6347 # Now renumber links
6350 # Should be good to go.
6352 }
6354 /**
6355 * Returns true if the given array, presumed to be generated by
6356 * serializeHalfParsedText(), is compatible with the current version of the
6357 * parser.
6358 *
6359 * @param array $data
6360 *
6361 * @return bool
6362 */
6365 }
6367 /**
6368 * Parsed a width param of imagelink like 300px or 200x300px
6369 *
6370 * @param string $value
6371 *
6372 * @return array
6373 * @since 1.20
6374 */
6379 }
6381 # (bug 13500) In both cases (width/height and width only),
6382 # permit trailing "px" for backward compatibility.
6391 }
6393 }
6395 /**
6396 * Lock the current instance of the parser.
6397 *
6398 * This is meant to stop someone from calling the parser
6399 * recursively and messing up all the strip state.
6400 *
6401 * @throws MWException If parser is in a parse
6402 * @return ScopedCallback The lock will be released once the return value goes out of scope.
6403 */
6408 }
6414 } );
6417 }
6419 /**
6420 * Strip outer <p></p> tag from the HTML source of a single paragraph.
6421 *
6422 * Returns original HTML if the <p/> tag has any attributes, if there's no wrapping <p/> tag,
6423 * or if there is more than one <p/> tag in the input HTML.
6424 *
6425 * @param string $html
6426 * @return string
6427 * @since 1.24
6428 */
6434 }
6435 }
6438 }
6440 /**
6441 * Return this parser if it is not doing anything, otherwise
6442 * get a fresh parser. You can use this method by doing
6443 * $myParser = $wgParser->getFreshParser(), or more simply
6444 * $wgParser->getFreshParser()->parse( ... );
6445 * if you're unsure if $wgParser is safe to use.
6446 *
6447 * @since 1.24
6448 * @return Parser A parser object that is not parsing anything
6449 */
6456 }
6457 }
6459 /**
6460 * Set's up the PHP implementation of OOUI for use in this request
6461 * and instructs OutputPage to enable OOUI for itself.
6462 *
6463 * @since 1.26
6464 */
6468 }
6469 }