*
* - Parser::parse()
* produces HTML output
- * - Parser::preSaveTransform().
- * produces altered wiki markup.
+ * - Parser::preSaveTransform()
+ * produces altered wiki markup
* - Parser::preprocess()
* removes HTML comments and expands templates
* - Parser::cleanSig() and Parser::cleanSigInSig()
- * Cleans a signature before saving it to preferences
+ * cleans a signature before saving it to preferences
* - Parser::getSection()
- * Return the content of a section from an article for section editing
+ * return the content of a section from an article for section editing
* - Parser::replaceSection()
- * Replaces a section by number inside an article
+ * replaces a section by number inside an article
* - Parser::getPreloadText()
- * Removes <noinclude> sections, and <includeonly> tags.
+ * removes <noinclude> sections and <includeonly> tags
*
* Globals used:
* object: $wgContLang
*/
public $mInParse = false;
+ /** @var SectionProfiler */
+ protected $mProfiler;
+
/**
* @param array $conf
*/
$this->mPreprocessor = null;
}
+ $this->mProfiler = new SectionProfiler();
+
wfRunHooks( 'ParserClearState', array( &$this ) );
wfProfileOut( __METHOD__ );
}
* to internalParse() which does all the real work.
*/
- global $wgUseTidy, $wgAlwaysUseTidy, $wgShowHostnames;
+ global $wgShowHostnames;
$fname = __METHOD__ . '-' . wfGetCaller();
wfProfileIn( __METHOD__ );
wfProfileIn( $fname );
$text = $this->internalParse( $text );
wfRunHooks( 'ParserAfterParse', array( &$this, &$text, &$this->mStripState ) );
- $text = $this->mStripState->unstripGeneral( $text );
-
- # Clean up special characters, only run once, next-to-last before doBlockLevels
- $fixtags = array(
- # french spaces, last one Guillemet-left
- # only if there is something before the space
- '/(.) (?=\\?|:|;|!|%|\\302\\273)/' => '\\1 ',
- # french spaces, Guillemet-right
- '/(\\302\\253) /' => '\\1 ',
- '/ (!\s*important)/' => ' \\1', # Beware of CSS magic word !important, bug #11874.
- );
- $text = preg_replace( array_keys( $fixtags ), array_values( $fixtags ), $text );
-
- $text = $this->doBlockLevels( $text, $linestart );
-
- $this->replaceLinkHolders( $text );
-
- /**
- * The input doesn't get language converted if
- * a) It's disabled
- * b) Content isn't converted
- * c) It's a conversion table
- * d) it is an interface message (which is in the user language)
- */
- if ( !( $options->getDisableContentConversion()
- || isset( $this->mDoubleUnderscores['nocontentconvert'] ) )
- ) {
- if ( !$this->mOptions->getInterfaceMessage() ) {
- # The position of the convert() call should not be changed. it
- # assumes that the links are all replaced and the only thing left
- # is the <nowiki> mark.
- $text = $this->getConverterLanguage()->convert( $text );
- }
- }
+ $text = $this->internalParseHalfParsed( $text, true, $linestart );
/**
* A converted title will be provided in the output object if title and
}
}
- $text = $this->mStripState->unstripNoWiki( $text );
-
- wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
-
- $text = $this->replaceTransparentTags( $text );
- $text = $this->mStripState->unstripGeneral( $text );
-
- $text = Sanitizer::normalizeCharReferences( $text );
-
- if ( ( $wgUseTidy && $this->mOptions->getTidy() ) || $wgAlwaysUseTidy ) {
- $text = MWTidy::tidy( $text );
- } else {
- # attempt to sanitize at least some nesting problems
- # (bug #2702 and quite a few others)
- $tidyregs = array(
- # ''Something [http://www.cool.com cool''] -->
- # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
- '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
- '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
- # fix up an anchor inside another anchor, only
- # at least for a single single nested link (bug 3695)
- '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
- '\\1\\2</a>\\3</a>\\1\\4</a>',
- # fix div inside inline elements- doBlockLevels won't wrap a line which
- # contains a div, so fix it up here; replace
- # div with escaped text
- '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
- '\\1\\3<div\\5>\\6</div>\\8\\9',
- # remove empty italic or bold tag pairs, some
- # introduced by rules above
- '/<([bi])><\/\\1>/' => '',
- );
-
- $text = preg_replace(
- array_keys( $tidyregs ),
- array_values( $tidyregs ),
- $text );
- }
-
if ( $this->mExpensiveFunctionCount > $this->mOptions->getExpensiveParserFunctionLimit() ) {
$this->limitationWarn( 'expensive-parserfunction',
$this->mExpensiveFunctionCount,
);
}
- wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
-
# Information on include size limits, for the benefit of users who try to skirt them
if ( $this->mOptions->getEnableLimitReport() ) {
$max = $this->mOptions->getMaxIncludeSize();
$limitReport = str_replace( array( '-', '&' ), array( '‐', '&' ), $limitReport );
$text .= "\n<!-- \n$limitReport-->\n";
+ // Add on template profiling data
+ $dataByFunc = $this->mProfiler->getFunctionStats();
+ uasort( $dataByFunc, function( $a, $b ) {
+ return $a['real'] < $b['real']; // descending order
+ } );
+ $profileReport = "Transclusion expansion time report (%,ms,calls,template)\n";
+ foreach ( array_slice( $dataByFunc, 0, 10 ) as $item ) {
+ $profileReport .= sprintf( "%6.2f%% %8.3f %6d - %s\n",
+ $item['%real'], $item['real'], $item['calls'],
+ htmlspecialchars($item['name'] ) );
+ }
+ $text .= "\n<!-- \n$profileReport-->\n";
+
if ( $this->mGeneratedPPNodeCount > $this->mOptions->getMaxGeneratedPPNodeCount() / 10 ) {
wfDebugLog( 'generated-pp-node-count', $this->mGeneratedPPNodeCount . ' ' .
$this->mTitle->getPrefixedDBkey() );
}
/**
- * Recursive parser entry point that can be called from an extension tag
- * hook.
+ * Half-parse wikitext to half-parsed HTML. This recursive parser entry point
+ * can be called from an extension tag hook.
+ *
+ * The output of this function IS NOT SAFE PARSED HTML; it is "half-parsed"
+ * instead, which means that lists and links have not been fully parsed yet,
+ * and strip markers are still present.
+ *
+ * Use recursiveTagParseFully() to fully parse wikitext to output-safe HTML.
*
- * If $frame is not provided, then template variables (e.g., {{{1}}}) within $text are not expanded
+ * Use this function if you're a parser tag hook and you want to parse
+ * wikitext before or after applying additional transformations, and you
+ * intend to *return the result as hook output*, which will cause it to go
+ * through the rest of parsing process automatically.
+ *
+ * If $frame is not provided, then template variables (e.g., {{{1}}}) within
+ * $text are not expanded
*
* @param string $text Text extension wants to have parsed
* @param bool|PPFrame $frame The frame to use for expanding any template variables
- *
- * @return string
+ * @return string UNSAFE half-parsed HTML
*/
public function recursiveTagParse( $text, $frame = false ) {
wfProfileIn( __METHOD__ );
return $text;
}
+ /**
+ * Fully parse wikitext to fully parsed HTML. This recursive parser entry
+ * point can be called from an extension tag hook.
+ *
+ * The output of this function is fully-parsed HTML that is safe for output.
+ * If you're a parser tag hook, you might want to use recursiveTagParse()
+ * instead.
+ *
+ * If $frame is not provided, then template variables (e.g., {{{1}}}) within
+ * $text are not expanded
+ *
+ * @since 1.25
+ *
+ * @param string $text Text extension wants to have parsed
+ * @param bool|PPFrame $frame The frame to use for expanding any template variables
+ * @return string Fully parsed HTML
+ */
+ public function recursiveTagParseFully( $text, $frame = false ) {
+ wfProfileIn( __METHOD__ );
+ $text = $this->recursiveTagParse( $text, $frame );
+ $text = $this->internalParseHalfParsed( $text, false );
+ wfProfileOut( __METHOD__ );
+ return $text;
+ }
+
/**
* Expand templates and variables in the text, producing valid, static wikitext.
* Also removes comments.
}
/**
- * Helper function for parse() that transforms wiki markup into
+ * Helper function for parse() that transforms wiki markup into half-parsed
* HTML. Only called for $mOutputType == self::OT_HTML.
*
* @private
return $text;
}
+ /**
+ * Helper function for parse() that transforms half-parsed HTML into fully
+ * parsed HTML.
+ *
+ * @param string $text
+ * @param bool $isMain
+ * @param bool $linestart
+ * @return string
+ */
+ private function internalParseHalfParsed( $text, $isMain = true, $linestart = true ) {
+ global $wgUseTidy, $wgAlwaysUseTidy;
+
+ $text = $this->mStripState->unstripGeneral( $text );
+
+ # Clean up special characters, only run once, next-to-last before doBlockLevels
+ $fixtags = array(
+ # french spaces, last one Guillemet-left
+ # only if there is something before the space
+ '/(.) (?=\\?|:|;|!|%|\\302\\273)/' => '\\1 ',
+ # french spaces, Guillemet-right
+ '/(\\302\\253) /' => '\\1 ',
+ '/ (!\s*important)/' => ' \\1', # Beware of CSS magic word !important, bug #11874.
+ );
+ $text = preg_replace( array_keys( $fixtags ), array_values( $fixtags ), $text );
+
+ $text = $this->doBlockLevels( $text, $linestart );
+
+ $this->replaceLinkHolders( $text );
+
+ /**
+ * The input doesn't get language converted if
+ * a) It's disabled
+ * b) Content isn't converted
+ * c) It's a conversion table
+ * d) it is an interface message (which is in the user language)
+ */
+ if ( !( $this->mOptions->getDisableContentConversion()
+ || isset( $this->mDoubleUnderscores['nocontentconvert'] ) )
+ ) {
+ if ( !$this->mOptions->getInterfaceMessage() ) {
+ # The position of the convert() call should not be changed. it
+ # assumes that the links are all replaced and the only thing left
+ # is the <nowiki> mark.
+ $text = $this->getConverterLanguage()->convert( $text );
+ }
+ }
+
+ $text = $this->mStripState->unstripNoWiki( $text );
+
+ if ( $isMain ) {
+ wfRunHooks( 'ParserBeforeTidy', array( &$this, &$text ) );
+ }
+
+ $text = $this->replaceTransparentTags( $text );
+ $text = $this->mStripState->unstripGeneral( $text );
+
+ $text = Sanitizer::normalizeCharReferences( $text );
+
+ if ( ( $wgUseTidy && $this->mOptions->getTidy() ) || $wgAlwaysUseTidy ) {
+ $text = MWTidy::tidy( $text );
+ } else {
+ # attempt to sanitize at least some nesting problems
+ # (bug #2702 and quite a few others)
+ $tidyregs = array(
+ # ''Something [http://www.cool.com cool''] -->
+ # <i>Something</i><a href="http://www.cool.com"..><i>cool></i></a>
+ '/(<([bi])>)(<([bi])>)?([^<]*)(<\/?a[^<]*>)([^<]*)(<\/\\4>)?(<\/\\2>)/' =>
+ '\\1\\3\\5\\8\\9\\6\\1\\3\\7\\8\\9',
+ # fix up an anchor inside another anchor, only
+ # at least for a single single nested link (bug 3695)
+ '/(<a[^>]+>)([^<]*)(<a[^>]+>[^<]*)<\/a>(.*)<\/a>/' =>
+ '\\1\\2</a>\\3</a>\\1\\4</a>',
+ # fix div inside inline elements- doBlockLevels won't wrap a line which
+ # contains a div, so fix it up here; replace
+ # div with escaped text
+ '/(<([aib]) [^>]+>)([^<]*)(<div([^>]*)>)(.*)(<\/div>)([^<]*)(<\/\\2>)/' =>
+ '\\1\\3<div\\5>\\6</div>\\8\\9',
+ # remove empty italic or bold tag pairs, some
+ # introduced by rules above
+ '/<([bi])><\/\\1>/' => '',
+ );
+
+ $text = preg_replace(
+ array_keys( $tidyregs ),
+ array_values( $tidyregs ),
+ $text );
+ }
+
+ if ( $isMain ) {
+ wfRunHooks( 'ParserAfterTidy', array( &$this, &$text ) );
+ }
+
+ return $text;
+ }
+
/**
* Replace special strings like "ISBN xxx" and "RFC xxx" with
* magic external links.
$t
);
- if ( $openmatch or $closematch ) {
+ if ( $openmatch || $closematch ) {
$paragraphStack = false;
# @todo bug 5718: paragraph closed
$output .= $this->closeParagraph();
- if ( $preOpenMatch and !$preCloseMatch ) {
+ if ( $preOpenMatch && !$preCloseMatch ) {
$this->mInPre = true;
}
$bqOffset = 0;
case 'numberofedits':
$value = $pageLang->formatNum( SiteStats::edits() );
break;
- case 'numberofviews':
- global $wgDisableCounters;
- $value = !$wgDisableCounters ? $pageLang->formatNum( SiteStats::views() ) : '';
- break;
case 'currenttimestamp':
$value = wfTimestamp( TS_MW, $ts );
break;
$args = ( null == $piece['parts'] ) ? array() : $piece['parts'];
wfProfileOut( __METHOD__ . '-setup' );
- $titleProfileIn = null; // profile templates
+ $profileSection = null; // profile templates
# SUBST
wfProfileIn( __METHOD__ . '-modifiers' );
# Load from database
if ( !$found && $title ) {
- if ( !Profiler::instance()->isPersistent() ) {
- # Too many unique items can kill profiling DBs/collectors
- $titleProfileIn = __METHOD__ . "-title-" . $title->getPrefixedDBkey();
- wfProfileIn( $titleProfileIn ); // template in
- }
+ $profileSection = $this->mProfiler->scopedProfileIn( $title->getPrefixedDBkey() );
wfProfileIn( __METHOD__ . '-loadtpl' );
if ( !$title->isExternal() ) {
if ( $title->isSpecialPage()
# Recover the source wikitext and return it
if ( !$found ) {
$text = $frame->virtualBracketedImplode( '{{', '|', '}}', $titleWithSpaces, $args );
- if ( $titleProfileIn ) {
- wfProfileOut( $titleProfileIn ); // template out
+ if ( $profileSection ) {
+ $this->mProfiler->scopedProfileOut( $profileSection );
}
wfProfileOut( __METHOD__ );
return array( 'object' => $text );
$isLocalObj = false;
}
- if ( $titleProfileIn ) {
- wfProfileOut( $titleProfileIn ); // template out
+ if ( $profileSection ) {
+ $this->mProfiler->scopedProfileOut( $profileSection );
}
# Replace raw HTML by a placeholder
/**
* Transform wiki markup when saving a page by doing "\r\n" -> "\n"
- * conversion, substitting signatures, {{subst:}} templates, etc.
+ * conversion, substituting signatures, {{subst:}} templates, etc.
*
* @param string $text The text to transform
* @param Title $title The Title object for the current article
*
* @param string $text
* @param int $options
- *
- * @return array Array of link CSS classes, indexed by PDBK.
*/
public function replaceLinkHolders( &$text, $options = 0 ) {
- return $this->mLinkHolders->replace( $text );
+ $this->mLinkHolders->replace( $text );
}
/**
dépôts / lhc / web / wiklou.git / blobdiff