parser.php

来自「php 开发的内容管理系统」· PHP 代码 · 共 2,095 行 · 第 1/5 页

						$y = "{$z}<{$l}>{$y[0]}" ;
					else {
						$attributes = $this->unstripForHTML( $y[0] );
						$y = "{$z}<{$l}".Sanitizer::fixTagAttributes($attributes, $l).">{$y[1]}" ;
					}
					$t[$k] .= $y ;
					array_push ( $td , true ) ;
				}
			}
		}

		# Closing open td, tr && table
		while ( count ( $td ) > 0 )
		{
			$l = array_pop ( $ltd ) ;
			if ( array_pop ( $td ) ) $t[] = '</td>' ;
			if ( array_pop ( $tr ) ) $t[] = '</tr>' ;
			if ( !array_pop ( $has_opened_tr ) ) $t[] = "<tr><td></td></tr>" ;
			$t[] = '</table>' ;
		}

		$t = implode ( "\n" , $t ) ;
		# special case: don't return empty table
		if($t == "<table>\n<tr><td></td></tr>\n</table>")
			$t = '';
		wfProfileOut( $fname );
		return $t ;
	}

	/**
	 * Helper function for parse() that transforms wiki markup into
	 * HTML. Only called for $mOutputType == OT_HTML.
	 *
	 * @private
	 */
	function internalParse( $text ) {
		$args = array();
		$isMain = true;
		$fname = 'Parser::internalParse';
		wfProfileIn( $fname );

		# Remove <noinclude> tags and <includeonly> sections
		$text = strtr( $text, array( '<onlyinclude>' => '' , '</onlyinclude>' => '' ) );
		$text = strtr( $text, array( '<noinclude>' => '', '</noinclude>' => '') );
		$text = preg_replace( '/<includeonly>.*?<\/includeonly>/s', '', $text );
		
		$text = Sanitizer::removeHTMLtags( $text, array( &$this, 'attributeStripCallback' ) );

		$text = $this->replaceVariables( $text, $args );

		// Tables need to come after variable replacement for things to work
		// properly; putting them before other transformations should keep
		// exciting things like link expansions from showing up in surprising
		// places.
		$text = $this->doTableStuff( $text );

		$text = preg_replace( '/(^|\n)-----*/', '\\1<hr />', $text );

		$text = $this->stripToc( $text );
		$this->stripNoGallery( $text );
		$text = $this->doHeadings( $text );
		if($this->mOptions->getUseDynamicDates()) {
			$df =& DateFormatter::getInstance();
			$text = $df->reformat( $this->mOptions->getDateFormat(), $text );
		}
		$text = $this->doAllQuotes( $text );
		$text = $this->replaceInternalLinks( $text );
		$text = $this->replaceExternalLinks( $text );

		# replaceInternalLinks may sometimes leave behind
		# absolute URLs, which have to be masked to hide them from replaceExternalLinks
		$text = str_replace($this->mUniqPrefix."NOPARSE", "", $text);

		$text = $this->doMagicLinks( $text );
		$text = $this->formatHeadings( $text, $isMain );

		wfProfileOut( $fname );
		return $text;
	}

	/**
	 * Replace special strings like "ISBN xxx" and "RFC xxx" with
	 * magic external links.
	 *
	 * @private
	 */
	function &doMagicLinks( &$text ) {
		$text = $this->magicISBN( $text );
		$text = $this->magicRFC( $text, 'RFC ', 'rfcurl' );
		$text = $this->magicRFC( $text, 'PMID ', 'pubmedurl' );
		return $text;
	}

	/**
	 * Parse headers and return html
	 *
	 * @private
	 */
	function doHeadings( $text ) {
		$fname = 'Parser::doHeadings';
		wfProfileIn( $fname );
		for ( $i = 6; $i >= 1; --$i ) {
			$h = str_repeat( '=', $i );
			$text = preg_replace( "/^{$h}(.+){$h}\\s*$/m",
			  "<h{$i}>\\1</h{$i}>\\2", $text );
		}
		wfProfileOut( $fname );
		return $text;
	}

	/**
	 * Replace single quotes with HTML markup
	 * @private
	 * @return string the altered text
	 */
	function doAllQuotes( $text ) {
		$fname = 'Parser::doAllQuotes';
		wfProfileIn( $fname );
		$outtext = '';
		$lines = explode( "\n", $text );
		foreach ( $lines as $line ) {
			$outtext .= $this->doQuotes ( $line ) . "\n";
		}
		$outtext = substr($outtext, 0,-1);
		wfProfileOut( $fname );
		return $outtext;
	}

	/**
	 * Helper function for doAllQuotes()
	 * @private
	 */
	function doQuotes( $text ) {
		$arr = preg_split( "/(''+)/", $text, -1, PREG_SPLIT_DELIM_CAPTURE );
		if ( count( $arr ) == 1 )
			return $text;
		else
		{
			# First, do some preliminary work. This may shift some apostrophes from
			# being mark-up to being text. It also counts the number of occurrences
			# of bold and italics mark-ups.
			$i = 0;
			$numbold = 0;
			$numitalics = 0;
			foreach ( $arr as $r )
			{
				if ( ( $i % 2 ) == 1 )
				{
					# If there are ever four apostrophes, assume the first is supposed to
					# be text, and the remaining three constitute mark-up for bold text.
					if ( strlen( $arr[$i] ) == 4 )
					{
						$arr[$i-1] .= "'";
						$arr[$i] = "'''";
					}
					# If there are more than 5 apostrophes in a row, assume they're all
					# text except for the last 5.
					else if ( strlen( $arr[$i] ) > 5 )
					{
						$arr[$i-1] .= str_repeat( "'", strlen( $arr[$i] ) - 5 );
						$arr[$i] = "'''''";
					}
					# Count the number of occurrences of bold and italics mark-ups.
					# We are not counting sequences of five apostrophes.
					if ( strlen( $arr[$i] ) == 2 ) $numitalics++;  else
					if ( strlen( $arr[$i] ) == 3 ) $numbold++;     else
					if ( strlen( $arr[$i] ) == 5 ) { $numitalics++; $numbold++; }
				}
				$i++;
			}

			# If there is an odd number of both bold and italics, it is likely
			# that one of the bold ones was meant to be an apostrophe followed
			# by italics. Which one we cannot know for certain, but it is more
			# likely to be one that has a single-letter word before it.
			if ( ( $numbold % 2 == 1 ) && ( $numitalics % 2 == 1 ) )
			{
				$i = 0;
				$firstsingleletterword = -1;
				$firstmultiletterword = -1;
				$firstspace = -1;
				foreach ( $arr as $r )
				{
					if ( ( $i % 2 == 1 ) and ( strlen( $r ) == 3 ) )
					{
						$x1 = substr ($arr[$i-1], -1);
						$x2 = substr ($arr[$i-1], -2, 1);
						if ($x1 == ' ') {
							if ($firstspace == -1) $firstspace = $i;
						} else if ($x2 == ' ') {
							if ($firstsingleletterword == -1) $firstsingleletterword = $i;
						} else {
							if ($firstmultiletterword == -1) $firstmultiletterword = $i;
						}
					}
					$i++;
				}

				# If there is a single-letter word, use it!
				if ($firstsingleletterword > -1)
				{
					$arr [ $firstsingleletterword ] = "''";
					$arr [ $firstsingleletterword-1 ] .= "'";
				}
				# If not, but there's a multi-letter word, use that one.
				else if ($firstmultiletterword > -1)
				{
					$arr [ $firstmultiletterword ] = "''";
					$arr [ $firstmultiletterword-1 ] .= "'";
				}
				# ... otherwise use the first one that has neither.
				# (notice that it is possible for all three to be -1 if, for example,
				# there is only one pentuple-apostrophe in the line)
				else if ($firstspace > -1)
				{
					$arr [ $firstspace ] = "''";
					$arr [ $firstspace-1 ] .= "'";
				}
			}

			# Now let's actually convert our apostrophic mush to HTML!
			$output = '';
			$buffer = '';
			$state = '';
			$i = 0;
			foreach ($arr as $r)
			{
				if (($i % 2) == 0)
				{
					if ($state == 'both')
						$buffer .= $r;
					else
						$output .= $r;
				}
				else
				{
					if (strlen ($r) == 2)
					{
						if ($state == 'i')
						{ $output .= '</i>'; $state = ''; }
						else if ($state == 'bi')
						{ $output .= '</i>'; $state = 'b'; }
						else if ($state == 'ib')
						{ $output .= '</b></i><b>'; $state = 'b'; }
						else if ($state == 'both')
						{ $output .= '<b><i>'.$buffer.'</i>'; $state = 'b'; }
						else # $state can be 'b' or ''
						{ $output .= '<i>'; $state .= 'i'; }
					}
					else if (strlen ($r) == 3)
					{
						if ($state == 'b')
						{ $output .= '</b>'; $state = ''; }
						else if ($state == 'bi')
						{ $output .= '</i></b><i>'; $state = 'i'; }
						else if ($state == 'ib')
						{ $output .= '</b>'; $state = 'i'; }
						else if ($state == 'both')
						{ $output .= '<i><b>'.$buffer.'</b>'; $state = 'i'; }
						else # $state can be 'i' or ''
						{ $output .= '<b>'; $state .= 'b'; }
					}
					else if (strlen ($r) == 5)
					{
						if ($state == 'b')
						{ $output .= '</b><i>'; $state = 'i'; }
						else if ($state == 'i')
						{ $output .= '</i><b>'; $state = 'b'; }
						else if ($state == 'bi')
						{ $output .= '</i></b>'; $state = ''; }
						else if ($state == 'ib')
						{ $output .= '</b></i>'; $state = ''; }
						else if ($state == 'both')
						{ $output .= '<i><b>'.$buffer.'</b></i>'; $state = ''; }
						else # ($state == '')
						{ $buffer = ''; $state = 'both'; }
					}
				}
				$i++;
			}
			# Now close all remaining tags.  Notice that the order is important.
			if ($state == 'b' || $state == 'ib')
				$output .= '</b>';
			if ($state == 'i' || $state == 'bi' || $state == 'ib')
				$output .= '</i>';
			if ($state == 'bi')
				$output .= '</b>';
			if ($state == 'both')
				$output .= '<b><i>'.$buffer.'</i></b>';
			return $output;
		}
	}

	/**
	 * Replace external links
	 *
 	 * Note: this is all very hackish and the order of execution matters a lot.
	 * Make sure to run maintenance/parserTests.php if you change this code.
	 *
	 * @private
	 */
	function replaceExternalLinks( $text ) {
		global $wgContLang;
		$fname = 'Parser::replaceExternalLinks';
		wfProfileIn( $fname );

		$sk =& $this->mOptions->getSkin();

		$bits = preg_split( EXT_LINK_BRACKETED, $text, -1, PREG_SPLIT_DELIM_CAPTURE );

		$s = $this->replaceFreeExternalLinks( array_shift( $bits ) );

		$i = 0;
		while ( $i<count( $bits ) ) {
			$url = $bits[$i++];
			$protocol = $bits[$i++];
			$text = $bits[$i++];
			$trail = $bits[$i++];

			# The characters '<' and '>' (which were escaped by
			# removeHTMLtags()) should not be included in
			# URLs, per RFC 2396.
			if (preg_match('/&(lt|gt);/', $url, $m2, PREG_OFFSET_CAPTURE)) {
				$text = substr($url, $m2[0][1]) . ' ' . $text;
				$url = substr($url, 0, $m2[0][1]);
			}

			# If the link text is an image URL, replace it with an <img> tag
			# This happened by accident in the original parser, but some people used it extensively
			$img = $this->maybeMakeExternalImage( $text );
			if ( $img !== false ) {
				$text = $img;
			}

			$dtrail = '';

			# Set linktype for CSS - if URL==text, link is essentially free
			$linktype = ($text == $url) ? 'free' : 'text';

			# No link text, e.g. [http://domain.tld/some.link]
			if ( $text == '' ) {
				# Autonumber if allowed. See bug #5918
				if ( strpos( wfUrlProtocols(), substr($protocol, 0, strpos($protocol, ':')) ) !== false ) {
					$text = '[' . ++$this->mAutonumber . ']';
					$linktype = 'autonumber';
				} else {
					# Otherwise just use the URL
					$text = htmlspecialchars( $url );
					$linktype = 'free';
				}
			} else {
				# Have link text, e.g. [http://domain.tld/some.link text]s
				# Check for trail
				list( $dtrail, $trail ) = Linker::splitTrail( $trail );
			}

			$text = $wgContLang->markNoConversion($text);

			# Normalize any HTML entities in input. They will be
			# re-escaped by makeExternalLink().
			$url = Sanitizer::decodeCharReferences( $url );
			
			# Escape any control characters introduced by the above step
			$url = preg_replace( '/[\][<>"\\x00-\\x20\\x7F]/e', "urlencode('\\0')", $url );

			# Process the trail (i.e. everything after this link up until start of the next link),
			# replacing any non-bracketed links
			$trail = $this->replaceFreeExternalLinks( $trail );

			# Use the encoded URL
			# This means that users can paste URLs directly into the text
			# Funny characters like &ouml; aren't valid in URLs anyway
			# This was changed in August 2004
			$s .= $sk->makeExternalLink( $url, $text, false, $linktype, $this->mTitle->getNamespace() ) . $dtrail . $trail;

			# Register link in the output object.
			# Replace unnecessary URL escape codes with the referenced character
			# This prevents spammers from hiding links from the filters
			$pasteurized = Parser::replaceUnusualEscapes( $url );
			$this->mOutput->addExternalLink( $pasteurized );
		}

		wfProfileOut( $fname );
		return $s;
	}

	/**
	 * Replace anything that looks like a URL with a link
	 * @private
	 */
	function replaceFreeExternalLinks( $text ) {
		global $wgContLang;
		$fname = 'Parser::replaceFreeExternalLinks';
		wfProfileIn( $fname );

		$bits = preg_split( '/(\b(?:' . wfUrlProtocols() . '))/S', $text, -1, PREG_SPLIT_DELIM_CAPTURE );
		$s = array_shift( $bits );
		$i = 0;

		$sk =& $this->mOptions->getSkin();

		while ( $i < count( $bits ) ){
			$protocol = $bits[$i++];
			$remainder = $bits[$i++];

			if ( preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $remainder, $m ) ) {
				# Found some characters after the protocol that look promising
				$url = $protocol . $m[1];
				$trail = $m[2];

				# special case: handle urls as url args:
				# http://www.example.com/foo?=http://www.example.com/bar
				if(strlen($trail) == 0 && 
					isset($bits[$i]) &&
					preg_match('/^'. wfUrlProtocols() . '$/S', $bits[$i]) &&
					preg_match( '/^('.EXT_LINK_URL_CLASS.'+)(.*)$/s', $bits[$i + 1], $m )) 
				{
					# add protocol, arg
					$url .= $bits[$i] . $m[1]; # protocol, url as arg to previous link