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

					$i += 2;
					$trail = $m[2];
				}

				# 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)) {
					$trail = substr($url, $m2[0][1]) . $trail;
					$url = substr($url, 0, $m2[0][1]);
				}

				# Move trailing punctuation to $trail
				$sep = ',;\.:!?';
				# If there is no left bracket, then consider right brackets fair game too
				if ( strpos( $url, '(' ) === false ) {
					$sep .= ')';
				}

				$numSepChars = strspn( strrev( $url ), $sep );
				if ( $numSepChars ) {
					$trail = substr( $url, -$numSepChars ) . $trail;
					$url = substr( $url, 0, -$numSepChars );
				}

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

				# Is this an external image?
				$text = $this->maybeMakeExternalImage( $url );
				if ( $text === false ) {
					# Not an image, make a link
					$text = $sk->makeExternalLink( $url, $wgContLang->markNoConversion($url), true, 'free', $this->mTitle->getNamespace() );
					# Register it in the output object...
					# Replace unnecessary URL escape codes with their equivalent characters
					$pasteurized = Parser::replaceUnusualEscapes( $url );
					$this->mOutput->addExternalLink( $pasteurized );
				}
				$s .= $text . $trail;
			} else {
				$s .= $protocol . $remainder;
			}
		}
		wfProfileOut( $fname );
		return $s;
	}

	/**
	 * Replace unusual URL escape codes with their equivalent characters
	 * @param string
	 * @return string
	 * @static
	 * @fixme This can merge genuinely required bits in the path or query string,
	 *        breaking legit URLs. A proper fix would treat the various parts of
	 *        the URL differently; as a workaround, just use the output for
	 *        statistical records, not for actual linking/output.
	 */
	function replaceUnusualEscapes( $url ) {
		return preg_replace_callback( '/%[0-9A-Fa-f]{2}/',
			array( 'Parser', 'replaceUnusualEscapesCallback' ), $url );
	}

	/**
	 * Callback function used in replaceUnusualEscapes().
	 * Replaces unusual URL escape codes with their equivalent character
	 * @static
	 * @private
	 */
	function replaceUnusualEscapesCallback( $matches ) {
		$char = urldecode( $matches[0] );
		$ord = ord( $char );
		// Is it an unsafe or HTTP reserved character according to RFC 1738?
		if ( $ord > 32 && $ord < 127 && strpos( '<>"#{}|\^~[]`;/?', $char ) === false ) {
			// No, shouldn't be escaped
			return $char;
		} else {
			// Yes, leave it escaped
			return $matches[0];
		}
	}

	/**
	 * make an image if it's allowed, either through the global
	 * option or through the exception
	 * @private
	 */
	function maybeMakeExternalImage( $url ) {
		$sk =& $this->mOptions->getSkin();
		$imagesfrom = $this->mOptions->getAllowExternalImagesFrom();
		$imagesexception = !empty($imagesfrom);
		$text = false;
		if ( $this->mOptions->getAllowExternalImages()
		     || ( $imagesexception && strpos( $url, $imagesfrom ) === 0 ) ) {
			if ( preg_match( EXT_IMAGE_REGEX, $url ) ) {
				# Image found
				$text = $sk->makeExternalImage( htmlspecialchars( $url ) );
			}
		}
		return $text;
	}

	/**
	 * Process [[ ]] wikilinks
	 *
	 * @private
	 */
	function replaceInternalLinks( $s ) {
		global $wgContLang;
		static $fname = 'Parser::replaceInternalLinks' ;

		wfProfileIn( $fname );

		wfProfileIn( $fname.'-setup' );
		static $tc = FALSE;
		# the % is needed to support urlencoded titles as well
		if ( !$tc ) { $tc = Title::legalChars() . '#%'; }

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

		#split the entire text string on occurences of [[
		$a = explode( '[[', ' ' . $s );
		#get the first element (all text up to first [[), and remove the space we added
		$s = array_shift( $a );
		$s = substr( $s, 1 );

		# Match a link having the form [[namespace:link|alternate]]trail
		static $e1 = FALSE;
		if ( !$e1 ) { $e1 = "/^([{$tc}]+)(?:\\|(.+?))?]](.*)\$/sD"; }
		# Match cases where there is no "]]", which might still be images
		static $e1_img = FALSE;
		if ( !$e1_img ) { $e1_img = "/^([{$tc}]+)\\|(.*)\$/sD"; }
		# Match the end of a line for a word that's not followed by whitespace,
		# e.g. in the case of 'The Arab al[[Razi]]', 'al' will be matched
		$e2 = wfMsgForContent( 'linkprefix' );

		$useLinkPrefixExtension = $wgContLang->linkPrefixExtension();

		if( is_null( $this->mTitle ) ) {
			throw new MWException( 'nooo' );
		}
		$nottalk = !$this->mTitle->isTalkPage();

		if ( $useLinkPrefixExtension ) {
			if ( preg_match( $e2, $s, $m ) ) {
				$first_prefix = $m[2];
			} else {
				$first_prefix = false;
			}
		} else {
			$prefix = '';
		}

		$selflink = $this->mTitle->getPrefixedText();
		wfProfileOut( $fname.'-setup' );

		$checkVariantLink = sizeof($wgContLang->getVariants())>1;
		$useSubpages = $this->areSubpagesAllowed();

		# Loop for each link
		for ($k = 0; isset( $a[$k] ); $k++) {
			$line = $a[$k];
			if ( $useLinkPrefixExtension ) {
				wfProfileIn( $fname.'-prefixhandling' );
				if ( preg_match( $e2, $s, $m ) ) {
					$prefix = $m[2];
					$s = $m[1];
				} else {
					$prefix='';
				}
				# first link
				if($first_prefix) {
					$prefix = $first_prefix;
					$first_prefix = false;
				}
				wfProfileOut( $fname.'-prefixhandling' );
			}

			$might_be_img = false;

			if ( preg_match( $e1, $line, $m ) ) { # page with normal text or alt
				$text = $m[2];
				# If we get a ] at the beginning of $m[3] that means we have a link that's something like:
				# [[Image:Foo.jpg|[http://example.com desc]]] <- having three ] in a row fucks up,
				# the real problem is with the $e1 regex
				# See bug 1300.
				#
				# Still some problems for cases where the ] is meant to be outside punctuation,
				# and no image is in sight. See bug 2095.
				#
				if( $text !== '' && 
					preg_match( "/^\](.*)/s", $m[3], $n ) && 
					strpos($text, '[') !== false 
				) 
				{
					$text .= ']'; # so that replaceExternalLinks($text) works later
					$m[3] = $n[1];
				}
				# fix up urlencoded title texts
				if(preg_match('/%/', $m[1] )) 
					# Should anchors '#' also be rejected?
					$m[1] = str_replace( array('<', '>'), array('&lt;', '&gt;'), urldecode($m[1]) );
				$trail = $m[3];
			} elseif( preg_match($e1_img, $line, $m) ) { # Invalid, but might be an image with a link in its caption
				$might_be_img = true;
				$text = $m[2];
				if(preg_match('/%/', $m[1] )) $m[1] = urldecode($m[1]);
				$trail = "";
			} else { # Invalid form; output directly
				$s .= $prefix . '[[' . $line ;
				continue;
			}

			# Don't allow internal links to pages containing
			# PROTO: where PROTO is a valid URL protocol; these
			# should be external links.
			if (preg_match('/^(\b(?:' . wfUrlProtocols() . '))/', $m[1])) {
				$s .= $prefix . '[[' . $line ;
				continue;
			}

			# Make subpage if necessary
			if( $useSubpages ) {
				$link = $this->maybeDoSubpageLink( $m[1], $text );
			} else {
				$link = $m[1];
			}

			$noforce = (substr($m[1], 0, 1) != ':');
			if (!$noforce) {
				# Strip off leading ':'
				$link = substr($link, 1);
			}

			$nt = Title::newFromText( $this->unstripNoWiki($link, $this->mStripState) );
			if( !$nt ) {
				$s .= $prefix . '[[' . $line;
				continue;
			}

			#check other language variants of the link
			#if the article does not exist
			if( $checkVariantLink
			    && $nt->getArticleID() == 0 ) {
				$wgContLang->findVariantLink($link, $nt);
			}

			$ns = $nt->getNamespace();
			$iw = $nt->getInterWiki();

			if ($might_be_img) { # if this is actually an invalid link
				if ($ns == NS_IMAGE && $noforce) { #but might be an image
					$found = false;
					while (isset ($a[$k+1]) ) {
						#look at the next 'line' to see if we can close it there
						$spliced = array_splice( $a, $k + 1, 1 );
						$next_line = array_shift( $spliced );
						if( preg_match("/^(.*?]].*?)]](.*)$/sD", $next_line, $m) ) {
						# the first ]] closes the inner link, the second the image
							$found = true;
							$text .= '[[' . $m[1];
							$trail = $m[2];
							break;
						} elseif( preg_match("/^.*?]].*$/sD", $next_line, $m) ) {
							#if there's exactly one ]] that's fine, we'll keep looking
							$text .= '[[' . $m[0];
						} else {
							#if $next_line is invalid too, we need look no further
							$text .= '[[' . $next_line;
							break;
						}
					}
					if ( !$found ) {
						# we couldn't find the end of this imageLink, so output it raw
						#but don't ignore what might be perfectly normal links in the text we've examined
						$text = $this->replaceInternalLinks($text);
						$s .= $prefix . '[[' . $link . '|' . $text;
						# note: no $trail, because without an end, there *is* no trail
						continue;
					}
				} else { #it's not an image, so output it raw
					$s .= $prefix . '[[' . $link . '|' . $text;
					# note: no $trail, because without an end, there *is* no trail
					continue;
				}
			}

			$wasblank = ( '' == $text );
			if( $wasblank ) $text = $link;


			# Link not escaped by : , create the various objects
			if( $noforce ) {

				# Interwikis
				if( $iw && $this->mOptions->getInterwikiMagic() && $nottalk && $wgContLang->getLanguageName( $iw ) ) {
					$this->mOutput->addLanguageLink( $nt->getFullText() );
					$s = rtrim($s . "\n");
					$s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;
					continue;
				}

				if ( $ns == NS_IMAGE ) {
					wfProfileIn( "$fname-image" );
					if ( !wfIsBadImage( $nt->getDBkey() ) ) {
						# recursively parse links inside the image caption
						# actually, this will parse them in any other parameters, too,
						# but it might be hard to fix that, and it doesn't matter ATM
						$text = $this->replaceExternalLinks($text);
						$text = $this->replaceInternalLinks($text);

						# cloak any absolute URLs inside the image markup, so replaceExternalLinks() won't touch them
						$s .= $prefix . $this->armorLinks( $this->makeImage( $nt, $text ) ) . $trail;
						$this->mOutput->addImage( $nt->getDBkey() );

						wfProfileOut( "$fname-image" );
						continue;
					} else {
						# We still need to record the image's presence on the page
						$this->mOutput->addImage( $nt->getDBkey() );
					}
					wfProfileOut( "$fname-image" );

				}

				if ( $ns == NS_CATEGORY ) {
					wfProfileIn( "$fname-category" );
					$s = rtrim($s . "\n"); # bug 87

					if ( $wasblank ) {
						if ( $this->mTitle->getNamespace() == NS_CATEGORY ) {
							$sortkey = $this->mTitle->getText();
						} else {
							$sortkey = $this->mTitle->getPrefixedText();
						}
					} else {
						$sortkey = $text;
					}
					$sortkey = Sanitizer::decodeCharReferences( $sortkey );
					$sortkey = str_replace( "\n", '', $sortkey );
					$sortkey = $wgContLang->convertCategoryKey( $sortkey );
					$this->mOutput->addCategory( $nt->getDBkey(), $sortkey );

					/**
					 * Strip the whitespace Category links produce, see bug 87
					 * @todo We might want to use trim($tmp, "\n") here.
					 */
					$s .= trim($prefix . $trail, "\n") == '' ? '': $prefix . $trail;

					wfProfileOut( "$fname-category" );
					continue;
				}
			}

			if( ( $nt->getPrefixedText() === $selflink ) &&
			    ( $nt->getFragment() === '' ) ) {
				# Self-links are handled specially; generally de-link and change to bold.
				$s .= $prefix . $sk->makeSelfLinkObj( $nt, $text, '', $trail );
				continue;
			}

			# Special and Media are pseudo-namespaces; no pages actually exist in them
			if( $ns == NS_MEDIA ) {
				$link = $sk->makeMediaLinkObj( $nt, $text );
				# Cloak with NOPARSE to avoid replacement in replaceExternalLinks
				$s .= $prefix . $this->armorLinks( $link ) . $trail;
				$this->mOutput->addImage( $nt->getDBkey() );
				continue;
			} elseif( $ns == NS_SPECIAL ) {
				$s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
				continue;
			} elseif( $ns == NS_IMAGE ) {
				$img = Image::newFromTitle( $nt );
				if( $img->exists() ) {
					// Force a blue link if the file exists; may be a remote
					// upload on the shared repository, and we want to see its
					// auto-generated page.
					$s .= $this->makeKnownLinkHolder( $nt, $text, '', $trail, $prefix );
					continue;
				}
			}
			$s .= $this->makeLinkHolder( $nt, $text, '', $trail, $prefix );
		}
		wfProfileOut( $fname );
		return $s;
	}

	/**
	 * Make a link placeholder. The text returned can be later resolved to a real link with
	 * replaceLinkHolders(). This is done for two reasons: firstly to avoid further
	 * parsing of interwiki links, and secondly to allow all extistence checks and
	 * article length checks (for stub links) to be bundled into a single query.
	 *
	 */
	function makeLinkHolder( &$nt, $text = '', $query = '', $trail = '', $prefix = '' ) {
		if ( ! is_object($nt) ) {
			# Fail gracefully
			$retVal = "<!-- ERROR -->{$prefix}{$text}{$trail}";
		} else {
			# Separate the link trail from the rest of the link
			list( $inside, $trail ) = Linker::splitTrail( $trail );

			if ( $nt->isExternal() ) {
				$nr = array_push( $this->mInterwikiLinkHolders['texts'], $prefix.$text.$inside );
				$this->mInterwikiLinkHolders['titles'][] = $nt;
				$retVal = '<!--IWLINK '. ($nr-1) ."-->{$trail}";
			} else {
				$nr = array_push( $this->mLinkHolders['namespaces'], $nt->getNamespace() );
				$this->mLinkHolders['dbkeys'][] = $nt->getDBkey();
				$this->mLinkHolders['queries'][] = $query;
				$this->mLinkHolders['texts'][] = $prefix.$text.$inside;
				$this->mLinkHolders['titles'][] = $nt;

				$retVal = '<!--LINK '. ($nr-1) ."-->{$trail}";
			}
		}