* @ingroup Parser
*/
// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
-class Preprocessor_Hash implements Preprocessor {
+class Preprocessor_Hash extends Preprocessor {
// @codingStandardsIgnoreEnd
/**
*/
public $parser;
- const CACHE_VERSION = 1;
+ const CACHE_PREFIX = 'preprocess-hash';
public function __construct( $parser ) {
$this->parser = $parser;
* @return PPNode_Hash_Tree
*/
public function preprocessToObj( $text, $flags = 0 ) {
- // Check cache.
- global $wgMemc, $wgPreprocessorCacheThreshold;
-
- $cacheable = $wgPreprocessorCacheThreshold !== false
- && strlen( $text ) > $wgPreprocessorCacheThreshold;
-
- if ( $cacheable ) {
- $cacheKey = wfMemcKey( 'preprocess-hash', md5( $text ), $flags );
- $cacheValue = $wgMemc->get( $cacheKey );
- if ( $cacheValue ) {
- $version = substr( $cacheValue, 0, 8 );
- if ( intval( $version ) == self::CACHE_VERSION ) {
- $hash = unserialize( substr( $cacheValue, 8 ) );
- // From the cache
- wfDebugLog( "Preprocessor",
- "Loaded preprocessor hash from memcached (key $cacheKey)" );
- return $hash;
- }
- }
+ $tree = $this->cacheGetTree( $text, $flags );
+ if ( $tree !== false ) {
+ return unserialize( $tree );
}
- $rules = array(
- '{' => array(
- 'end' => '}',
- 'names' => array(
- 2 => 'template',
- 3 => 'tplarg',
- ),
- 'min' => 2,
- 'max' => 3,
- ),
- '[' => array(
- 'end' => ']',
- 'names' => array( 2 => null ),
- 'min' => 2,
- 'max' => 2,
- )
- );
-
$forInclusion = $flags & Parser::PTD_FOR_INCLUSION;
$xmlishElements = $this->parser->getStripList();
$inHeading = false;
// True if there are no more greater-than (>) signs right of $i
$noMoreGT = false;
+ // Map of tag name => true if there are no more closing tags of given type right of $i
+ $noMoreClosingTag = array();
// True to ignore all input up to the next <onlyinclude>
$findOnlyinclude = $enableOnlyinclude;
// Do a line-start run without outputting an LF character
}
} elseif ( $curChar == $currentClosing ) {
$found = 'close';
- } elseif ( isset( $rules[$curChar] ) ) {
+ } elseif ( isset( $this->rules[$curChar] ) ) {
$found = 'open';
- $rule = $rules[$curChar];
+ $rule = $this->rules[$curChar];
} else {
# Some versions of PHP have a strcspn which stops on null characters
# Ignore and continue
} else {
$attrEnd = $tagEndPos;
// Find closing tag
- if ( preg_match( "/<\/" . preg_quote( $name, '/' ) . "\s*>/i",
+ if (
+ !isset( $noMoreClosingTag[$name] ) &&
+ preg_match( "/<\/" . preg_quote( $name, '/' ) . "\s*>/i",
$text, $matches, PREG_OFFSET_CAPTURE, $tagEndPos + 1 )
) {
$inner = substr( $text, $tagEndPos + 1, $matches[0][1] - $tagEndPos - 1 );
$i = $matches[0][1] + strlen( $matches[0][0] );
$close = $matches[0][0];
} else {
- // No end tag -- let it run out to the end of the text.
- $inner = substr( $text, $tagEndPos + 1 );
- $i = $lengthText;
- $close = null;
+ // No end tag -- don't match the tag, treat opening tag as literal and resume parsing.
+ $i = $tagEndPos + 1;
+ $accum->addLiteral( substr( $text, $tagStartPos, $tagEndPos + 1 - $tagStartPos ) );
+ // Cache results, otherwise we have O(N^2) performance for input like <foo><foo><foo>...
+ $noMoreClosingTag[$name] = true;
+ continue;
}
}
// <includeonly> and <noinclude> just become <ignore> tags
# check for maximum matching characters (if there are 5 closing
# characters, we will probably need only 3 - depending on the rules)
- $rule = $rules[$piece->open];
+ $rule = $this->rules[$piece->open];
if ( $count > $rule['max'] ) {
# The specified maximum exists in the callback array, unless the caller
# has made an error
$lastNode = $node;
}
if ( !$node ) {
- if ( $cacheable ) {
- }
+ // if ( $cacheable ) { ... }
throw new MWException( __METHOD__ . ': eqpos not found' );
}
if ( $node->name !== 'equals' ) {
- if ( $cacheable ) {
- }
+ // if ( $cacheable ) { ... }
throw new MWException( __METHOD__ . ': eqpos is not equals' );
}
$equalsNode = $node;
$piece->parts = array( new PPDPart_Hash );
$piece->count -= $matchingCount;
# do we still qualify for any callback with remaining count?
- $min = $rules[$piece->open]['min'];
+ $min = $this->rules[$piece->open]['min'];
if ( $piece->count >= $min ) {
$stack->push( $piece );
$accum =& $stack->getAccum();
$rootNode->lastChild = $stack->rootAccum->lastNode;
// Cache
- if ( $cacheable ) {
- $cacheValue = sprintf( "%08d", self::CACHE_VERSION ) . serialize( $rootNode );
-
- // T111289: Cache values should not exceed 1 Mb, but they do.
- if ( strlen( $cacheValue ) <= 1e6 ) {
- $wgMemc->set( $cacheKey, $cacheValue, 86400 );
- wfDebugLog( "Preprocessor", "Saved preprocessor Hash to memcached (key $cacheKey)" );
- }
- }
+ $this->cacheSetTree( $text, $flags, serialize( $rootNode ) );
return $rootNode;
}
$children[] = $child;
}
}
- return $children;
+ return new PPNode_Hash_Array( $children );
}
/**