<?php
+/**
+ * Preprocessor using PHP's dom extension
+ *
+ * @file
+ * @ingroup Parser
+ */
/**
* @ingroup Parser
function newPartNodeArray( $values ) {
//NOTE: DOM manipulation is slower than building & parsing XML! (or so Tim sais)
- $xml = "";
- $xml .= "<list>";
+ $xml = "<list>";
foreach ( $values as $k => $val ) {
-
+
if ( is_int( $k ) ) {
$xml .= "<part><name index=\"$k\"/><value>" . htmlspecialchars( $val ) ."</value></part>";
} else {
function preprocessToObj( $text, $flags = 0 ) {
wfProfileIn( __METHOD__ );
global $wgMemc, $wgPreprocessorCacheThreshold;
-
+
$xml = false;
- $cacheable = strlen( $text ) > $wgPreprocessorCacheThreshold;
+ $cacheable = $wgPreprocessorCacheThreshold !== false && strlen( $text ) > $wgPreprocessorCacheThreshold;
if ( $cacheable ) {
wfProfileIn( __METHOD__.'-cacheable' );
wfProfileOut( __METHOD__ );
return $obj;
}
-
+
function preprocessToXml( $text, $flags = 0 ) {
wfProfileIn( __METHOD__ );
$rules = array(
// Search backwards for leading whitespace
$wsStart = $i ? ( $i - strspn( $revText, ' ', strlen( $text ) - $i ) ) : 0;
// Search forwards for trailing whitespace
- // $wsEnd will be the position of the last space
+ // $wsEnd will be the position of the last space (or the '>' if there's none)
$wsEnd = $endPos + 2 + strspn( $text, ' ', $endPos + 3 );
// Eat the line if possible
// TODO: This could theoretically be done if $wsStart == 0, i.e. for comments at
if ( $stack->top ) {
$part = $stack->top->getCurrentPart();
- if ( isset( $part->commentEnd ) && $part->commentEnd == $wsStart - 1 ) {
- // Comments abutting, no change in visual end
- $part->commentEnd = $wsEnd;
- } else {
+ if ( ! (isset( $part->commentEnd ) && $part->commentEnd == $wsStart - 1 )) {
$part->visualEnd = $wsStart;
- $part->commentEnd = $endPos;
}
+ // Else comments abutting, no change in visual end
+ $part->commentEnd = $endPos;
}
$i = $endPos + 1;
$inner = substr( $text, $startPos, $endPos - $startPos + 1 );
} else {
$attrEnd = $tagEndPos;
// Find closing tag
- if ( preg_match( "/<\/" . preg_quote( $name, '/' ) . "\s*>/i",
- $text, $matches, PREG_OFFSET_CAPTURE, $tagEndPos + 1 ) )
+ if ( 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] );
$count = $piece->count;
$equalsLength = strspn( $revText, '=', strlen( $text ) - $searchStart );
if ( $equalsLength > 0 ) {
- if ( $i - $equalsLength == $piece->startPos ) {
+ if ( $searchStart - $equalsLength == $piece->startPos ) {
// This is just a single string of equals signs on its own line
// Replicate the doHeadings behaviour /={count}(.+)={count}/
// First find out how many equals signs there really are (don't stop at 6)
// another heading. Infinite loops are avoided because the next iteration MUST
// hit the heading open case above, which unconditionally increments the
// input pointer.
- }
-
- elseif ( $found == 'open' ) {
+ } elseif ( $found == 'open' ) {
# count opening brace characters
$count = strspn( $text, $curChar, $i );
'open' => $curChar,
'close' => $rule['end'],
'count' => $count,
- 'lineStart' => ($i > 0 && $text[$i-1] == "\n"),
+ 'lineStart' => ($i == 0 || $text[$i-1] == "\n"),
);
$stack->push( $piece );
$accum .= htmlspecialchars( str_repeat( $curChar, $count ) );
}
$i += $count;
- }
-
- elseif ( $found == 'close' ) {
+ } elseif ( $found == 'close' ) {
$piece = $stack->top;
# lets check if there are enough characters for closing brace
$maxCount = $piece->count;
# check for maximum matching characters (if there are 5 closing
# characters, we will probably need only 3 - depending on the rules)
- $matchingCount = 0;
$rule = $rules[$piece->open];
if ( $count > $rule['max'] ) {
# The specified maximum exists in the callback array, unless the caller
$element = "<$name$attr>";
$element .= "<title>$title</title>";
$argIndex = 1;
- foreach ( $parts as $partIndex => $part ) {
+ foreach ( $parts as $part ) {
if ( isset( $part->eqpos ) ) {
$argName = substr( $part->out, 0, $part->eqpos );
$argValue = substr( $part->out, $part->eqpos + 1 );
$xml = $stack->rootAccum;
wfProfileOut( __METHOD__ );
-
+
return $xml;
}
}
return $root;
}
- if ( ++$this->parser->mPPNodeCount > $this->parser->mOptions->mMaxPPNodeCount )
+ if ( ++$this->parser->mPPNodeCount > $this->parser->mOptions->getMaxPPNodeCount() )
{
return '<span class="error">Node-count limit exceeded</span>';
}
- if ( $expansionDepth > $this->parser->mOptions->mMaxPPExpandDepth ) {
+ if ( $expansionDepth > $this->parser->mOptions->getMaxPPExpandDepth() ) {
return '<span class="error">Expansion depth limit exceeded</span>';
}
wfProfileIn( __METHOD__ );
$iteratorStack[$level] = false;
}
- if ( $contextNode instanceof PPNode_DOM ) $contextNode = $contextNode->node;
+ if ( $contextNode instanceof PPNode_DOM ) {
+ $contextNode = $contextNode->node;
+ }
$newIterator = false;
$titles = $xpath->query( 'title', $contextNode );
$title = $titles->item( 0 );
$parts = $xpath->query( 'part', $contextNode );
- if ( $flags & self::NO_TEMPLATES ) {
+ if ( $flags & PPFrame::NO_TEMPLATES ) {
$newIterator = $this->virtualBracketedImplode( '{{', '|', '}}', $title, $parts );
} else {
$lineStart = $contextNode->getAttribute( 'lineStart' );
$titles = $xpath->query( 'title', $contextNode );
$title = $titles->item( 0 );
$parts = $xpath->query( 'part', $contextNode );
- if ( $flags & self::NO_ARGS ) {
+ if ( $flags & PPFrame::NO_ARGS ) {
$newIterator = $this->virtualBracketedImplode( '{{{', '|', '}}}', $title, $parts );
} else {
$params = array(
# Remove it in HTML, pre+remove and STRIP_COMMENTS modes
if ( $this->parser->ot['html']
|| ( $this->parser->ot['pre'] && $this->parser->mOptions->getRemoveComments() )
- || ( $flags & self::STRIP_COMMENTS ) )
+ || ( $flags & PPFrame::STRIP_COMMENTS ) )
{
$out .= '';
}
# Add a strip marker in PST mode so that pstPass2() can run some old-fashioned regexes on the result
# Not in RECOVER_COMMENTS mode (extractSections) though
- elseif ( $this->parser->ot['wiki'] && ! ( $flags & self::RECOVER_COMMENTS ) ) {
+ elseif ( $this->parser->ot['wiki'] && ! ( $flags & PPFrame::RECOVER_COMMENTS ) ) {
$out .= $this->parser->insertStripItem( $contextNode->textContent );
}
# Recover the literal comment in RECOVER_COMMENTS and pre+no-remove
# OT_WIKI will only respect <ignore> in substed templates.
# The other output types respect it unless NO_IGNORE is set.
# extractSections() sets NO_IGNORE and so never respects it.
- if ( ( !isset( $this->parent ) && $this->parser->ot['wiki'] ) || ( $flags & self::NO_IGNORE ) ) {
+ if ( ( !isset( $this->parent ) && $this->parser->ot['wiki'] ) || ( $flags & PPFrame::NO_IGNORE ) ) {
$out .= $contextNode->textContent;
} else {
$out .= '';
# Heading
$s = $this->expand( $contextNode->childNodes, $flags );
- # Insert a heading marker only for <h> children of <root>
- # This is to stop extractSections from going over multiple tree levels
- if ( $contextNode->parentNode->nodeName == 'root'
- && $this->parser->ot['html'] )
- {
+ # Insert a heading marker only for <h> children of <root>
+ # This is to stop extractSections from going over multiple tree levels
+ if ( $contextNode->parentNode->nodeName == 'root'
+ && $this->parser->ot['html'] )
+ {
# Insert heading index marker
$headingIndex = $contextNode->getAttribute( 'i' );
$titleText = $this->title->getPrefixedDBkey();
$marker = "{$this->parser->mUniqPrefix}-h-$serial-" . Parser::MARKER_SUFFIX;
$count = $contextNode->getAttribute( 'level' );
$s = substr( $s, 0, $count ) . $marker . substr( $s, $count );
- $this->parser->mStripState->general->setPair( $marker, '' );
+ $this->parser->mStripState->addGeneral( $marker, '' );
}
$out .= $s;
} else {
$first = true;
$s = '';
foreach ( $args as $root ) {
- if ( $root instanceof PPNode_DOM ) $root = $root->node;
+ if ( $root instanceof PPNode_DOM ) {
+ $root = $root->node;
+ }
if ( !is_array( $root ) && !( $root instanceof DOMNodeList ) ) {
$root = array( $root );
}
$args = array_slice( func_get_args(), 1 );
$out = array();
$first = true;
- if ( $root instanceof PPNode_DOM ) $root = $root->node;
foreach ( $args as $root ) {
+ if ( $root instanceof PPNode_DOM ) {
+ $root = $root->node;
+ }
if ( !is_array( $root ) && !( $root instanceof DOMNodeList ) ) {
$root = array( $root );
}
$first = true;
foreach ( $args as $root ) {
- if ( $root instanceof PPNode_DOM ) $root = $root->node;
+ if ( $root instanceof PPNode_DOM ) {
+ $root = $root->node;
+ }
if ( !is_array( $root ) && !( $root instanceof DOMNodeList ) ) {
$root = array( $root );
}
var $numberedExpansionCache, $namedExpansionCache;
function __construct( $preprocessor, $parent = false, $numberedArgs = array(), $namedArgs = array(), $title = false ) {
- PPFrame_DOM::__construct( $preprocessor );
+ parent::__construct( $preprocessor );
+
$this->parent = $parent;
$this->numberedArgs = $numberedArgs;
$this->namedArgs = $namedArgs;
}
return $arguments;
}
-
+
function getNumberedArguments() {
$arguments = array();
foreach ( array_keys($this->numberedArgs) as $key ) {
}
return $arguments;
}
-
+
function getNamedArguments() {
$arguments = array();
foreach ( array_keys($this->namedArgs) as $key ) {
}
if ( !isset( $this->numberedExpansionCache[$index] ) ) {
# No trimming for unnamed arguments
- $this->numberedExpansionCache[$index] = $this->parent->expand( $this->numberedArgs[$index], self::STRIP_COMMENTS );
+ $this->numberedExpansionCache[$index] = $this->parent->expand( $this->numberedArgs[$index], PPFrame::STRIP_COMMENTS );
}
return $this->numberedExpansionCache[$index];
}
if ( !isset( $this->namedExpansionCache[$name] ) ) {
# Trim named arguments post-expand, for backwards compatibility
$this->namedExpansionCache[$name] = trim(
- $this->parent->expand( $this->namedArgs[$name], self::STRIP_COMMENTS ) );
+ $this->parent->expand( $this->namedArgs[$name], PPFrame::STRIP_COMMENTS ) );
}
return $this->namedExpansionCache[$name];
}
var $args;
function __construct( $preprocessor, $args ) {
- PPFrame_DOM::__construct( $preprocessor );
+ parent::__construct( $preprocessor );
$this->args = $args;
}
* @ingroup Parser
*/
class PPNode_DOM implements PPNode {
- var $node;
+ var $node, $xpath;
function __construct( $node, $xpath = false ) {
$this->node = $node;
}
- function __get( $name ) {
- if ( $name == 'xpath' ) {
+ function getXPath() {
+ if ( $this->xpath === null ) {
$this->xpath = new DOMXPath( $this->node->ownerDocument );
}
return $this->xpath;
}
function getChildrenOfType( $type ) {
- return new self( $this->xpath->query( $type, $this->node ) );
+ return new self( $this->getXPath()->query( $type, $this->node ) );
}
function getLength() {
* value PPNode value
*/
function splitArg() {
- $names = $this->xpath->query( 'name', $this->node );
- $values = $this->xpath->query( 'value', $this->node );
+ $xpath = $this->getXPath();
+ $names = $xpath->query( 'name', $this->node );
+ $values = $xpath->query( 'value', $this->node );
if ( !$names->length || !$values->length ) {
throw new MWException( 'Invalid brace node passed to ' . __METHOD__ );
}
* All values in the resulting array are PPNodes. Inner and close are optional.
*/
function splitExt() {
- $names = $this->xpath->query( 'name', $this->node );
- $attrs = $this->xpath->query( 'attr', $this->node );
- $inners = $this->xpath->query( 'inner', $this->node );
- $closes = $this->xpath->query( 'close', $this->node );
+ $xpath = $this->getXPath();
+ $names = $xpath->query( 'name', $this->node );
+ $attrs = $xpath->query( 'attr', $this->node );
+ $inners = $xpath->query( 'inner', $this->node );
+ $closes = $xpath->query( 'close', $this->node );
if ( !$names->length || !$attrs->length ) {
throw new MWException( 'Invalid ext node passed to ' . __METHOD__ );
}
* Split a <h> node
*/
function splitHeading() {
- if ( !$this->nodeName == 'h' ) {
+ if ( $this->getName() !== 'h' ) {
throw new MWException( 'Invalid h node passed to ' . __METHOD__ );
}
return array(