/**
* Preprocessor using PHP arrays
*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ * http://www.gnu.org/copyleft/gpl.html
+ *
* @file
* @ingroup Parser
*/
-
+
/**
* Differences from DOM schema:
* * attribute nodes are children
- * * <h> nodes that aren't at the top are replaced with <possible-h>
+ * * "<h>" nodes that aren't at the top are replaced with <possible-h>
+ *
+ * Nodes are stored in a recursive array data structure. A node store is an
+ * array where each element may be either a scalar (representing a text node)
+ * or a "descriptor", which is a two-element array where the first element is
+ * the node name and the second element is the node store for the children.
+ *
+ * Attributes are represented as children that have a node name starting with
+ * "@", and a single text node child.
+ *
+ * @todo: Consider replacing descriptor arrays with objects of a new class.
+ * Benchmark and measure resulting memory impact.
+ *
* @ingroup Parser
*/
-class Preprocessor_Hash implements Preprocessor {
- var $parser;
-
- const CACHE_VERSION = 1;
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
+class Preprocessor_Hash extends Preprocessor {
+ // @codingStandardsIgnoreEnd
+
+ /**
+ * @var Parser
+ */
+ public $parser;
+
+ const CACHE_PREFIX = 'preprocess-hash';
+ const CACHE_VERSION = 2;
- function __construct( $parser ) {
+ public function __construct( $parser ) {
$this->parser = $parser;
}
- function newFrame() {
+ /**
+ * @return PPFrame_Hash
+ */
+ public function newFrame() {
return new PPFrame_Hash( $this );
}
- function newCustomFrame( $args ) {
+ /**
+ * @param array $args
+ * @return PPCustomFrame_Hash
+ */
+ public function newCustomFrame( $args ) {
return new PPCustomFrame_Hash( $this, $args );
}
- function newPartNodeArray( $values ) {
- $list = array();
+ /**
+ * @param array $values
+ * @return PPNode_Hash_Array
+ */
+ public function newPartNodeArray( $values ) {
+ $list = [];
foreach ( $values as $k => $val ) {
- $partNode = new PPNode_Hash_Tree( 'part' );
- $nameNode = new PPNode_Hash_Tree( 'name' );
-
if ( is_int( $k ) ) {
- $nameNode->addChild( new PPNode_Hash_Attr( 'index', $k ) );
- $partNode->addChild( $nameNode );
+ $store = [ [ 'part', [
+ [ 'name', [ [ '@index', [ $k ] ] ] ],
+ [ 'value', [ strval( $val ) ] ],
+ ] ] ];
} else {
- $nameNode->addChild( new PPNode_Hash_Text( $k ) );
- $partNode->addChild( $nameNode );
- $partNode->addChild( new PPNode_Hash_Text( '=' ) );
+ $store = [ [ 'part', [
+ [ 'name', [ strval( $k ) ] ],
+ '=',
+ [ 'value', [ strval( $val ) ] ],
+ ] ] ];
}
- $valueNode = new PPNode_Hash_Tree( 'value' );
- $valueNode->addChild( new PPNode_Hash_Text( $val ) );
- $partNode->addChild( $valueNode );
-
- $list[] = $partNode;
+ $list[] = new PPNode_Hash_Tree( $store, 0 );
}
$node = new PPNode_Hash_Array( $list );
/**
* Preprocess some wikitext and return the document tree.
- * This is the ghost of Parser::replace_variables().
*
- * @param $text String: the text to parse
- * @param $flags Integer: bitwise combination of:
- * Parser::PTD_FOR_INCLUSION Handle <noinclude>/<includeonly> as if the text is being
- * included. Default is to assume a direct page view.
+ * @param string $text The text to parse
+ * @param int $flags Bitwise combination of:
+ * Parser::PTD_FOR_INCLUSION Handle "<noinclude>" and "<includeonly>" as if the text is being
+ * included. Default is to assume a direct page view.
*
* The generated DOM tree must depend only on the input text and the flags.
- * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of bug 4899.
+ * The DOM tree must be the same in OT_HTML and OT_WIKI mode, to avoid a regression of T6899.
*
* Any flag added to the $flags parameter here, or any other parameter liable to cause a
* change in the DOM tree for a given text, must be passed through the section identifier
* in the section edit link and thus back to extractSections().
*
- * The output of this function is currently only cached in process memory, but a persistent
- * cache may be implemented at a later date which takes further advantage of these strict
- * dependency requirements.
- *
- * @private
- */
- function preprocessToObj( $text, $flags = 0 ) {
- wfProfileIn( __METHOD__ );
-
-
- // Check cache.
- global $wgMemc, $wgPreprocessorCacheThreshold;
-
- $cacheable = strlen( $text ) > $wgPreprocessorCacheThreshold;
- if ( $cacheable ) {
- wfProfileIn( __METHOD__.'-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)" );
- wfProfileOut( __METHOD__.'-cacheable' );
- wfProfileOut( __METHOD__ );
- return $hash;
- }
+ * @throws MWException
+ * @return PPNode_Hash_Tree
+ */
+ public function preprocessToObj( $text, $flags = 0 ) {
+ global $wgDisableLangConversion;
+
+ $tree = $this->cacheGetTree( $text, $flags );
+ if ( $tree !== false ) {
+ $store = json_decode( $tree );
+ if ( is_array( $store ) ) {
+ return new PPNode_Hash_Tree( $store, 0 );
}
- wfProfileIn( __METHOD__.'-cache-miss' );
}
- $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();
+ $xmlishAllowMissingEndTag = [ 'includeonly', 'noinclude', 'onlyinclude' ];
$enableOnlyinclude = false;
if ( $forInclusion ) {
- $ignoredTags = array( 'includeonly', '/includeonly' );
- $ignoredElements = array( 'noinclude' );
+ $ignoredTags = [ 'includeonly', '/includeonly' ];
+ $ignoredElements = [ 'noinclude' ];
$xmlishElements[] = 'noinclude';
- if ( strpos( $text, '<onlyinclude>' ) !== false && strpos( $text, '</onlyinclude>' ) !== false ) {
+ if ( strpos( $text, '<onlyinclude>' ) !== false
+ && strpos( $text, '</onlyinclude>' ) !== false
+ ) {
$enableOnlyinclude = true;
}
} else {
- $ignoredTags = array( 'noinclude', '/noinclude', 'onlyinclude', '/onlyinclude' );
- $ignoredElements = array( 'includeonly' );
+ $ignoredTags = [ 'noinclude', '/noinclude', 'onlyinclude', '/onlyinclude' ];
+ $ignoredElements = [ 'includeonly' ];
$xmlishElements[] = 'includeonly';
}
$xmlishRegex = implode( '|', array_merge( $xmlishElements, $ignoredTags ) );
$stack = new PPDStack_Hash;
$searchBase = "[{<\n";
- $revText = strrev( $text ); // For fast reverse searches
+ if ( !$wgDisableLangConversion ) {
+ $searchBase .= '-';
+ }
- $i = 0; # Input pointer, starts out pointing to a pseudo-newline before the start
- $accum =& $stack->getAccum(); # Current accumulator
- $findEquals = false; # True to find equals signs in arguments
- $findPipe = false; # True to take notice of pipe characters
+ // For fast reverse searches
+ $revText = strrev( $text );
+ $lengthText = strlen( $text );
+
+ // Input pointer, starts out pointing to a pseudo-newline before the start
+ $i = 0;
+ // Current accumulator. See the doc comment for Preprocessor_Hash for the format.
+ $accum =& $stack->getAccum();
+ // True to find equals signs in arguments
+ $findEquals = false;
+ // True to take notice of pipe characters
+ $findPipe = false;
$headingIndex = 1;
- $inHeading = false; # True if $i is inside a possible heading
- $noMoreGT = false; # True if there are no more greater-than (>) signs right of $i
- $findOnlyinclude = $enableOnlyinclude; # True to ignore all input up to the next <onlyinclude>
- $fakeLineStart = true; # Do a line-start run without outputting an LF character
+ // True if $i is inside a possible heading
+ $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 = [];
+ // True to ignore all input up to the next <onlyinclude>
+ $findOnlyinclude = $enableOnlyinclude;
+ // Do a line-start run without outputting an LF character
+ $fakeLineStart = true;
while ( true ) {
- //$this->memCheck();
+ // $this->memCheck();
if ( $findOnlyinclude ) {
// Ignore all input up to the next <onlyinclude>
$startPos = strpos( $text, '<onlyinclude>', $i );
if ( $startPos === false ) {
// Ignored section runs to the end
- $accum->addNodeWithText( 'ignore', substr( $text, $i ) );
+ $accum[] = [ 'ignore', [ substr( $text, $i ) ] ];
break;
}
$tagEndPos = $startPos + strlen( '<onlyinclude>' ); // past-the-end
- $accum->addNodeWithText( 'ignore', substr( $text, $i, $tagEndPos - $i ) );
+ $accum[] = [ 'ignore', [ substr( $text, $i, $tagEndPos - $i ) ] ];
$i = $tagEndPos;
$findOnlyinclude = false;
}
$search = $searchBase;
if ( $stack->top === false ) {
$currentClosing = '';
+ } elseif (
+ $stack->top->close === '}-' &&
+ $stack->top->count > 2
+ ) {
+ # adjust closing for -{{{...{{
+ $currentClosing = '}';
+ $search .= $currentClosing;
} else {
$currentClosing = $stack->top->close;
$search .= $currentClosing;
# Output literal section, advance input counter
$literalLength = strcspn( $text, $search, $i );
if ( $literalLength > 0 ) {
- $accum->addLiteral( substr( $text, $i, $literalLength ) );
+ self::addLiteral( $accum, substr( $text, $i, $literalLength ) );
$i += $literalLength;
}
- if ( $i >= strlen( $text ) ) {
+ if ( $i >= $lengthText ) {
if ( $currentClosing == "\n" ) {
// Do a past-the-end run to finish off the heading
$curChar = '';
break;
}
} else {
- $curChar = $text[$i];
+ $curChar = $curTwoChar = $text[$i];
+ if ( ( $i + 1 ) < $lengthText ) {
+ $curTwoChar .= $text[$i + 1];
+ }
if ( $curChar == '|' ) {
$found = 'pipe';
} elseif ( $curChar == '=' ) {
} else {
$found = 'line-start';
}
+ } elseif ( $curTwoChar == $currentClosing ) {
+ $found = 'close';
+ $curChar = $curTwoChar;
} elseif ( $curChar == $currentClosing ) {
$found = 'close';
- } elseif ( isset( $rules[$curChar] ) ) {
+ } elseif ( isset( $this->rules[$curTwoChar] ) ) {
+ $curChar = $curTwoChar;
$found = 'open';
- $rule = $rules[$curChar];
+ $rule = $this->rules[$curChar];
+ } elseif ( isset( $this->rules[$curChar] ) ) {
+ $found = 'open';
+ $rule = $this->rules[$curChar];
} else {
- # Some versions of PHP have a strcspn which stops on null characters
- # Ignore and continue
+ # Some versions of PHP have a strcspn which stops on
+ # null characters; ignore these and continue.
+ # We also may get '-' and '}' characters here which
+ # don't match -{ or $currentClosing. Add these to
+ # output and continue.
+ if ( $curChar == '-' || $curChar == '}' ) {
+ self::addLiteral( $accum, $curChar );
+ }
++$i;
continue;
}
if ( $found == 'angle' ) {
$matches = false;
// Handle </onlyinclude>
- if ( $enableOnlyinclude && substr( $text, $i, strlen( '</onlyinclude>' ) ) == '</onlyinclude>' ) {
+ if ( $enableOnlyinclude
+ && substr( $text, $i, strlen( '</onlyinclude>' ) ) == '</onlyinclude>'
+ ) {
$findOnlyinclude = true;
continue;
}
// Determine element name
if ( !preg_match( $elementsRegex, $text, $matches, 0, $i + 1 ) ) {
// Element name missing or not listed
- $accum->addLiteral( '<' );
+ self::addLiteral( $accum, '<' );
++$i;
continue;
}
// Handle comments
if ( isset( $matches[2] ) && $matches[2] == '!--' ) {
- // To avoid leaving blank lines, when a comment is both preceded
- // and followed by a newline (ignoring spaces), trim leading and
- // trailing spaces and one of the newlines.
+ // To avoid leaving blank lines, when a sequence of
+ // space-separated comments is both preceded and followed by
+ // a newline (ignoring spaces), then
+ // trim leading and trailing spaces and the trailing newline.
// Find the end
$endPos = strpos( $text, '-->', $i + 4 );
if ( $endPos === false ) {
// Unclosed comment in input, runs to end
$inner = substr( $text, $i );
- $accum->addNodeWithText( 'comment', $inner );
- $i = strlen( $text );
+ $accum[] = [ 'comment', [ $inner ] ];
+ $i = $lengthText;
} else {
// Search backwards for leading whitespace
- $wsStart = $i ? ( $i - strspn( $revText, ' ', strlen( $text ) - $i ) ) : 0;
+ $wsStart = $i ? ( $i - strspn( $revText, " \t", $lengthText - $i ) ) : 0;
+
// Search forwards for trailing whitespace
- // $wsEnd will be the position of the last space
- $wsEnd = $endPos + 2 + strspn( $text, ' ', $endPos + 3 );
+ // $wsEnd will be the position of the last space (or the '>' if there's none)
+ $wsEnd = $endPos + 2 + strspn( $text, " \t", $endPos + 3 );
+
+ // Keep looking forward as long as we're finding more
+ // comments.
+ $comments = [ [ $wsStart, $wsEnd ] ];
+ while ( substr( $text, $wsEnd + 1, 4 ) == '<!--' ) {
+ $c = strpos( $text, '-->', $wsEnd + 4 );
+ if ( $c === false ) {
+ break;
+ }
+ $c = $c + 2 + strspn( $text, " \t", $c + 3 );
+ $comments[] = [ $wsEnd + 1, $c ];
+ $wsEnd = $c;
+ }
+
// Eat the line if possible
// TODO: This could theoretically be done if $wsStart == 0, i.e. for comments at
// the overall start. That's not how Sanitizer::removeHTMLcomments() did it, but
// it's a possible beneficial b/c break.
if ( $wsStart > 0 && substr( $text, $wsStart - 1, 1 ) == "\n"
- && substr( $text, $wsEnd + 1, 1 ) == "\n" )
- {
- $startPos = $wsStart;
- $endPos = $wsEnd + 1;
+ && substr( $text, $wsEnd + 1, 1 ) == "\n"
+ ) {
// Remove leading whitespace from the end of the accumulator
- // Sanity check first though
$wsLength = $i - $wsStart;
+ $endIndex = count( $accum ) - 1;
+
+ // Sanity check
if ( $wsLength > 0
- && $accum->lastNode instanceof PPNode_Hash_Text
- && substr( $accum->lastNode->value, -$wsLength ) === str_repeat( ' ', $wsLength ) )
- {
- $accum->lastNode->value = substr( $accum->lastNode->value, 0, -$wsLength );
+ && $endIndex >= 0
+ && is_string( $accum[$endIndex] )
+ && strspn( $accum[$endIndex], " \t", -$wsLength ) === $wsLength
+ ) {
+ $accum[$endIndex] = substr( $accum[$endIndex], 0, -$wsLength );
+ }
+
+ // Dump all but the last comment to the accumulator
+ foreach ( $comments as $j => $com ) {
+ $startPos = $com[0];
+ $endPos = $com[1] + 1;
+ if ( $j == ( count( $comments ) - 1 ) ) {
+ break;
+ }
+ $inner = substr( $text, $startPos, $endPos - $startPos );
+ $accum[] = [ 'comment', [ $inner ] ];
}
+
// Do a line-start run next time to look for headings after the comment
$fakeLineStart = true;
} else {
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 );
- $accum->addNodeWithText( 'comment', $inner );
+ $accum[] = [ 'comment', [ $inner ] ];
}
continue;
}
// Infinite backtrack
// Disable tag search to prevent worst-case O(N^2) performance
$noMoreGT = true;
- $accum->addLiteral( '<' );
+ self::addLiteral( $accum, '<' );
++$i;
continue;
}
// Handle ignored tags
if ( in_array( $lowerName, $ignoredTags ) ) {
- $accum->addNodeWithText( 'ignore', substr( $text, $i, $tagEndPos - $i + 1 ) );
+ $accum[] = [ 'ignore', [ substr( $text, $i, $tagEndPos - $i + 1 ) ] ];
$i = $tagEndPos + 1;
continue;
}
$tagStartPos = $i;
- if ( $text[$tagEndPos-1] == '/' ) {
+ if ( $text[$tagEndPos - 1] == '/' ) {
// Short end tag
$attrEnd = $tagEndPos - 1;
$inner = null;
} else {
$attrEnd = $tagEndPos;
// Find closing tag
- if ( preg_match( "/<\/" . preg_quote( $name, '/' ) . "\s*>/i",
- $text, $matches, PREG_OFFSET_CAPTURE, $tagEndPos + 1 ) )
- {
+ 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 = strlen( $text );
- $close = null;
+ // No end tag
+ if ( in_array( $name, $xmlishAllowMissingEndTag ) ) {
+ // Let it run out to the end of the text.
+ $inner = substr( $text, $tagEndPos + 1 );
+ $i = $lengthText;
+ $close = null;
+ } else {
+ // Don't match the tag, treat opening tag as literal and resume parsing.
+ $i = $tagEndPos + 1;
+ self::addLiteral( $accum,
+ 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
if ( in_array( $lowerName, $ignoredElements ) ) {
- $accum->addNodeWithText( 'ignore', substr( $text, $tagStartPos, $i - $tagStartPos ) );
+ $accum[] = [ 'ignore', [ substr( $text, $tagStartPos, $i - $tagStartPos ) ] ];
continue;
}
$attr = substr( $text, $attrStart, $attrEnd - $attrStart );
}
- $extNode = new PPNode_Hash_Tree( 'ext' );
- $extNode->addChild( PPNode_Hash_Tree::newWithText( 'name', $name ) );
- $extNode->addChild( PPNode_Hash_Tree::newWithText( 'attr', $attr ) );
+ $children = [
+ [ 'name', [ $name ] ],
+ [ 'attr', [ $attr ] ] ];
if ( $inner !== null ) {
- $extNode->addChild( PPNode_Hash_Tree::newWithText( 'inner', $inner ) );
+ $children[] = [ 'inner', [ $inner ] ];
}
if ( $close !== null ) {
- $extNode->addChild( PPNode_Hash_Tree::newWithText( 'close', $close ) );
+ $children[] = [ 'close', [ $close ] ];
}
- $accum->addNode( $extNode );
- }
-
- elseif ( $found == 'line-start' ) {
+ $accum[] = [ 'ext', $children ];
+ } elseif ( $found == 'line-start' ) {
// Is this the start of a heading?
// Line break belongs before the heading element in any case
if ( $fakeLineStart ) {
$fakeLineStart = false;
} else {
- $accum->addLiteral( $curChar );
+ self::addLiteral( $accum, $curChar );
$i++;
}
$count = strspn( $text, '=', $i, 6 );
if ( $count == 1 && $findEquals ) {
- // DWIM: This looks kind of like a name/value separator
- // Let's let the equals handler have it and break the potential heading
- // This is heuristic, but AFAICT the methods for completely correct disambiguation are very complex.
+ // DWIM: This looks kind of like a name/value separator.
+ // Let's let the equals handler have it and break the potential
+ // heading. This is heuristic, but AFAICT the methods for
+ // completely correct disambiguation are very complex.
} elseif ( $count > 0 ) {
- $piece = array(
+ $piece = [
'open' => "\n",
'close' => "\n",
- 'parts' => array( new PPDPart_Hash( str_repeat( '=', $count ) ) ),
+ 'parts' => [ new PPDPart_Hash( str_repeat( '=', $count ) ) ],
'startPos' => $i,
- 'count' => $count );
+ 'count' => $count ];
$stack->push( $piece );
$accum =& $stack->getAccum();
extract( $stack->getFlags() );
$i += $count;
}
- }
-
- elseif ( $found == 'line-end' ) {
+ } elseif ( $found == 'line-end' ) {
$piece = $stack->top;
// A heading must be open, otherwise \n wouldn't have been in the search list
- assert( $piece->open == "\n" );
+ assert( $piece->open === "\n" );
$part = $piece->getCurrentPart();
- // Search back through the input to see if it has a proper close
- // Do this using the reversed string since the other solutions (end anchor, etc.) are inefficient
- $wsLength = strspn( $revText, " \t", strlen( $text ) - $i );
+ // Search back through the input to see if it has a proper close.
+ // Do this using the reversed string since the other solutions
+ // (end anchor, etc.) are inefficient.
+ $wsLength = strspn( $revText, " \t", $lengthText - $i );
$searchStart = $i - $wsLength;
if ( isset( $part->commentEnd ) && $searchStart - 1 == $part->commentEnd ) {
// Comment found at line end
// Search for equals signs before the comment
$searchStart = $part->visualEnd;
- $searchStart -= strspn( $revText, " \t", strlen( $text ) - $searchStart );
+ $searchStart -= strspn( $revText, " \t", $lengthText - $searchStart );
}
$count = $piece->count;
- $equalsLength = strspn( $revText, '=', strlen( $text ) - $searchStart );
+ $equalsLength = strspn( $revText, '=', $lengthText - $searchStart );
if ( $equalsLength > 0 ) {
if ( $searchStart - $equalsLength == $piece->startPos ) {
// This is just a single string of equals signs on its own line
- // Replicate the doHeadings behaviour /={count}(.+)={count}/
+ // Replicate the doHeadings behavior /={count}(.+)={count}/
// First find out how many equals signs there really are (don't stop at 6)
$count = $equalsLength;
if ( $count < 3 ) {
}
if ( $count > 0 ) {
// Normal match, output <h>
- $element = new PPNode_Hash_Tree( 'possible-h' );
- $element->addChild( new PPNode_Hash_Attr( 'level', $count ) );
- $element->addChild( new PPNode_Hash_Attr( 'i', $headingIndex++ ) );
- $element->lastChild->nextSibling = $accum->firstNode;
- $element->lastChild = $accum->lastNode;
+ $element = [ [ 'possible-h',
+ array_merge(
+ [
+ [ '@level', [ $count ] ],
+ [ '@i', [ $headingIndex++ ] ]
+ ],
+ $accum
+ )
+ ] ];
} else {
// Single equals sign on its own line, count=0
$element = $accum;
extract( $stack->getFlags() );
// Append the result to the enclosing accumulator
- if ( $element instanceof PPNode ) {
- $accum->addNode( $element );
- } else {
- $accum->addAccum( $element );
- }
+ array_splice( $accum, count( $accum ), 0, $element );
+
// Note that we do NOT increment the input pointer.
// This is because the closing linebreak could be the opening linebreak of
// 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 );
+ $curLen = strlen( $curChar );
+ $count = ( $curLen > 1 ) ?
+ # allow the final character to repeat
+ strspn( $text, $curChar[$curLen - 1], $i + 1 ) + 1 :
+ strspn( $text, $curChar, $i );
# we need to add to stack only if opening brace count is enough for one of the rules
if ( $count >= $rule['min'] ) {
# Add it to the stack
- $piece = array(
+ $piece = [
'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 =& $stack->getAccum();
extract( $stack->getFlags() );
} else {
# Add literal brace(s)
- $accum->addLiteral( str_repeat( $curChar, $count ) );
+ self::addLiteral( $accum, 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;
- $count = strspn( $text, $curChar, $i, $maxCount );
+ if ( $piece->close === '}-' && $curChar === '}' ) {
+ $maxCount--; # don't try to match closing '-' as a '}'
+ }
+ $curLen = strlen( $curChar );
+ $count = ( $curLen > 1 ) ? $curLen :
+ strspn( $text, $curChar, $i, $maxCount );
# 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 ( $piece->close === '}-' && $piece->count > 2 ) {
+ # tweak for -{..{{ }}..}-
+ $rule = $this->rules['{'];
+ }
if ( $count > $rule['max'] ) {
# The specified maximum exists in the callback array, unless the caller
# has made an error
}
}
- if ($matchingCount <= 0) {
+ if ( $matchingCount <= 0 ) {
# No matching element found in callback array
# Output a literal closing brace and continue
- $accum->addLiteral( str_repeat( $curChar, $count ) );
+ $endText = substr( $text, $i, $count );
+ self::addLiteral( $accum, $endText );
$i += $count;
continue;
}
$name = $rule['names'][$matchingCount];
if ( $name === null ) {
// No element, just literal text
+ $endText = substr( $text, $i, $matchingCount );
$element = $piece->breakSyntax( $matchingCount );
- $element->addLiteral( str_repeat( $rule['end'], $matchingCount ) );
+ self::addLiteral( $element, $endText );
} else {
# Create XML element
- # Note: $parts is already XML, does not need to be encoded further
$parts = $piece->parts;
$titleAccum = $parts[0]->out;
unset( $parts[0] );
- $element = new PPNode_Hash_Tree( $name );
+ $children = [];
# The invocation is at the start of the line if lineStart is set in
# the stack, and all opening brackets are used up.
if ( $maxCount == $matchingCount && !empty( $piece->lineStart ) ) {
- $element->addChild( new PPNode_Hash_Attr( 'lineStart', 1 ) );
+ $children[] = [ '@lineStart', [ 1 ] ];
}
- $titleNode = new PPNode_Hash_Tree( 'title' );
- $titleNode->firstChild = $titleAccum->firstNode;
- $titleNode->lastChild = $titleAccum->lastNode;
- $element->addChild( $titleNode );
+ $titleNode = [ 'title', $titleAccum ];
+ $children[] = $titleNode;
$argIndex = 1;
foreach ( $parts as $part ) {
if ( isset( $part->eqpos ) ) {
- // Find equals
- $lastNode = false;
- for ( $node = $part->out->firstNode; $node; $node = $node->nextSibling ) {
- if ( $node === $part->eqpos ) {
- break;
- }
- $lastNode = $node;
- }
- if ( !$node ) {
- throw new MWException( __METHOD__. ': eqpos not found' );
- }
- if ( $node->name !== 'equals' ) {
- throw new MWException( __METHOD__ .': eqpos is not equals' );
- }
- $equalsNode = $node;
-
- // Construct name node
- $nameNode = new PPNode_Hash_Tree( 'name' );
- if ( $lastNode !== false ) {
- $lastNode->nextSibling = false;
- $nameNode->firstChild = $part->out->firstNode;
- $nameNode->lastChild = $lastNode;
- }
-
- // Construct value node
- $valueNode = new PPNode_Hash_Tree( 'value' );
- if ( $equalsNode->nextSibling !== false ) {
- $valueNode->firstChild = $equalsNode->nextSibling;
- $valueNode->lastChild = $part->out->lastNode;
- }
- $partNode = new PPNode_Hash_Tree( 'part' );
- $partNode->addChild( $nameNode );
- $partNode->addChild( $equalsNode->firstChild );
- $partNode->addChild( $valueNode );
- $element->addChild( $partNode );
+ $equalsNode = $part->out[$part->eqpos];
+ $nameNode = [ 'name', array_slice( $part->out, 0, $part->eqpos ) ];
+ $valueNode = [ 'value', array_slice( $part->out, $part->eqpos + 1 ) ];
+ $partNode = [ 'part', [ $nameNode, $equalsNode, $valueNode ] ];
+ $children[] = $partNode;
} else {
- $partNode = new PPNode_Hash_Tree( 'part' );
- $nameNode = new PPNode_Hash_Tree( 'name' );
- $nameNode->addChild( new PPNode_Hash_Attr( 'index', $argIndex++ ) );
- $valueNode = new PPNode_Hash_Tree( 'value' );
- $valueNode->firstChild = $part->out->firstNode;
- $valueNode->lastChild = $part->out->lastNode;
- $partNode->addChild( $nameNode );
- $partNode->addChild( $valueNode );
- $element->addChild( $partNode );
+ $nameNode = [ 'name', [ [ '@index', [ $argIndex++ ] ] ] ];
+ $valueNode = [ 'value', $part->out ];
+ $partNode = [ 'part', [ $nameNode, $valueNode ] ];
+ $children[] = $partNode;
}
}
+ $element = [ [ $name, $children ] ];
}
# Advance input pointer
$accum =& $stack->getAccum();
# Re-add the old stack element if it still has unmatched opening characters remaining
- if ($matchingCount < $piece->count) {
- $piece->parts = array( new PPDPart_Hash );
+ if ( $matchingCount < $piece->count ) {
+ $piece->parts = [ new PPDPart_Hash ];
$piece->count -= $matchingCount;
# do we still qualify for any callback with remaining count?
- $names = $rules[$piece->open]['names'];
- $skippedBraces = 0;
- $enclosingAccum =& $accum;
- while ( $piece->count ) {
- if ( array_key_exists( $piece->count, $names ) ) {
- $stack->push( $piece );
- $accum =& $stack->getAccum();
- break;
- }
- --$piece->count;
- $skippedBraces ++;
+ $min = $this->rules[$piece->open]['min'];
+ if ( $piece->count >= $min ) {
+ $stack->push( $piece );
+ $accum =& $stack->getAccum();
+ } else {
+ $s = substr( $piece->open, 0, -1 );
+ $s .= str_repeat(
+ substr( $piece->open, -1 ),
+ $piece->count - strlen( $s )
+ );
+ self::addLiteral( $accum, $s );
}
- $enclosingAccum->addLiteral( str_repeat( $piece->open, $skippedBraces ) );
}
extract( $stack->getFlags() );
# Add XML element to the enclosing accumulator
- if ( $element instanceof PPNode ) {
- $accum->addNode( $element );
- } else {
- $accum->addAccum( $element );
- }
- }
-
- elseif ( $found == 'pipe' ) {
+ array_splice( $accum, count( $accum ), 0, $element );
+ } elseif ( $found == 'pipe' ) {
$findEquals = true; // shortcut for getFlags()
$stack->addPart();
$accum =& $stack->getAccum();
++$i;
- }
-
- elseif ( $found == 'equals' ) {
+ } elseif ( $found == 'equals' ) {
$findEquals = false; // shortcut for getFlags()
- $accum->addNodeWithText( 'equals', '=' );
- $stack->getCurrentPart()->eqpos = $accum->lastNode;
+ $accum[] = [ 'equals', [ '=' ] ];
+ $stack->getCurrentPart()->eqpos = count( $accum ) - 1;
+ ++$i;
+ } elseif ( $found == 'dash' ) {
+ self::addLiteral( $accum, '-' );
++$i;
}
}
# Output any remaining unclosed brackets
foreach ( $stack->stack as $piece ) {
- $stack->rootAccum->addAccum( $piece->breakSyntax() );
+ array_splice( $stack->rootAccum, count( $stack->rootAccum ), 0, $piece->breakSyntax() );
}
# Enable top-level headings
- for ( $node = $stack->rootAccum->firstNode; $node; $node = $node->nextSibling ) {
- if ( isset( $node->name ) && $node->name === 'possible-h' ) {
- $node->name = 'h';
+ foreach ( $stack->rootAccum as &$node ) {
+ if ( is_array( $node ) && $node[PPNode_Hash_Tree::NAME] === 'possible-h' ) {
+ $node[PPNode_Hash_Tree::NAME] = 'h';
}
}
- $rootNode = new PPNode_Hash_Tree( 'root' );
- $rootNode->firstChild = $stack->rootAccum->firstNode;
- $rootNode->lastChild = $stack->rootAccum->lastNode;
-
+ $rootStore = [ [ 'root', $stack->rootAccum ] ];
+ $rootNode = new PPNode_Hash_Tree( $rootStore, 0 );
+
// Cache
- if ($cacheable) {
- $cacheValue = sprintf( "%08d", self::CACHE_VERSION ) . serialize( $rootNode );
- $wgMemc->set( $cacheKey, $cacheValue, 86400 );
- wfProfileOut( __METHOD__.'-cache-miss' );
- wfProfileOut( __METHOD__.'-cacheable' );
- wfDebugLog( "Preprocessor", "Saved preprocessor Hash to memcached (key $cacheKey)" );
+ $tree = json_encode( $rootStore, JSON_UNESCAPED_SLASHES | JSON_UNESCAPED_UNICODE );
+ if ( $tree !== false ) {
+ $this->cacheSetTree( $text, $flags, $tree );
}
-
- wfProfileOut( __METHOD__ );
+
return $rootNode;
}
+
+ private static function addLiteral( array &$accum, $text ) {
+ $n = count( $accum );
+ if ( $n && is_string( $accum[$n - 1] ) ) {
+ $accum[$n - 1] .= $text;
+ } else {
+ $accum[] = $text;
+ }
+ }
}
/**
* Stack class to help Preprocessor::preprocessToObj()
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPDStack_Hash extends PPDStack {
- function __construct() {
+ // @codingStandardsIgnoreEnd
+
+ public function __construct() {
$this->elementClass = 'PPDStackElement_Hash';
parent::__construct();
- $this->rootAccum = new PPDAccum_Hash;
+ $this->rootAccum = [];
}
}
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPDStackElement_Hash extends PPDStackElement {
- function __construct( $data = array() ) {
+ // @codingStandardsIgnoreEnd
+
+ public function __construct( $data = [] ) {
$this->partClass = 'PPDPart_Hash';
parent::__construct( $data );
}
/**
* Get the accumulator that would result if the close is not found.
+ *
+ * @param int|bool $openingCount
+ * @return array
*/
- function breakSyntax( $openingCount = false ) {
+ public function breakSyntax( $openingCount = false ) {
if ( $this->open == "\n" ) {
$accum = $this->parts[0]->out;
} else {
if ( $openingCount === false ) {
$openingCount = $this->count;
}
- $accum = new PPDAccum_Hash;
- $accum->addLiteral( str_repeat( $this->open, $openingCount ) );
+ $s = substr( $this->open, 0, -1 );
+ $s .= str_repeat(
+ substr( $this->open, -1 ),
+ $openingCount - strlen( $s )
+ );
+ $accum = [ $s ];
+ $lastIndex = 0;
$first = true;
foreach ( $this->parts as $part ) {
if ( $first ) {
$first = false;
+ } elseif ( is_string( $accum[$lastIndex] ) ) {
+ $accum[$lastIndex] .= '|';
} else {
- $accum->addLiteral( '|' );
+ $accum[++$lastIndex] = '|';
+ }
+ foreach ( $part->out as $node ) {
+ if ( is_string( $node ) && is_string( $accum[$lastIndex] ) ) {
+ $accum[$lastIndex] .= $node;
+ } else {
+ $accum[++$lastIndex] = $node;
+ }
}
- $accum->addAccum( $part->out );
}
}
return $accum;
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPDPart_Hash extends PPDPart {
- function __construct( $out = '' ) {
- $accum = new PPDAccum_Hash;
+ // @codingStandardsIgnoreEnd
+
+ public function __construct( $out = '' ) {
if ( $out !== '' ) {
- $accum->addLiteral( $out );
+ $accum = [ $out ];
+ } else {
+ $accum = [];
}
parent::__construct( $accum );
}
}
/**
+ * An expansion frame, used as a context to expand the result of preprocessToObj()
* @ingroup Parser
*/
-class PPDAccum_Hash {
- var $firstNode, $lastNode;
-
- function __construct() {
- $this->firstNode = $this->lastNode = false;
- }
-
- /**
- * Append a string literal
- */
- function addLiteral( $s ) {
- if ( $this->lastNode === false ) {
- $this->firstNode = $this->lastNode = new PPNode_Hash_Text( $s );
- } elseif ( $this->lastNode instanceof PPNode_Hash_Text ) {
- $this->lastNode->value .= $s;
- } else {
- $this->lastNode->nextSibling = new PPNode_Hash_Text( $s );
- $this->lastNode = $this->lastNode->nextSibling;
- }
- }
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
+class PPFrame_Hash implements PPFrame {
+ // @codingStandardsIgnoreEnd
/**
- * Append a PPNode
+ * @var Parser
*/
- function addNode( PPNode $node ) {
- if ( $this->lastNode === false ) {
- $this->firstNode = $this->lastNode = $node;
- } else {
- $this->lastNode->nextSibling = $node;
- $this->lastNode = $node;
- }
- }
+ public $parser;
/**
- * Append a tree node with text contents
+ * @var Preprocessor
*/
- function addNodeWithText( $name, $value ) {
- $node = PPNode_Hash_Tree::newWithText( $name, $value );
- $this->addNode( $node );
- }
+ public $preprocessor;
/**
- * Append a PPAccum_Hash
- * Takes over ownership of the nodes in the source argument. These nodes may
- * subsequently be modified, especially nextSibling.
+ * @var Title
*/
- function addAccum( $accum ) {
- if ( $accum->lastNode === false ) {
- // nothing to add
- } elseif ( $this->lastNode === false ) {
- $this->firstNode = $accum->firstNode;
- $this->lastNode = $accum->lastNode;
- } else {
- $this->lastNode->nextSibling = $accum->firstNode;
- $this->lastNode = $accum->lastNode;
- }
- }
-}
-
-/**
- * An expansion frame, used as a context to expand the result of preprocessToObj()
- * @ingroup Parser
- */
-class PPFrame_Hash implements PPFrame {
- var $preprocessor, $parser, $title;
- var $titleCache;
+ public $title;
+ public $titleCache;
/**
* Hashtable listing templates which are disallowed for expansion in this frame,
* having been encountered previously in parent frames.
*/
- var $loopCheckHash;
+ public $loopCheckHash;
/**
* Recursion depth of this frame, top = 0
* Note that this is NOT the same as expansion depth in expand()
*/
- var $depth;
+ public $depth;
+
+ private $volatile = false;
+ private $ttl = null;
+ /**
+ * @var array
+ */
+ protected $childExpansionCache;
/**
* Construct a new preprocessor frame.
- * @param $preprocessor Preprocessor: the parent preprocessor
+ * @param Preprocessor $preprocessor The parent preprocessor
*/
- function __construct( $preprocessor ) {
+ public function __construct( $preprocessor ) {
$this->preprocessor = $preprocessor;
$this->parser = $preprocessor->parser;
$this->title = $this->parser->mTitle;
- $this->titleCache = array( $this->title ? $this->title->getPrefixedDBkey() : false );
- $this->loopCheckHash = array();
+ $this->titleCache = [ $this->title ? $this->title->getPrefixedDBkey() : false ];
+ $this->loopCheckHash = [];
$this->depth = 0;
+ $this->childExpansionCache = [];
}
/**
* Create a new child frame
* $args is optionally a multi-root PPNode or array containing the template arguments
+ *
+ * @param array|bool|PPNode_Hash_Array $args
+ * @param Title|bool $title
+ * @param int $indexOffset
+ * @throws MWException
+ * @return PPTemplateFrame_Hash
*/
- function newChild( $args = false, $title = false ) {
- $namedArgs = array();
- $numberedArgs = array();
+ public function newChild( $args = false, $title = false, $indexOffset = 0 ) {
+ $namedArgs = [];
+ $numberedArgs = [];
if ( $title === false ) {
$title = $this->title;
}
$bits = $arg->splitArg();
if ( $bits['index'] !== '' ) {
// Numbered parameter
- $numberedArgs[$bits['index']] = $bits['value'];
- unset( $namedArgs[$bits['index']] );
+ $index = $bits['index'] - $indexOffset;
+ if ( isset( $namedArgs[$index] ) || isset( $numberedArgs[$index] ) ) {
+ $this->parser->getOutput()->addWarning( wfMessage( 'duplicate-args-warning',
+ wfEscapeWikiText( $this->title ),
+ wfEscapeWikiText( $title ),
+ wfEscapeWikiText( $index ) )->text() );
+ $this->parser->addTrackingCategory( 'duplicate-args-category' );
+ }
+ $numberedArgs[$index] = $bits['value'];
+ unset( $namedArgs[$index] );
} else {
// Named parameter
$name = trim( $this->expand( $bits['name'], PPFrame::STRIP_COMMENTS ) );
+ if ( isset( $namedArgs[$name] ) || isset( $numberedArgs[$name] ) ) {
+ $this->parser->getOutput()->addWarning( wfMessage( 'duplicate-args-warning',
+ wfEscapeWikiText( $this->title ),
+ wfEscapeWikiText( $title ),
+ wfEscapeWikiText( $name ) )->text() );
+ $this->parser->addTrackingCategory( 'duplicate-args-category' );
+ }
$namedArgs[$name] = $bits['value'];
unset( $numberedArgs[$name] );
}
return new PPTemplateFrame_Hash( $this->preprocessor, $this, $numberedArgs, $namedArgs, $title );
}
- function expand( $root, $flags = 0 ) {
+ /**
+ * @throws MWException
+ * @param string|int $key
+ * @param string|PPNode $root
+ * @param int $flags
+ * @return string
+ */
+ public function cachedExpand( $key, $root, $flags = 0 ) {
+ // we don't have a parent, so we don't have a cache
+ return $this->expand( $root, $flags );
+ }
+
+ /**
+ * @throws MWException
+ * @param string|PPNode $root
+ * @param int $flags
+ * @return string
+ */
+ public function expand( $root, $flags = 0 ) {
static $expansionDepth = 0;
if ( is_string( $root ) ) {
return $root;
}
- if ( ++$this->parser->mPPNodeCount > $this->parser->mOptions->getMaxPPNodeCount() )
- {
+ if ( ++$this->parser->mPPNodeCount > $this->parser->mOptions->getMaxPPNodeCount() ) {
+ $this->parser->limitationWarn( 'node-count-exceeded',
+ $this->parser->mPPNodeCount,
+ $this->parser->mOptions->getMaxPPNodeCount()
+ );
return '<span class="error">Node-count limit exceeded</span>';
}
if ( $expansionDepth > $this->parser->mOptions->getMaxPPExpandDepth() ) {
+ $this->parser->limitationWarn( 'expansion-depth-exceeded',
+ $expansionDepth,
+ $this->parser->mOptions->getMaxPPExpandDepth()
+ );
return '<span class="error">Expansion depth limit exceeded</span>';
}
++$expansionDepth;
+ if ( $expansionDepth > $this->parser->mHighestExpansionDepth ) {
+ $this->parser->mHighestExpansionDepth = $expansionDepth;
+ }
- $outStack = array( '', '' );
- $iteratorStack = array( false, $root );
- $indexStack = array( 0, 0 );
+ $outStack = [ '', '' ];
+ $iteratorStack = [ false, $root ];
+ $indexStack = [ 0, 0 ];
while ( count( $iteratorStack ) > 1 ) {
$level = count( $outStack ) - 1;
- $iteratorNode =& $iteratorStack[ $level ];
+ $iteratorNode =& $iteratorStack[$level];
$out =& $outStack[$level];
$index =& $indexStack[$level];
}
$newIterator = false;
+ $contextName = false;
+ $contextChildren = false;
if ( $contextNode === false ) {
// nothing to do
} elseif ( is_string( $contextNode ) ) {
$out .= $contextNode;
- } elseif ( is_array( $contextNode ) || $contextNode instanceof PPNode_Hash_Array ) {
+ } elseif ( $contextNode instanceof PPNode_Hash_Array ) {
$newIterator = $contextNode;
} elseif ( $contextNode instanceof PPNode_Hash_Attr ) {
// No output
} elseif ( $contextNode instanceof PPNode_Hash_Text ) {
$out .= $contextNode->value;
} elseif ( $contextNode instanceof PPNode_Hash_Tree ) {
- if ( $contextNode->name == 'template' ) {
- # Double-brace expansion
- $bits = $contextNode->splitTemplate();
- if ( $flags & PPFrame::NO_TEMPLATES ) {
- $newIterator = $this->virtualBracketedImplode( '{{', '|', '}}', $bits['title'], $bits['parts'] );
+ $contextName = $contextNode->name;
+ $contextChildren = $contextNode->getRawChildren();
+ } elseif ( is_array( $contextNode ) ) {
+ // Node descriptor array
+ if ( count( $contextNode ) !== 2 ) {
+ throw new MWException( __METHOD__.
+ ': found an array where a node descriptor should be' );
+ }
+ list( $contextName, $contextChildren ) = $contextNode;
+ } else {
+ throw new MWException( __METHOD__ . ': Invalid parameter type' );
+ }
+
+ // Handle node descriptor array or tree object
+ if ( $contextName === false ) {
+ // Not a node, already handled above
+ } elseif ( $contextName[0] === '@' ) {
+ // Attribute: no output
+ } elseif ( $contextName === 'template' ) {
+ # Double-brace expansion
+ $bits = PPNode_Hash_Tree::splitRawTemplate( $contextChildren );
+ if ( $flags & PPFrame::NO_TEMPLATES ) {
+ $newIterator = $this->virtualBracketedImplode(
+ '{{', '|', '}}',
+ $bits['title'],
+ $bits['parts']
+ );
+ } else {
+ $ret = $this->parser->braceSubstitution( $bits, $this );
+ if ( isset( $ret['object'] ) ) {
+ $newIterator = $ret['object'];
} else {
- $ret = $this->parser->braceSubstitution( $bits, $this );
- if ( isset( $ret['object'] ) ) {
- $newIterator = $ret['object'];
- } else {
- $out .= $ret['text'];
- }
+ $out .= $ret['text'];
}
- } elseif ( $contextNode->name == 'tplarg' ) {
- # Triple-brace expansion
- $bits = $contextNode->splitTemplate();
- if ( $flags & PPFrame::NO_ARGS ) {
- $newIterator = $this->virtualBracketedImplode( '{{{', '|', '}}}', $bits['title'], $bits['parts'] );
+ }
+ } elseif ( $contextName === 'tplarg' ) {
+ # Triple-brace expansion
+ $bits = PPNode_Hash_Tree::splitRawTemplate( $contextChildren );
+ if ( $flags & PPFrame::NO_ARGS ) {
+ $newIterator = $this->virtualBracketedImplode(
+ '{{{', '|', '}}}',
+ $bits['title'],
+ $bits['parts']
+ );
+ } else {
+ $ret = $this->parser->argSubstitution( $bits, $this );
+ if ( isset( $ret['object'] ) ) {
+ $newIterator = $ret['object'];
} else {
- $ret = $this->parser->argSubstitution( $bits, $this );
- if ( isset( $ret['object'] ) ) {
- $newIterator = $ret['object'];
- } else {
- $out .= $ret['text'];
- }
- }
- } elseif ( $contextNode->name == 'comment' ) {
- # HTML-style comment
- # Remove it in HTML, pre+remove and STRIP_COMMENTS modes
- if ( $this->parser->ot['html']
- || ( $this->parser->ot['pre'] && $this->parser->mOptions->getRemoveComments() )
- || ( $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 & PPFrame::RECOVER_COMMENTS ) ) {
- $out .= $this->parser->insertStripItem( $contextNode->firstChild->value );
+ $out .= $ret['text'];
}
+ }
+ } elseif ( $contextName === 'comment' ) {
+ # HTML-style comment
+ # Remove it in HTML, pre+remove and STRIP_COMMENTS modes
+ # Not in RECOVER_COMMENTS mode (msgnw) though.
+ if ( ( $this->parser->ot['html']
+ || ( $this->parser->ot['pre'] && $this->parser->mOptions->getRemoveComments() )
+ || ( $flags & PPFrame::STRIP_COMMENTS )
+ ) && !( $flags & PPFrame::RECOVER_COMMENTS )
+ ) {
+ $out .= '';
+ } elseif ( $this->parser->ot['wiki'] && !( $flags & PPFrame::RECOVER_COMMENTS ) ) {
+ # 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.
+ $out .= $this->parser->insertStripItem( $contextChildren[0] );
+ } else {
# Recover the literal comment in RECOVER_COMMENTS and pre+no-remove
- else {
- $out .= $contextNode->firstChild->value;
- }
- } elseif ( $contextNode->name == 'ignore' ) {
- # Output suppression used by <includeonly> etc.
- # 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 & PPFrame::NO_IGNORE ) ) {
- $out .= $contextNode->firstChild->value;
- } else {
- //$out .= '';
+ $out .= $contextChildren[0];
+ }
+ } elseif ( $contextName === 'ignore' ) {
+ # Output suppression used by <includeonly> etc.
+ # 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 & PPFrame::NO_IGNORE )
+ ) {
+ $out .= $contextChildren[0];
+ } else {
+ // $out .= '';
+ }
+ } elseif ( $contextName === 'ext' ) {
+ # Extension tag
+ $bits = PPNode_Hash_Tree::splitRawExt( $contextChildren ) +
+ [ 'attr' => null, 'inner' => null, 'close' => null ];
+ if ( $flags & PPFrame::NO_TAGS ) {
+ $s = '<' . $bits['name']->getFirstChild()->value;
+ if ( $bits['attr'] ) {
+ $s .= $bits['attr']->getFirstChild()->value;
}
- } elseif ( $contextNode->name == 'ext' ) {
- # Extension tag
- $bits = $contextNode->splitExt() + array( 'attr' => null, 'inner' => null, 'close' => null );
- $out .= $this->parser->extensionSubstitution( $bits, $this );
- } elseif ( $contextNode->name == 'h' ) {
- # Heading
- if ( $this->parser->ot['html'] ) {
- # Expand immediately and insert heading index marker
- $s = '';
- for ( $node = $contextNode->firstChild; $node; $node = $node->nextSibling ) {
- $s .= $this->expand( $node, $flags );
+ if ( $bits['inner'] ) {
+ $s .= '>' . $bits['inner']->getFirstChild()->value;
+ if ( $bits['close'] ) {
+ $s .= $bits['close']->getFirstChild()->value;
}
-
- $bits = $contextNode->splitHeading();
- $titleText = $this->title->getPrefixedDBkey();
- $this->parser->mHeadings[] = array( $titleText, $bits['i'] );
- $serial = count( $this->parser->mHeadings ) - 1;
- $marker = "{$this->parser->mUniqPrefix}-h-$serial-" . Parser::MARKER_SUFFIX;
- $s = substr( $s, 0, $bits['level'] ) . $marker . substr( $s, $bits['level'] );
- $this->parser->mStripState->general->setPair( $marker, '' );
- $out .= $s;
} else {
- # Expand in virtual stack
- $newIterator = $contextNode->getChildren();
+ $s .= '/>';
}
+ $out .= $s;
+ } else {
+ $out .= $this->parser->extensionSubstitution( $bits, $this );
+ }
+ } elseif ( $contextName === 'h' ) {
+ # Heading
+ if ( $this->parser->ot['html'] ) {
+ # Expand immediately and insert heading index marker
+ $s = $this->expand( $contextChildren, $flags );
+ $bits = PPNode_Hash_Tree::splitRawHeading( $contextChildren );
+ $titleText = $this->title->getPrefixedDBkey();
+ $this->parser->mHeadings[] = [ $titleText, $bits['i'] ];
+ $serial = count( $this->parser->mHeadings ) - 1;
+ $marker = Parser::MARKER_PREFIX . "-h-$serial-" . Parser::MARKER_SUFFIX;
+ $s = substr( $s, 0, $bits['level'] ) . $marker . substr( $s, $bits['level'] );
+ $this->parser->mStripState->addGeneral( $marker, '' );
+ $out .= $s;
} else {
- # Generic recursive expansion
- $newIterator = $contextNode->getChildren();
+ # Expand in virtual stack
+ $newIterator = $contextChildren;
}
} else {
- throw new MWException( __METHOD__.': Invalid parameter type' );
+ # Generic recursive expansion
+ $newIterator = $contextChildren;
}
if ( $newIterator !== false ) {
return $outStack[0];
}
- function implodeWithFlags( $sep, $flags /*, ... */ ) {
+ /**
+ * @param string $sep
+ * @param int $flags
+ * @param string|PPNode $args,...
+ * @return string
+ */
+ public function implodeWithFlags( $sep, $flags /*, ... */ ) {
$args = array_slice( func_get_args(), 2 );
$first = true;
$root = $root->value;
}
if ( !is_array( $root ) ) {
- $root = array( $root );
+ $root = [ $root ];
}
foreach ( $root as $node ) {
if ( $first ) {
/**
* Implode with no flags specified
* This previously called implodeWithFlags but has now been inlined to reduce stack depth
+ * @param string $sep
+ * @param string|PPNode $args,...
+ * @return string
*/
- function implode( $sep /*, ... */ ) {
+ public function implode( $sep /*, ... */ ) {
$args = array_slice( func_get_args(), 1 );
$first = true;
$root = $root->value;
}
if ( !is_array( $root ) ) {
- $root = array( $root );
+ $root = [ $root ];
}
foreach ( $root as $node ) {
if ( $first ) {
/**
* Makes an object that, when expand()ed, will be the same as one obtained
* with implode()
+ *
+ * @param string $sep
+ * @param string|PPNode $args,...
+ * @return PPNode_Hash_Array
*/
- function virtualImplode( $sep /*, ... */ ) {
+ public function virtualImplode( $sep /*, ... */ ) {
$args = array_slice( func_get_args(), 1 );
- $out = array();
+ $out = [];
$first = true;
foreach ( $args as $root ) {
$root = $root->value;
}
if ( !is_array( $root ) ) {
- $root = array( $root );
+ $root = [ $root ];
}
foreach ( $root as $node ) {
if ( $first ) {
/**
* Virtual implode with brackets
+ *
+ * @param string $start
+ * @param string $sep
+ * @param string $end
+ * @param string|PPNode $args,...
+ * @return PPNode_Hash_Array
*/
- function virtualBracketedImplode( $start, $sep, $end /*, ... */ ) {
+ public function virtualBracketedImplode( $start, $sep, $end /*, ... */ ) {
$args = array_slice( func_get_args(), 3 );
- $out = array( $start );
+ $out = [ $start ];
$first = true;
foreach ( $args as $root ) {
$root = $root->value;
}
if ( !is_array( $root ) ) {
- $root = array( $root );
+ $root = [ $root ];
}
foreach ( $root as $node ) {
if ( $first ) {
return new PPNode_Hash_Array( $out );
}
- function __toString() {
+ public function __toString() {
return 'frame{}';
}
- function getPDBK( $level = false ) {
+ /**
+ * @param bool $level
+ * @return array|bool|string
+ */
+ public function getPDBK( $level = false ) {
if ( $level === false ) {
return $this->title->getPrefixedDBkey();
} else {
}
}
- function getArguments() {
- return array();
+ /**
+ * @return array
+ */
+ public function getArguments() {
+ return [];
}
- function getNumberedArguments() {
- return array();
+ /**
+ * @return array
+ */
+ public function getNumberedArguments() {
+ return [];
}
- function getNamedArguments() {
- return array();
+ /**
+ * @return array
+ */
+ public function getNamedArguments() {
+ return [];
}
/**
* Returns true if there are no arguments in this frame
+ *
+ * @return bool
*/
- function isEmpty() {
+ public function isEmpty() {
return true;
}
- function getArgument( $name ) {
+ /**
+ * @param int|string $name
+ * @return bool Always false in this implementation.
+ */
+ public function getArgument( $name ) {
return false;
}
/**
* Returns true if the infinite loop check is OK, false if a loop is detected
+ *
+ * @param Title $title
+ *
+ * @return bool
*/
- function loopCheck( $title ) {
+ public function loopCheck( $title ) {
return !isset( $this->loopCheckHash[$title->getPrefixedDBkey()] );
}
/**
* Return true if the frame is a template frame
+ *
+ * @return bool
*/
- function isTemplate() {
+ public function isTemplate() {
return false;
}
+
+ /**
+ * Get a title of frame
+ *
+ * @return Title
+ */
+ public function getTitle() {
+ return $this->title;
+ }
+
+ /**
+ * Set the volatile flag
+ *
+ * @param bool $flag
+ */
+ public function setVolatile( $flag = true ) {
+ $this->volatile = $flag;
+ }
+
+ /**
+ * Get the volatile flag
+ *
+ * @return bool
+ */
+ public function isVolatile() {
+ return $this->volatile;
+ }
+
+ /**
+ * Set the TTL
+ *
+ * @param int $ttl
+ */
+ public function setTTL( $ttl ) {
+ if ( $ttl !== null && ( $this->ttl === null || $ttl < $this->ttl ) ) {
+ $this->ttl = $ttl;
+ }
+ }
+
+ /**
+ * Get the TTL
+ *
+ * @return int|null
+ */
+ public function getTTL() {
+ return $this->ttl;
+ }
}
/**
* Expansion frame with template arguments
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPTemplateFrame_Hash extends PPFrame_Hash {
- var $numberedArgs, $namedArgs, $parent;
- var $numberedExpansionCache, $namedExpansionCache;
+ // @codingStandardsIgnoreEnd
+
+ public $numberedArgs, $namedArgs, $parent;
+ public $numberedExpansionCache, $namedExpansionCache;
- function __construct( $preprocessor, $parent = false, $numberedArgs = array(), $namedArgs = array(), $title = false ) {
+ /**
+ * @param Preprocessor $preprocessor
+ * @param bool|PPFrame $parent
+ * @param array $numberedArgs
+ * @param array $namedArgs
+ * @param bool|Title $title
+ */
+ public function __construct( $preprocessor, $parent = false, $numberedArgs = [],
+ $namedArgs = [], $title = false
+ ) {
parent::__construct( $preprocessor );
$this->parent = $parent;
$this->loopCheckHash[$pdbk] = true;
}
$this->depth = $parent->depth + 1;
- $this->numberedExpansionCache = $this->namedExpansionCache = array();
+ $this->numberedExpansionCache = $this->namedExpansionCache = [];
}
- function __toString() {
+ public function __toString() {
$s = 'tplframe{';
$first = true;
$args = $this->numberedArgs + $this->namedArgs;
$s .= '}';
return $s;
}
+
+ /**
+ * @throws MWException
+ * @param string|int $key
+ * @param string|PPNode $root
+ * @param int $flags
+ * @return string
+ */
+ public function cachedExpand( $key, $root, $flags = 0 ) {
+ if ( isset( $this->parent->childExpansionCache[$key] ) ) {
+ return $this->parent->childExpansionCache[$key];
+ }
+ $retval = $this->expand( $root, $flags );
+ if ( !$this->isVolatile() ) {
+ $this->parent->childExpansionCache[$key] = $retval;
+ }
+ return $retval;
+ }
+
/**
* Returns true if there are no arguments in this frame
+ *
+ * @return bool
*/
- function isEmpty() {
+ public function isEmpty() {
return !count( $this->numberedArgs ) && !count( $this->namedArgs );
}
- function getArguments() {
- $arguments = array();
+ /**
+ * @return array
+ */
+ public function getArguments() {
+ $arguments = [];
foreach ( array_merge(
- array_keys($this->numberedArgs),
- array_keys($this->namedArgs)) as $key ) {
- $arguments[$key] = $this->getArgument($key);
+ array_keys( $this->numberedArgs ),
+ array_keys( $this->namedArgs ) ) as $key ) {
+ $arguments[$key] = $this->getArgument( $key );
}
return $arguments;
}
-
- function getNumberedArguments() {
- $arguments = array();
- foreach ( array_keys($this->numberedArgs) as $key ) {
- $arguments[$key] = $this->getArgument($key);
+
+ /**
+ * @return array
+ */
+ public function getNumberedArguments() {
+ $arguments = [];
+ foreach ( array_keys( $this->numberedArgs ) as $key ) {
+ $arguments[$key] = $this->getArgument( $key );
}
return $arguments;
}
-
- function getNamedArguments() {
- $arguments = array();
- foreach ( array_keys($this->namedArgs) as $key ) {
- $arguments[$key] = $this->getArgument($key);
+
+ /**
+ * @return array
+ */
+ public function getNamedArguments() {
+ $arguments = [];
+ foreach ( array_keys( $this->namedArgs ) as $key ) {
+ $arguments[$key] = $this->getArgument( $key );
}
return $arguments;
}
- function getNumberedArgument( $index ) {
+ /**
+ * @param int $index
+ * @return string|bool
+ */
+ public function getNumberedArgument( $index ) {
if ( !isset( $this->numberedArgs[$index] ) ) {
return false;
}
if ( !isset( $this->numberedExpansionCache[$index] ) ) {
# No trimming for unnamed arguments
- $this->numberedExpansionCache[$index] = $this->parent->expand( $this->numberedArgs[$index], PPFrame::STRIP_COMMENTS );
+ $this->numberedExpansionCache[$index] = $this->parent->expand(
+ $this->numberedArgs[$index],
+ PPFrame::STRIP_COMMENTS
+ );
}
return $this->numberedExpansionCache[$index];
}
- function getNamedArgument( $name ) {
+ /**
+ * @param string $name
+ * @return string|bool
+ */
+ public function getNamedArgument( $name ) {
if ( !isset( $this->namedArgs[$name] ) ) {
return false;
}
return $this->namedExpansionCache[$name];
}
- function getArgument( $name ) {
+ /**
+ * @param int|string $name
+ * @return string|bool
+ */
+ public function getArgument( $name ) {
$text = $this->getNumberedArgument( $name );
if ( $text === false ) {
$text = $this->getNamedArgument( $name );
/**
* Return true if the frame is a template frame
+ *
+ * @return bool
*/
- function isTemplate() {
+ public function isTemplate() {
return true;
}
+
+ public function setVolatile( $flag = true ) {
+ parent::setVolatile( $flag );
+ $this->parent->setVolatile( $flag );
+ }
+
+ public function setTTL( $ttl ) {
+ parent::setTTL( $ttl );
+ $this->parent->setTTL( $ttl );
+ }
}
/**
* Expansion frame with custom arguments
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPCustomFrame_Hash extends PPFrame_Hash {
- var $args;
+ // @codingStandardsIgnoreEnd
- function __construct( $preprocessor, $args ) {
+ public $args;
+
+ public function __construct( $preprocessor, $args ) {
parent::__construct( $preprocessor );
$this->args = $args;
}
- function __toString() {
+ public function __toString() {
$s = 'cstmframe{';
$first = true;
foreach ( $this->args as $name => $value ) {
return $s;
}
- function isEmpty() {
+ /**
+ * @return bool
+ */
+ public function isEmpty() {
return !count( $this->args );
}
- function getArgument( $index ) {
+ /**
+ * @param int|string $index
+ * @return string|bool
+ */
+ public function getArgument( $index ) {
if ( !isset( $this->args[$index] ) ) {
return false;
}
return $this->args[$index];
}
+
+ public function getArguments() {
+ return $this->args;
+ }
}
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPNode_Hash_Tree implements PPNode {
- var $name, $firstChild, $lastChild, $nextSibling;
+ // @codingStandardsIgnoreEnd
+
+ public $name;
+
+ /**
+ * The store array for children of this node. It is "raw" in the sense that
+ * nodes are two-element arrays ("descriptors") rather than PPNode_Hash_*
+ * objects.
+ */
+ private $rawChildren;
+
+ /**
+ * The store array for the siblings of this node, including this node itself.
+ */
+ private $store;
+
+ /**
+ * The index into $this->store which contains the descriptor of this node.
+ */
+ private $index;
- function __construct( $name ) {
- $this->name = $name;
- $this->firstChild = $this->lastChild = $this->nextSibling = false;
+ /**
+ * The offset of the name within descriptors, used in some places for
+ * readability.
+ */
+ const NAME = 0;
+
+ /**
+ * The offset of the child list within descriptors, used in some places for
+ * readability.
+ */
+ const CHILDREN = 1;
+
+ /**
+ * Construct an object using the data from $store[$index]. The rest of the
+ * store array can be accessed via getNextSibling().
+ *
+ * @param array $store
+ * @param int $index
+ */
+ public function __construct( array $store, $index ) {
+ $this->store = $store;
+ $this->index = $index;
+ list( $this->name, $this->rawChildren ) = $this->store[$index];
+ }
+
+ /**
+ * Construct an appropriate PPNode_Hash_* object with a class that depends
+ * on what is at the relevant store index.
+ *
+ * @param array $store
+ * @param int $index
+ * @return PPNode_Hash_Tree|PPNode_Hash_Attr|PPNode_Hash_Text
+ */
+ public static function factory( array $store, $index ) {
+ if ( !isset( $store[$index] ) ) {
+ return false;
+ }
+
+ $descriptor = $store[$index];
+ if ( is_string( $descriptor ) ) {
+ $class = 'PPNode_Hash_Text';
+ } elseif ( is_array( $descriptor ) ) {
+ if ( $descriptor[self::NAME][0] === '@' ) {
+ $class = 'PPNode_Hash_Attr';
+ } else {
+ $class = 'PPNode_Hash_Tree';
+ }
+ } else {
+ throw new MWException( __METHOD__.': invalid node descriptor' );
+ }
+ return new $class( $store, $index );
}
- function __toString() {
+ /**
+ * Convert a node to XML, for debugging
+ */
+ public function __toString() {
$inner = '';
$attribs = '';
- for ( $node = $this->firstChild; $node; $node = $node->nextSibling ) {
+ for ( $node = $this->getFirstChild(); $node; $node = $node->getNextSibling() ) {
if ( $node instanceof PPNode_Hash_Attr ) {
$attribs .= ' ' . $node->name . '="' . htmlspecialchars( $node->value ) . '"';
} else {
}
}
- static function newWithText( $name, $text ) {
- $obj = new self( $name );
- $obj->addChild( new PPNode_Hash_Text( $text ) );
- return $obj;
+ /**
+ * @return PPNode_Hash_Array
+ */
+ public function getChildren() {
+ $children = [];
+ foreach ( $this->rawChildren as $i => $child ) {
+ $children[] = self::factory( $this->rawChildren, $i );
+ }
+ return new PPNode_Hash_Array( $children );
}
- function addChild( $node ) {
- if ( $this->lastChild === false ) {
- $this->firstChild = $this->lastChild = $node;
+ /**
+ * Get the first child, or false if there is none. Note that this will
+ * return a temporary proxy object: different instances will be returned
+ * if this is called more than once on the same node.
+ *
+ * @return PPNode_Hash_Tree|PPNode_Hash_Attr|PPNode_Hash_Text|bool
+ */
+ public function getFirstChild() {
+ if ( !isset( $this->rawChildren[0] ) ) {
+ return false;
} else {
- $this->lastChild->nextSibling = $node;
- $this->lastChild = $node;
+ return self::factory( $this->rawChildren, 0 );
}
}
- function getChildren() {
- $children = array();
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- $children[] = $child;
+ /**
+ * Get the next sibling, or false if there is none. Note that this will
+ * return a temporary proxy object: different instances will be returned
+ * if this is called more than once on the same node.
+ *
+ * @return PPNode_Hash_Tree|PPNode_Hash_Attr|PPNode_Hash_Text|bool
+ */
+ public function getNextSibling() {
+ return self::factory( $this->store, $this->index + 1 );
+ }
+
+ /**
+ * Get an array of the children with a given node name
+ *
+ * @param string $name
+ * @return PPNode_Hash_Array
+ */
+ public function getChildrenOfType( $name ) {
+ $children = [];
+ foreach ( $this->rawChildren as $i => $child ) {
+ if ( is_array( $child ) && $child[self::NAME] === $name ) {
+ $children[] = self::factory( $this->rawChildren, $i );
+ }
}
return new PPNode_Hash_Array( $children );
}
- function getFirstChild() {
- return $this->firstChild;
+ /**
+ * Get the raw child array. For internal use.
+ * @return array
+ */
+ public function getRawChildren() {
+ return $this->rawChildren;
}
- function getNextSibling() {
- return $this->nextSibling;
+ /**
+ * @return bool
+ */
+ public function getLength() {
+ return false;
}
- function getChildrenOfType( $name ) {
- $children = array();
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- if ( isset( $child->name ) && $child->name === $name ) {
- $children[] = $name;
- }
- }
- return $children;
+ /**
+ * @param int $i
+ * @return bool
+ */
+ public function item( $i ) {
+ return false;
}
- function getLength() { return false; }
- function item( $i ) { return false; }
-
- function getName() {
+ /**
+ * @return string
+ */
+ public function getName() {
return $this->name;
}
/**
- * Split a <part> node into an associative array containing:
- * name PPNode name
- * index String index
- * value PPNode value
+ * Split a "<part>" node into an associative array containing:
+ * - name PPNode name
+ * - index String index
+ * - value PPNode value
+ *
+ * @throws MWException
+ * @return array
+ */
+ public function splitArg() {
+ return self::splitRawArg( $this->rawChildren );
+ }
+
+ /**
+ * Like splitArg() but for a raw child array. For internal use only.
+ * @param array $children
+ * @return array
*/
- function splitArg() {
- $bits = array();
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- if ( !isset( $child->name ) ) {
+ public static function splitRawArg( array $children ) {
+ $bits = [];
+ foreach ( $children as $i => $child ) {
+ if ( !is_array( $child ) ) {
continue;
}
- if ( $child->name === 'name' ) {
- $bits['name'] = $child;
- if ( $child->firstChild instanceof PPNode_Hash_Attr
- && $child->firstChild->name === 'index' )
- {
- $bits['index'] = $child->firstChild->value;
+ if ( $child[self::NAME] === 'name' ) {
+ $bits['name'] = new self( $children, $i );
+ if ( isset( $child[self::CHILDREN][0][self::NAME] )
+ && $child[self::CHILDREN][0][self::NAME] === '@index'
+ ) {
+ $bits['index'] = $child[self::CHILDREN][0][self::CHILDREN][0];
}
- } elseif ( $child->name === 'value' ) {
- $bits['value'] = $child;
+ } elseif ( $child[self::NAME] === 'value' ) {
+ $bits['value'] = new self( $children, $i );
}
}
}
/**
- * Split an <ext> node into an associative array containing name, attr, inner and close
+ * Split an "<ext>" node into an associative array containing name, attr, inner and close
* All values in the resulting array are PPNodes. Inner and close are optional.
+ *
+ * @throws MWException
+ * @return array
+ */
+ public function splitExt() {
+ return self::splitRawExt( $this->rawChildren );
+ }
+
+ /**
+ * Like splitExt() but for a raw child array. For internal use only.
+ * @param array $children
+ * @return array
*/
- function splitExt() {
- $bits = array();
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- if ( !isset( $child->name ) ) {
+ public static function splitRawExt( array $children ) {
+ $bits = [];
+ foreach ( $children as $i => $child ) {
+ if ( !is_array( $child ) ) {
continue;
}
- if ( $child->name == 'name' ) {
- $bits['name'] = $child;
- } elseif ( $child->name == 'attr' ) {
- $bits['attr'] = $child;
- } elseif ( $child->name == 'inner' ) {
- $bits['inner'] = $child;
- } elseif ( $child->name == 'close' ) {
- $bits['close'] = $child;
+ switch ( $child[self::NAME] ) {
+ case 'name':
+ $bits['name'] = new self( $children, $i );
+ break;
+ case 'attr':
+ $bits['attr'] = new self( $children, $i );
+ break;
+ case 'inner':
+ $bits['inner'] = new self( $children, $i );
+ break;
+ case 'close':
+ $bits['close'] = new self( $children, $i );
+ break;
}
}
if ( !isset( $bits['name'] ) ) {
}
/**
- * Split an <h> node
+ * Split an "<h>" node
+ *
+ * @throws MWException
+ * @return array
*/
- function splitHeading() {
+ public function splitHeading() {
if ( $this->name !== 'h' ) {
throw new MWException( 'Invalid h node passed to ' . __METHOD__ );
}
- $bits = array();
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- if ( !isset( $child->name ) ) {
+ return self::splitRawHeading( $this->rawChildren );
+ }
+
+ /**
+ * Like splitHeading() but for a raw child array. For internal use only.
+ * @param array $children
+ * @return array
+ */
+ public static function splitRawHeading( array $children ) {
+ $bits = [];
+ foreach ( $children as $i => $child ) {
+ if ( !is_array( $child ) ) {
continue;
}
- if ( $child->name == 'i' ) {
- $bits['i'] = $child->value;
- } elseif ( $child->name == 'level' ) {
- $bits['level'] = $child->value;
+ if ( $child[self::NAME] === '@i' ) {
+ $bits['i'] = $child[self::CHILDREN][0];
+ } elseif ( $child[self::NAME] === '@level' ) {
+ $bits['level'] = $child[self::CHILDREN][0];
}
}
if ( !isset( $bits['i'] ) ) {
}
/**
- * Split a <template> or <tplarg> node
+ * Split a "<template>" or "<tplarg>" node
+ *
+ * @throws MWException
+ * @return array
+ */
+ public function splitTemplate() {
+ return self::splitRawTemplate( $this->rawChildren );
+ }
+
+ /**
+ * Like splitTemplate() but for a raw child array. For internal use only.
+ * @param array $children
+ * @return array
*/
- function splitTemplate() {
- $parts = array();
- $bits = array( 'lineStart' => '' );
- for ( $child = $this->firstChild; $child; $child = $child->nextSibling ) {
- if ( !isset( $child->name ) ) {
+ public static function splitRawTemplate( array $children ) {
+ $parts = [];
+ $bits = [ 'lineStart' => '' ];
+ foreach ( $children as $i => $child ) {
+ if ( !is_array( $child ) ) {
continue;
}
- if ( $child->name == 'title' ) {
- $bits['title'] = $child;
- }
- if ( $child->name == 'part' ) {
- $parts[] = $child;
- }
- if ( $child->name == 'lineStart' ) {
+ switch ( $child[self::NAME] ) {
+ case 'title':
+ $bits['title'] = new self( $children, $i );
+ break;
+ case 'part':
+ $parts[] = new self( $children, $i );
+ break;
+ case '@lineStart':
$bits['lineStart'] = '1';
+ break;
}
}
if ( !isset( $bits['title'] ) ) {
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPNode_Hash_Text implements PPNode {
- var $value, $nextSibling;
+ // @codingStandardsIgnoreEnd
+
+ public $value;
+ private $store, $index;
- function __construct( $value ) {
- if ( is_object( $value ) ) {
+ /**
+ * Construct an object using the data from $store[$index]. The rest of the
+ * store array can be accessed via getNextSibling().
+ *
+ * @param array $store
+ * @param int $index
+ */
+ public function __construct( array $store, $index ) {
+ $this->value = $store[$index];
+ if ( !is_scalar( $this->value ) ) {
throw new MWException( __CLASS__ . ' given object instead of string' );
}
- $this->value = $value;
+ $this->store = $store;
+ $this->index = $index;
}
- function __toString() {
+ public function __toString() {
return htmlspecialchars( $this->value );
}
- function getNextSibling() {
- return $this->nextSibling;
+ public function getNextSibling() {
+ return PPNode_Hash_Tree::factory( $this->store, $this->index + 1 );
+ }
+
+ public function getChildren() {
+ return false;
+ }
+
+ public function getFirstChild() {
+ return false;
+ }
+
+ public function getChildrenOfType( $name ) {
+ return false;
+ }
+
+ public function getLength() {
+ return false;
+ }
+
+ public function item( $i ) {
+ return false;
+ }
+
+ public function getName() {
+ return '#text';
}
- function getChildren() { return false; }
- function getFirstChild() { return false; }
- function getChildrenOfType( $name ) { return false; }
- function getLength() { return false; }
- function item( $i ) { return false; }
- function getName() { return '#text'; }
- function splitArg() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitExt() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitHeading() { throw new MWException( __METHOD__ . ': not supported' ); }
+ public function splitArg() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
+
+ public function splitExt() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
+
+ public function splitHeading() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
}
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPNode_Hash_Array implements PPNode {
- var $value, $nextSibling;
+ // @codingStandardsIgnoreEnd
+
+ public $value;
- function __construct( $value ) {
+ public function __construct( $value ) {
$this->value = $value;
}
- function __toString() {
+ public function __toString() {
return var_export( $this, true );
}
- function getLength() {
+ public function getLength() {
return count( $this->value );
}
- function item( $i ) {
+ public function item( $i ) {
return $this->value[$i];
}
- function getName() { return '#nodelist'; }
+ public function getName() {
+ return '#nodelist';
+ }
+
+ public function getNextSibling() {
+ return false;
+ }
+
+ public function getChildren() {
+ return false;
+ }
+
+ public function getFirstChild() {
+ return false;
+ }
+
+ public function getChildrenOfType( $name ) {
+ return false;
+ }
+
+ public function splitArg() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
- function getNextSibling() {
- return $this->nextSibling;
+ public function splitExt() {
+ throw new MWException( __METHOD__ . ': not supported' );
}
- function getChildren() { return false; }
- function getFirstChild() { return false; }
- function getChildrenOfType( $name ) { return false; }
- function splitArg() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitExt() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitHeading() { throw new MWException( __METHOD__ . ': not supported' ); }
+ public function splitHeading() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
}
/**
* @ingroup Parser
*/
+// @codingStandardsIgnoreStart Squiz.Classes.ValidClassName.NotCamelCaps
class PPNode_Hash_Attr implements PPNode {
- var $name, $value, $nextSibling;
+ // @codingStandardsIgnoreEnd
- function __construct( $name, $value ) {
- $this->name = $name;
- $this->value = $value;
+ public $name, $value;
+ private $store, $index;
+
+ /**
+ * Construct an object using the data from $store[$index]. The rest of the
+ * store array can be accessed via getNextSibling().
+ *
+ * @param array $store
+ * @param int $index
+ */
+ public function __construct( array $store, $index ) {
+ $descriptor = $store[$index];
+ if ( $descriptor[PPNode_Hash_Tree::NAME][0] !== '@' ) {
+ throw new MWException( __METHOD__.': invalid name in attribute descriptor' );
+ }
+ $this->name = substr( $descriptor[PPNode_Hash_Tree::NAME], 1 );
+ $this->value = $descriptor[PPNode_Hash_Tree::CHILDREN][0];
+ $this->store = $store;
+ $this->index = $index;
}
- function __toString() {
+ public function __toString() {
return "<@{$this->name}>" . htmlspecialchars( $this->value ) . "</@{$this->name}>";
}
- function getName() {
+ public function getName() {
return $this->name;
}
- function getNextSibling() {
- return $this->nextSibling;
+ public function getNextSibling() {
+ return PPNode_Hash_Tree::factory( $this->store, $this->index + 1 );
+ }
+
+ public function getChildren() {
+ return false;
+ }
+
+ public function getFirstChild() {
+ return false;
+ }
+
+ public function getChildrenOfType( $name ) {
+ return false;
+ }
+
+ public function getLength() {
+ return false;
+ }
+
+ public function item( $i ) {
+ return false;
+ }
+
+ public function splitArg() {
+ throw new MWException( __METHOD__ . ': not supported' );
}
- function getChildren() { return false; }
- function getFirstChild() { return false; }
- function getChildrenOfType( $name ) { return false; }
- function getLength() { return false; }
- function item( $i ) { return false; }
- function splitArg() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitExt() { throw new MWException( __METHOD__ . ': not supported' ); }
- function splitHeading() { throw new MWException( __METHOD__ . ': not supported' ); }
+ public function splitExt() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
+
+ public function splitHeading() {
+ throw new MWException( __METHOD__ . ': not supported' );
+ }
}