| Index: trunk/phase3/includes/Parser.php |
| — | — | @@ -132,6 +132,7 @@ |
| 133 | 133 | $this->mTagHooks = array(); |
| 134 | 134 | $this->mFunctionHooks = array(); |
| 135 | 135 | $this->clearState(); |
| | 136 | + $this->setHook( 'pre', array( $this, 'renderPreTag' ) ); |
| 136 | 137 | } |
| 137 | 138 | |
| 138 | 139 | /** |
| — | — | @@ -422,7 +423,7 @@ |
| 423 | 424 | $commentState = array(); |
| 424 | 425 | |
| 425 | 426 | $elements = array_merge( |
| 426 | | - array( 'nowiki', 'pre', 'gallery' ), |
| | 427 | + array( 'nowiki', 'gallery' ), |
| 427 | 428 | array_keys( $this->mTagHooks ) ); |
| 428 | 429 | global $wgRawHtml; |
| 429 | 430 | if( $wgRawHtml ) { |
| — | — | @@ -463,11 +464,6 @@ |
| 464 | 465 | case 'math': |
| 465 | 466 | $output = MathRenderer::renderMath( $content ); |
| 466 | 467 | break; |
| 467 | | - case 'pre': |
| 468 | | - // Backwards-compatibility hack |
| 469 | | - $content = preg_replace( '!<nowiki>(.*?)</nowiki>!is', '\\1', $content ); |
| 470 | | - $output = '<pre>' . wfEscapeHTMLTagsOnly( $content ) . '</pre>'; |
| 471 | | - break; |
| 472 | 468 | case 'gallery': |
| 473 | 469 | $output = $this->renderImageGallery( $content ); |
| 474 | 470 | break; |
| — | — | @@ -4032,6 +4028,19 @@ |
| 4033 | 4029 | } |
| 4034 | 4030 | |
| 4035 | 4031 | /** |
| | 4032 | + * Tag hook handler for 'pre'. |
| | 4033 | + */ |
| | 4034 | + function renderPreTag( $text, $attribs, $parser ) { |
| | 4035 | + // Backwards-compatibility hack |
| | 4036 | + $content = preg_replace( '!<nowiki>(.*?)</nowiki>!is', '\\1', $text ); |
| | 4037 | + |
| | 4038 | + $attribs = Sanitizer::validateTagAttributes( $attribs, 'pre' ); |
| | 4039 | + return wfOpenElement( 'pre', $attribs ) . |
| | 4040 | + wfEscapeHTMLTagsOnly( $content ) . |
| | 4041 | + '</pre>'; |
| | 4042 | + } |
| | 4043 | + |
| | 4044 | + /** |
| 4036 | 4045 | * Renders an image gallery from a text with one line per image. |
| 4037 | 4046 | * text labels may be given by using |-style alternative text. E.g. |
| 4038 | 4047 | * Image:one.jpg|The number "1" |
| Index: trunk/phase3/includes/Sanitizer.php |
| — | — | @@ -538,53 +538,26 @@ |
| 539 | 539 | } |
| 540 | 540 | |
| 541 | 541 | /** |
| 542 | | - * Take a tag soup fragment listing an HTML element's attributes |
| 543 | | - * and normalize it to well-formed XML, discarding unwanted attributes. |
| | 542 | + * Take an array of attribute names and values and normalize or discard |
| | 543 | + * illegal values for the given element type. |
| 544 | 544 | * |
| 545 | | - * - Normalizes attribute names to lowercase |
| 546 | 545 | * - Discards attributes not on a whitelist for the given element |
| 547 | | - * - Turns broken or invalid entities into plaintext |
| 548 | | - * - Double-quotes all attribute values |
| 549 | | - * - Attributes without values are given the name as attribute |
| 550 | | - * - Double attributes are discarded |
| 551 | 546 | * - Unsafe style attributes are discarded |
| 552 | | - * - Prepends space if there are attributes. |
| 553 | 547 | * |
| 554 | | - * @param string $text |
| | 548 | + * @param array $attribs |
| 555 | 549 | * @param string $element |
| 556 | | - * @return string |
| | 550 | + * @return array |
| 557 | 551 | * |
| 558 | 552 | * @todo Check for legal values where the DTD limits things. |
| 559 | 553 | * @todo Check for unique id attribute :P |
| 560 | 554 | */ |
| 561 | | - function fixTagAttributes( $text, $element ) { |
| 562 | | - if( trim( $text ) == '' ) { |
| 563 | | - return ''; |
| 564 | | - } |
| 565 | | - |
| 566 | | - # Unquoted attribute |
| 567 | | - # Since we quote this later, this can be anything distinguishable |
| 568 | | - # from the end of the attribute |
| 569 | | - $pairs = array(); |
| 570 | | - if( !preg_match_all( |
| 571 | | - MW_ATTRIBS_REGEX, |
| 572 | | - $text, |
| 573 | | - $pairs, |
| 574 | | - PREG_SET_ORDER ) ) { |
| 575 | | - return ''; |
| 576 | | - } |
| 577 | | - |
| | 555 | + function validateTagAttributes( $attribs, $element ) { |
| 578 | 556 | $whitelist = array_flip( Sanitizer::attributeWhitelist( $element ) ); |
| 579 | | - $attribs = array(); |
| 580 | | - foreach( $pairs as $set ) { |
| 581 | | - $attribute = strtolower( $set[1] ); |
| | 557 | + $out = array(); |
| | 558 | + foreach( $attribs as $attribute => $value ) { |
| 582 | 559 | if( !isset( $whitelist[$attribute] ) ) { |
| 583 | 560 | continue; |
| 584 | 561 | } |
| 585 | | - |
| 586 | | - $raw = Sanitizer::getTagAttributeCallback( $set ); |
| 587 | | - $value = Sanitizer::normalizeAttributeValue( $raw ); |
| 588 | | - |
| 589 | 562 | # Strip javascript "expression" from stylesheets. |
| 590 | 563 | # http://msdn.microsoft.com/workshop/author/dhtml/overview/recalc.asp |
| 591 | 564 | if( $attribute == 'style' ) { |
| — | — | @@ -592,7 +565,7 @@ |
| 593 | 566 | |
| 594 | 567 | // Remove any comments; IE gets token splitting wrong |
| 595 | 568 | $stripped = preg_replace( '!/\\*.*?\\*/!S', ' ', $stripped ); |
| 596 | | - $value = htmlspecialchars( $stripped ); |
| | 569 | + $value = $stripped; |
| 597 | 570 | |
| 598 | 571 | // ... and continue checks |
| 599 | 572 | $stripped = preg_replace( '!\\\\([0-9A-Fa-f]{1,6})[ \\n\\r\\t\\f]?!e', |
| — | — | @@ -608,9 +581,48 @@ |
| 609 | 582 | if ( $attribute === 'id' ) |
| 610 | 583 | $value = Sanitizer::escapeId( $value ); |
| 611 | 584 | |
| | 585 | + // If this attribute was previously set, override it. |
| | 586 | + // Output should only have one attribute of each name. |
| | 587 | + $out[$attribute] = $value; |
| | 588 | + } |
| | 589 | + return $out; |
| | 590 | + } |
| | 591 | + |
| | 592 | + /** |
| | 593 | + * Take a tag soup fragment listing an HTML element's attributes |
| | 594 | + * and normalize it to well-formed XML, discarding unwanted attributes. |
| | 595 | + * Output is safe for further wikitext processing, with escaping of |
| | 596 | + * values that could trigger problems. |
| | 597 | + * |
| | 598 | + * - Normalizes attribute names to lowercase |
| | 599 | + * - Discards attributes not on a whitelist for the given element |
| | 600 | + * - Turns broken or invalid entities into plaintext |
| | 601 | + * - Double-quotes all attribute values |
| | 602 | + * - Attributes without values are given the name as attribute |
| | 603 | + * - Double attributes are discarded |
| | 604 | + * - Unsafe style attributes are discarded |
| | 605 | + * - Prepends space if there are attributes. |
| | 606 | + * |
| | 607 | + * @param string $text |
| | 608 | + * @param string $element |
| | 609 | + * @return string |
| | 610 | + */ |
| | 611 | + function fixTagAttributes( $text, $element ) { |
| | 612 | + if( trim( $text ) == '' ) { |
| | 613 | + return ''; |
| | 614 | + } |
| | 615 | + |
| | 616 | + $stripped = Sanitizer::validateTagAttributes( |
| | 617 | + Sanitizer::decodeTagAttributes( $text ), $element ); |
| | 618 | + |
| | 619 | + $attribs = array(); |
| | 620 | + foreach( $stripped as $attribute => $value ) { |
| | 621 | + $encAttribute = htmlspecialchars( $attribute ); |
| | 622 | + |
| | 623 | + $encValue = htmlspecialchars( $value ); |
| 612 | 624 | # Templates and links may be expanded in later parsing, |
| 613 | 625 | # creating invalid or dangerous output. Suppress this. |
| 614 | | - $value = strtr( $value, array( |
| | 626 | + $encValue = strtr( $encValue, array( |
| 615 | 627 | '<' => '<', // This should never happen, |
| 616 | 628 | '>' => '>', // we've received invalid input |
| 617 | 629 | '"' => '"', // which should have been escaped. |
| — | — | @@ -625,16 +637,13 @@ |
| 626 | 638 | ) ); |
| 627 | 639 | |
| 628 | 640 | # Stupid hack |
| 629 | | - $value = preg_replace_callback( |
| | 641 | + $encValue = preg_replace_callback( |
| 630 | 642 | '/(' . wfUrlProtocols() . ')/', |
| 631 | 643 | array( 'Sanitizer', 'armorLinksCallback' ), |
| 632 | | - $value ); |
| 633 | | - |
| 634 | | - // If this attribute was previously set, override it. |
| 635 | | - // Output should only have one attribute of each name. |
| 636 | | - $attribs[$attribute] = "$attribute=\"$value\""; |
| | 644 | + $encValue ); |
| | 645 | + |
| | 646 | + $attribs[] = "$encAttribute=\"$encValue\""; |
| 637 | 647 | } |
| 638 | | - |
| 639 | 648 | return count( $attribs ) ? ' ' . implode( ' ', $attribs ) : ''; |
| 640 | 649 | } |
| 641 | 650 | |
| Index: trunk/phase3/RELEASE-NOTES |
| — | — | @@ -443,6 +443,9 @@ |
| 444 | 444 | * (bug 6175) Improvement to German translation (de) |
| 445 | 445 | * Redirect Special:Logs to Special:Log |
| 446 | 446 | * (bug 6206) Linktrail for Swedish localization (se) |
| | 447 | +* (bug 3202) Attributes now allowed on <pre> tags |
| | 448 | +* Sanitizer::validateTagAttributes now available to discard illegal/unsafe |
| | 449 | + attribute values from an array. |
| 447 | 450 | |
| 448 | 451 | |
| 449 | 452 | == Compatibility == |