HTML API: Implement active format reconstruction

src/wp-includes/html-api/class-wp-html-active-formatting-elements.php

@@ -43,6 +43,18 @@ class WP_HTML_Active_Formatting_Elements {
 	 */
 	private $stack = array();
 
+	/**
+	 * Returns the node at the given 1-offset index in the list of active formatting elements.
+	 *
+	 * @since 7.0.0
+	 *
+	 * @param int $index Number of nodes from the top node to return.
+	 * @return WP_HTML_Token|null Node at the given index in the stack, if one exists, otherwise null.
+	 */
+	public function at( $nth ) {
+		return $this->stack[ $nth - 1 ];
+	}
+
 	/**
 	 * Reports if a specific node is in the stack of active formatting elements.
 	 *
@@ -110,8 +122,9 @@ public function insert_marker(): void {
 	 * @see https://html.spec.whatwg.org/#push-onto-the-list-of-active-formatting-elements
 	 *
 	 * @param WP_HTML_Token $token Push this node onto the stack.
+	 * @return bool Whether a node was pushed onto the stack of active formatting elements.
 	 */
-	public function push( WP_HTML_Token $token ) {
+	public function push( WP_HTML_Token $token ): bool {
 		/*
 		 * > If there are already three elements in the list of active formatting elements after the last marker,
 		 * > if any, or anywhere in the list if there are no markers, that have the same tag name, namespace, and
@@ -120,11 +133,32 @@ public function push( WP_HTML_Token $token ) {
 		 * > created by the parser; two elements have the same attributes if all their parsed attributes can be
 		 * > paired such that the two attributes in each pair have identical names, namespaces, and values
 		 * > (the order of the attributes does not matter).
-		 *
-		 * @todo Implement the "Noah's Ark clause" to only add up to three of any given kind of formatting elements to the stack.
 		 */
+
+		if ( 'marker' !== $token->node_name ) {
+			$existing_count = 0;
+			foreach ( $this->walk_up() as $item ) {
+				if ( 'marker' === $item->node_name ) {
+					break;
+				}
+
+				if (
+					$item->node_name === $token->node_name &&
+					$item->namespace === $token->namespace
+					// @todo Compare attributes. For now, bail if there are three matching tag names + namespaces.
+				) {
+					++$existing_count;
+					if ( $existing_count >= 3 ) {
+						// @todo Implement removing the earliest element and moving forward.
+						return false;
+					}
+				}
+			}
+		}
+
 		// > Add element to the list of active formatting elements.
 		$this->stack[] = $token;
+		return true;
 	}
 
 	/**

src/wp-includes/html-api/class-wp-html-processor.php

@@ -256,6 +256,18 @@ class WP_HTML_Processor extends WP_HTML_Tag_Processor {
 	 */
 	private $context_node = null;
 
+	/**
+	 * If a formatting element has been reconstructed, this will hold
+	 * the parsed attributes from the original format, once requested.
+	 *
+	 * These attributes are not modifiable.
+	 *
+	 * @since 7.0.0
+	 *
+	 * @var array
+	 */
+	protected $actively_reconstructed_formatting_attributes = array();
+
 	/*
 	 * Public Interface Functions
 	 */
@@ -2766,7 +2778,10 @@ private function step_in_body(): bool {
 
 				$this->reconstruct_active_formatting_elements();
 				$this->insert_html_element( $this->state->current_token );
-				$this->state->active_formatting_elements->push( $this->state->current_token );
+				if ( false === $this->state->active_formatting_elements->push( $this->state->current_token ) ) {
+					$this->bail( 'Cannot track formatting elements when encountering a fourth identical token.' );
+				}
+				$this->actively_reconstructed_formatting_attributes[ $this->state->current_token->bookmark_name ] = $this->attributes;
 				return true;
 
 			/*
@@ -2787,7 +2802,10 @@ private function step_in_body(): bool {
 			case '+U':
 				$this->reconstruct_active_formatting_elements();
 				$this->insert_html_element( $this->state->current_token );
-				$this->state->active_formatting_elements->push( $this->state->current_token );
+				if ( false === $this->state->active_formatting_elements->push( $this->state->current_token ) ) {
+					$this->bail( 'Cannot track formatting elements when encountering a fourth identical token.' );
+				}
+				$this->actively_reconstructed_formatting_attributes[ $this->state->current_token->bookmark_name ] = $this->attributes;
 				return true;
 
 			/*
@@ -2803,7 +2821,10 @@ private function step_in_body(): bool {
 				}
 
 				$this->insert_html_element( $this->state->current_token );
-				$this->state->active_formatting_elements->push( $this->state->current_token );
+				if ( false === $this->state->active_formatting_elements->push( $this->state->current_token ) ) {
+					$this->bail( 'Cannot track formatting elements when encountering a fourth identical token.' );
+				}
+				$this->actively_reconstructed_formatting_attributes[ $this->state->current_token->bookmark_name ] = $this->attributes;
 				return true;
 
 			/*
@@ -5284,7 +5305,46 @@ public function get_token_type(): ?string {
 	 * @return string|true|null Value of attribute or `null` if not available. Boolean attributes return `true`.
 	 */
 	public function get_attribute( $name ) {
-		return $this->is_virtual() ? null : parent::get_attribute( $name );
+		if ( $this->is_virtual() ) {
+			$virtual_attributes = $this->actively_reconstructed_formatting_attributes[ $this->current_element->token->bookmark_name ?? '' ] ?? null;
+			if ( null === $virtual_attributes ) {
+				return null;
+			}
+
+			$current_attributes    = $this->attributes;
+			$current_updates       = $this->lexical_updates;
+			$this->lexical_updates = array();
+			$this->attributes      = $virtual_attributes;
+			$parser_state          = $this->parser_state;
+			$this->parser_state    = WP_HTML_Tag_Processor::STATE_MATCHED_TAG;
+			$attribute_names       = parent::get_attribute( $name );
+			$this->attributes      = $current_attributes;
+			$this->parser_state    = $parser_state;
+			$this->lexical_updates = $current_updates;
+
+			return $attribute_names;
+		}
+
+		return parent::get_attribute( $name );
+	}
+
+	/**
+	 * Returns the adjusted attribute name for a given attribute, taking into
+	 * account the current parsing context, whether HTML, SVG, or MathML.
+	 *
+	 * @since 7.0.0 Subclassed for the HTML Processor.
+	 *
+	 * @param string       $attribute_name Which attribute name to adjust.
+	 *
+	 * @return string|null The qualified attribute name or null if not on matched tag.
+	 */
+	public function get_qualified_attribute_name( $attribute_name ): ?string {
+		if ( $this->is_virtual() ) {
+			$namespace = $this->current_element->token->namespace;
+			return self::lookup_qualified_attribute_name( $namespace, $attribute_name );
+		}
+
+		return parent::get_qualified_attribute_name( $attribute_name );
 	}
 
 	/**
@@ -5362,7 +5422,24 @@ public function remove_attribute( $name ): bool {
 	 * @return array|null List of attribute names, or `null` when no tag opener is matched.
 	 */
 	public function get_attribute_names_with_prefix( $prefix ): ?array {
-		return $this->is_virtual() ? null : parent::get_attribute_names_with_prefix( $prefix );
+		if ( $this->is_virtual() ) {
+			$virtual_attributes = $this->actively_reconstructed_formatting_attributes[ $this->current_element->token->bookmark_name ?? '' ] ?? null;
+			if ( null === $virtual_attributes ) {
+				return null;
+			}
+
+			$current_attributes = $this->attributes;
+			$this->attributes   = $virtual_attributes;
+			$parser_state       = $this->parser_state;
+			$this->parser_state = WP_HTML_Tag_Processor::STATE_MATCHED_TAG;
+			$attribute_names    = parent::get_attribute_names_with_prefix( $prefix );
+			$this->attributes   = $current_attributes;
+			$this->parser_state = $parser_state;
+
+			return $attribute_names;
+		}
+
+		return parent::get_attribute_names_with_prefix( $prefix );
 	}
 
 	/**
@@ -5865,6 +5942,7 @@ private function get_adjusted_current_node(): ?WP_HTML_Token {
 	 * > been explicitly closed.
 	 *
 	 * @since 6.4.0
+	 * @since 7.0.0 Added additional support.
 	 *
 	 * @throws WP_HTML_Unsupported_Exception When encountering unsupported HTML input.
 	 *
@@ -5873,34 +5951,89 @@ private function get_adjusted_current_node(): ?WP_HTML_Token {
 	 * @return bool Whether any formatting elements needed to be reconstructed.
 	 */
 	private function reconstruct_active_formatting_elements(): bool {
+		$count = $this->state->active_formatting_elements->count();
 		/*
-		 * > If there are no entries in the list of active formatting elements, then there is nothing
-		 * > to reconstruct; stop this algorithm.
+		 * > 1. If there are no entries in the list of active formatting elements,
+		 * >    then there is nothing to reconstruct; stop this algorithm.
 		 */
-		if ( 0 === $this->state->active_formatting_elements->count() ) {
+		if ( 0 === $count ) {
 			return false;
 		}
 
-		$last_entry = $this->state->active_formatting_elements->current_node();
+		$currently_at = $count;
+		$last_entry   = $this->state->active_formatting_elements->at( $currently_at );
+		/*
+		 * > 2. If the last (most recently added) entry in the list of active formatting
+		 * >    elements is a marker, or if it is an element that is in the stack of open
+		 * >    elements, then there is nothing to reconstruct; stop this algorithm.
+		 */
 		if (
-
-			/*
-			 * > If the last (most recently added) entry in the list of active formatting elements is a marker;
-			 * > stop this algorithm.
-			 */
 			'marker' === $last_entry->node_name ||
-
-			/*
-			 * > If the last (most recently added) entry in the list of active formatting elements is an
-			 * > element that is in the stack of open elements, then there is nothing to reconstruct;
-			 * > stop this algorithm.
-			 */
 			$this->state->stack_of_open_elements->contains_node( $last_entry )
 		) {
 			return false;
 		}
 
-		$this->bail( 'Cannot reconstruct active formatting elements when advancing and rewinding is required.' );
+		/*
+		 * > 3. Let entry be the last (most recently added) element
+		 * >    in the list of active formatting elements.
+		 */
+		$entry = $last_entry;
+
+		/*
+		 * > 4. Rewind: If there are no entries before entry in the list of active
+		 * >    formatting elements, then jump to the step labeled create.
+		 */
+		rewind:
+		if ( 1 === $currently_at ) {
+			goto create;
+		}
+
+		/*
+		 * > 5. Let entry be the entry one earlier than entry
+		 * >    in the list of active formatting elements.
+		 */
+		$entry = $this->state->active_formatting_elements->at( --$currently_at );
+
+		/*
+		 * > 6. If entry is neither a marker nor an element that is also in
+		 * >    the stack of open elements, go to the step labeled rewind.
+		 */
+		if (
+			'marker' !== $entry->node_name &&
+			! $this->state->stack_of_open_elements->contains_node( $entry )
+		) {
+			goto rewind;
+		}
+
+		/*
+		 * > 7. Advance: Let entry be the element one later than entry
+		 * >    in the list of active formatting elements.
+		 */
+		advance:
+		$entry = $this->state->active_formatting_elements->at( ++$currently_at );
+
+		/*
+		 * > 8. Create: Insert an HTML element for the token for which the
+		 * >    element entry was created, to obtain new element.
+		 */
+		create:
+		$this->insert_html_element( $entry );
+
+		/*
+		 * > 9. Replace the entry for _entry_ in the list with an entry for new element.
+		 * >    This doesn't need to happen here since no DOM is being created.
+		 */
+
+		/*
+		 * > 10. If the entry for new element in the list of active formatting elements
+		 * >     is not the last entry in the list, return to the step labeled advance.
+		 */
+		if ( $count !== $currently_at ) {
+			goto advance;
+		}
+
+		return true;
 	}
 
 	/**

src/wp-includes/html-api/class-wp-html-tag-processor.php

@@ -708,7 +708,7 @@ class WP_HTML_Tag_Processor {
 	 * @since 6.2.0
 	 * @var WP_HTML_Attribute_Token[]
 	 */
-	private $attributes = array();
+	protected $attributes = array();
 
 	/**
 	 * Tracks spans of duplicate attributes on a given tag, used for removing
@@ -3051,23 +3051,37 @@ public function get_qualified_tag_name(): ?string {
 	 *
 	 * @since 6.7.0
 	 *
-	 * @param string $attribute_name Which attribute to adjust.
+	 * @param string       $attribute_name Which attribute name to adjust.
 	 *
-	 * @return string|null
+	 * @return string|null The qualified attribute name or null if not on matched tag.
 	 */
 	public function get_qualified_attribute_name( $attribute_name ): ?string {
 		if ( self::STATE_MATCHED_TAG !== $this->parser_state ) {
 			return null;
 		}
+		$namespace = $this->get_namespace();
+		return self::lookup_qualified_attribute_name( $namespace, $attribute_name );
+	}
 
-		$namespace  = $this->get_namespace();
+	/**
+	 * Returns the adjusted attribute name for a given attribute, taking into
+	 * account the provided namespace.
+	 *
+	 * @since 7.0.0
+	 *
+	 * @param string  $ns             The namespace to use: 'html', 'svg', or 'math'.
+	 * @param string  $attribute_name Which attribute to adjust.
+	 *
+	 * @return string The qualified attribute name.
+	 */
+	final protected static function lookup_qualified_attribute_name( string $ns, string $attribute_name ): string {
 		$lower_name = strtolower( $attribute_name );
 
-		if ( 'math' === $namespace && 'definitionurl' === $lower_name ) {
+		if ( 'math' === $ns && 'definitionurl' === $lower_name ) {
 			return 'definitionURL';
 		}
 
-		if ( 'svg' === $this->get_namespace() ) {
+		if ( 'svg' === $ns ) {
 			switch ( $lower_name ) {
 				case 'attributename':
 					return 'attributeName';
@@ -3245,7 +3259,7 @@ public function get_qualified_attribute_name( $attribute_name ): ?string {
 			}
 		}
 
-		if ( 'html' !== $namespace ) {
+		if ( 'html' !== $ns ) {
 			switch ( $lower_name ) {
 				case 'xlink:actuate':
 					return 'xlink actuate';

tests/phpunit/tests/html-api/wpHtmlProcessor.php

@@ -185,18 +185,23 @@ public function test_clear_to_navigate_after_seeking() {
 	}
 
 	/**
-	 * Ensures that support is added for reconstructing active formatting elements
-	 * before the HTML Processor handles situations with unclosed formats requiring it.
+	 * Ensures that support is added for reconstructing active formatting elements.
 	 *
 	 * @ticket 58517
 	 *
 	 * @covers WP_HTML_Processor::reconstruct_active_formatting_elements
 	 */
-	public function test_fails_to_reconstruct_formatting_elements() {
-		$processor = WP_HTML_Processor::create_fragment( '<p><em>One<p><em>Two<p><em>Three<p><em>Four' );
+	public function test_reconstructs_formatting_elements() {
+		$processor = WP_HTML_Processor::create_fragment( '<p><em>One<p><em><span>Two<p><em>Three<p><em>Four' );
 
 		$this->assertTrue( $processor->next_tag( 'EM' ), 'Could not find first EM.' );
-		$this->assertFalse( $processor->next_tag( 'EM' ), 'Should have aborted before finding second EM as it required reconstructing the first EM.' );
+		$this->assertSame( array( 'HTML', 'BODY', 'P', 'EM' ), $processor->get_breadcrumbs(), 'Found incorrect breadcrumbs for first EM.' );
+		$this->assertTrue( $processor->next_tag( 'SPAN' ), 'Could not find test span.' );
+		$this->assertSame(
+			array( 'HTML', 'BODY', 'P', 'EM', 'EM', 'SPAN' ),
+			$processor->get_breadcrumbs(),
+			'Found incorrect breadcrumbs for test SPAN; should have created two EMs.'
+		);
 	}
 
 	/**