HTML API: Ensure bookmark exaustion does not error

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

@@ -989,8 +989,6 @@ public function expects_closer( ?WP_HTML_Token $node = null ): ?bool {
 	 *
 	 * @since 6.4.0
 	 *
-	 * @throws Exception When unable to allocate a bookmark for the next token in the input HTML document.
-	 *
 	 * @see self::PROCESS_NEXT_NODE
 	 * @see self::REPROCESS_CURRENT_NODE
 	 *
@@ -1040,8 +1038,13 @@ public function step( $node_to_process = self::PROCESS_NEXT_NODE ): bool {
 		$token_name            = $this->get_token_name();
 
 		if ( self::REPROCESS_CURRENT_NODE !== $node_to_process ) {
+			$bookmark_name = $this->bookmark_token();
+			if ( false === $bookmark_name ) {
+				return false;
+			}
+
 			$this->state->current_token = new WP_HTML_Token(
-				$this->bookmark_token(),
+				$bookmark_name,
 				$token_name,
 				$this->has_self_closing_flag(),
 				$this->release_internal_bookmark_on_destruct
@@ -1600,7 +1603,9 @@ private function step_before_html(): bool {
 		 * > Switch the insertion mode to "before head", then reprocess the token.
 		 */
 		before_html_anything_else:
-		$this->insert_virtual_node( 'HTML' );
+		if ( ! $this->insert_virtual_node( 'HTML' ) ) {
+			return false;
+		}
 		$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_BEFORE_HEAD;
 		return $this->step( self::REPROCESS_CURRENT_NODE );
 	}
@@ -1697,7 +1702,11 @@ private function step_before_head(): bool {
 		 * > Insert an HTML element for a "head" start tag token with no attributes.
 		 */
 		before_head_anything_else:
-		$this->state->head_element   = $this->insert_virtual_node( 'HEAD' );
+		$head_element = $this->insert_virtual_node( 'HEAD' );
+		if ( ! $head_element ) {
+			return false;
+		}
+		$this->state->head_element   = $head_element;
 		$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_IN_HEAD;
 		return $this->step( self::REPROCESS_CURRENT_NODE );
 	}
@@ -2166,7 +2175,9 @@ private function step_after_head(): bool {
 		 * > Insert an HTML element for a "body" start tag token with no attributes.
 		 */
 		after_head_anything_else:
-		$this->insert_virtual_node( 'BODY' );
+		if ( ! $this->insert_virtual_node( 'BODY' ) ) {
+			return false;
+		}
 		$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_IN_BODY;
 		return $this->step( self::REPROCESS_CURRENT_NODE );
 	}
@@ -3318,7 +3329,9 @@ private function step_in_table(): bool {
 				 * > Insert an HTML element for a "colgroup" start tag token with no attributes,
 				 * > then switch the insertion mode to "in column group".
 				 */
-				$this->insert_virtual_node( 'COLGROUP' );
+				if ( ! $this->insert_virtual_node( 'COLGROUP' ) ) {
+					return false;
+				}
 				$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_IN_COLUMN_GROUP;
 				return $this->step( self::REPROCESS_CURRENT_NODE );
 
@@ -3344,7 +3357,9 @@ private function step_in_table(): bool {
 				 * > Insert an HTML element for a "tbody" start tag token with no attributes,
 				 * > then switch the insertion mode to "in table body".
 				 */
-				$this->insert_virtual_node( 'TBODY' );
+				if ( ! $this->insert_virtual_node( 'TBODY' ) ) {
+					return false;
+				}
 				$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_IN_TABLE_BODY;
 				return $this->step( self::REPROCESS_CURRENT_NODE );
 
@@ -3699,7 +3714,9 @@ private function step_in_table_body(): bool {
 			case '+TD':
 				// @todo Indicate a parse error once it's possible.
 				$this->state->stack_of_open_elements->clear_to_table_body_context();
-				$this->insert_virtual_node( 'TR' );
+				if ( ! $this->insert_virtual_node( 'TR' ) ) {
+					return false;
+				}
 				$this->state->insertion_mode = WP_HTML_Processor_State::INSERTION_MODE_IN_ROW;
 				return $this->step( self::REPROCESS_CURRENT_NODE );
 
@@ -5100,14 +5117,12 @@ private function step_in_foreign_content(): bool {
 	 * @since 6.4.0
 	 * @since 6.5.0 Renamed from bookmark_tag() to bookmark_token().
 	 *
-	 * @throws Exception When unable to allocate requested bookmark.
-	 *
 	 * @return string|false Name of created bookmark, or false if unable to create.
 	 */
 	private function bookmark_token() {
 		if ( ! parent::set_bookmark( ++$this->bookmark_counter ) ) {
 			$this->last_error = self::ERROR_EXCEEDED_MAX_BOOKMARKS;
-			throw new Exception( 'could not allocate bookmark' );
+			return false;
 		}
 
 		return "{$this->bookmark_counter}";
@@ -6290,11 +6305,14 @@ private function insert_foreign_element( WP_HTML_Token $token, bool $only_add_to
 	 * @param string      $token_name    Name of token to create and insert into the stack of open elements.
 	 * @param string|null $bookmark_name Optional. Name to give bookmark for created virtual node.
 	 *                                   Defaults to auto-creating a bookmark name.
-	 * @return WP_HTML_Token Newly-created virtual token.
+	 * @return WP_HTML_Token|false Newly-created virtual token or false on failure.
 	 */
-	private function insert_virtual_node( $token_name, $bookmark_name = null ): WP_HTML_Token {
+	private function insert_virtual_node( $token_name, $bookmark_name = null ) {
 		$here = $this->bookmarks[ $this->state->current_token->bookmark_name ];
 		$name = $bookmark_name ?? $this->bookmark_token();
+		if ( false === $name ) {
+			return false;
+		}
 
 		$this->bookmarks[ $name ] = new WP_HTML_Span( $here->start, 0 );
 

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

@@ -1068,7 +1068,7 @@ public function test_ensure_next_token_method_extensibility( $html, $expected_to
 	/**
 	 * Ensure that lowercased tag_name query matches tags case-insensitively.
 	 *
-	 * @group 62427
+	 * @ticket 62427
 	 */
 	public function test_next_tag_lowercase_tag_name() {
 		// The upper case <DIV> is irrelevant but illustrates the case-insentivity.
@@ -1079,4 +1079,67 @@ public function test_next_tag_lowercase_tag_name() {
 		$processor = WP_HTML_Processor::create_fragment( '<svg><RECT>' );
 		$this->assertTrue( $processor->next_tag( array( 'tag_name' => 'rect' ) ) );
 	}
+
+	/**
+	 * Ensure that the processor does not throw errors in cases of extreme HTML nesting.
+	 *
+	 * @ticket 64394
+	 *
+	 * @expectedIncorrectUsage WP_HTML_Tag_Processor::set_bookmark
+	 */
+	public function test_deep_nesting_fails_process_without_error() {
+		$html      = str_repeat( '<i>', WP_HTML_Processor::MAX_BOOKMARKS * 2 );
+		$processor = WP_HTML_Processor::create_fragment( $html );
+
+		// The fragment parser starts with a few context tokens already bookmarked.
+		$reached_tokens = ( fn() => count( $this->bookmarks ) )->call( $processor );
+		while ( $processor->next_token() ) {
+			++$reached_tokens;
+		}
+		$this->assertSame( WP_HTML_Processor::MAX_BOOKMARKS, $reached_tokens );
+		$this->assertSame(
+			WP_HTML_Processor::ERROR_EXCEEDED_MAX_BOOKMARKS,
+			$processor->get_last_error(),
+			'Failed to report exceeded-max-bookmarks error.'
+		);
+	}
+
+	/**
+	 * @ticket 64394
+	 *
+	 * @expectedIncorrectUsage WP_HTML_Tag_Processor::set_bookmark
+	 */
+	public function test_deep_nesting_fails_processing_virtual_tokens_without_error() {
+		$html      = str_repeat( '<table><td>', WP_HTML_Processor::MAX_BOOKMARKS * 2 );
+		$processor = WP_HTML_Processor::create_fragment( $html );
+
+		// The fragment parser starts with a few context tokens already bookmarked.
+		$reached_tokens = ( fn() => count( $this->bookmarks ) )->call( $processor );
+		while ( $processor->next_token() ) {
+			++$reached_tokens;
+		}
+
+		/*
+		 * This test has some variability depending on how the virtual tokens align.
+		 * It will produce 1 real, 2 virtual, 1 real.
+		 *
+		 * "<table><td><table><td>…" produces:
+		 * └─TABLE
+		 *   └─TBODY (virtual)
+		 *     └─TR (virtual)
+		 *       └─TD
+		 *         └─TABLE
+		 *           └─TBODY (virtual)
+		 *             └─TR (virtual)
+		 *               └─TD
+		 *                 └─…
+		 */
+		$this->assertGreaterThanOrEqual( WP_HTML_Processor::MAX_BOOKMARKS - 1, $reached_tokens );
+		$this->assertLessThanOrEqual( WP_HTML_Processor::MAX_BOOKMARKS + 1, $reached_tokens );
+		$this->assertSame(
+			WP_HTML_Processor::ERROR_EXCEEDED_MAX_BOOKMARKS,
+			$processor->get_last_error(),
+			'Failed to report exceeded-max-bookmarks error.'
+		);
+	}
 }