HTML API: Handle adoption agency fallback end tags

Author: sirreal

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

@@ -2913,8 +2913,7 @@ private function step_in_body(): bool {
 			case '-STRONG':
 			case '-TT':
 			case '-U':
-				$this->run_adoption_agency_algorithm();
-				return true;
+				return $this->run_adoption_agency_algorithm();
 
 			/*
 			 * > A start tag whose tag name is one of: "applet", "marquee", "object"
@@ -3246,41 +3245,57 @@ private function step_in_body(): bool {
 			$this->insert_html_element( $this->state->current_token );
 			return true;
 		} else {
-			/*
-			 * > Any other end tag
-			 */
+			return $this->step_in_body_any_other_end_tag();
+		}
 
-			/*
-			 * Find the corresponding tag opener in the stack of open elements, if
-			 * it exists before reaching a special element, which provides a kind
-			 * of boundary in the stack. For example, a `</custom-tag>` should not
-			 * close anything beyond its containing `P` or `DIV` element.
-			 */
-			foreach ( $this->state->stack_of_open_elements->walk_up() as $node ) {
-				if ( 'html' === $node->namespace && $token_name === $node->node_name ) {
-					break;
-				}
+		$this->bail( 'Should not have been able to reach end of IN BODY processing. Check HTML API code.' );
+		// This unnecessary return prevents tools from inaccurately reporting type errors.
+		return false;
+	}
 
-				if ( self::is_special( $node ) ) {
-					// This is a parse error, ignore the token.
-					return $this->step();
-				}
+	/**
+	 * Parses an "any other end tag" token in the "in body" insertion mode.
+	 *
+	 * @since 6.9.0
+	 * @ignore
+	 *
+	 * @throws WP_HTML_Unsupported_Exception When encountering unsupported HTML input.
+	 *
+	 * @return bool Whether an element was found.
+	 */
+	private function step_in_body_any_other_end_tag(): bool {
+		$token_name = $this->get_token_name();
+
+		/*
+		 * Find the corresponding tag opener in the stack of open elements, if
+		 * it exists before reaching a special element, which provides a kind
+		 * of boundary in the stack. For example, a `</custom-tag>` should not
+		 * close anything beyond its containing `P` or `DIV` element.
+		 */
+		foreach ( $this->state->stack_of_open_elements->walk_up() as $node ) {
+			if ( 'html' === $node->namespace && $token_name === $node->node_name ) {
+				break;
 			}
 
-			$this->generate_implied_end_tags( $token_name );
-			if ( $node !== $this->state->stack_of_open_elements->current_node() ) {
-				// @todo Record parse error: this error doesn't impact parsing.
+			if ( self::is_special( $node ) ) {
+				// This is a parse error, ignore the token.
+				return $this->step();
 			}
+		}
 
-			foreach ( $this->state->stack_of_open_elements->walk_up() as $item ) {
-				$this->state->stack_of_open_elements->pop();
-				if ( $node === $item ) {
-					return true;
-				}
+		$this->generate_implied_end_tags( $token_name );
+		if ( $node !== $this->state->stack_of_open_elements->current_node() ) {
+			// @todo Record parse error: this error doesn't impact parsing.
+		}
+
+		foreach ( $this->state->stack_of_open_elements->walk_up() as $item ) {
+			$this->state->stack_of_open_elements->pop();
+			if ( $node === $item ) {
+				return true;
 			}
 		}
 
-		$this->bail( 'Should not have been able to reach end of IN BODY processing. Check HTML API code.' );
+		$this->bail( 'Should not have been able to reach end of IN BODY "any other end tag" processing. Check HTML API code.' );
 		// This unnecessary return prevents tools from inaccurately reporting type errors.
 		return false;
 	}
@@ -6216,8 +6231,10 @@ private function reset_insertion_mode_appropriately(): void {
 	 * @throws WP_HTML_Unsupported_Exception When encountering unsupported HTML input.
 	 *
 	 * @see https://html.spec.whatwg.org/#adoption-agency-algorithm
+	 *
+	 * @return bool Whether an element was found.
 	 */
-	private function run_adoption_agency_algorithm(): void {
+	private function run_adoption_agency_algorithm(): bool {
 		$budget       = 1000;
 		$subject      = $this->get_tag();
 		$current_node = $this->state->stack_of_open_elements->current_node();
@@ -6229,13 +6246,13 @@ private function run_adoption_agency_algorithm(): void {
 			! $this->state->active_formatting_elements->contains_node( $current_node )
 		) {
 			$this->state->stack_of_open_elements->pop();
-			return;
+			return true;
 		}
 
 		$outer_loop_counter = 0;
 		while ( $budget-- > 0 ) {
 			if ( $outer_loop_counter++ >= 8 ) {
-				return;
+				return true;
 			}
 
 			/*
@@ -6258,18 +6275,18 @@ private function run_adoption_agency_algorithm(): void {
 
 			// > If there is no such element, then return and instead act as described in the "any other end tag" entry above.
 			if ( null === $formatting_element ) {
-				$this->bail( 'Cannot run adoption agency when "any other end tag" is required.' );
+				return $this->step_in_body_any_other_end_tag();
 			}
 
 			// > If formatting element is not in the stack of open elements, then this is a parse error; remove the element from the list, and return.
 			if ( ! $this->state->stack_of_open_elements->contains_node( $formatting_element ) ) {
 				$this->state->active_formatting_elements->remove_node( $formatting_element );
-				return;
+				return true;
 			}
 
 			// > If formatting element is in the stack of open elements, but the element is not in scope, then this is a parse error; return.
 			if ( ! $this->state->stack_of_open_elements->has_element_in_scope( $formatting_element->node_name ) ) {
-				return;
+				return true;
 			}
 
 			/*
@@ -6305,7 +6322,7 @@ private function run_adoption_agency_algorithm(): void {
 
 					if ( $formatting_element->bookmark_name === $item->bookmark_name ) {
 						$this->state->active_formatting_elements->remove_node( $formatting_element );
-						return;
+						return true;
 					}
 				}
 			}
@@ -6314,6 +6331,8 @@ private function run_adoption_agency_algorithm(): void {
 		}
 
 		$this->bail( 'Cannot run adoption agency when looping required.' );
+		// This unnecessary return prevents tools from inaccurately reporting type errors.
+		return false;
 	}
 
 	/**

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

@@ -142,6 +142,36 @@ public function test_unexpected_closing_tags_are_removed() {
 		);
 	}
 
+	/**
+	 * Ensures that unexpected closing formatting tags are ignored.
+	 *
+	 * @ticket 65372
+	 *
+	 * @dataProvider data_unexpected_closing_formatting_tags
+	 *
+	 * @param string $html     HTML containing an unexpected closing formatting tag.
+	 * @param string $expected Expected normalized output.
+	 */
+	public function test_unexpected_closing_formatting_tags_are_ignored( string $html, string $expected ) {
+		$this->assertSame(
+			$expected,
+			WP_HTML_Processor::normalize( $html ),
+			'Should have ignored unexpected closing formatting tags.'
+		);
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array[]
+	 */
+	public static function data_unexpected_closing_formatting_tags() {
+		return array(
+			'Unexpected A end tag' => array( 'one</a>two', 'onetwo' ),
+			'Unexpected B end tag' => array( 'one</b>two', 'onetwo' ),
+		);
+	}
+
 	/**
 	 * Ensures that self-closing elements in foreign content retain their self-closing flag.
 	 *

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

@@ -405,6 +405,66 @@ public function test_in_body_any_other_end_tag_with_unclosed_non_special_element
 		$this->assertSame( array( 'HTML', 'BODY', 'DIV', 'DIV' ), $processor->get_breadcrumbs(), 'Failed to produce expected DOM nesting: SPAN should be closed and DIV should be its sibling.' );
 	}
 
+	/**
+	 * Verifies that when the adoption agency algorithm finds no matching
+	 * active formatting element, it acts like "any other end tag".
+	 *
+	 * @covers WP_HTML_Processor::step_in_body
+	 *
+	 * @ticket 65372
+	 *
+	 * @dataProvider data_in_body_adoption_agency_falls_back_to_any_other_end_tag
+	 *
+	 * @param string $formatting_tag_name Formatting tag name with no active formatting element.
+	 */
+	public function test_in_body_adoption_agency_falls_back_to_any_other_end_tag( string $formatting_tag_name ) {
+		$processor = WP_HTML_Processor::create_fragment( "<div><span></{$formatting_tag_name}><code target></code></span></div>" );
+
+		$processor->next_tag( 'SPAN' );
+		$this->assertSame( 'SPAN', $processor->get_tag(), "Expected to start test on SPAN element but found {$processor->get_tag()} instead." );
+		$this->assertSame( array( 'HTML', 'BODY', 'DIV', 'SPAN' ), $processor->get_breadcrumbs(), 'Failed to produce expected DOM nesting.' );
+
+		$this->assertTrue( $processor->next_tag( 'CODE' ), "Failed to ignore unexpected {$formatting_tag_name} closer and advance to CODE opener." );
+		$this->assertSame( 'CODE', $processor->get_tag(), "Expected to find CODE element, but found {$processor->get_tag()} instead." );
+		$this->assertSame( array( 'HTML', 'BODY', 'DIV', 'SPAN', 'CODE' ), $processor->get_breadcrumbs(), 'Failed to keep SPAN open after unexpected formatting closer.' );
+	}
+
+	/**
+	 * Verifies that the adoption agency fallback preserves the "any other end tag"
+	 * step result when the ignored token is followed by EOF.
+	 *
+	 * @covers WP_HTML_Processor::step_in_body
+	 *
+	 * @ticket 65372
+	 *
+	 * @dataProvider data_in_body_adoption_agency_falls_back_to_any_other_end_tag
+	 *
+	 * @param string $formatting_tag_name Formatting tag name with no active formatting element.
+	 */
+	public function test_in_body_adoption_agency_fallback_preserves_any_other_end_tag_step_result( string $formatting_tag_name ) {
+		$ordinary_processor = WP_HTML_Processor::create_fragment( '<span></x>' );
+		$this->assertTrue( $ordinary_processor->step(), 'Failed to find the SPAN opener before an ordinary unexpected end tag.' );
+		$this->assertSame( 'SPAN', $ordinary_processor->get_tag(), "Expected to start test on SPAN element but found {$ordinary_processor->get_tag()} instead." );
+		$this->assertFalse( $ordinary_processor->step(), 'Expected ordinary unexpected end tag followed by EOF to return false.' );
+
+		$formatting_processor = WP_HTML_Processor::create_fragment( "<span></{$formatting_tag_name}>" );
+		$this->assertTrue( $formatting_processor->step(), 'Failed to find the SPAN opener before an unexpected formatting end tag.' );
+		$this->assertSame( 'SPAN', $formatting_processor->get_tag(), "Expected to start test on SPAN element but found {$formatting_processor->get_tag()} instead." );
+		$this->assertFalse( $formatting_processor->step(), 'Expected unexpected formatting end tag followed by EOF to return false.' );
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array[]
+	 */
+	public static function data_in_body_adoption_agency_falls_back_to_any_other_end_tag() {
+		return array(
+			'Unexpected A end tag' => array( 'a' ),
+			'Unexpected B end tag' => array( 'b' ),
+		);
+	}
+
 	/**
 	 * Ensures that closing `</br>` tags are appropriately treated as opening tags with no attributes.
 	 *