HTML API: Handle "adoption agency" any other end tag case

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

@@ -3256,38 +3256,55 @@ private function step_in_body(): bool {
 			/*
 			 * > Any other end tag
 			 */
+			return $this->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();
-				}
+	/**
+	 * In body insertion mode's "any other end tag" logic can be invoked from different places
+	 * and may require additional processing.
+	 *
+	 * @param $should_step bool Set to `false` to prevent advancing the parser in case the token
+	 *                          is ignored.
+	 * @return bool Whether an element was found.
+	 */
+	private function in_body_any_other_end_tag( bool $should_step = true ): 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 $should_step ? $this->step() : false;
 			}
+		}
 
-			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;
 	}
@@ -6265,7 +6282,14 @@ 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 ) {
-				$this->bail( 'Cannot run adoption agency when "any other end tag" is required.' );
+				/*
+				 * The Adoption Agency Algorithm is not responsible for advancing the parser state,
+				 * so `in_body_any_other_end_tag` is invoked with `$should_step = false` argument.
+				 * This algorithm will return to the context that called it, which is responsible
+				 * for advancing the parser.
+				 */
+				$this->in_body_any_other_end_tag( false );
+				return;
 			}
 
 			// > 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.

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

@@ -265,6 +265,47 @@ 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 $formatting_tag_name Formatting tag name with no active formatting element.
+	 */
+	public function test_unexpected_closing_formatting_tags_are_ignored( string $formatting_tag_name ) {
+		$this->assertSame(
+			'onetwo',
+			WP_HTML_Processor::normalize( "one</{$formatting_tag_name}>two" ),
+			"Should have ignored unexpected {$formatting_tag_name} closer."
+		);
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array[]
+	 */
+	public static function data_unexpected_closing_formatting_tags() {
+		return array(
+			'Unexpected A end tag'      => array( 'a' ),
+			'Unexpected B end tag'      => array( 'b' ),
+			'Unexpected BIG end tag'    => array( 'big' ),
+			'Unexpected CODE end tag'   => array( 'code' ),
+			'Unexpected EM end tag'     => array( 'em' ),
+			'Unexpected FONT end tag'   => array( 'font' ),
+			'Unexpected I end tag'      => array( 'i' ),
+			'Unexpected NOBR end tag'   => array( 'nobr' ),
+			'Unexpected S end tag'      => array( 's' ),
+			'Unexpected SMALL end tag'  => array( 'small' ),
+			'Unexpected STRIKE end tag' => array( 'strike' ),
+			'Unexpected STRONG end tag' => array( 'strong' ),
+			'Unexpected TT end tag'     => array( 'tt' ),
+			'Unexpected U end tag'      => array( 'u' ),
+		);
+	}
+
 	/**
 	 * Ensures that self-closing elements in foreign content retain their self-closing flag.
 	 *

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

@@ -577,6 +577,7 @@ public static function data_virtual_nodes_breadcrumbs() {
 			'Implied P tag opener on unmatched closer'    => array( '</p>', 1, 'P', 'open', array( 'HTML', 'BODY', 'P' ) ),
 			'Implied heading tag closer on heading child' => array( '<h1><h2>', 2, 'H1', 'close', array( 'HTML', 'BODY' ) ),
 			'Implied A tag closer on A tag child'         => array( '<a><a>', 2, 'A', 'close', array( 'HTML', 'BODY' ) ),
+			'Explicit A closer after sibling A'           => array( '<a><a></a></a>', 4, 'A', 'close', array( 'HTML', 'BODY' ) ),
 			'Implied A tag closer on A tag descendent'    => array( '<a><span><a>', 4, 'A', 'close', array( 'HTML', 'BODY' ) ),
 		);
 	}

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

@@ -405,6 +405,78 @@ 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_fallback_end_tags
+	 *
+	 * @param string $formatting_tag_name Formatting tag name with no active formatting element.
+	 */
+	public function test_in_body_adoption_agency_fallback_ignores_unexpected_formatting_end_tag( string $formatting_tag_name ) {
+		$processor = WP_HTML_Processor::create_fragment( "<div><span></{$formatting_tag_name}><code target></code></span></div>" );
+
+		$this->assertTrue( $processor->next_tag( 'SPAN' ), 'Failed to find the SPAN opener before an unexpected formatting end tag.' );
+		$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 before unexpected formatting closer.' );
+
+		$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_fallback_end_tags
+	 *
+	 * @param string $formatting_tag_name Formatting tag name with no active formatting element.
+	 */
+	public function test_in_body_adoption_agency_fallback_preserves_ignored_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_tag_name} end tag followed by EOF to return false." );
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array[]
+	 */
+	public static function data_in_body_adoption_agency_fallback_end_tags() {
+		return array(
+			'Unexpected A end tag'      => array( 'a' ),
+			'Unexpected B end tag'      => array( 'b' ),
+			'Unexpected BIG end tag'    => array( 'big' ),
+			'Unexpected CODE end tag'   => array( 'code' ),
+			'Unexpected EM end tag'     => array( 'em' ),
+			'Unexpected FONT end tag'   => array( 'font' ),
+			'Unexpected I end tag'      => array( 'i' ),
+			'Unexpected NOBR end tag'   => array( 'nobr' ),
+			'Unexpected S end tag'      => array( 's' ),
+			'Unexpected SMALL end tag'  => array( 'small' ),
+			'Unexpected STRIKE end tag' => array( 'strike' ),
+			'Unexpected STRONG end tag' => array( 'strong' ),
+			'Unexpected TT end tag'     => array( 'tt' ),
+			'Unexpected U end tag'      => array( 'u' ),
+		);
+	}
+
 	/**
 	 * Ensures that closing `</br>` tags are appropriately treated as opening tags with no attributes.
 	 *