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

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

@@ -2921,8 +2921,7 @@
 			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"
@@ -6246,8 +6245,10 @@
 	 * @throws WP_HTML_Unsupported_Exception When encountering unsupported HTML input.
 	 *
 	 * @see https://html.spec.whatwg.org/#adoption-agency-algorithm
+	 *
+	 * @return bool Whether the current token was handled without exhausting input.
 	 */
-	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();
@@ -6259,13 +6260,13 @@
 			! $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;
 			}
 
 			/*
@@ -6286,20 +6287,20 @@
 				}
 			}
 
-			// > If there is no such element, then return and instead act as described in the "any other end tag" entry above.
+			// > If there is no such element, then act as described in the "any other end tag" entry above and return.
 			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;
 			}
 
 			/*
@@ -6335,7 +6336,7 @@
 
 					if ( $formatting_element->bookmark_name === $item->bookmark_name ) {
 						$this->state->active_formatting_elements->remove_node( $formatting_element );
-						return;
+						return true;
 					}
 				}
 			}

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

@@ -281,6 +281,47 @@ public function test_unexpected_closing_tags_are_removed() {
 		);
 	}
 
+	/**
+	 * Ensures that unexpected closing formatting tags are ignored.
+	 *
+	 * @ticket 65383
+	 *
+	 * @dataProvider data_formatting_tag_names
+	 *
+	 * @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<string, array{0: string}>
+	 */
+	public static function data_formatting_tag_names(): array {
+		return array(
+			'A tag'      => array( 'a' ),
+			'B tag'      => array( 'b' ),
+			'BIG tag'    => array( 'big' ),
+			'CODE tag'   => array( 'code' ),
+			'EM tag'     => array( 'em' ),
+			'FONT tag'   => array( 'font' ),
+			'I tag'      => array( 'i' ),
+			'NOBR tag'   => array( 'nobr' ),
+			'S tag'      => array( 's' ),
+			'SMALL tag'  => array( 'small' ),
+			'STRIKE tag' => array( 'strike' ),
+			'STRONG tag' => array( 'strong' ),
+			'TT tag'     => array( 'tt' ),
+			'U 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' ) ),
+			'Redundant 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 +405,92 @@ 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 65383
+	 *
+	 * @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 65383
+	 *
+	 * @dataProvider data_in_body_adoption_agency_fallback_end_tags
+	 *
+	 * @param string $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 $tag_name ): void {
+		$processor = WP_HTML_Processor::create_fragment( "</{$tag_name}>" );
+		$this->assertFalse( $processor->step(), "Expected unexpected {$tag_name} end tag followed by EOF to return false." );
+	}
+
+	/**
+	 * Verifies that when the adoption agency algorithm returns after removing
+	 * a formatting element from the active formatting elements list, it does
+	 * not report the current token as EOF.
+	 *
+	 * @covers WP_HTML_Processor::step_in_body
+	 *
+	 * @ticket 65383
+	 *
+	 * @dataProvider data_in_body_adoption_agency_fallback_end_tags
+	 *
+	 * @param string $tag_name Formatting tag name with no open element.
+	 */
+	public function test_in_body_adoption_agency_removes_inactive_formatting_element_and_continues( string $tag_name ): void {
+		$processor = WP_HTML_Processor::create_fragment( "<p><{$tag_name}></p></{$tag_name}><span target></span>" );
+
+		$this->assertTrue( $processor->next_tag( $tag_name ), "Failed to find the {$tag_name} opener before it is popped by the P closer." );
+		$this->assertTrue( $processor->next_tag( 'SPAN' ), "Failed to advance past the inactive {$tag_name} closer to the following SPAN opener." );
+		$this->assertSame( array( 'HTML', 'BODY', 'SPAN' ), $processor->get_breadcrumbs(), "Expected SPAN to be a BODY child after the inactive {$tag_name} closer." );
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array<string, array{0: string}>
+	 */
+	public static function data_in_body_adoption_agency_fallback_end_tags(): array {
+		return array(
+			'A tag'      => array( 'a' ),
+			'B tag'      => array( 'b' ),
+			'BIG tag'    => array( 'big' ),
+			'CODE tag'   => array( 'code' ),
+			'EM tag'     => array( 'em' ),
+			'FONT tag'   => array( 'font' ),
+			'I tag'      => array( 'i' ),
+			'NOBR tag'   => array( 'nobr' ),
+			'S tag'      => array( 's' ),
+			'SMALL tag'  => array( 'small' ),
+			'STRIKE tag' => array( 'strike' ),
+			'STRONG tag' => array( 'strong' ),
+			'TT tag'     => array( 'tt' ),
+			'U tag'      => array( 'u' ),
+		);
+	}
+
 	/**
 	 * Ensures that closing `</br>` tags are appropriately treated as opening tags with no attributes.
 	 *