HTML API: Track self-closing flag during tokenization

Author: sirreal

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

@@ -627,6 +627,15 @@ class WP_HTML_Tag_Processor {
 	 */
 	private $token_length;
 
+	/**
+	 * Whether the current tag token has the self-closing flag.
+	 *
+	 * @since 6.9.0
+	 *
+	 * @var bool
+	 */
+	private $self_closing_flag = false;
+
 	/**
 	 * Byte offset in input document where current tag name starts.
 	 *
@@ -1079,6 +1088,7 @@ private function base_class_next_token(): bool {
 		$tag_name_starts_at   = $this->tag_name_starts_at;
 		$tag_name_length      = $this->tag_name_length;
 		$tag_ends_at          = $this->token_starts_at + $this->token_length;
+		$self_closing_flag    = $this->self_closing_flag;
 		$attributes           = $this->attributes;
 		$duplicate_attributes = $this->duplicate_attributes;
 
@@ -1136,6 +1146,7 @@ private function base_class_next_token(): bool {
 		$this->text_length          = $this->tag_name_starts_at - $this->text_starts_at;
 		$this->tag_name_starts_at   = $tag_name_starts_at;
 		$this->tag_name_length      = $tag_name_length;
+		$this->self_closing_flag    = $self_closing_flag;
 		$this->attributes           = $attributes;
 		$this->duplicate_attributes = $duplicate_attributes;
 
@@ -2143,7 +2154,19 @@ private function parse_next_attribute(): bool {
 		$doc_length = strlen( $this->html );
 
 		// Skip whitespace and slashes.
-		$this->bytes_already_parsed += strspn( $this->html, " \t\f\r\n/", $this->bytes_already_parsed );
+		$skipped_start               = $this->bytes_already_parsed;
+		$this->bytes_already_parsed += strspn( $this->html, " \t\f\r\n/", $skipped_start );
+
+		// A slash inside an unquoted attribute value will not have been skipped here.
+		if (
+			$this->bytes_already_parsed < $doc_length &&
+			$this->bytes_already_parsed > $skipped_start &&
+			'/' === $this->html[ $this->bytes_already_parsed - 1 ] &&
+			'>' === $this->html[ $this->bytes_already_parsed ]
+		) {
+			$this->self_closing_flag = true;
+		}
+
 		if ( $this->bytes_already_parsed >= $doc_length ) {
 			$this->parser_state = self::STATE_INCOMPLETE_INPUT;
 
@@ -2327,6 +2350,7 @@ private function after_tag(): void {
 
 		$this->token_starts_at          = null;
 		$this->token_length             = null;
+		$this->self_closing_flag        = false;
 		$this->tag_name_starts_at       = null;
 		$this->tag_name_length          = null;
 		$this->text_starts_at           = 0;
@@ -3335,42 +3359,7 @@ public function has_self_closing_flag(): bool {
 			return false;
 		}
 
-		/*
-		 * The self-closing flag is the solidus at the _end_ of the tag, not the beginning.
-		 *
-		 * Example:
-		 *
-		 *     <figure />
-		 *             ^ this appears one character before the end of the closing ">".
-		 */
-		$self_closing_flag_at = $this->token_starts_at + $this->token_length - 2;
-		if ( '/' !== $this->html[ $self_closing_flag_at ] ) {
-			return false;
-		}
-
-		foreach ( $this->attributes as $attribute ) {
-			$attribute_ends_at = $attribute->start + $attribute->length;
-			if (
-				$self_closing_flag_at >= $attribute->start &&
-				$self_closing_flag_at < $attribute_ends_at
-			) {
-				return false;
-			}
-		}
-
-		foreach ( $this->duplicate_attributes ?? array() as $duplicate_attributes ) {
-			foreach ( $duplicate_attributes as $attribute ) {
-				$attribute_ends_at = $attribute->start + $attribute->length;
-				if (
-					$self_closing_flag_at >= $attribute->start &&
-					$self_closing_flag_at < $attribute_ends_at
-				) {
-					return false;
-				}
-			}
-		}
-
-		return true;
+		return $this->self_closing_flag;
 	}
 
 	/**

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

@@ -119,6 +119,45 @@ public static function data_has_self_closing_flag() {
 		);
 	}
 
+	/**
+	 * Ensures internally consumed special-element closers do not affect the opener's
+	 * self-closing flag.
+	 *
+	 * @ticket 61576
+	 *
+	 * @covers WP_HTML_Tag_Processor::has_self_closing_flag
+	 *
+	 * @dataProvider data_special_atomic_self_closing_flags
+	 *
+	 * @param string $html        Input HTML whose first tag might contain the self-closing flag `/`.
+	 * @param bool   $flag_is_set Whether the input HTML's first tag contains the self-closing flag.
+	 */
+	public function test_special_atomic_elements_report_opening_tag_self_closing_flag( string $html, bool $flag_is_set ) {
+		$processor = new WP_HTML_Tag_Processor( $html );
+
+		$this->assertTrue( $processor->next_token(), 'Expected to find complete special atomic tag.' );
+
+		if ( $flag_is_set ) {
+			$this->assertTrue( $processor->has_self_closing_flag(), 'Did not find the self-closing flag on the opening tag.' );
+		} else {
+			$this->assertFalse( $processor->has_self_closing_flag(), 'Reported the internally consumed closing tag self-closing flag on the opening tag.' );
+		}
+	}
+
+	/**
+	 * Data provider.
+	 *
+	 * @return array[]
+	 */
+	public static function data_special_atomic_self_closing_flags() {
+		return array(
+			'SCRIPT closer self-closing flag' => array( '<script>x</script/>', false ),
+			'STYLE closer self-closing flag'  => array( '<style>x</style/>', false ),
+			'TITLE closer self-closing flag'  => array( '<title>x</title/>', false ),
+			'TITLE opener self-closing flag'  => array( '<title/>x</title>', true ),
+		);
+	}
+
 	/**
 	 * Ensures a trailing slash in an unquoted attribute value is part of the value.
 	 *