From 54d644ff4c6a907d0d33e2cc26fe9ba195edf324 Mon Sep 17 00:00:00 2001
From: Dennis Snell <dennis.snell@automattic.com>
Date: Mon, 22 Jan 2024 22:10:29 -0700
Subject: [PATCH] Normalize PI-Node lookalike behavior

---
 .../html-api/class-wp-html-tag-processor.php  | 17 +++++++++---
 .../wpHtmlTagProcessor-token-scanning.php     | 27 ++++++++++++-------
 2 files changed, 30 insertions(+), 14 deletions(-)

diff --git a/src/wp-includes/html-api/class-wp-html-tag-processor.php b/src/wp-includes/html-api/class-wp-html-tag-processor.php
index bb1f8dff7115d..e7c71bada4e9f 100644
--- a/src/wp-includes/html-api/class-wp-html-tag-processor.php
+++ b/src/wp-includes/html-api/class-wp-html-tag-processor.php
@@ -1818,7 +1818,6 @@ private function parse_next_tag() {
 					if ( 0 < $pi_target_length ) {
 						$pi_target_length += strspn( $comment_text, 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789:_-.', $pi_target_length );
 
-						$this->parser_state       = self::STATE_COMMENT;
 						$this->comment_type       = self::COMMENT_AS_PI_NODE_LOOKALIKE;
 						$this->tag_name_starts_at = $this->token_starts_at + 2;
 						$this->tag_name_length    = $pi_target_length;
@@ -2545,13 +2544,24 @@ public function get_attribute_names_with_prefix( $prefix ) {
 	 * @return string|null Name of currently matched tag in input HTML, or `null` if none found.
 	 */
 	public function get_tag() {
-		if ( self::STATE_MATCHED_TAG !== $this->parser_state ) {
+		if ( null === $this->tag_name_starts_at ) {
 			return null;
 		}
 
 		$tag_name = substr( $this->html, $this->tag_name_starts_at, $this->tag_name_length );
 
-		return strtoupper( $tag_name );
+		if ( self::STATE_MATCHED_TAG === $this->parser_state ) {
+			return strtoupper( $tag_name );
+		}
+
+		if (
+			self::STATE_COMMENT === $this->parser_state &&
+			self::COMMENT_AS_PI_NODE_LOOKALIKE === $this->get_comment_type()
+		) {
+			return $tag_name;
+		}
+
+		return null;
 	}
 
 	/**
@@ -3397,7 +3407,6 @@ private function matches() {
 	 *
 	 *     <!-->
 	 *     <!--->
-	 *     <!---->
 	 *
 	 * @since 6.5.0
 	 */
diff --git a/tests/phpunit/tests/html-api/wpHtmlTagProcessor-token-scanning.php b/tests/phpunit/tests/html-api/wpHtmlTagProcessor-token-scanning.php
index 2f59b430f21c3..538144910550f 100644
--- a/tests/phpunit/tests/html-api/wpHtmlTagProcessor-token-scanning.php
+++ b/tests/phpunit/tests/html-api/wpHtmlTagProcessor-token-scanning.php
@@ -319,9 +319,15 @@ public function test_basic_assertion_cdata_section() {
 		$processor->next_token();
 
 		$this->assertSame(
-			'#cdata-section',
+			'#comment',
 			$processor->get_token_name(),
-			"Should have found CDATA section name but found {$processor->get_token_name()} instead."
+			"Should have found comment token but found {$processor->get_token_name()} instead."
+		);
+
+		$this->assertSame(
+			WP_HTML_Processor::COMMENT_AS_CDATA_LOOKALIKE,
+			$processor->get_comment_type(),
+			'Should have detected a CDATA-like invalid comment.'
 		);
 
 		$this->assertNull(
@@ -405,20 +411,21 @@ public function test_basic_assertion_processing_instruction() {
 		$processor->next_token();
 
 		$this->assertSame(
-			'#processing-instruction',
-			$processor->get_token_type(),
-			"Should have found PI node but found {$processor->get_token_type()} instead."
+			'#comment',
+			$processor->get_token_name(),
+			"Should have found comment token but found {$processor->get_token_name()} instead."
 		);
 
 		$this->assertSame(
-			'wp-bit',
-			$processor->get_token_name(),
-			"Should have found PI target as name but found {$processor->get_token_name()} instead."
+			WP_HTML_Processor::COMMENT_AS_PI_NODE_LOOKALIKE,
+			$processor->get_comment_type(),
+			'Should have detected a Processing Instruction-like invalid comment.'
 		);
 
-		$this->assertNull(
+		$this->assertSame(
+			'wp-bit',
 			$processor->get_tag(),
-			'Should not have been able to query tag name on non-element token.'
+			"Should have found PI target as tag name but found {$processor->get_tag()} instead."
 		);
 
 		$this->assertNull(