From bf800af9b3392fb733da7524f77d38a5b36f892e Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Thu, 9 Apr 2026 08:50:22 +0000
Subject: [PATCH 1/2] Initial plan


From 55cd8ec0506c684761e5e031e65f7bfa4afb2208 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Thu, 9 Apr 2026 09:41:58 +0000
Subject: [PATCH 2/2] Fix strcspn/strspn boundary scans in decode_string_or_url
 and decode_escape_at; add targeted tests

Agent-Logs-Url: https://github.com/WordPress/php-toolkit/sessions/c4d1289b-697f-44ee-9750-fac7d627e1b5

Co-authored-by: sirreal <841763+sirreal@users.noreply.github.com>
---
 .../DataLiberation/CSS/class-cssprocessor.php | 16 +++----
 .../DataLiberation/Tests/CSSProcessorTest.php | 42 +++++++++++++++++++
 2 files changed, 48 insertions(+), 10 deletions(-)

diff --git a/components/DataLiberation/CSS/class-cssprocessor.php b/components/DataLiberation/CSS/class-cssprocessor.php
index e7f50adb..adcdd087 100644
--- a/components/DataLiberation/CSS/class-cssprocessor.php
+++ b/components/DataLiberation/CSS/class-cssprocessor.php
@@ -1564,13 +1564,11 @@ private function decode_string_or_url( int $start, int $length ): string {
 		$end     = $start + $length;
 
 		while ( $at < $end ) {
-			// Find next special character.
-			$normal_len = strcspn( $this->css, $special_chars, $at );
+			// Find next special character within the token boundary.
+			$normal_len = strcspn( $this->css, $special_chars, $at, $end - $at );
 			if ( $normal_len > 0 ) {
-				// Clamp to not exceed the end boundary.
-				$normal_len = min( $normal_len, $end - $at );
-				$decoded   .= substr( $this->css, $at, $normal_len );
-				$at        += $normal_len;
+				$decoded .= substr( $this->css, $at, $normal_len );
+				$at      += $normal_len;
 			}
 
 			if ( $at >= $end ) {
@@ -1648,11 +1646,9 @@ private function decode_escape_at( int $offset, &$bytes_consumed ): string {
 			return "\u{FFFD}";
 		}
 
-		// Hex digits.
-		$hex_len = strspn( $this->css, '0123456789ABCDEFabcdef', $at );
+		// Hex digits (CSS spec allows at most 6).
+		$hex_len = strspn( $this->css, '0123456789ABCDEFabcdef', $at, 6 );
 		if ( $hex_len > 0 ) {
-			// Consume up to 6 hex digits.
-			$hex_len = min( $hex_len, 6 );
 			$hex     = substr( $this->css, $at, $hex_len );
 			$at     += $hex_len;
 
diff --git a/components/DataLiberation/Tests/CSSProcessorTest.php b/components/DataLiberation/Tests/CSSProcessorTest.php
index a6f7f7f3..113af1b0 100644
--- a/components/DataLiberation/Tests/CSSProcessorTest.php
+++ b/components/DataLiberation/Tests/CSSProcessorTest.php
@@ -1556,4 +1556,46 @@ public function test_bad_string_token_value_is_null(): void {
 		$this->assertSame( CSSProcessor::TOKEN_BAD_STRING, $processor->get_token_type() );
 		$this->assertNull( $processor->get_token_value() );
 	}
+
+	/**
+	 * Tests that decode_string_or_url() respects the token's length boundary
+	 * and does not include content from beyond the token end.
+	 *
+	 * The escape sequence \41 (= 'A') triggers the slow path in
+	 * decode_string_or_url(). The CSS string continues with "; color: red;"
+	 * after the closing quote, which must not appear in the token value.
+	 */
+	public function test_decode_string_or_url_respects_length_boundary(): void {
+		// \41 = 'A' — triggers the slow path; "; color: red;" follows the token.
+		$css = '"hello\\41 world"; color: red;';
+
+		$processor = CSSProcessor::create( $css );
+		$processor->next_token();
+
+		$this->assertSame( CSSProcessor::TOKEN_STRING, $processor->get_token_type() );
+		$this->assertSame( 'helloAworld', $processor->get_token_value() );
+		$this->assertSame( '"helloAworld"', $processor->get_normalized_token() );
+	}
+
+	/**
+	 * Tests that decode_escape_at() consumes at most 6 hex digits, as required
+	 * by the CSS Syntax Level 3 specification.
+	 *
+	 * A hex escape with 7 consecutive hex digits must only consume the first 6,
+	 * leaving the 7th as a literal character in the string value.
+	 *
+	 * @see https://www.w3.org/TR/css-syntax-3/#consume-escaped-code-point
+	 */
+	public function test_decode_escape_at_hex_limit_is_six_digits(): void {
+		// \000041 is 6 hex digits → U+0041 = 'A'; the trailing '1' is literal.
+		// Without the length limit, strspn() would scan 7 hex digits (0000411),
+		// giving U+0411 = 'Б' (Cyrillic), which is incorrect.
+		$css = '"\\0000411rest"';
+
+		$processor = CSSProcessor::create( $css );
+		$processor->next_token();
+
+		$this->assertSame( CSSProcessor::TOKEN_STRING, $processor->get_token_type() );
+		$this->assertSame( 'A1rest', $processor->get_token_value() );
+	}
 }