Fix multibyte character corruption in post summaries

.github/changelog/1995-from-description

@@ -0,0 +1,4 @@
+Significance: patch
+Type: fixed
+
+Fix multibyte character corruption in post summaries, preventing Greek and other non-ASCII text from being garbled during text processing.

includes/functions.php

@@ -1201,7 +1201,7 @@ function generate_post_summary( $post, $length = 500 ) {
 	 * @param string $excerpt_more The excerpt more.
 	 */
 	$excerpt_more = \apply_filters( 'activitypub_excerpt_more', '[…]' );
-	$length       = $length - strlen( $excerpt_more );
+	$length       = $length - \mb_strlen( $excerpt_more, 'UTF-8' );
 
 	$content = \sanitize_post_field( 'post_excerpt', $post->post_excerpt, $post->ID );
 
@@ -1223,13 +1223,13 @@ function generate_post_summary( $post, $length = 500 ) {
 	}
 
 	$content = \strip_shortcodes( $content );
-	$content = \html_entity_decode( $content );
 	$content = \wp_strip_all_tags( $content );
+	$content = \html_entity_decode( $content, ENT_QUOTES, 'UTF-8' );
 	$content = \trim( $content );
-	$content = \preg_replace( '/\R+/m', "\n\n", $content );
-	$content = \preg_replace( '/[\r\t]/', '', $content );
+	$content = \preg_replace( '/\R+/mu', "\n\n", $content );
+	$content = \preg_replace( '/[\r\t]/u', '', $content );
 
-	if ( $length && \strlen( $content ) > $length ) {
+	if ( $length && \mb_strlen( $content, 'UTF-8' ) > $length ) {
 		$content = \wordwrap( $content, $length, '</activitypub-summary>' );
 		$content = \explode( '</activitypub-summary>', $content, 2 );
 		$content = $content[0] . ' ' . $excerpt_more;

includes/transformer/class-post.php

@@ -208,7 +208,7 @@ protected function get_image() {
 
 		$alt = \get_post_meta( $id, '_wp_attachment_image_alt', true );
 		if ( $alt ) {
-			$image['name'] = \wp_strip_all_tags( \html_entity_decode( $alt ) );
+			$image['name'] = \html_entity_decode( \wp_strip_all_tags( $alt ), ENT_QUOTES, 'UTF-8' );
 		}
 
 		return $image;
@@ -264,7 +264,7 @@ protected function get_icon() {
 
 		$alt = \get_post_meta( $id, '_wp_attachment_image_alt', true );
 		if ( $alt ) {
-			$image['name'] = \wp_strip_all_tags( \html_entity_decode( $alt ) );
+			$image['name'] = \html_entity_decode( \wp_strip_all_tags( $alt ), ENT_QUOTES, 'UTF-8' );
 		}
 
 		return $image;
@@ -1013,11 +1013,11 @@ public function wp_attachment_to_activity_attachment( $media ) {
 					);
 
 					if ( ! empty( $media['alt'] ) ) {
-						$image['name'] = \wp_strip_all_tags( \html_entity_decode( $media['alt'] ) );
+						$image['name'] = \html_entity_decode( \wp_strip_all_tags( $media['alt'] ), ENT_QUOTES, 'UTF-8' );
 					} else {
 						$alt = \get_post_meta( $id, '_wp_attachment_image_alt', true );
 						if ( $alt ) {
-							$image['name'] = \wp_strip_all_tags( \html_entity_decode( $alt ) );
+							$image['name'] = \html_entity_decode( \wp_strip_all_tags( $alt ), ENT_QUOTES, 'UTF-8' );
 						}
 					}
 

tests/includes/class-test-functions.php

@@ -607,6 +607,13 @@ public function get_post_summary_data() {
 				),
 				'<p>Hello World</p>' . PHP_EOL,
 			),
+			array(
+				'Greek Excerpt',
+				array(
+					'post_excerpt' => 'Τι μπορεί να σου συμβεί σε μια βόλτα για να αγοράσεις μια βαλίτσα για τα ταξίδια σου; Όλα είναι πιθανά αν έχεις ανοιχτές τις "κεραίες" σου!',
+				),
+				'<p>Τι μπορεί να σου συμβεί σε μια βόλτα για να αγοράσεις μια βαλίτσα για τα ταξίδια σου; Όλα είναι πιθανά αν έχεις ανοιχτές τις &#8220;κεραίες&#8221; σου!</p>' . PHP_EOL,
+			),
 			array(
 				'Content',
 				array(

tests/includes/class-test-shortcodes.php

@@ -106,7 +106,7 @@ public function test_excerpt() {
 		$content = do_shortcode( '[ap_excerpt length="25"]' );
 		wp_reset_postdata();
 
-		$this->assertEquals( "<p>Lorem ipsum dolor […]</p>\n", $content );
+		$this->assertEquals( "<p>Lorem ipsum dolor sit […]</p>\n", $content );
 	}
 
 	/**