Ticket #33106: 33106.2.patch

src/wp-includes/class-wp-embed.php

@@ -318 +318 @@
          * @return string Potentially modified $content.
          */
         public function autoembed( $content ) {
-                // Strip newlines from all elements.
-                $content = wp_replace_in_html_tags( $content, array( "\n" => " " ) );
+                // Avoid multiline elements and comments.
+                $content = wp_html_split( $content );
+                for ( $i = 0, $c = count( $content ); $i < $c; $i += 2 ) { 
+                
+                        if ( '' == $content[$i] ) continue;
 
-                // Find URLs that are on their own line.
-                return preg_replace_callback( '|^(\s*)(https?://[^\s"]+)(\s*)$|im', array( $this, 'autoembed_callback' ), $content );
+                        // Find URLs that are on their own line.
+                        $content[$i] = preg_replace_callback( '|^(\s*)(https?://[^\s"]+)(\s*)$|im', array( $this, 'autoembed_callback' ), $content[$i] );
+
+                }
+                
+                return implode( $content );
         }
 
         /**

src/wp-includes/formatting.php

@@ -596 +596 @@
 }
 
 /**
- * Replace characters or phrases within HTML elements only.
+ * Separate HTML elements and comments from the text.
  *
- * @since 4.2.3
+ * @since 4.2.4
  *
- * @param string $haystack The text which has to be formatted.
- * @param array $replace_pairs In the form array('from' => 'to', ...).
+ * @param string $input The text which has to be formatted.
  * @return string The formatted text.
  */
-function wp_replace_in_html_tags( $haystack, $replace_pairs ) {
+function wp_html_split( $input ) {
         // Find all elements.
         $comments =
                   '!'           // Start of comment, after the <.
@@ -614 +613 @@
                 . ')*+'         // Loop possessively.
                 . '(?:-->)?';   // End of comment. If not found, match all input.
 
+        $cdata =
+                  '!\[CDATA\['  // Start of comment, after the <.
+                . '[^\]]*+'     // Consume non-].
+                . '(?:'         // Unroll the loop: Consume everything until ]]> is found.
+                .     '](?!]>)' // One ] not followed by end of comment.
+                .     '[^\]]*+' // Consume non-].
+                . ')*+'         // Loop possessively.
+                . '(?:]]>)?';   // End of comment. If not found, match all input.
+
         $regex =
                   '/('              // Capture the entire match.
                 .     '<'           // Find start of element.
@@ -620 +628 @@
                 .     '(?(?=!--)'   // Is this a comment?
                 .         $comments // Find end of comment.
                 .     '|'
-                .         '[^>]*>?' // Find end of element. If not found, match all input.
+                .         '(?(?=!\[CDATA\[)' // Is this a comment?
+                .             $cdata // Find end of comment.
+                .         '|'
+                .             '[^>]*>?' // Find end of element. If not found, match all input.
+                .         ')'
                 .     ')'
                 . ')/s';
 
-        $textarr = preg_split( $regex, $haystack, -1, PREG_SPLIT_DELIM_CAPTURE );
+        return preg_split( $regex, $input, -1, PREG_SPLIT_DELIM_CAPTURE );
+}
+
+/**
+ * Replace characters or phrases within HTML elements only.
+ *
+ * @since 4.2.3
+ *
+ * @param string $haystack The text which has to be formatted.
+ * @param array $replace_pairs In the form array('from' => 'to', ...).
+ * @return string The formatted text.
+ */
+function wp_replace_in_html_tags( $haystack, $replace_pairs ) {
+        // Find all elements.
+        $textarr = wp_html_split( $haystack );
         $changed = false;
 
         // Optimize when searching for one item.