Changeset 29748 for trunk/src/wp-includes/formatting.php

Timestamp:

09/17/2014 03:13:24 PM (10 years ago)

Author:

wonderboymusic

Message:

wptexturize() improvements:

• Expand the wptexturize() RegEx to include the list of registered shortcodes.
• Avoid backtracking after [ chars by not filtering params in registered shortcodes. This will cause escaped shortcodes and their params to become texturized if not registered.
• Registered shortcode params will never be texturized, even when escaped.
• Move all tests involving unregistered shortcodes to a new and improved unit.
• Update one test involving HTML within shortcode params.

Props miqrogroove.
See #29557.

File:

• trunk/src/wp-includes/formatting.php (modified) (5 diffs)

trunk/src/wp-includes/formatting.php

@@ -29 +29 @@
  */
 function wptexturize($text, $reset = false) {
-    global $wp_cockneyreplace;
+    global $wp_cockneyreplace, $shortcode_tags;
     static $static_characters, $static_replacements, $dynamic_characters, $dynamic_replacements,
         $default_no_texturize_tags, $default_no_texturize_shortcodes, $run_texturize = true;
@@ -206 +206 @@
     // Look for shortcodes and HTML elements.
 
+    $tagnames = array_keys( $shortcode_tags );
+    $tagregexp = join( '|', array_map( 'preg_quote', $tagnames ) );
+    $tagregexp = "(?:$tagregexp)(?![\\w-])"; // Excerpt of get_shortcode_regex().
+    
     $regex =  '/('          // Capture the entire match.
         .   '<'     // Find start of element.
@@ -216 +220 @@
         .   '\['        // Find start of shortcode.
         .   '\[?'       // Shortcodes may begin with [[
-        .   '(?:'
-        .       '[^\[\]<>]' // Shortcodes do not contain other shortcodes.
-        .   '|'
-        .       '<[^>]+>'   // HTML elements permitted. Prevents matching ] before >.
-        .   ')++'
+        .   '\/?'       // Closing slash may precede name.
+        .   $tagregexp  // Only match registered shortcodes, because performance.
+        .   '[^\[\]]*'  // Shortcodes do not contain other shortcodes.
         .   '\]'        // Find end of shortcode.
         .   '\]?'       // Shortcodes may end with ]]
@@ -242 +244 @@
             continue;
 
-        } elseif ( '[' === $first && 1 === preg_match( '/^\[(?:[^\[\]<>]|<[^>]+>)++\]$/', $curl ) ) {
+        } elseif ( '[' === $first && 1 === preg_match( '/^\[\[?\/?' . $tagregexp . '[^\[\]]*\]\]?$/', $curl ) ) {
             // This is a shortcode delimiter.
 
-            _wptexturize_pushpop_element( $curl, $no_texturize_shortcodes_stack, $no_texturize_shortcodes );
-
-        } elseif ( '[' === $first && 1 === preg_match( '/^\[\[?(?:[^\[\]<>]|<[^>]+>)++\]\]?$/', $curl ) ) {
-            // This is an escaped shortcode delimiter.
-
-            // Do not texturize.
-            // Do not push to the shortcodes stack.
-
-            continue;
+            if ( '[[' !== substr( $curl, 0, 2 ) && ']]' !== substr( $curl, -2 ) ) {
+                // Looks like a normal shortcode.
+                _wptexturize_pushpop_element( $curl, $no_texturize_shortcodes_stack, $no_texturize_shortcodes );
+            } else {
+                // Looks like an escaped shortcode.
+                // Do not texturize.
+                // Do not push to the shortcodes stack.
+                continue;
+            }
 
         } elseif ( empty( $no_texturize_shortcodes_stack ) && empty( $no_texturize_tags_stack ) ) {
@@ -314 +316 @@
     // Parse out the tag name.
     $space = strpos( $text, ' ' );
-    if ( FALSE === $space ) {
+    if ( false === $space ) {
         $space = -1;
     } else {