Changeset 425

Timestamp:

Oct 31, 2013 7:34:46 PM (11 years ago)

Author:

anonymous

Message:

Added unicdoe case to word split.

File:

• trunk/lib/SpellCheck.inc.php (modified) (1 diff)

trunk/lib/SpellCheck.inc.php

@@ -266 +266 @@
     {
         $corrections = array();
-        $words = preg_split('/([\W]+?)/', $string, -1, PREG_SPLIT_DELIM_CAPTURE);
+        // Split words on punctuation except apostrophes
+        // http://stackoverflow.com/questions/790596/split-a-text-into-single-words
+        $words = preg_split("/((^\p{P}+)|(\p{P}*\s+\p{P}*)|[\p{Pd}-–—]+|(\+|(\p{P}+$))/", $string);
         // Remove non-word elements.
         $words = preg_grep('/\w+/', $words);