No reason to consider % dangerous twice...

Also i retract my pcre statement from before, i doubt a perl pack and hex call are going to work in pcre.  The regex should give plenty of idea what you need to do in your language of choice though, thats the point :)

git-svn-id: http://svn.freeswitch.org/svn/freeswitch/trunk@3482 d0543943-73ff-0310-b7d9-9358b9ac24b2

diff --git a/src/switch_utils.c b/src/switch_utils.c
index 64b610a6212c2820a5646d5cea3aeeefd2edba07..623942898cda8246d0ca02bb452831ed95417596 100644 (file)
--- a/src/switch_utils.c
+++ b/src/switch_utils.c
@@ -297,7 +297,7 @@ SWITCH_DECLARE(size_t) switch_url_encode(char *url, char *buf, size_t len)
 {
     char *p;
     size_t x = 0;
-    const char urlunsafe[] = "\r\n %\"#%&+:;<=>?@[\\]^`{|}";
+    const char urlunsafe[] = "\r\n \"#%&+:;<=>?@[\\]^`{|}";
     const char hex[] = "0123456789ABCDEF";
 
     memset(buf, 0, len);