|
From: SourceForge.net <no...@so...> - 2003-02-25 02:43:53
|
Bugs item #692528, was opened at 2003-02-24 15:51 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=403611&aid=692528&group_id=31885 Category: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: Sven Eichler (koala_s) Assigned to: Nobody/Anonymous (nobody) Summary: affected_rows() and update in new_user_md5.php Initial Comment: I have found a bug in new_user_md5.php at this line: $query = "update auth_user_md5 set username='$username', $password_query perms='$permlist' where user_id='$u_id'"; $db->query($query); if ($db->affected_rows() == 0) { .... From http://www.php.net/manual/en/function.mysql-affected-rows.php: "When using UPDATE, MySQL will not update columns where the new value is the same as the old value. This creates the possiblity that mysql_affected_rows() may not actually equal the number of rows matched, only the number of rows that were literally affected by the query." I mean that the 0 comparison with $db->affected_rows() is not enough for a correct error test. A test with the output of mysql_errno() can fix this problem. But it give in db_mysql.inc, at this time, not a function to return direkt this result. ---------------------------------------------------------------------- Comment By: Adam Thompson (athompso) Date: 2003-02-24 20:53 Message: Logged In: YES user_id=339498 My original fix for this problem appears to still be present in db_mysql.inc(? I think - maybe ct_sql.inc?); the same theory of operation can be adapted to this case. Unfortunately, it triples the size of the error-checking code... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=403611&aid=692528&group_id=31885 |
