Mercurial > pidgin
comparison libpurple/util.c @ 18600:321658bf0e03
Valgrind was showing "Invalid read of size 1" when a string is passed
to this function that doesn't have an @ in it. I think this bug exists
in the O'Reilly version, too
| author | Mark Doliner <mark@kingant.net> |
|---|---|
| date | Sun, 22 Jul 2007 06:53:13 +0000 |
| parents | 481eb37730bf |
| children | 450090c68351 |
comparison
equal
deleted
inserted
replaced
| 18599:4efd240ffaa0 | 18600:321658bf0e03 |
|---|---|
| 3992 } | 3992 } |
| 3993 if (*c == '@') break; | 3993 if (*c == '@') break; |
| 3994 if (*c <= ' ' || *c >= 127) return FALSE; | 3994 if (*c <= ' ' || *c >= 127) return FALSE; |
| 3995 if (strchr(rfc822_specials, *c)) return FALSE; | 3995 if (strchr(rfc822_specials, *c)) return FALSE; |
| 3996 } | 3996 } |
| 3997 | |
| 3998 /* It's obviously not an email address if we didn't find an '@' above */ | |
| 3999 if (*c == '\0') return FALSE; | |
| 4000 | |
| 3997 /* strictly we should return false if (*(c - 1) == '.') too, but I think | 4001 /* strictly we should return false if (*(c - 1) == '.') too, but I think |
| 3998 * we should permit user.@domain type addresses - they do work :) */ | 4002 * we should permit user.@domain type addresses - they do work :) */ |
| 3999 if (c == address) return FALSE; | 4003 if (c == address) return FALSE; |
| 4000 | 4004 |
| 4001 /* next we validate the domain portion (name@domain) (rfc1035 & rfc1011) */ | 4005 /* next we validate the domain portion (name@domain) (rfc1035 & rfc1011) */ |