On Mon, 2005-02-28 at 14:19 -0500, Jeffrey Stedfast wrote: > On Mon, 2005-02-28 at 14:04 -0500, Jeffrey Stedfast wrote: > > On Mon, 2005-02-28 at 12:22 +0800, Not Zed wrote: > > > > > > So, how many valid mails does this break? > > > > > > e.g. > > > > > > =?iso-8859-1?b?foo this is a hidden message not for evolution users > > > bar?= > > > > probably would decode into garbage... or whatever "foo this is a hidden > > message not for evolution users bar" is base64 decoded. > > turns out, this subject would be decoded into: > > Subject: =?iso-8859-1?b?foo this is a hidden message not for evolution > users bar?= > > the base64 decoder fails and so rfc2047_decode_word() fails, at which > point we fall back to treating it like normal text n/m, it decodes into garbage. I was right the first time. the reason it didn't decode before was because of an off-by-1 bug in my transition to memchr() (which is now fixed). I'll finish writing up a test suite and send a new patch, hopefully by the end of today... gonna work on more pressing issues first. Jeff > -- Jeffrey Stedfast Evolution Hacker - Novell, Inc. fejj ximian com - www.novell.com
Attachment:
smime.p7s
Description: S/MIME cryptographic signature