Re: [Evolution] Evolution & KlamAV

[Jeffrey Stedfast <fejj novell com>]

I think only the Pipe-to-Program *action* (not rule) reads back data
from the piped program... That might explain what you are seeing?

Well, that and outgoing filters are applied *after* the message is sent
across the wire (but before the message is saved into the Sent folder)

Jeff

On Mon, 2006-06-19 at 12:06 -0400, Thomas R. Krishan wrote:
> On Mon, 2006-06-19 at 09:38 -0400, Jeffrey Stedfast wrote:
> > According to the code in camel-filter-driver.c:pipe_to_system(),
> > Evolution does in fact re-read the message from the piped program and
> > replaces the pre-filtered message with it.
>
> I'll have to admit, when I sent my first message, I had not looked at
> the code but I had done some testing which led me to the conclusion that
> the piped message is not making its way back to Evolution.
>
> Here's what I did:
>
> I created a message filter for outgoing messages.  The filter used the
> pipe to command to send the message to klammail.  I then told the filter
> under all conditions to save the message to a unique folder.
>
> To test this, I created a message and attached various versions of the
> eicar test virus (eicar.com, eicar.txt, and eicar_com.zip).  I sent the
> message.  The message I find in the folder (and the one that gets sent)
> does not contain the klammail inserted headers.
>
> If I perform the same test using KMail (pipe the message to klammail,
> save it to a special folder), I find the following inserted into the
> header:
>
>    X-Virus-Status: Yes
>    X-Virus-Checker: Scanned by KlamAV 0.37 on Linux (virus-found
>    Eicar-Test-Signature); Mon, 19 Jun 2006 11:57:52 -0400
>
> How do you explain the behavioural differences?
>
> Regards,
> Tom
>   
-- 
Jeffrey Stedfast
Evolution Hacker - Novell, Inc.
fejj novell com  - www.novell.com