Content-Id problem with Anomy Sanitizer

From: Kimmo Suominen (
Date: Sun 18 Jan 2004 - 23:04:51 GMT

  • Next message: APPANAH Ravi: "Problem removing file"


    I've been using Anomy Sanitizer for quite some time, and I'm also
    the maintainer of the pkgsrc package for it (

    According to the CHANGELOG.sanitizer.txt in version 1.61 a change
    was made that causes the Content-Id header to be interpreted as
    a (possible) filename for the attachment.

    The problem with this is that the Content-Id is often formed with
    the FQDN of the machine appended, such as "". When
    most people (me included) filter all ".com" attachments, then all
    messages from ".com" domains would also get dropped, when the
    sender uses an MUA adding Content-Id headers.

    I think the change is in bin/Anomy/ the line that

        $reader->ParseContentHeader("; _id=", "content-id", $header_log);

    What would you think about removing that call altogether (thus
    reverting this part of version 1.61)? Alternatively, it could be
    made optional, perhaps? (Although I think most people would need
    to disable this in practice.)

    + Kim

    hosted by