[Mimedefang] Interesting anti-HTML filter side-effect

[G. Roderick Singleton]

On Sat, 2003-09-20 at 23:15, David F. Skoll wrote:
> On Sat, 20 Sep 2003, G. Roderick Singleton wrote:
> 
> > Care to post the rules? I for one am struggling to find a solution in a
> > hurry.
> 
> The filter looks something like this (edited to protect the innocent/guilty).
> Not all functions are included; you'll have to figure out how to implement
> the missing stuff.  This (very complicated) filter_begin function sets
> $HTMLOK to 1 if we'll accept HTML, or 0 otherwise.  Later on, in filter,
> if we find a text/html part and $HTMLOK is 0, we bounce the message.
> 
> Note:  Believe it or not, this is a *simplified* version of the actual
> filter. :-)
 [snip]

Thanks very much. Still not quite what I'm after. I suspect that my perl
skills et cetera are not up to the task but what I want is to bounce any
and all html messages even those that are masqueraded with funny lines.
I don't want to reject messages with attachments. Now I see that
spamassassin detects spam messages yet the default filter doesn't seem
to bounce them based on the value returned. Like I said my perl skills
are rudimentary at best and so far reading the manpages hasn't helped.
So ...

Can anyone offer a snippet that will reject ALL html encoded messages? I
used to use the linux filter until Thursday when I switched back to the
default. I liked what the linux example did under normal conditions but
it proved inadequate for the onslaught whereas the default was able to
cope. In the meantime I'll get out my trusty "camel" and get reading.
Maybe I'll understand the answers when they arrive :-)
-- 
Gerry Roderick Singleton <gerry at pathtech.org>
416-452-4583