[Mailman-Users] Attachments

[Roger Espiritu]

Hi Richard:

I noticed that you were working on pdf issues/attachments etc... I was
wondering if you can help me in regards to attachments in the archive
lists...

On my content filtering settings, these are my current settings:
 
filter content on "no"
nothing on "filter mime types" 
pass mime types:
multipart/mixed
multipart/alternative
text/plain
convert html to plaintext: "no"
filter action: "preserve"

How do I have all attachments be accepted by the ListServ. I noticed
that sometimes in the archive sections, excel files or word docs cannot
be displayed and sometimes some messages with attachments can be viewed.
Is there something that I need to set to have ALL attachments be
accepted?

I appreciate your help but if someone does know the solution to
this...please help?!!

 -Roger



-----Original Message-----
From: Richard Barrett [mailto:r.barrett at openinfo.co.uk] 
Sent: Wednesday, November 05, 2003 8:02 AM
To: tech-mail at prupref.com
Cc: mailman-users at python.org
Subject: Re: [Mailman-Users] Attachments - PDF Issue

Chris

On Wednesday, November 5, 2003, at 03:25  pm, McKeever Chris wrote:

>
>
> On Wed, 5 Nov 2003 08:03 , Richard Barrett <r.barrett at ftel.co.uk>
sent:
>
>>
>> On Wednesday, November 5, 2003, at 12:58  am, McKeever Chris wrote:
>>
>>> I am trying to figure out how to resolve PDF's being converted into
>>> .obj extensions in the archive (they get sent to list subscribers
>>> still in .pdf
>>> form).
>>>
>>> I have tried sending from multiple clients (yahoo, our webmail, and
>>> Outlook Express) and I keep getting this issue.
>>>
>>> Ultimately, I would like the .obj to just be .PDF so end-users do
not
>>> get the 'open with' dialog box.
>>>
>>> Here is a sample archived message:
>>>
>>> A non-text attachment was scrubbed...
>>> Name: x-4.pdf
>>> Type: application/octet-stream
>>
>> You problem is the MIME type the MUA attaching the PDF file is
putting
>> in the Content-type. It appears to be saying it is
>> application/octet-stream rather than application/pdf
>>
>> It is questionable practice to rely on the file suffix to ascribe
>> content type and Mailman does not appear to so do, which is why a
file
>> originally called xxx.pdf becomes yyy.obj
>>
>> If the Content-type is corrected I think you will find Mailman 
>> extracts
>> the attachments into .pdf suffixed filenames.
>>
>>> Size: 55233 bytes
>>> Desc: not available
>>> Url :
>>> http://list.domain.com/pipermail/filefax/attachments/20031104/
>>> ea6ccb1e/x-4-0001.obj
>>>
>
> I understand that it is questionable, but there isnt _anything_ I can 
> do on the MUA side, and seeing that it isnt a public list/archive
> that really only gets .pdf files (sent via a fax server) I dont think 
> it is all that dangerous.
>
> I need to figure out how to have the code rely on the extension over 
> the content type.  If anyone has knowledge of modifying the
> Scrubber.py (I think that is where it is) to allow this, please let me

> know
>

Attached is a crude little patch that may solve your problem. You can 
apply the patch from  within either the MM build directory and rerun 
./configure and make install, or to the $prefix directory, using the 
following command:

     patch -p1 < path-to-patchfile

Particularly if you decide to apply the patch to the MM $prefix 
directory then do first take a backup of 
$prefix/Mailman/Handlers/Scrubber.py in case the patched code doesn't 
work and you need to revert.

You will need to restart mailmanctl after applying the patch.

Let me know how you get on.

Richard