[Mailman-Developers] Mailman archive and MIME

[Harald Meland]

[Tomas Fasth]

> From: Harald Meland <Harald.Meland@usit.uio.no>
> 
> >I haven't tried this, but wouldn't simply squeezing the mail text
> >through the .unmimify() method of the `mimify' standard Python module
> >before formatting it into HTML work (for the QP/base64 MIME issue,
> >multipart messages is a whole other can of worms)?

In the general case it wouldn't, unless there also was some kind of
8bit-character-set -> HTML representation thingy, i.e. something to
convert the ISO-8859-1 character "å" into the HTML element "&aring;".
Unless there is some standard Python module for doing this, I opt for
leaving it all out until after 1.0 is out -- we might as well do this
Right from the start.

> That's probably one way to go. A good thing trying to use existing code. I'm
> not sure what exactly .unmimify() does, though.

It converts quoted-printable (and optionally base64) encoded parts to
8bit parts.  I does not format it as HTML, though.  It also does
nothing with regard to "Content-Type:"-type things.

> The minimum requirements for messages written in western european
> languages would be for the archiver's email-to-html conversion to
> honor the MIME header encoding (RFC 2047) and the basics of MIME
> body format (RFC 2045).

Do include all of RFC2045 when you say "the basics of MIME body
format"?  I really can't see the benefit of e.g. parsing
multipart/alternative messages and such when formatting them as HTML.

> >Is there a problem with what headers are retained in the mbox-format
> >version of the archive (i.e. are the headers needed for proper MIME
> >decoding stripped from the archive)?
> 
> Not as far as I can see.  Both Content-* and MIME-Version headers
> seem to be retained.

Duh, just me looking in the wrong place -- indeed, the
listname.mbox/listname.mbox archive does include all headers.  BTW, is
there any way of configuring which headers (if any) should be removed
when generating the static downloadable text files?
-- 
Harald