Building email threads from unix mailboxes

> Is it enough to get the In-Reply-To header for each message and build a

dictionary of { Message-ID: message } pairs? Or is it more complicated
than that?

To be RFC 2822 compliant, In-Reply-To and References are sufficient.
Other clients may add more headers, and not all clients are RFC 2822
compliant.

- Josiah

Thanks.

Is the References header a running list of all the In-Reply-To headers
so far in the thread?

Jed Parsons wrote:

Is the References header a running list of all the In-Reply-To headers
so far in the thread?

It depends on the service. Some only keep a few of the last references,
some only one, some retain the full list from the very beginning (at
least as far as the RFC will allow).

Probably if you wanted to handle robust threading, you'd want to go by
In-Reply-To and References, backtracking manually (rather than relying
on any given References list to be complete), and then, for systems like
mail-to-news gateways which may break the In-Reply-To/References chain,
group by similar subjects posted around the same time.

--
__ Erik Max Francis && ma*@alcyone.com && http://www.alcyone.com/max/
/ \ San Jose, CA, USA && 37 20 N 121 53 W && AIM erikmaxfrancis
\__/ Love is the triumph of imagination over intelligence.
-- H.L. Mencken

On Oct 19, 2004, at 11:24 AM, Jed Parsons wrote:

What headers to I have to know about to build thread trees from Unix
mailboxes?

Is it enough to get the In-Reply-To header for each message and build a
dictionary of { Message-ID: message } pairs? Or is it more complicated
than that?
<http://www.jwz.org/doc/threading.html> has a good write-up about the
threading algorithm used by Netscape Mail and News 2.0 and 3.0, and
Grendel (<http://www.mozilla.org/projects/grendel/>). Jamie Zawinski
was responsible for the design of Netscape Mail and News 2.0 and 3.0.
If there isn't already a module to do this (and apologies if there is
one and I don't know about it), are the current tools of choice the
'email' and 'mailbox' modules? (And I guess I'd want to use the mime
decoding tools in 'email' to deal with messages that come with
attachments or html or other stuff.)
A.M. Kuchling has made a Python implementation of JWZ's algorithm
available at <http://www.amk.ca/python/code/jwz>.
Thanks for any tips,

Jed

Regards,

Mark Rowe
<http://bdash.net.nz/>

Awesome! Thanks so much.

j

In article <10**********************@z14g2000cwz.googlegroups .com>,
"Jed Parsons" <je**@ilm.com> wrote:

[Apologies if this arrives twice; my news server appears to have dropped
my initial post.]

The best algorithm for threading email messages I've seen is Jamie
Zawinski's. It's at:

http://www.jwz.org/doc/threading.html

I have an implementation in Python that's linked from:

http://www.mondoinfo.com/blog/C18226...988/index.html

Regards,
Matt