I wrote: Rod> The first part of the patch makes the bbdb-extract-field Rod> function stop at the start of the real message (i.e. past any Rod> pseudo-headers such as "Archive-name:") instead of at the Rod> formal start of the message. Jamie writes: Jamie> I think this is a bad idea. Lots of people forward Jamie> messages such that their headers look like this: Jamie> From: Jamie Zawinski <[EMAIL PROTECTED]> Jamie> To: [EMAIL PROTECTED] (Rod Whitby) Jamie> Subject: fwd Jamie> From: some weenie Jamie> To: some other weenie Jamie> Subject: something else Jamie> the forwarded message body Good point. I didn't think of that. I certainly agree that you shouldn't break BBDB to take into account some other broken behaviour. I still want to get at that "Archive-name: " pseudo-header, though. I guess the only way for me to do this is to set fieldval to the message body when a magic field is named in the auto-notes-alist, and then use the user-supplied function that I added as the second part of my patch to extract the relevant part of the fieldval (i.e. extract out the "Archive-name: " pseudo-header). Jamie, do you intend to include the second part of the patch in BBDB distribution (the part that allows the STRING field of the auto-notes-alist record to be a string, integer or function) ? If so, I will hack some more on the extract-field function to return the message body contents when a magic field is requested (is "Body" a good name for the magic field ? Can you suggest a better name ?). Is this the way I should go about this ? Anyone have a better way to get at the message body ? -- --------------------------------------------- _--_|\ | Rod Whitby ([EMAIL PROTECTED]) / \ | Canon Information Systems Research Australia \_.--._/ | 1 Thomas Holt Drive, North Ryde, N.S.W., 2113. v |