Re: [PATCH] nnmail-fancy-expiry-target

[Simon Josefsson <jas@extundo.com>]

"Davide G.M.Salvetti" <salve@debian.org> writes:

> [Nevin, I seem to have problems posting to ding.  Will you please check
> if this message also shows there, as it should?  Thanks.]
>
>>>>>> "Nevin" == Nevin Kapur <nevin@jhu.edu> writes:
>
> Nevin> The patch I posted yesterday will work around this by ignoring
> Nevin> such empty buffers and more generally, non-existent headers.
>
> Ok.  Yet there are two problems still.
>
> 1. If the target (nnimap) group where expired articles should go doesn't
>    already exists, the articles will be deleted instead (with no error
>    signaled to the user).

Does this patch fix it?

--- nnimap.el.~6.37.~	2002-07-10 18:19:30.000000000 +0200
+++ nnimap.el	2002-07-26 18:42:11.000000000 +0200
@@ -1296,6 +1296,8 @@
 (defun nnimap-expiry-target (arts group server)
   (unless (eq nnmail-expiry-target 'delete)
     (with-temp-buffer
+      (or (imap-mailbox-status group 'UIDVALIDITY)
+	  (imap-mailbox-create group))
       (dolist (art (gnus-uncompress-sequence arts))
 	(nnimap-request-article art group server  (current-buffer))
 	;; hints for optimization in `nnimap-request-accept-article'

> 2. If you have a large hole in the article number sequence (say, you
>    have an IMAP group with a three months old ticked article, then a
>    large gap of expired articles and then some new articles), the
>    expiring process takes for ever.  I guess that's because Gnus request
>    a lot of non existing articles from the IMAP server, even if it needs
>    not to.

Could you profile this to find out what is taking time?  It is
difficult to improve it otherwise.  It may be range calculations that
is taking time, then I'm afraid there may be little to do right now.