This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Update on bugzilla stuff
- From: Daniel Berlin <dan at cgsoftware dot com>
- To: Phil Edwards <pedwards at disaster dot jaj dot com>
- Cc: Wolfgang Bangerth <wolfgang dot bangerth at iwr dot uni-heidelberg dot de>, <gcc at gcc dot gnu dot org>
- Date: Thu, 3 Jan 2002 23:45:26 -0500 (EST)
- Subject: Re: Update on bugzilla stuff
On Wed, 2 Jan 2002, Phil Edwards wrote:
> On Wed, Jan 02, 2002 at 10:02:35AM -0500, Daniel Berlin wrote:
> > This is because it was only an email followup.
> > I've not yet handled comments that appear *only* as email, as it's tricky,
> > because of how GNATS intermixes real email messages with status changes.
> > I've started to code it, and it seems to work okay, but i haven't made
> > the results appear on that server yet.
>
> Some weird things are also happening to comments made as part of a change,
> not just as an email followup.
>
> I just pulled up a PR I'm working on, 1853. Take a look at comment #2.
> The changes by Benjmain only cover down through the line beginning
> with "Still an issue." Then, about six months later, there was a third
> change from me. That begins at the "Responsible-Changed-" lines, and the
> accompanying comment text is mine.
This has been fixed.
It should now properly handled *-Changed.
In addition, it should properly add followup emails to the list of
comments.
In addition to that, it should decode uuencoded attachments appearing in
either the How-To-Repeat or Fix (attachments meaning multiple uuencoded
attachments are handled right),
replacing each with a "See attachment #<id> for <filename>", which will be
a hyperlink to the attachment.
It might miss replacing a little bit of the text.
This was done at the expense of recognizing more uuencoded attachments.
Some of them have end lines that really are in an unformatted field at the
end of the PR for some reason (I got bit by this).
So we'll still decode them properly (because the decoder will skip the
lines in between until it hits the right lines, rather than abort), but we
might leave a line or two in the how-to-repeat/fix..
This can't really be fixed without manually putting the end lines back
where they belong.
It won't decode attachments appearing in the followup emails, i'm
working on that tommorrow, though i don't expect it to take long.
Whee.
>
>
> Phil
>
>