Re: limiting articles and tagging

[David Carlton]

On Wed, 02 Apr 2008 15:43:14 -0700, David Carlton <carlton@bactrian.org> said:

> On Sun, 30 Mar 2008 17:42:39 -0700, David Carlton <carlton@bactrian.org> said:
>> I'd like a way to mark certain articles in an nnmail group with a tag
>> (e.g. articles I want to respond to, articles where I'm waiting for a
>> specific response), and then to have a way to only see the matching
>> articles in that group.

> I looked at the existing marks; I don't think they'll do the trick,
> and unfortunately the marks mechanism doesn't seem like it's
> extensible at all.  Browsing through the other gnus-summary-limit
> functions, though, "gnus-summary-limit-to-extra" caught my eye.

> Just what are the consequences of adding a header to
> gnus-extra-headers?  I'm thinking that I could add in another header
> there ("X-Tag", say), write some simple commands to edit messages to
> add/modify the X-Tag: header in the current article, and write some
> wrappers to gnus-summary-limit-to-extra to pick out the articles that
> I want.

> Does that sound like it should work?  Are there any gotchas I should
> know about before I start tinkering with it?  I'm worried that there
> might be some index of the extra headers for articles that could cause
> problems if I go this route, or something like that.

> Also, gnus-summary-limit-to-extra doesn't look that complicated; could
> I write a function that does something similar to it but doesn't care
> about gnus-extra-headers?  gnus-summary-limit-to-extra calls
> gnus-summary-find-matching; when looking through the latter function,
> at first I was afraid that it depends on there being a
> narrowly-defined set of mail-header-XXX functions, but will
> mail-header-extra work with an arbitrary header, will gnus-data-header
> return all the headers?

I poked around some more; the bad news is that, no, gnus-data-header
only returns some of the headers (including ones in
gnus-extra-headers), but the good news is that now the code makes a
bit more sense to me, and I'm less worried about some sort of weird
cache being there that I'm not aware of.  So I think I'll try that
route when I have a few spare minutes, unless somebody knows why it's
a bad idea...

David Carlton
carlton@bactrian.org