Using gnus-list-identifiers

Using gnus-list-identifiers

[Rene H. Larsen]

I'm subscribed to a few mailing lists which use a list identifier in
the subject header (e.g. "[Masq]") and I'd like Gnus to hide it for
me.  I discovered the gnus-list-identifiers while browsing through the 
Gnus customization and it was just what I was looking for, or so I
thought.

The problem is that this identifier hiding only works if I set
gnus-list-identifiers globally.  What I'd like to do is to set it as a 
group-local variable so that it is only applied to those lists that
actually use identifiers.  I checked with C-h v that the variable is
set correctly in the summary buffer but it is ignored in both the
summary display and by the corresponding article treatment function.

This makes list identifier hiding *a lot* less useful than it could
be.  I don't like having some catch-all regexp applied to the subject
of all the articles that I read.  Instead, I'd like to be able to
specify exactly the right regexp on just the groups that need it.

Any suggestions?
-- 
#!/usr/bin/perl -w
for(0..12){$b[$_]="|".($_==6?"-":" ")x78}for(0..78){substr($b[-
(sin($_*3.1415/22)-1)*6.499],$_,1)="*"}print join "\n",@b,"";

Re: Using gnus-list-identifiers

[Karl Kleinpaste]

gnus-list-identifiers isn't a variable which is applied on a per-group
basis for display purpoess.; it is a list of identifiers which should
be removed from mail *when the mail is stored in your mail groups*.
That is, it doesn't work on mail you've already got; it works on mail
that arrives after you set the variable.  It actually modifies the
mail content before storing to disc.

It is indeed a "global" variable, in that its effect is seen outside
any particular mail group.

`C-h i' -> Gnus -> Select Methods -> Getting Mail -> Washing Mail

Re: Using gnus-list-identifiers

[Rene H. Larsen]

Karl Kleinpaste <karl@justresearch.com> writes:

> gnus-list-identifiers isn't a variable which is applied on a per-group
> basis for display purpoess.; it is a list of identifiers which should
> be removed from mail *when the mail is stored in your mail groups*.

That may be true, but it is hard to tell from the available
documentation; gnus-list-identifiers isn't mentioned anywhere in the
Gnus manual and its docstring simply states:

     Regexp that matches list identifiers to be removed from subject.
     This can also be a list of regexps.

> That is, it doesn't work on mail you've already got; it works on mail
> that arrives after you set the variable.  It actually modifies the
> mail content before storing to disc.

That is not my experience from my tinkering with it.  The variable
shows up in M-x customize RET gnus-summary-format RET as a setting
that modifies the subject in the summary listing.  Also, if you set
gnus-treat-strip-list-identifiers to 'head, article treatment will
remove the prefix from article subjects as they are displayed.

All this works just fine if I set gnus-list-identifiers globally (with
setq or set-variable).  What I would like is a way to set this on a
per-group basis so that I don't risk hiding parts of the subject in
other, unrelated, news group and mail folders.  What I was pointing
out in my original message was that this does not seem to work if you
just set it as a group-local variable (through the group properties).

If this is not currently possible, please regard this posting as a
feature request.

> It is indeed a "global" variable, in that its effect is seen outside
> any particular mail group.

However, I would like for it to apply to only a select group of
folders.

> `C-h i' -> Gnus -> Select Methods -> Getting Mail -> Washing Mail

This won't work for my setup, though, since I read all my mail from an 
Imap server and do my mail splitting with procmail at the server end.
-- 
#!/usr/bin/perl -w
for(0..12){$b[$_]="|".($_==6?"-":" ")x78}for(0..78){substr($b[-
(sin($_*3.1415/22)-1)*6.499],$_,1)="*"}print join "\n",@b,"";

Re: Using gnus-list-identifiers

[Karl Kleinpaste]

"Rene H. Larsen" <renehl@post1.tele.dk> writes:
> ...but it is hard to tell from the available
> documentation; gnus-list-identifiers isn't mentioned anywhere in the
> Gnus manual

First, I made a mistake in what I showed you before, because I simply
searched the manual index for "list-identifiers" and relayed how to
reach that, which happens to be the entry for nnmail-list-identifiers,
which has been part of Gnus for quite some time.  But that was in my
0.95 at home, not the 0.96 I'm running at the office, which I have now
checked.

Second, gnus-list-identifiers *is* in the manual.  Since you have the
variable (which is evidently new for 0.96, it not being found in the
0.95 manual), then you have the manual to describe it.  Do this, and
see the discussion of `W W l' there:

`C-h i' -> Gnus -> The Summary Buffer -> Article Treatment -> Article Hiding

That's an explicit road map to the page that discusses the variable.

> That is not my experience from my tinkering with it.  

That's because I was discussing nnmail-list-identifiers, which is
different from what you asked.  My apologies.

However, you have manifestly not checked the manual yourself.  Again,
it is in the section named above.

Re: Using gnus-list-identifiers

[Simon Josefsson]

Rene H. Larsen <renehl@post1.tele.dk> writes:

> This makes list identifier hiding *a lot* less useful than it could
> be.  I don't like having some catch-all regexp applied to the subject
> of all the articles that I read.  Instead, I'd like to be able to
> specify exactly the right regexp on just the groups that need it.

What about using list of regexp's? I use

(setq gnus-list-identifiers '("\\[zebra [0-9]+\\]" "\\[gtk-list\\]"
			      "LAM:" "linux-ipsec:" "DENTS:"
			      "\\[linux-security\\]" "FVWM:"))

I think the removal of list identifiers are done before summary local
variables are bound.

Re: Using gnus-list-identifiers

[Rene H. Larsen]

Karl Kleinpaste <karl@justresearch.com> writes:

> "Rene H. Larsen" <renehl@post1.tele.dk> writes:
> > ...but it is hard to tell from the available
> > documentation; gnus-list-identifiers isn't mentioned anywhere in the
> > Gnus manual
> 
> First, I made a mistake in what I showed you before, because I simply
> searched the manual index for "list-identifiers" and relayed how to
> reach that, which happens to be the entry for nnmail-list-identifiers,
> which has been part of Gnus for quite some time.  But that was in my
> 0.95 at home, not the 0.96 I'm running at the office, which I have now
> checked.

Fair enough.

> Second, gnus-list-identifiers *is* in the manual.  Since you have the
> variable (which is evidently new for 0.96, it not being found in the
> 0.95 manual), then you have the manual to describe it.  Do this, and
> see the discussion of `W W l' there:
> 
> `C-h i' -> Gnus -> The Summary Buffer -> Article Treatment -> Article Hiding
> 
> That's an explicit road map to the page that discusses the variable.

Thank you for your patience.  As it turns out, due to a faulty
infopath setting, the Emacs Info reader picked up the manual for the
version of Gnus shipped with my XEmacs 20.4 (Gnus 5.4.66, I believe)
and not the manual shipped with pGnus, and I didn't notice it until
today when I tried to follow your directions and didn't find what you
were claiming to be there.  It is all fixed now and yes, it is in the
manual.  No wonder I always found the Gnus manual a bit lacking...

> However, you have manifestly not checked the manual yourself.  Again,
> it is in the section named above.

Well, I /did/ check the manual.  Just the wrong version of it...
-- 
#!/usr/bin/perl -w
for(0..12){$b[$_]="|".($_==6?"-":" ")x78}for(0..78){substr($b[-
(sin($_*3.1415/22)-1)*6.499],$_,1)="*"}print join "\n",@b,"";

Re: Using gnus-list-identifiers

[Rene H. Larsen]

Simon Josefsson <jas@pdc.kth.se> writes:

> Rene H. Larsen <renehl@post1.tele.dk> writes:
> 
> > This makes list identifier hiding *a lot* less useful than it could
> > be.  I don't like having some catch-all regexp applied to the subject
> > of all the articles that I read.  Instead, I'd like to be able to
> > specify exactly the right regexp on just the groups that need it.
> 
> What about using list of regexp's? I use

Yes, I'll probably end up doing just that.  I was just hoping that
there was a way to set it on a per-group basis, to minimize the risk
of hiding something that shouldn't be hidden.

> I think the removal of list identifiers are done before summary local
> variables are bound.

That sounds reasonable.  I had more-or-less come to the same
conclusion myself.
-- 
#!/usr/bin/perl -w
for(0..12){$b[$_]="|".($_==6?"-":" ")x78}for(0..78){substr($b[-
(sin($_*3.1415/22)-1)*6.499],$_,1)="*"}print join "\n",@b,"";

Re: Using gnus-list-identifiers

[Per Abrahamsen]

"Rene H. Larsen" <renehl@post1.tele.dk> writes:

> Yes, I'll probably end up doing just that.  I was just hoping that
> there was a way to set it on a per-group basis, to minimize the risk
> of hiding something that shouldn't be hidden.

You can set group specific variables as group parameters.  Type M-C-g
when reading the folder, and look for the "Variables:" entry near the
bottom of the buffer.

Re: Using gnus-list-identifiers

[Rene H. Larsen]

Per Abrahamsen <abraham@dina.kvl.dk> writes:

> You can set group specific variables as group parameters.  Type M-C-g
> when reading the folder, and look for the "Variables:" entry near the
> bottom of the buffer.

Yes, I know about group-local variables, but for some reason this
doesn't work for gnus-list-identifiers.  Simon Josefsson said earlier
in the thread that he thinks that the removal of list identifiers are
done before summary local variables are bound.
-- 
#!/usr/bin/perl -w
for(0..12){$b[$_]="|".($_==6?"-":" ")x78}for(0..78){substr($b[-
(sin($_*3.1415/22)-1)*6.499],$_,1)="*"}print join "\n",@b,"";