Patch: Allow nnrss to be verbose

Patch: Allow nnrss to be verbose

[Mark Plaksin]

This patch adds nnrss-verbose.  When t, nnrss writes to the minibuffer when
it requests a feed.

Sometimes fetching an RSS feed takes too long and I want to interrupt it.
Without this patch, I can't tell which feed it's trying to fetch.  I also
can't tell that Gnus is trying to fetch an RSS feed--it just looks stuck
and I wonder what I'm interrupting when I hit C-g :)

--- nnrss.el	2005-01-30 08:16:34.000000000 -0500
+++ ../../gnus/lisp/nnrss.el	2005-02-19 13:17:00.000000000 -0500
@@ -66,6 +66,9 @@
 
 (defvar nnrss-use-local nil)
 
+(defvar nnrss-verbose t
+  "Write messages when requesting group.")
+
 (defvar nnrss-description-field 'X-Gnus-Description
   "Field name used for DESCRIPTION.
 To use the description in headers, put this name into `nnmail-extra-headers'.")
@@ -145,6 +148,8 @@
   'nov)
 
 (deffoo nnrss-request-group (group &optional server dont-check)
+  (if nnrss-verbose
+      (message (concat "nnrss requesting " group "...")))
   (setq group (nnrss-decode-group-name group))
   (nnrss-possibly-change-group group server)
   (if dont-check
@@ -154,7 +159,10 @@
     (nnheader-insert
      "211 %d %d %d %s\n" nnrss-group-max nnrss-group-min nnrss-group-max
      (prin1-to-string group)
-     t)))
+     t))
+  (if nnrss-verbose
+      (message (concat "nnrss done with " group "."))
+    t))
 
 (deffoo nnrss-close-group (group &optional server)
   t)

Re: Patch: Allow nnrss to be verbose

[Katsumi Yamaoka]

>>>>> In <m3y8dko1db.fsf@water.tss.usg.edu> 
>>>>>	Mark Plaksin <happy@usg.edu> wrote:

> This patch adds nnrss-verbose.  When t, nnrss writes to the minibuffer when
> it requests a feed.

> Sometimes fetching an RSS feed takes too long and I want to interrupt it.
> Without this patch, I can't tell which feed it's trying to fetch.  I also
> can't tell that Gnus is trying to fetch an RSS feed--it just looks stuck
> and I wonder what I'm interrupting when I hit C-g :)

Thanks.  It seems to be useful.  Is there any reason we're
unable to make nnrss always verbose? (isn't the nnrss-verbose
variable needless?)

Re: Patch: Allow nnrss to be verbose

[Mark Plaksin]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

> Thanks.  It seems to be useful.  Is there any reason we're
> unable to make nnrss always verbose? (isn't the nnrss-verbose
> variable needless?)

*I* always want it verbose.  I figured somebody would want it to be silent
:)

Re: Patch: Allow nnrss to be verbose

[Katsumi Yamaoka]

>>>>> In <m37jl3wlzu.fsf@water.tss.usg.edu> 
>>>>>	Mark Plaksin <happy@usg.edu> wrote:
> Katsumi Yamaoka <yamaoka@jpl.org> writes:

> > Thanks.  It seems to be useful.  Is there any reason we're
> > unable to make nnrss always verbose? (isn't the nnrss-verbose
> > variable needless?)

> *I* always want it verbose.  I figured somebody would want it to be silent
> :)

I see.  Nobody isn't able to assert it should always be verbose.
I've committed your change in the Gnus CVS trunk.  Thank you.

Re: Patch: Allow nnrss to be verbose

[Arne Jørgensen]

[-- Attachment #1: Type: text/plain, Size: 336 bytes --]

Mark Plaksin <happy@usg.edu> writes:

> This patch adds nnrss-verbose.  When t, nnrss writes to the minibuffer when
> it requests a feed.

Nice, but I think changing it to

      Requesting nnrss:Foo...
      Requesting nnrss:Foo...done

would be more emacsish. Patch attached.

Kind regards,
-- 
Arne Jørgensen <http://arnested.dk/>


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: nnrss.patch --]
[-- Type: text/x-patch, Size: 960 bytes --]

Index: lisp/nnrss.el
===================================================================
RCS file: /usr/local/cvsroot/gnus/lisp/nnrss.el,v
retrieving revision 7.25
diff -u -p -r7.25 nnrss.el
--- lisp/nnrss.el	20 Feb 2005 11:42:56 -0000	7.25
+++ lisp/nnrss.el	21 Feb 2005 13:49:10 -0000
@@ -149,7 +149,7 @@ ARTICLE is the article number of the cur
 
 (deffoo nnrss-request-group (group &optional server dont-check)
   (if nnrss-verbose
-      (message (concat "nnrss requesting " group "...")))
+      (message (concat "Requesting nnrss:" group "...")))
   (setq group (nnrss-decode-group-name group))
   (nnrss-possibly-change-group group server)
   (prog1
@@ -162,7 +162,7 @@ ARTICLE is the article number of the cur
 	 (prin1-to-string group)
 	 t))
     (when nnrss-verbose
-      (message (concat "nnrss done with " group ".")))))
+      (message (concat "Requesting nnrss:" group "...done")))))
 
 (deffoo nnrss-close-group (group &optional server)
   t)

Re: Patch: Allow nnrss to be verbose

[Reiner Steib]

On Mon, Feb 21 2005, Arne Jørgensen wrote:

>       Requesting nnrss:Foo...
>       Requesting nnrss:Foo...done
>
> would be more emacsish.

ACK.

Additionally, I would suggest to use `nnheader-message' (i.e. use the
variable `gnus-verbose-backends') instead of introducing
`nnrss-verbose'.

As I don't use nnrss, I don't know which LEVEL would be suitable here.

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Patch: Allow nnrss to be verbose

[Arne Jørgensen]

Reiner Steib <reinersteib+gmane@imap.cc> writes:

> On Mon, Feb 21 2005, Arne Jørgensen wrote:
>
>>       Requesting nnrss:Foo...
>>       Requesting nnrss:Foo...done
>>
>> would be more emacsish.
>
> ACK.
>
> Additionally, I would suggest to use `nnheader-message' (i.e. use the
> variable `gnus-verbose-backends') instead of introducing
> `nnrss-verbose'.

Didn't know about `nnheader-message' or `gnus-verbose-backends' before
now, but I agree.

> As I don't use nnrss, I don't know which LEVEL would be suitable here.

I use nnrss, but have no clue about which level to use either. What
are other backends using for similar messages?

Kind regards,
-- 
Arne Jørgensen <http://arnested.dk/>

Re: Patch: Allow nnrss to be verbose

[Arne Jørgensen]

[-- Attachment #1: Type: text/plain, Size: 373 bytes --]

Arne Jørgensen <arne@arnested.dk> writes:

> I use nnrss, but have no clue about which level to use either. What
> are other backends using for similar messages?

5 or 6 it seems.

And it looks like the normal thing to write would be

    nnrss: Requesting Foo...
    nnrss: Requesting Foo...done

Patch attached.

Kind regards,
-- 
Arne Jørgensen <http://arnested.dk/>


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: nnrss.patch --]
[-- Type: text/x-patch, Size: 1350 bytes --]

Index: lisp/nnrss.el
===================================================================
RCS file: /usr/local/cvsroot/gnus/lisp/nnrss.el,v
retrieving revision 7.25
diff -u -p -r7.25 nnrss.el
--- lisp/nnrss.el	20 Feb 2005 11:42:56 -0000	7.25
+++ lisp/nnrss.el	21 Feb 2005 14:53:11 -0000
@@ -66,9 +66,6 @@
 
 (defvar nnrss-use-local nil)
 
-(defvar nnrss-verbose t
-  "Write messages when requesting group.")
-
 (defvar nnrss-description-field 'X-Gnus-Description
   "Field name used for DESCRIPTION.
 To use the description in headers, put this name into `nnmail-extra-headers'.")
@@ -148,8 +145,7 @@ ARTICLE is the article number of the cur
   'nov)
 
 (deffoo nnrss-request-group (group &optional server dont-check)
-  (if nnrss-verbose
-      (message (concat "nnrss requesting " group "...")))
+  (nnheader-message 6 (concat "nnrss: Requesting " group "..."))
   (setq group (nnrss-decode-group-name group))
   (nnrss-possibly-change-group group server)
   (prog1
@@ -161,8 +157,7 @@ ARTICLE is the article number of the cur
 	 "211 %d %d %d %s\n" nnrss-group-max nnrss-group-min nnrss-group-max
 	 (prin1-to-string group)
 	 t))
-    (when nnrss-verbose
-      (message (concat "nnrss done with " group ".")))))
+    (nnheader-message 6 (concat "nnrss: Requesting " group "...done"))))
 
 (deffoo nnrss-close-group (group &optional server)
   t)

Re: Patch: Allow nnrss to be verbose

[Reiner Steib]

On Mon, Feb 21 2005, Arne Jørgensen wrote:

> Arne Jørgensen <arne@arnested.dk> writes:
>
>> I use nnrss, but have no clue about which level to use either. What
>> are other backends using for similar messages?
>
> 5 or 6 it seems.

As this is below the default of `gnus-verbose-backends' (7), it is
okay.

> And it looks like the normal thing to write would be
>
>     nnrss: Requesting Foo...
>     nnrss: Requesting Foo...done
>
> Patch attached.

Committed (on the trunk), thanks.

I think such changes should go in the v5-10 branch, too.  But applying
the patch (cvs diff -r7.24 -r7.26) fails on v5-10.

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Patch: Allow nnrss to be verbose

[Katsumi Yamaoka]

>>>>> In <v93bvpyk3d.fsf@marauder.physik.uni-ulm.de> Reiner Steib wrote:

> Committed (on the trunk), thanks.

Thanks for putting it right.

> I think such changes should go in the v5-10 branch, too.  But applying
> the patch (cvs diff -r7.24 -r7.26) fails on v5-10.


I can make it support multilingual text as in the trunk (it
includes the verbose patch, of course), although it might not be
a bugfix.  Am I able to do it?  Actually, nnrss in the v5-10
branch cannot be used at least for Japanese sites, for example:

http://www3.asahi.com/rss/index.rdf

P.S.
Though, in reality, I'm neither interested in nnrss nor using it
because it only shows subjects, links and possibly brief texts.
nnshimbun provided by emacs-w3m is much better for me.

Re: Patch: Allow nnrss to be verbose

[Katsumi Yamaoka]

>>>>> In <b9ybradv84l.fsf@jpl.org> Katsumi Yamaoka wrote:

>> I think such changes should go in the v5-10 branch, too.  But applying
>> the patch (cvs diff -r7.24 -r7.26) fails on v5-10.

> I can make it support multilingual text as in the trunk (it
> includes the verbose patch, of course), although it might not be
> a bugfix.  Am I able to do it?

Ouch!  It requires an overall replacing of rfc2047.el with the
one in the trunk.  The main load to me is to write the ChangeLog,
though.

Fixes for nnrss.el (and rfc2047.el) in the v5-10 branch (was: Patch: Allow nnrss to be verbose)

[Reiner Steib]

On Mon, Feb 21 2005, Katsumi Yamaoka wrote:

>>>>>> In <v93bvpyk3d.fsf@marauder.physik.uni-ulm.de> Reiner Steib wrote:
>> I think such changes should go in the v5-10 branch, too.  But applying
>> the patch (cvs diff -r7.24 -r7.26) fails on v5-10.
>
> I can make it support multilingual text as in the trunk (it
> includes the verbose patch, of course), although it might not be
> a bugfix.  Am I able to do it?  Actually, nnrss in the v5-10
> branch cannot be used at least for Japanese sites, for example:
>
> http://www3.asahi.com/rss/index.rdf
[...]

On Tue, Feb 22 2005, Katsumi Yamaoka wrote:

> Ouch!  It requires an overall replacing of rfc2047.el with the
> one in the trunk.  The main load to me is to write the ChangeLog,
> though.

Katsumi told me that he will try to look into this after releasing the
next version of emacs-w3m.

But if anyone else want to work on it, here are some remarks when
trying to sync `rfc2047.el' and `nnrss.el':

- It seems that nobody has merged the bugfixes from the trunk yet.  We
  should surely do this[1].  Jesper Harder has committed most of them
  on the trunk, but probably he didn't have time for merging them to
  v5-10 (the stable branch) yet.

- We cannot apply all changes blindly (see: "cvs diff -r7.1 -r7.26
  nnrss.el" or better `C-u C-x v = nnrss.el RET 7.1 RET 7.26 RET')
  since we want to keep the v5-10 compatible with Emacs 20.7 and
  XEmacs 21.1.  (E.g. `make-hash-table' is not available in Emacs
  20.7.)

  OTOH, `nnrss.el' is comparable to `gnus-registry.el' and
  `spam-stat.el' where we already use `make-hash-table' there in
  v5-10.  Opinions?

Bye, Reiner.

[1] See a recent bug report from Gaetan Leurent on emacs-devel:
    http://thread.gmane.org/qlkoee2ernr.fsf%40clipper.ens.fr
    <news:qlkoee2ernr.fsf@clipper.ens.fr>
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: Fixes for nnrss.el (and rfc2047.el) in the v5-10 branch

[Katsumi Yamaoka]

>>>>> In <v93bveouya.fsf_-_@marauder.physik.uni-ulm.de> Reiner Steib wrote:

> Katsumi told me that he will try to look into this after releasing the
> next version of emacs-w3m.

I'm now working on merging nnrss.el into the v5-10 branch and I
noticed there are some German web sites using the windows-1252
encoding even if they declare that the encoding is iso-8859-1.
For instance, the following rss feed declares
encoding="ISO-8859-1" and uses characters in 0x80-0x9f which are
out of the iso-8859-1 specification.

http://www.laut.de/partner/allgemein/news.rdf

I'm groping for the way (which can be used with old Emacsen too)
to solve the problem.

[...]

>   OTOH, `nnrss.el' is comparable to `gnus-registry.el' and
>   `spam-stat.el' where we already use `make-hash-table' there in
>   v5-10.  Opinions?

I've changed nnrss.el for the v5-10 branch so that it may use
`gnus-make-hashtable'.  In addition, there are other problems
which I've fixed in the local system:

time-date.el uses `#x10000' and `#xffff' which cannot be read by
Emacs 20.

`rfc2047-decode' won't work in XEmacs because
`mm-subst-char-in-string' doesn't allow the fourth argument.

I will commit those fixes together with nnrss.el and others.

Regards,

Re: Fixes for nnrss.el (and rfc2047.el) in the v5-10 branch

[Katsumi Yamaoka]

I've committed the changes in both the Gnus trunk and the v5-10
branch.  The changes in the v5-10 branch are for only Gnus v5.11,
so please don't merge them back into the Gnus trunk.

>>>>> In <b9ymzseaw9m.fsf@jpl.org> Katsumi Yamaoka wrote:

> I'm now working on merging nnrss.el into the v5-10 branch and I
> noticed there are some German web sites using the windows-1252
> encoding even if they declare that the encoding is iso-8859-1.

I solved it by using windows-1252 if it is available, or making
a text/plain article instead of a MIME article for old Emacsen
which sometimes caused an error.

FYI: you can use the windows-1252 coding system even in Emacs 21
by using ARISAWA Akihiro's codepage-ex.el, which is available at:

http://www.nijino.com/ari/emacs/codepage-ex.html

I've put the following in my ~/.emacs file.

(if (and (not (featurep 'xemacs))
	 (= emacs-major-version 21)
	 (not (require 'code-pages nil t)))
    (progn
      (require 'codepage-ex)
      (mapc (lambda (cp) (codepage-ex-setup (car cp)))
	    (cp-ex-supported-codepages))
      (define-coding-system-alias 'windows-1250 'cp1250)
      (define-coding-system-alias 'windows-1251 'cp1251)
      (define-coding-system-alias 'windows-1252 'cp1252)
      (define-coding-system-alias 'windows-1253 'cp1253)
      (define-coding-system-alias 'windows-1254 'cp1254)
      (define-coding-system-alias 'windows-1255 'cp1255)
      (define-coding-system-alias 'windows-1257 'cp1257)))

> I will commit those fixes together with nnrss.el and others.

Done.