`C-c C-v' should keep citation prefix

`C-c C-v' should keep citation prefix

[Kai Großjohann]

Suppose I cite a message.  Suppose I get this:

> foo bar baz

Now further suppose I mark the word bar and hit C-c C-v.  The `>' will
be gone.  I think it would be a good idea if C-c C-v would keep it.

Thoughts?

kai
-- 
~/.signature: No such file or directory

Re: `C-c C-v' should keep citation prefix

[Steven E. Harris]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> I think it would be a good idea if C-c C-v would keep it.

I agree wholeheartedly. In fact, while quoting the line above, I had
to add the '>' to the beginning of the line after using C-c
C-v. Adding that leading '>' is part of "muscle memory" by now, but it
would be nice to not have to do it.

-- 
Steven E. Harris        :: seh@speakeasy.org
GnuPG                   :: 0x70248E67

Re: `C-c C-v' should keep citation prefix

[Simon Josefsson]

"Steven E. Harris" <sharris@speakeasy.org> writes:

>> I think it would be a good idea if C-c C-v would keep it.
> 
> I agree wholeheartedly. In fact, while quoting the line above, I had
> to add the '>' to the beginning of the line after using C-c C-v.

Does this do anything useful?

Index: message.el
===================================================================
RCS file: /usr/local/cvsroot/gnus/lisp/message.el,v
retrieving revision 6.83
diff -u -u -w -r6.83 message.el
--- message.el	2001/05/26 09:25:13	6.83
+++ message.el	2001/06/05 18:35:51
@@ -1786,7 +1786,13 @@
 (defun message-delete-not-region (beg end)
   "Delete everything in the body of the current message outside of the region."
   (interactive "r")
+  (let (citeprefix)
   (save-excursion
+      (goto-char beg)
+      ;; snarf citation prefix, if appropriate
+      (unless (eq (point) (progn (beginning-of-line) (point)))
+	(when (looking-at message-cite-prefix-regexp)
+	  (setq citeprefix (match-string 0))))
     (goto-char end)
     (delete-region (point) (if (not (message-goto-signature))
 			       (point)
@@ -1796,7 +1802,9 @@
     (goto-char beg)
     (delete-region beg (progn (message-goto-body)
 			      (forward-line 2)
-			      (point))))
+				(point)))
+      (when citeprefix
+	(insert citeprefix))))
   (when (message-goto-signature)
     (forward-line -2)))
 

Re: `C-c C-v' should keep citation prefix

[Steven E. Harris]

Simon Josefsson <jas@extundo.com> writes:

> Does this do anything useful?

When I try to use the patched version, I get:

,----
| Symbol's value as variable is void: message-cite-prefix-regexp
`----

That may be because I'm running Gnus v5.8.8, though.

-- 
Steven E. Harris        :: seh@speakeasy.org
GnuPG                   :: 0x70248E67

Re: `C-c C-v' should keep citation prefix

[Simon Josefsson]

"Steven E. Harris" <sharris@speakeasy.org> writes:

> When I try to use the patched version, I get:
> 
> ,----
>| Symbol's value as variable is void: message-cite-prefix-regexp
> `----
> 
> That may be because I'm running Gnus v5.8.8, though.

Yes.  I committed my patch, so if you try OGnus C-c C-v should keep
citation prefixes.

Re: `C-c C-v' should keep citation prefix

[Josh Huber]

[...]

> >| Symbol's value as variable is void: message-cite-prefix-regexp

I'm getting this error with the latest oort gnus (just cvs updated a
couple minutes ago) whenever I view a message...

Signaling: (void-variable message-cite-prefix-regexp)
  (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +==")
  eval((concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +=="))
  custom-initialize-reset(gnus-supercite-regexp (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +=="))
  custom-declare-variable(gnus-supercite-regexp (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +==") "*Regexp matching normal Supercite attribution lines.\nThe first grouping must match prefixes added by other packages." :group gnus-cite :type regexp)
  load-internal("gnus-cite" nil nil nil nil nil)
  load("gnus-cite" nil nil nil)
  gnus-article-highlight-citation()
  gnus-treat-article(nil 1 1)
  gnus-display-mime()
  gnus-article-prepare-display()
  gnus-article-prepare(32880 nil)
  gnus-summary-display-article(32880 nil)
  gnus-summary-select-article(nil nil pseudo)
  gnus-summary-scroll-up(1)
  call-interactively(gnus-summary-scroll-up)

-- 
Josh Huber

Re: `C-c C-v' should keep citation prefix

[Simon Josefsson]

On 19 Jun 2001, Josh Huber wrote:

> > >| Symbol's value as variable is void: message-cite-prefix-regexp
>
> I'm getting this error with the latest oort gnus (just cvs updated a
> couple minutes ago) whenever I view a message...
>
> Signaling: (void-variable message-cite-prefix-regexp)
>   (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +==")
>   eval((concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +=="))
>   custom-initialize-reset(gnus-supercite-regexp (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +=="))
>   custom-declare-variable(gnus-supercite-regexp (concat "^\\(" message-cite-prefix-regexp "\\)? *" ">>>>> +\"\\([^\"\n]+\\)\" +==") "*Regexp matching normal Supercite attribution lines.\nThe first grouping must match prefixes added by other packages." :group gnus-cite :type regexp)
>   load-internal("gnus-cite" nil nil nil nil nil)
>   load("gnus-cite" nil nil nil)

This is weird, `message-cite-prefix-regexp' is defined in message.el and
gnus-cite.el requires message.el.

Are you sure gnus-cite.el isn't shadowed by some older version somehow?
Or that your gnus-cite.el isn't up to date somehow.  Or perhaps there is
some circular requiring going on. If this doesn't help, perhaps some
custom guru have more insight.

Re: `C-c C-v' should keep citation prefix

[Josh Huber]

Simon Josefsson <jas@extundo.com> writes:

> This is weird, `message-cite-prefix-regexp' is defined in message.el and
> gnus-cite.el requires message.el.

> Are you sure gnus-cite.el isn't shadowed by some older version somehow?

This was it...I did a rm -rf /usr/share/xemacs21/packages/lisp/gnus/
and the problem went away.

The real question is this.  How can I prevent this in the future?
Next time I upgrade xemacs this directory with an old version of Gnus
will come back.  Should I remove it from my load path?  Will that do
the trick, or is it too late? (has it already been loaded?)

I'm going to try that out, we'll see how it goes...

Thanks for the tip,

-- 
Josh Huber

Re: `C-c C-v' should keep citation prefix

[Simon Josefsson]

On 20 Jun 2001, Josh Huber wrote:

> This was it...I did a rm -rf /usr/share/xemacs21/packages/lisp/gnus/
> and the problem went away.
>
> The real question is this.  How can I prevent this in the future?
> Next time I upgrade xemacs this directory with an old version of Gnus
> will come back.  Should I remove it from my load path?  Will that do
> the trick, or is it too late? (has it already been loaded?)

Put a statement in your .emacs that makes `load-path' contain your own
Gnus directory before the system-wide one.  If you're on a single-user
system, perhaps also remove the Gnus xemacs-package completely.

Re: `C-c C-v' should keep citation prefix

[Josh Huber]

(note, please don't cc me via email also, I'm reading the newsgroup)

Simon Josefsson <jas@extundo.com> writes:

> Put a statement in your .emacs that makes `load-path' contain your own
> Gnus directory before the system-wide one.  If you're on a single-user
> system, perhaps also remove the Gnus xemacs-package completely.

This isn't enough apparently...I even went so far as to remove the old
gnus from my load path:

;; get rid of old gnus dirs in the load path
(delete-if
 (lambda (dir) (string-match "gnus" dir))
 load-path)

;; I definately want to use Oort Gnus
(add-to-list 'load-path "~/cvs/gnus/lisp")
(add-to-list 'load-path "~/cvs/gnus/contrib")
(add-to-list 'load-path "~/elisp")

The only two directories in my load path that contain "gnus" are the
two I've added.

It's almost as if the old gnus gets loaded before I get a chance to
remove it from my load path.

Starting xemacs with -no-site-file also solves the problem...perhaps I
should move this to the xemacs group?

Thanks again,

-- 
Josh Huber

Re: `C-c C-v' should keep citation prefix

[David S. Goldberg]

I had this problem with bbdb, (and probably had it with gnus too but
somehow didn't notice :-).  My solution is to make bbdb (and gnus
among others) "personal packages" by putting them under
~/.xemacs/packages/lisp.  The problem is that the package directories
get sucked in before load-path customizations take effect.  I imagine
that XEmacs gives a better option, but since the above worked I didn't
bother digging.

FWIW, for bbdb I also had to do:

(unload-feature 'bbdb-autoloads t)

But haven't had a need to do anything similar for gnus (so far).
-- 
Dave Goldberg
dsg@world.std.com

Re: `C-c C-v' should keep citation prefix

[Kai Großjohann]

On 20 Jun 2001, Josh Huber wrote:

> (note, please don't cc me via email also, I'm reading the newsgroup)
> 
> Simon Josefsson <jas@extundo.com> writes:
> 
>> Put a statement in your .emacs that makes `load-path' contain your
>> own Gnus directory before the system-wide one.  If you're on a
>> single-user system, perhaps also remove the Gnus xemacs-package
>> completely.
> 
> This isn't enough apparently...I even went so far as to remove the
> old gnus from my load path:

Are you sure you put it in .emacs and not .gnus?  Load-path frobbing
can't be done from .gnus -- then parts of Gnus have already been
loaded and it's too late.

And if something in .emacs causes parts of Gnus to be loaded, even
somewhere low in .emacs it could be too late.

Just before frobbing load-path, you could have a look at the variable
load-history to see if parts of Gnus have already been loaded.

kai
-- 
~/.signature: No such file or directory

Re: `C-c C-v' should keep citation prefix

[Josh Huber]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> Are you sure you put it in .emacs and not .gnus?  Load-path frobbing
> can't be done from .gnus -- then parts of Gnus have already been
> loaded and it's too late.

It's definitely before gnus is loaded.

> And if something in .emacs causes parts of Gnus to be loaded, even
> somewhere low in .emacs it could be too late.

It's the first thing that's done in my initialization code.

> Just before frobbing load-path, you could have a look at the variable
> load-history to see if parts of Gnus have already been loaded.

This is a good idea :)

There are may parts of gnus loaded before I do anything in my init
code:

...
gnus-define-keys-safe gnus-define-keymap gnus-define-keys-1
gnus-completing-read gnus-y-or-n-p gnus-yes-or-no-p gnus-dd-mmm
gnus-date-get-time gnus-time-iso8601 gnus-date-iso8601
gnus-mode-string-quote gnus-make-hashtable gnus-create-hash-size
gnus-verbose gnus-message gnus-error gnus-split-references
gnus-parent-id gnus-buffer-live-p gnus-horizontal-recenter
gnus-read-event-char gnus-sortable-date gnus-copy-file
gnus-work-buffer gnus-set-work-buffer gnus-group-real-name
gnus-make-sort-function gnus-make-sort-function-1
gnus-turn-off-edit-menu gnus-prin1 gnus-prin1-to-string
gnus-make-directory gnus-write-buffer gnus-delete-file
gnus-strip-whitespace gnus-put-text-property-excluding-newlines
gnus-put-text-property-excluding-characters-with-faces
gnus-atomic-be-safe gnus-atomic-progn gnus-atomic-progn-assign
gnus-atomic-setq gnus-output-to-rmail gnus-output-to-mail
gnus-convert-article-to-rmail gnus-map-function gnus-run-hooks
gnus-parse-netrc gnus-netrc-machine gnus-netrc-get gnus-alive-p
gnus-remove-duplicates gnus-delete-if gnus-delete-alist gnus-pull
gnus-globalify-regexp gnus-set-window-start
...

hmmm...

I can just remove the gnus directory for the old version, but I really
would like to solve this the right way.  that way, the next time I
upgrade my xemacs packages, it doesn't copy a new (old) gnus directory
in /usr/share/xemacs21/packages/lisp/.

hmm!

-- 
Josh Huber

Re: `C-c C-v' should keep citation prefix

[Paul Jarc]

Josh Huber <huber@alum.wpi.edu> writes:
> I can just remove the gnus directory for the old version, but I really
> would like to solve this the right way.

I'd use --no-site-file and then load site-start.el explicitly in your
.emacs after you've tweaked load-path.


paul

Re: `C-c C-v' should keep citation prefix

[Josh Huber]

prj@po.cwru.edu (Paul Jarc) writes:

> I'd use --no-site-file and then load site-start.el explicitly in your
> .emacs after you've tweaked load-path.

That's a great idea...

and it works!

Thanks, I've finally done it... :)

here's the relevant section, btw:

;; get rid of old gnus dirs in the load path
(delete-if
 (lambda (dir) (string-match "gnus" dir))
 load-path)

;; I definately want to use Oort Gnus
(add-to-list 'load-path "~/cvs/gnus/lisp")
(add-to-list 'load-path "~/cvs/gnus/contrib")
(add-to-list 'load-path "~/elisp")

;; now run the site-start file (we started with -no-site-file)
(load "/usr/share/xemacs21/site-packages/lisp/site-start")

...cool.

-- 
Josh Huber