Skeleton pair insert not working after upgrade.

[S. McKay Curtis]

Hi all,

I recently upgraded my Emacs and ESS using Vincent Goulet's very convenient
package and installer.  But now that I have upgraded, the skeleton pair
insert for the parenthesis "(" is not working in R files.  That is, I have
the following code in my .emacs file

(setq skeleton-pair t)
;;(setq skeleton-pair-on-word t)
(global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "<") 'skeleton-pair-insert-maybe)

which, before upgrade, would automatically type a closing parenthesis ")"
every time I typed an opening parenthesis "(" [and would do the same for the
other characters "[", "{", etc.].  Now that I have upgraded, I do not get a
closing parenthesis after I type an open parenthesis when editing R files,
even though it works for other characters like "[".  Also, the skeleton pair
insert for "(" works in the R command buffer, just not in an R file.

Any thoughts on how to fix this?


Thanks
McKay

        [[alternative HTML version deleted]]

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

On Aug 27, 2009, at 12:21 PM, S. McKay Curtis wrote:

> Hi all,
>
> I recently upgraded my Emacs and ESS using Vincent Goulet's very  
> convenient
> package and installer.  But now that I have upgraded, the skeleton  
> pair
> insert for the parenthesis "(" is not working in R files.  That is,  
> I have
> the following code in my .emacs file
>
> (setq skeleton-pair t)
> ;;(setq skeleton-pair-on-word t)
> (global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
> (global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
> (global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
> (global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
> (global-set-key (kbd "<") 'skeleton-pair-insert-maybe)
>
> which, before upgrade, would automatically type a closing  
> parenthesis ")"
> every time I typed an opening parenthesis "(" [and would do the same  
> for the
> other characters "[", "{", etc.].  Now that I have upgraded, I do  
> not get a
> closing parenthesis after I type an open parenthesis when editing R  
> files,
> even though it works for other characters like "[".  Also, the  
> skeleton pair
> insert for "(" works in the R command buffer, just not in an R file.
>
> Any thoughts on how to fix this?
>
>
> Thanks
> McKay

Yep...I can confirm the behavior on OSX 10.5.8, however it is not just  
with Vincent's version of Emacs/ESS. I can replicate it here with my  
own Emacs 23 compiled directly from the Emacs cvs and also with  
Aquamacs downloaded from their site. Aquamacs is based upon Emacs 22,  
so this is not just with Emacs 23.

McKay, I presume that you are on OSX?  If so, have you installed the  
10.5.8 update or are you still on 10.5.7? I wonder if that update,  
which was pushed earlier this month on the 5th may have anything to do  
with this. There was also a security update (2009-004) on the 12th.

Is anyone seeing this on other OSs?

It also happens with a brace '{}' pair, where the closing brace is not  
automatically inserted.

This happens in ESS[S] mode, either when editing an R file, or when in  
an R chunk in a .Rnw file. In the LaTeX chunks all is well.

It all works properly in iESS [R] mode and if I disable ESS completely  
in my .emacs file and then open an R file or a .Rnw file, the behavior  
is correct.

It also works properly in Rd mode.

I see incomplete behavior when opening a Perl script file, where in  
that case (using CPerl mode), double and single quotes will be  
properly inserted, but parens, braces and brackets are not.

I am not sure why I had not noticed this until now and am scratching  
my head to figure it out. I have not changed anything in my .emacs and  
have not updated other emacs lisp packages that I load.

One thing I did note. If I revert to ESS 5.3.11 or 5.3.10 from 5.4,  
then the paren pairing happens OK in ESS[S] modes, but brace ('{}')  
pairing still does not.

I am confuzzled....I am not ready to say this is ESS itself yet,  
unless others on non-OSX systems are seeing it as well.

Regards,

Marc Schwartz

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Vinh Nguyen-3]

i'm running Mac OS X 10.5.8 with emacs compiled from the cvs source
using the nextstep method.  i have ESS 5.4, and I DO NOT HAVE the
problem u stated.  i have the following in my .emacs:

;; following from Marc Schwartz from ESS list, works globally
;; enable skeleton-pair insert globally
(setq skeleton-pair t)
;(setq skeleton-pair-on-word t)
(global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
(global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
;;(global-set-key (kbd "\'") 'skeleton-pair-insert-maybe) ;; apostrophe
(global-set-key (kbd "\`") 'skeleton-pair-insert-maybe)
;;(global-set-key (kbd "<") 'skeleton-pair-insert-maybe)
;; Delete empty pairs like '()' taken from
;; http://www.emacswiki.org/cgi-bin/wiki/SkeletonMode#toc14 ##Deletion section
(defvar skeletons-alist
  '((?\( . ?\))
    (?\' . ?\')
    (?\" . ?\")
    (?[  . ?])
    (?{  . ?})
    (?<  . ?>)
    (?$  . ?$)))
(defadvice delete-backward-char (before delete-empty-pair activate)
  (if (eq (cdr (assq (char-before) skeletons-alist)) (char-after))
      (and (char-after) (delete-char 1))))

vinh
--
This e-mail/fax message, including any attachments, is for the sole
use of the intended recipient(s) and may contain confidential and
privileged information. Any unauthorized review, use, disclosure or
distribution is prohibited. If you are not the intended recipient,
please contact the sender by reply e-mail/fax and destroy all copies
of the original message.



On Thu, Aug 27, 2009 at 4:57 PM, Marc Schwartz<marc_schwartz@...> wrote:

> On Aug 27, 2009, at 12:21 PM, S. McKay Curtis wrote:
>
>> Hi all,
>>
>> I recently upgraded my Emacs and ESS using Vincent Goulet's very
>> convenient
>> package and installer.  But now that I have upgraded, the skeleton pair
>> insert for the parenthesis "(" is not working in R files.  That is, I have
>> the following code in my .emacs file
>>
>> (setq skeleton-pair t)
>> ;;(setq skeleton-pair-on-word t)
>> (global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "<") 'skeleton-pair-insert-maybe)
>>
>> which, before upgrade, would automatically type a closing parenthesis ")"
>> every time I typed an opening parenthesis "(" [and would do the same for
>> the
>> other characters "[", "{", etc.].  Now that I have upgraded, I do not get
>> a
>> closing parenthesis after I type an open parenthesis when editing R files,
>> even though it works for other characters like "[".  Also, the skeleton
>> pair
>> insert for "(" works in the R command buffer, just not in an R file.
>>
>> Any thoughts on how to fix this?
>>
>>
>> Thanks
>> McKay
>
>
>
> Yep...I can confirm the behavior on OSX 10.5.8, however it is not just with
> Vincent's version of Emacs/ESS. I can replicate it here with my own Emacs 23
> compiled directly from the Emacs cvs and also with Aquamacs downloaded from
> their site. Aquamacs is based upon Emacs 22, so this is not just with Emacs
> 23.
>
> McKay, I presume that you are on OSX?  If so, have you installed the 10.5.8
> update or are you still on 10.5.7? I wonder if that update, which was pushed
> earlier this month on the 5th may have anything to do with this. There was
> also a security update (2009-004) on the 12th.
>
> Is anyone seeing this on other OSs?
>
> It also happens with a brace '{}' pair, where the closing brace is not
> automatically inserted.
>
> This happens in ESS[S] mode, either when editing an R file, or when in an R
> chunk in a .Rnw file. In the LaTeX chunks all is well.
>
> It all works properly in iESS [R] mode and if I disable ESS completely in my
> .emacs file and then open an R file or a .Rnw file, the behavior is correct.
>
> It also works properly in Rd mode.
>
> I see incomplete behavior when opening a Perl script file, where in that
> case (using CPerl mode), double and single quotes will be properly inserted,
> but parens, braces and brackets are not.
>
> I am not sure why I had not noticed this until now and am scratching my head
> to figure it out. I have not changed anything in my .emacs and have not
> updated other emacs lisp packages that I load.
>
> One thing I did note. If I revert to ESS 5.3.11 or 5.3.10 from 5.4, then the
> paren pairing happens OK in ESS[S] modes, but brace ('{}') pairing still
> does not.
>
> I am confuzzled....I am not ready to say this is ESS itself yet, unless
> others on non-OSX systems are seeing it as well.
>
> Regards,
>
> Marc Schwartz
>
> ______________________________________________
> ESS-help@... mailing list
> https://stat.ethz.ch/mailman/listinfo/ess-help
>

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[S. McKay Curtis]

Marc,

Misery loves company, so I'm glad you were able to replicate my troubles!

I am having the problem on Windows XP.

I'm not an Emacs guru, but the file ess-mode.el contains this statement

  (define-key ess-mode-map "("  ;; allow to toggle after customization:
    (if ess-r-args-electric-paren 'ess-r-args-auto-show 'self-insert-command))

which looks like it redefines the parenthesis.  Do the lines of code
above negate what I have in my .emacs file?

Puzzled,
McKay


On Thu, Aug 27, 2009 at 4:57 PM, Marc Schwartz<marc_schwartz@...> wrote:

> On Aug 27, 2009, at 12:21 PM, S. McKay Curtis wrote:
>
>> Hi all,
>>
>> I recently upgraded my Emacs and ESS using Vincent Goulet's very
>> convenient
>> package and installer.  But now that I have upgraded, the skeleton pair
>> insert for the parenthesis "(" is not working in R files.  That is, I have
>> the following code in my .emacs file
>>
>> (setq skeleton-pair t)
>> ;;(setq skeleton-pair-on-word t)
>> (global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
>> (global-set-key (kbd "<") 'skeleton-pair-insert-maybe)
>>
>> which, before upgrade, would automatically type a closing parenthesis ")"
>> every time I typed an opening parenthesis "(" [and would do the same for
>> the
>> other characters "[", "{", etc.].  Now that I have upgraded, I do not get
>> a
>> closing parenthesis after I type an open parenthesis when editing R files,
>> even though it works for other characters like "[".  Also, the skeleton
>> pair
>> insert for "(" works in the R command buffer, just not in an R file.
>>
>> Any thoughts on how to fix this?
>>
>>
>> Thanks
>> McKay
>
>
>
> Yep...I can confirm the behavior on OSX 10.5.8, however it is not just with
> Vincent's version of Emacs/ESS. I can replicate it here with my own Emacs 23
> compiled directly from the Emacs cvs and also with Aquamacs downloaded from
> their site. Aquamacs is based upon Emacs 22, so this is not just with Emacs
> 23.
>
> McKay, I presume that you are on OSX?  If so, have you installed the 10.5.8
> update or are you still on 10.5.7? I wonder if that update, which was pushed
> earlier this month on the 5th may have anything to do with this. There was
> also a security update (2009-004) on the 12th.
>
> Is anyone seeing this on other OSs?
>
> It also happens with a brace '{}' pair, where the closing brace is not
> automatically inserted.
>
> This happens in ESS[S] mode, either when editing an R file, or when in an R
> chunk in a .Rnw file. In the LaTeX chunks all is well.
>
> It all works properly in iESS [R] mode and if I disable ESS completely in my
> .emacs file and then open an R file or a .Rnw file, the behavior is correct.
>
> It also works properly in Rd mode.
>
> I see incomplete behavior when opening a Perl script file, where in that
> case (using CPerl mode), double and single quotes will be properly inserted,
> but parens, braces and brackets are not.
>
> I am not sure why I had not noticed this until now and am scratching my head
> to figure it out. I have not changed anything in my .emacs and have not
> updated other emacs lisp packages that I load.
>
> One thing I did note. If I revert to ESS 5.3.11 or 5.3.10 from 5.4, then the
> paren pairing happens OK in ESS[S] modes, but brace ('{}') pairing still
> does not.
>
> I am confuzzled....I am not ready to say this is ESS itself yet, unless
> others on non-OSX systems are seeing it as well.
>
> Regards,
>
> Marc Schwartz
>
>

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Vitalie S.]

On Fri, 28 Aug 2009 06:19:07 +0200, S. McKay Curtis  
<smcurtis@...> wrote:

> Marc,
>
> Misery loves company, so I'm glad you were able to replicate my troubles!
>
> I am having the problem on Windows XP.
>
> I'm not an Emacs guru, but the file ess-mode.el contains this statement
>
>   (define-key ess-mode-map "("  ;; allow to toggle after customization:
>     (if ess-r-args-electric-paren 'ess-r-args-auto-show  
> 'self-insert-command))
>

ess-r-args-auto-show  uses (insert "(") to trigger  
(skeletion-pair-insert-maybe nil) which is bound globally to "(". But in  
emacs 23 that does not work any more.

So, just as a quick fix :

  (define-key ess-mode-map (kbd "(") '(lambda () (interactive)
                                        (skeleton-pair-insert-maybe nil)
                                        (ess-r-args-show)))

> which looks like it redefines the parenthesis.  Do the lines of code
> above negate what I have in my .emacs file?
>
> Puzzled,
> McKay
>
>
> On Thu, Aug 27, 2009 at 4:57 PM, Marc Schwartz<marc_schwartz@...>  
> wrote:
>> On Aug 27, 2009, at 12:21 PM, S. McKay Curtis wrote:
>>
>>> Hi all,
>>>
>>> I recently upgraded my Emacs and ESS using Vincent Goulet's very
>>> convenient
>>> package and installer.  But now that I have upgraded, the skeleton pair
>>> insert for the parenthesis "(" is not working in R files.  That is, I  
>>> have
>>> the following code in my .emacs file
>>>
>>> (setq skeleton-pair t)
>>> ;;(setq skeleton-pair-on-word t)
>>> (global-set-key (kbd "(") 'skeleton-pair-insert-maybe)
>>> (global-set-key (kbd "[") 'skeleton-pair-insert-maybe)
>>> (global-set-key (kbd "{") 'skeleton-pair-insert-maybe)
>>> (global-set-key (kbd "\"") 'skeleton-pair-insert-maybe)
>>> (global-set-key (kbd "<") 'skeleton-pair-insert-maybe)
>>>
>>> which, before upgrade, would automatically type a closing parenthesis  
>>> ")"
>>> every time I typed an opening parenthesis "(" [and would do the same  
>>> for
>>> the
>>> other characters "[", "{", etc.].  Now that I have upgraded, I do not  
>>> get
>>> a
>>> closing parenthesis after I type an open parenthesis when editing R  
>>> files,
>>> even though it works for other characters like "[".  Also, the skeleton
>>> pair
>>> insert for "(" works in the R command buffer, just not in an R file.
>>>
>>> Any thoughts on how to fix this?
>>>
>>>
>>> Thanks
>>> McKay
>>
>>
>>
>> Yep...I can confirm the behavior on OSX 10.5.8, however it is not just  
>> with
>> Vincent's version of Emacs/ESS. I can replicate it here with my own  
>> Emacs 23
>> compiled directly from the Emacs cvs and also with Aquamacs downloaded  
>> from
>> their site. Aquamacs is based upon Emacs 22, so this is not just with  
>> Emacs
>> 23.
>>
>> McKay, I presume that you are on OSX?  If so, have you installed the  
>> 10.5.8
>> update or are you still on 10.5.7? I wonder if that update, which was  
>> pushed
>> earlier this month on the 5th may have anything to do with this. There  
>> was
>> also a security update (2009-004) on the 12th.
>>
>> Is anyone seeing this on other OSs?
>>
>> It also happens with a brace '{}' pair, where the closing brace is not
>> automatically inserted.
>>
>> This happens in ESS[S] mode, either when editing an R file, or when in  
>> an R
>> chunk in a .Rnw file. In the LaTeX chunks all is well.
>>
>> It all works properly in iESS [R] mode and if I disable ESS completely  
>> in my
>> .emacs file and then open an R file or a .Rnw file, the behavior is  
>> correct.
>>
>> It also works properly in Rd mode.
>>
>> I see incomplete behavior when opening a Perl script file, where in that
>> case (using CPerl mode), double and single quotes will be properly  
>> inserted,
>> but parens, braces and brackets are not.
>>
>> I am not sure why I had not noticed this until now and am scratching my  
>> head
>> to figure it out. I have not changed anything in my .emacs and have not
>> updated other emacs lisp packages that I load.
>>
>> One thing I did note. If I revert to ESS 5.3.11 or 5.3.10 from 5.4,  
>> then the
>> paren pairing happens OK in ESS[S] modes, but brace ('{}') pairing still
>> does not.
>>
>> I am confuzzled....I am not ready to say this is ESS itself yet, unless
>> others on non-OSX systems are seeing it as well.
>>
>> Regards,
>>
>> Marc Schwartz
>>
>>
>
> ______________________________________________
> ESS-help@... mailing list
> https://stat.ethz.ch/mailman/listinfo/ess-help

--

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

On Aug 28, 2009, at 8:34 AM, Vitalie S. wrote:

> On Fri, 28 Aug 2009 06:19:07 +0200, S. McKay Curtis <smcurtis@...
> > wrote:
>
>> Marc,
>>
>> Misery loves company, so I'm glad you were able to replicate my  
>> troubles!
>>
>> I am having the problem on Windows XP.
>>
>> I'm not an Emacs guru, but the file ess-mode.el contains this  
>> statement
>>
>>  (define-key ess-mode-map "("  ;; allow to toggle after  
>> customization:
>>    (if ess-r-args-electric-paren 'ess-r-args-auto-show 'self-insert-
>> command))

>>
> ess-r-args-auto-show  uses (insert "(") to trigger (skeletion-pair-
> insert-maybe nil) which is bound globally to "(". But in emacs 23  
> that does not work any more.
>
> So, just as a quick fix :
>
> (define-key ess-mode-map (kbd "(") '(lambda () (interactive)
> (skeleton-pair-insert-maybe nil)
> (ess-r-args-show)))

<snip>

That does not work for me.  However, commenting out the two lines that  
McKay referenced in ess-mode.el DOES solve the problem with parens.

I have not yet tried to do anything with electric braces along the  
same lines, as the skeleton pairing does not work in ESS[S] mode for  
braces.

 From what I can tell the referenced code in ess-mode.el has been  
around since at least ESS 5.3.6, so that is not even close to being new.

FWIW, I completely stripped my .emacs file of everything except  
loading ESS and the skeleton pair code and no joy. So there are no  
other Emacs related settings or other "stuff" that seems to be  
effectual here.

So the questions:

1. Why does Vinh not have this same problem on OSX with emacs 23? What  
is different on his system?

2. What has happened recently to alter this behavior, which given that  
McKay is on Windows, is clearly not limited to OSX or any of the  
recent updates released by Apple.

3. Since I was able to replicate the problematic behavior with  
Aquamacs, which is based on Emacs 22, this is not limited to Emacs 23.


I am running:

   GNU Emacs 23.1.50.1 (i386-apple-darwin9.8.0, NS apple-
appkit-949.54) of 2009-08-27

with ESS 5.4.


Still head scratching....especially since for the life of me, I can't  
figure out why I had not noticed this change before now...

Marc

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Vinh Nguyen-3]

--
This e-mail/fax message, including any attachments, is for the sole
use of the intended recipient(s) and may contain confidential and
privileged information. Any unauthorized review, use, disclosure or
distribution is prohibited. If you are not the intended recipient,
please contact the sender by reply e-mail/fax and destroy all copies
of the original message.



On Fri, Aug 28, 2009 at 7:00 AM, Marc Schwartz<marc_schwartz@...> wrote:
>
> 1. Why does Vinh not have this same problem on OSX with emacs 23? What is
> different on his system?

I apologize, i just ran emacs with the -q option and ran the require
ess-site and the skeleton stuff to investigate, and I DID REPLICATE
the mentioned issue.  of the following list,
(defvar skeletons-alist
  '((?\( . ?\))
    (?\' . ?\')
    (?\" . ?\")
    (?[  . ?])
    (?{  . ?})
    (?<  . ?>)
    (?$  . ?$)))

"(" and "{" does not work for me.

the reason my parentheses worked before was i had the following in my
.emacs as well:
(defun my-ess-comma-hook ()
 (local-set-key "," (lambda () (interactive) (insert ", "))))
(add-hook 'ess-mode-hook 'my-ess-comma-hook)
(add-hook 'inferior-ess-mode-hook 'my-ess-comma-hook)

(defun my-ess-parens-hook ()
   (local-set-key "(" (lambda () (interactive)
                    (insert "(  )")
                    (forward-char -2))))

(add-hook 'ess-mode-hook 'my-ess-parens-hook)
(add-hook 'inferior-ess-mode-hook 'my-ess-parens-hook)

i apologize for the confusion.

vinh

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

On Aug 28, 2009, at 2:06 PM, Vinh Nguyen wrote:

> On Fri, Aug 28, 2009 at 7:00 AM, Marc Schwartz<marc_schwartz@...>  
> wrote:
>>
>> 1. Why does Vinh not have this same problem on OSX with emacs 23?  
>> What is
>> different on his system?
>
> I apologize, i just ran emacs with the -q option and ran the require
> ess-site and the skeleton stuff to investigate, and I DID REPLICATE
> the mentioned issue.  of the following list,
> (defvar skeletons-alist
>  '((?\( . ?\))
>    (?\' . ?\')
>    (?\" . ?\")
>    (?[  . ?])
>    (?{  . ?})
>    (?<  . ?>)
>    (?$  . ?$)))
>
> "(" and "{" does not work for me.
>
> the reason my parentheses worked before was i had the following in my
> .emacs as well:
> (defun my-ess-comma-hook ()
> (local-set-key "," (lambda () (interactive) (insert ", "))))
> (add-hook 'ess-mode-hook 'my-ess-comma-hook)
> (add-hook 'inferior-ess-mode-hook 'my-ess-comma-hook)
>
> (defun my-ess-parens-hook ()
>   (local-set-key "(" (lambda () (interactive)
>                    (insert "(  )")
>                    (forward-char -2))))
>
> (add-hook 'ess-mode-hook 'my-ess-parens-hook)
> (add-hook 'inferior-ess-mode-hook 'my-ess-parens-hook)
>
> i apologize for the confusion.
>
> vinh

Not a problem Vinh, thanks for the clarification.

I am still stumped by the change in the original behavior and am not  
clear on the best approach for fixing this at the emacs lisp level.  
Perhaps members of ESS Core will have some more refined thoughts here.

BTW, just as an aside, I tried this whole thing using Emacs within a  
terminal on OSX, which brings up the CLI (non-GUI) version of Emacs,  
which is GNU Emacs 22.1.1. I can replicate the problematic behavior  
there as well, so this is not limited to GUI versions, whether 22 or 23.

That being said and because I am leaving for vacation on Sunday for  
the week during which I will have limited e-mail access, I have been  
trying a work around for this and seem to have a possibility. I have  
not yet fully tested it and there may be other untoward effects of  
this particular approach.

Rather than using skeleton mode, I located the AutoPairs page at the  
Emacs Wiki:

   http://www.emacswiki.org/emacs/AutoPairs

and decide to give the textmate.el approach a try since it claims some  
additional functionality.  textmate.el is available from:

   http://code.google.com/p/emacs-textmate/

Once can download and un-tar the gz file there to someplace convenient  
on your system.

Once you have done that, add the following to your .emacs file, which  
will load textmate-mode at Emacs startup, rather than having to use 'M-
x textmate-mode' every time:

;; Modify the path to textmate.el as appropriate!!!
(load "/Users/marcschwartz/Downloads/Emacs/textmate/textmate")

(mapc (lambda (mode)
        (let ((hook (intern (concat (symbol-name mode)
                                    "-mode-hook"))))
          (add-hook hook (lambda () (textmate-mode +1)))))
       '(c cperl emacs-lisp inferior-lisp ess inferior-ess lisp noweb  
perl Rd text tex))


The final line above determines which Emacs modes this will work in  
automatically, so for other modes not indicated, add them as you  
require. I located the above code in the textmate issues area: http://code.google.com/p/emacs-textmate/issues/detail?id=5

Be sure to remove or comment the original skeleton mode related code  
in your .emacs.

Once you have that in place, this all should work, but again, I have  
not fully tested it for either functionality or conflicts with other  
ESS functions.

HTH,

Marc Schwartz

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Vitalie S.]

>
> Rather than using skeleton mode, I located the AutoPairs page at the  
> Emacs Wiki:
>
>    http://www.emacswiki.org/emacs/AutoPairs
>
> and decide to give the textmate.el approach a try since it claims some  
> additional functionality.  textmate.el is available from:
>
>    http://code.google.com/p/emacs-textmate/
>
> Once can download and un-tar the gz file there to someplace convenient  
> on your system.
>
> Once you have done that, add the following to your .emacs file, which  
> will load textmate-mode at Emacs startup, rather than having to use 'M-x  
> textmate-mode' every time:
>
> ;; Modify the path to textmate.el as appropriate!!!
> (load "/Users/marcschwartz/Downloads/Emacs/textmate/textmate")
>
> (mapc (lambda (mode)
> (let ((hook (intern (concat (symbol-name mode)
>    "-mode-hook"))))
>  (add-hook hook (lambda () (textmate-mode +1)))))
>        '(c cperl emacs-lisp inferior-lisp ess inferior-ess lisp noweb  
> perl Rd text tex))
>
>

Did not work for me in the beggining. The code uses internally  
python-backspace
so I placed (require 'python) in my .emacs and apparently it started  
working.

But what is exactly the functionality it brings above skeletion??
I can not see anything fancy when deleting parenthesis.

Vitalie.

> The final line above determines which Emacs modes this will work in  
> automatically, so for other modes not indicated, add them as you  
> require. I located the above code in the textmate issues area:  
> http://code.google.com/p/emacs-textmate/issues/detail?id=5
>
> Be sure to remove or comment the original skeleton mode related code in  
> your .emacs.
>
> Once you have that in place, this all should work, but again, I have not  
> fully tested it for either functionality or conflicts with other ESS  
> functions.
>
> HTH,
>
> Marc Schwartz
>

--

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Vitalie S.]

On Sat, 29 Aug 2009 11:31:14 +0200, Vitalie S. <vitosmail@...>  
wrote:

>
>>
>> Rather than using skeleton mode, I located the AutoPairs page at the  
>> Emacs Wiki:
>>
>>    http://www.emacswiki.org/emacs/AutoPairs
>>
>> and decide to give the textmate.el approach a try since it claims some  
>> additional functionality.  textmate.el is available from:
>>
>>    http://code.google.com/p/emacs-textmate/
>>
>> Once can download and un-tar the gz file there to someplace convenient  
>> on your system.
>>
>> Once you have done that, add the following to your .emacs file, which  
>> will load textmate-mode at Emacs startup, rather than having to use  
>> 'M-x textmate-mode' every time:
>>
>> ;; Modify the path to textmate.el as appropriate!!!
>> (load "/Users/marcschwartz/Downloads/Emacs/textmate/textmate")
>>
>> (mapc (lambda (mode)
>> (let ((hook (intern (concat (symbol-name mode)
>>    "-mode-hook"))))
>>  (add-hook hook (lambda () (textmate-mode +1)))))
>>        '(c cperl emacs-lisp inferior-lisp ess inferior-ess lisp noweb  
>> perl Rd text tex))
>>
>>
> Did not work for me in the beggining. The code uses internally  
> python-backspace
> so I placed (require 'python) in my .emacs and apparently it started  
> working.
>
> But what is exactly the functionality it brings above skeletion??
> I can not see anything fancy when deleting parenthesis.
>

In addition it does not solve the original problem (at least for me)  
ess-arg-show is not executed after pressing "(".

> Vitalie.
>> The final line above determines which Emacs modes this will work in  
>> automatically, so for other modes not indicated, add them as you  
>> require. I located the above code in the textmate issues area:  
>> http://code.google.com/p/emacs-textmate/issues/detail?id=5
>>
>> Be sure to remove or comment the original skeleton mode related code in  
>> your .emacs.
>>
>> Once you have that in place, this all should work, but again, I have  
>> not fully tested it for either functionality or conflicts with other  
>> ESS functions.
>>
>> HTH,
>>
>> Marc Schwartz
>>
>
>

--

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

On Aug 29, 2009, at 4:31 AM, Vitalie S. wrote:
 >> Marc Schwartz wrote:

>> Rather than using skeleton mode, I located the AutoPairs page at  
>> the Emacs Wiki:
>>
>>   http://www.emacswiki.org/emacs/AutoPairs
>>
>> and decide to give the textmate.el approach a try since it claims  
>> some additional functionality.  textmate.el is available from:
>>
>>   http://code.google.com/p/emacs-textmate/
>>
>> Once can download and un-tar the gz file there to someplace  
>> convenient on your system.
>>
>> Once you have done that, add the following to your .emacs file,  
>> which will load textmate-mode at Emacs startup, rather than having  
>> to use 'M-x textmate-mode' every time:
>>
>> ;; Modify the path to textmate.el as appropriate!!!
>> (load "/Users/marcschwartz/Downloads/Emacs/textmate/textmate")
>>
>> (mapc (lambda (mode)
>> (let ((hook (intern (concat (symbol-name mode)
>>    "-mode-hook"))))
>>  (add-hook hook (lambda () (textmate-mode +1)))))
>>       '(c cperl emacs-lisp inferior-lisp ess inferior-ess lisp  
>> noweb perl Rd text tex))
>>
>>
> Did not work for me in the beggining. The code uses internally  
> python-backspace
> so I placed (require 'python) in my .emacs and apparently it started  
> working.
>
> But what is exactly the functionality it brings above skeletion??
> I can not see anything fancy when deleting parenthesis.
>
> Vitalie.

1. I noted after the fact, the problem with the python backspace.  
There is a reference to a fix here:

   http://code.google.com/p/emacs-textmate/issues/detail?id=7

however, I could not get the empty pair deletion to work in the modes  
that I use. Thus, there does not seem to be any additional  
functionality using textmate mode that I can see (save perhaps in  
python mode, which I do not use). The only benefit may be that it  
seems to function in the relevant ESS modes, which the default  
skeleton mode does not. However, that functionality appears to be at  
the expense of others.

2. The generic AutoPairs code (http://www.emacswiki.org/emacs/ 
AutoPairs) does not appear to solve the ESS mode interactions any more  
than using regular skeleton mode.

3. I have not been using the ess-arg-show functionality, which is why  
I did not note the loss of it. However, it would seem logical on the  
face of it, that there is an interaction here between ess-arg-show,  
electric braces/parens and skeleton mode, since each functions upon  
the entry of common characters and only some of the three functions  
may be compatible with the others as presently implemented.


At this point, I am at the limit of what I know about emacs lisp. For  
the time being, it may be that there is a tradeoff between these  
various functions.

I would have to defer to others with a deeper knowledge of emacs lisp  
as to what might be a more inclusive solution to this problem, which  
might perhaps require the addition of the skeleton mode pairing  
functionality within the framework of existing ESS functions to  
facilitate compatibility.

Marc

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[rsparapa]

S. McKay Curtis wrote:

> Marc,
>
> Misery loves company, so I'm glad you were able to replicate my troubles!
>
> I am having the problem on Windows XP.
>
> I'm not an Emacs guru, but the file ess-mode.el contains this statement
>
>   (define-key ess-mode-map "("  ;; allow to toggle after customization:
>     (if ess-r-args-electric-paren 'ess-r-args-auto-show 'self-insert-command))
>
> which looks like it redefines the parenthesis.  Do the lines of code
> above negate what I have in my .emacs file?
>
> Puzzled,
> McKay

Sorry, I was on vacation.  But, you will see that this has been fixed
in SVN:
;;this is bass-ackwards
;;  (define-key ess-mode-map "("  ;; allow to toggle after customization:
;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
'self-insert-command))
   (if ess-r-args-electric-paren (define-key ess-mode-map "("
'ess-r-args-auto-show))

--
Rodney Sparapani      Center for Patient Care & Outcomes Research (PCOR)
Sr. Biostatistician              http://www.mcw.edu/pcor
4 wheels good, 2 wheels better!  Medical College of Wisconsin (MCW)
WWLD?:  What Would Lombardi Do?  Milwaukee, WI, USA

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[rsparapa]

Rodney Sparapani wrote:

>
> Sorry, I was on vacation.  But, you will see that this has been fixed
> in SVN:
> ;;this is bass-ackwards
> ;;  (define-key ess-mode-map "("  ;; allow to toggle after customization:
> ;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
> 'self-insert-command))
>   (if ess-r-args-electric-paren (define-key ess-mode-map "("
> 'ess-r-args-auto-show))
>

See the ess-help thread "skeleton-pair and XEmacs" for some details.
Note that there is an XEmacs bug (Issues 551 and 554) at the beginning
of the thread that has some workarounds, fixes in the pipeline.  But,
the end of the thread discusses this same ESS bug.  Fix to appear in ESS
5.5 RSN.

Rodney

Did the Great Recession cure Global Warming?

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[rsparapa]

Rodney Sparapani wrote:

>
> See the ess-help thread "skeleton-pair and XEmacs" for some details.
> Note that there is an XEmacs bug (Issues 551 and 554) at the beginning
> of the thread that has some workarounds, fixes in the pipeline.  But,
> the end of the thread discusses this same ESS bug.  Fix to appear in ESS
> 5.5 RSN.
>
> Rodney
>
> Did the Great Recession cure Global Warming?
>

I guess silence is a good thing, but I would like to hear confirmation
that this fixes it for "(".  How many people also need "{" working?
Note that you need to have ess-r-args-electric-paren set to nil and that
must occur before you load ess-site.  Check it's value afterward to be sure.

Rodney

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[S. McKay Curtis]

I changed the lines in my ess-mode.el file, but this did not fix the
problem.  Has anyone else had success?

On Mon, Aug 31, 2009 at 6:35 AM, Rodney Sparapani <rsparapa@...> wrote:

> S. McKay Curtis wrote:
>
>> Marc,
>>
>> Misery loves company, so I'm glad you were able to replicate my troubles!
>>
>> I am having the problem on Windows XP.
>>
>> I'm not an Emacs guru, but the file ess-mode.el contains this statement
>>
>>  (define-key ess-mode-map "("  ;; allow to toggle after customization:
>>    (if ess-r-args-electric-paren 'ess-r-args-auto-show
>> 'self-insert-command))
>>
>> which looks like it redefines the parenthesis.  Do the lines of code
>> above negate what I have in my .emacs file?
>>
>> Puzzled,
>> McKay
>>
>
> Sorry, I was on vacation.  But, you will see that this has been fixed
> in SVN:
> ;;this is bass-ackwards
> ;;  (define-key ess-mode-map "("  ;; allow to toggle after customization:
> ;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
> 'self-insert-command))
>  (if ess-r-args-electric-paren (define-key ess-mode-map "("
> 'ess-r-args-auto-show))
>
> --
> Rodney Sparapani      Center for Patient Care & Outcomes Research (PCOR)
> Sr. Biostatistician              http://www.mcw.edu/pcor
> 4 wheels good, 2 wheels better!  Medical College of Wisconsin (MCW)
> WWLD?:  What Would Lombardi Do?  Milwaukee, WI, USA
>

        [[alternative HTML version deleted]]

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

Hi all,

Apologies for the delay in my reply here, as I too was on vacation  
last week and am getting back up to date on e-mails and tasks.

As with McKay, simply changing the lines in ess-mode.el did not help.

However, I decided to checkout the svn version of ESS tonight, under  
the presumption that other code has sufficiently changed as to require  
needing the fully updated code base, rather than just changing two  
lines.

Upon doing that and making the appropriate adjustments to my ESS path  
in .emacs, I can say that the current svn code base of ESS does  
properly work (with respect to skeleton-pair parens) in ESS mode. Note  
that I make no adjustment to the default value of ess-r-args-electric-
paren, which in my case is nil as per Rodney's earlier reply.

However, as Rodney also noted, brace "{}" skeleton-pair does not work  
in ESS mode and that would be desirable.

HTH,

Marc

On Sep 7, 2009, at 4:03 PM, S. McKay Curtis wrote:

> I changed the lines in my ess-mode.el file, but this did not fix the
> problem.  Has anyone else had success?
>
> On Mon, Aug 31, 2009 at 6:35 AM, Rodney Sparapani <rsparapa@...>  
> wrote:
>
>> S. McKay Curtis wrote:
>>
>>> Marc,
>>>
>>> Misery loves company, so I'm glad you were able to replicate my  
>>> troubles!
>>>
>>> I am having the problem on Windows XP.
>>>
>>> I'm not an Emacs guru, but the file ess-mode.el contains this  
>>> statement
>>>
>>> (define-key ess-mode-map "("  ;; allow to toggle after  
>>> customization:
>>>   (if ess-r-args-electric-paren 'ess-r-args-auto-show
>>> 'self-insert-command))
>>>
>>> which looks like it redefines the parenthesis.  Do the lines of code
>>> above negate what I have in my .emacs file?
>>>
>>> Puzzled,
>>> McKay
>>>
>>
>> Sorry, I was on vacation.  But, you will see that this has been fixed
>> in SVN:
>> ;;this is bass-ackwards
>> ;;  (define-key ess-mode-map "("  ;; allow to toggle after  
>> customization:
>> ;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
>> 'self-insert-command))
>> (if ess-r-args-electric-paren (define-key ess-mode-map "("
>> 'ess-r-args-auto-show))

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[rsparapa]

It could be that you just needed to M-x byte-compile-file, but glad to hear
( is now working.
________________________________________
From: Marc Schwartz [marc_schwartz@...]
Sent: Wednesday, September 09, 2009 10:47 PM
To: S. McKay Curtis
Cc: Sparapani, Rodney; ess-help@...
Subject: Re: [ESS] Skeleton pair insert not working after upgrade.

Hi all,

Apologies for the delay in my reply here, as I too was on vacation
last week and am getting back up to date on e-mails and tasks.

As with McKay, simply changing the lines in ess-mode.el did not help.

However, I decided to checkout the svn version of ESS tonight, under
the presumption that other code has sufficiently changed as to require
needing the fully updated code base, rather than just changing two
lines.

Upon doing that and making the appropriate adjustments to my ESS path
in .emacs, I can say that the current svn code base of ESS does
properly work (with respect to skeleton-pair parens) in ESS mode. Note
that I make no adjustment to the default value of ess-r-args-electric-
paren, which in my case is nil as per Rodney's earlier reply.

However, as Rodney also noted, brace "{}" skeleton-pair does not work
in ESS mode and that would be desirable.

HTH,

Marc

On Sep 7, 2009, at 4:03 PM, S. McKay Curtis wrote:

> I changed the lines in my ess-mode.el file, but this did not fix the
> problem.  Has anyone else had success?
>
> On Mon, Aug 31, 2009 at 6:35 AM, Rodney Sparapani <rsparapa@...>
> wrote:
>
>> S. McKay Curtis wrote:
>>
>>> Marc,
>>>
>>> Misery loves company, so I'm glad you were able to replicate my
>>> troubles!
>>>
>>> I am having the problem on Windows XP.
>>>
>>> I'm not an Emacs guru, but the file ess-mode.el contains this
>>> statement
>>>
>>> (define-key ess-mode-map "("  ;; allow to toggle after
>>> customization:
>>>   (if ess-r-args-electric-paren 'ess-r-args-auto-show
>>> 'self-insert-command))
>>>
>>> which looks like it redefines the parenthesis.  Do the lines of code
>>> above negate what I have in my .emacs file?
>>>
>>> Puzzled,
>>> McKay
>>>
>>
>> Sorry, I was on vacation.  But, you will see that this has been fixed
>> in SVN:
>> ;;this is bass-ackwards
>> ;;  (define-key ess-mode-map "("  ;; allow to toggle after
>> customization:
>> ;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
>> 'self-insert-command))
>> (if ess-r-args-electric-paren (define-key ess-mode-map "("
>> 'ess-r-args-auto-show))

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[Marc Schwartz-3]

Rodney,

Just an FYI, that none of my ESS installation has been byte compiled,  
so there would not be a presumptive conflict between the .el and .elc  
versions of a given emacs lisp file. So modifying the .el file should  
have been sufficient and the lack of a behavior change after doing so,  
gave rise to my thinking that using the fully modified code base might  
be required given dependencies elsewhere.

The historical reason for this, in my case, is that when I was on  
Fedora Linux, I went between Emacs version 22 and 23 until 23 became  
stable enough to use all the time. Since byte compiling was not binary  
compatible between the two versions, I simply did not byte compile. I  
have, for better or worse, continued that habit, even now on OSX as I  
have not found a material performance impact, at least in my typical  
use.

Regards,

Marc

On Sep 10, 2009, at 12:09 PM, Sparapani, Rodney wrote:

> It could be that you just needed to M-x byte-compile-file, but glad  
> to hear
> ( is now working.
> ________________________________________
> From: Marc Schwartz [marc_schwartz@...]
> Sent: Wednesday, September 09, 2009 10:47 PM
> To: S. McKay Curtis
> Cc: Sparapani, Rodney; ess-help@...
> Subject: Re: [ESS] Skeleton pair insert not working after upgrade.
>
> Hi all,
>
> Apologies for the delay in my reply here, as I too was on vacation
> last week and am getting back up to date on e-mails and tasks.
>
> As with McKay, simply changing the lines in ess-mode.el did not help.
>
> However, I decided to checkout the svn version of ESS tonight, under
> the presumption that other code has sufficiently changed as to require
> needing the fully updated code base, rather than just changing two
> lines.
>
> Upon doing that and making the appropriate adjustments to my ESS path
> in .emacs, I can say that the current svn code base of ESS does
> properly work (with respect to skeleton-pair parens) in ESS mode. Note
> that I make no adjustment to the default value of ess-r-args-electric-
> paren, which in my case is nil as per Rodney's earlier reply.
>
> However, as Rodney also noted, brace "{}" skeleton-pair does not work
> in ESS mode and that would be desirable.
>
> HTH,
>
> Marc
>
> On Sep 7, 2009, at 4:03 PM, S. McKay Curtis wrote:
>
>> I changed the lines in my ess-mode.el file, but this did not fix the
>> problem.  Has anyone else had success?
>>
>> On Mon, Aug 31, 2009 at 6:35 AM, Rodney Sparapani <rsparapa@...>
>> wrote:
>>
>>> S. McKay Curtis wrote:
>>>
>>>> Marc,
>>>>
>>>> Misery loves company, so I'm glad you were able to replicate my
>>>> troubles!
>>>>
>>>> I am having the problem on Windows XP.
>>>>
>>>> I'm not an Emacs guru, but the file ess-mode.el contains this
>>>> statement
>>>>
>>>> (define-key ess-mode-map "("  ;; allow to toggle after
>>>> customization:
>>>>  (if ess-r-args-electric-paren 'ess-r-args-auto-show
>>>> 'self-insert-command))
>>>>
>>>> which looks like it redefines the parenthesis.  Do the lines of  
>>>> code
>>>> above negate what I have in my .emacs file?
>>>>
>>>> Puzzled,
>>>> McKay
>>>>
>>>
>>> Sorry, I was on vacation.  But, you will see that this has been  
>>> fixed
>>> in SVN:
>>> ;;this is bass-ackwards
>>> ;;  (define-key ess-mode-map "("  ;; allow to toggle after
>>> customization:
>>> ;;    (if ess-r-args-electric-paren 'ess-r-args-auto-show
>>> 'self-insert-command))
>>> (if ess-r-args-electric-paren (define-key ess-mode-map "("
>>> 'ess-r-args-auto-show))
>

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help

[rsparapa]

Marc Schwartz wrote:

> Rodney,
>
> Just an FYI, that none of my ESS installation has been byte compiled, so
> there would not be a presumptive conflict between the .el and .elc
> versions of a given emacs lisp file. So modifying the .el file should
> have been sufficient and the lack of a behavior change after doing so,
> gave rise to my thinking that using the fully modified code base might
> be required given dependencies elsewhere.
>
> The historical reason for this, in my case, is that when I was on Fedora
> Linux, I went between Emacs version 22 and 23 until 23 became stable
> enough to use all the time. Since byte compiling was not binary
> compatible between the two versions, I simply did not byte compile. I
> have, for better or worse, continued that habit, even now on OSX as I
> have not found a material performance impact, at least in my typical use.
>
> Regards,
>
> Marc

Weird.  I don't byte-compile under OS X either (I use the simple
installation instructions).  Anyways, I think I see how to fix {
with skeleton-pair.  It's just an either-or thing:

(defun ess-electric-brace (arg)
   "Insert character and correct line's indentation."
   (interactive "P")
;; skeleton-pair takes precedence
(if (and (boundp 'skeleton-pair) skeleton-pair (fboundp
'skeleton-pair-insert-maybe))
   (skeleton-pair-insert-maybe "{")
;; else
   (let (insertpos)
     (if (and (not arg)
             (eolp)
             (or (save-excursion
                   (skip-chars-backward " \t")
                   (bolp))
                 (if ess-auto-newline (progn (ess-indent-line) (newline) t) nil)))
        (progn
          (insert last-command-char)
          (ess-indent-line)
          (if ess-auto-newline
              (progn
                (newline)
                ;; (newline) may have done auto-fill
                (setq insertpos (- (point) 2))
                (ess-indent-line)))
          (save-excursion
            (if insertpos (goto-char (1+ insertpos)))
            (delete-char -1))))
     (if insertpos
        (save-excursion
          (goto-char insertpos)
          (self-insert-command (prefix-numeric-value arg)))
       (self-insert-command (prefix-numeric-value arg))))))

Now, checked in.  Will appear in ESS 5.5 RSN :o)

--
Rodney Sparapani      Center for Patient Care & Outcomes Research (PCOR)
Sr. Biostatistician              http://www.mcw.edu/pcor
4 wheels good, 2 wheels better!  Medical College of Wisconsin (MCW)
WWLD?:  What Would Lombardi Do?  Milwaukee, WI, USA

______________________________________________
ESS-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/ess-help