autoc-apitalize blocking call to accept-process-output quit inhibited

autoc-apitalize blocking call to accept-process-output quit inhibited

[Uwe Brauer]

Hello

I have used till emacs 23 successfully auto-capitalize (by Kevin
Rodgers) but now it stopped working and I obtain messages blocking call
to accept-process-output quit inhibited

Also Dave Love's ;;; autocap.el 

does not work anymore.

Anybody has an idea?

thanks

Uwe Brauer 

Re: autoc-apitalize blocking call to accept-process-output quit inhibited

[Stefan Monnier]

> I have used till emacs 23 successfully auto-capitalize (by Kevin
> Rodgers) but now it stopped working and I obtain messages blocking call
> to accept-process-output quit inhibited

While the two may be related, the "blocking call to
accept-process-output" is just a warning message (i.e. shouldn't affect
the actual behavior) used to indicate that the code could potentially
freeze Emacs


        Stefan

Re: autoc-apitalize blocking call to accept-process-output quit inhibited

[Uwe Brauer]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

      > > I have used till emacs 23 successfully auto-capitalize (by Kevin
      > > Rodgers) but now it stopped working and I obtain messages blocking call
      > > to accept-process-output quit inhibited
>
       > While the two may be related, the "blocking call to
       > accept-process-output" is just a warning message (i.e. shouldn't affect
       > the actual behavior) used to indicate that the code could potentially
       > freeze Emacs

OK, but as a matter of fact the auto-cap does not work, neither for
auto-capitalize.el nor for the one based on abbrevs.
(Strangely enough it still works in Xemacs 21.5.X, although both authors
were GNU emacs users).

I found this a wonderful feature, which I miss in most other
applications, but I seem to be the only one....
capitalize.

Re: autoc-apitalize blocking call to accept-process-output quit inhibited

[Uwe Brauer]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>
       > While the two may be related, the "blocking call to
       > accept-process-output" is just a warning message
       > (i.e. shouldn't affect the actual behavior) used to indicate
       > that the code could potentially freeze Emacs

Concerning auto-capitalize.el another error Iobtain is the following:

Debugger entered--Lisp error: (wrong-type-argument stringp nil)
  re-search-backward(nil nil t)
  auto-capitalize(28 29 0)
  self-insert-command(1)
  call-interactively(self-insert-command nil nil)
 

Re: auto-capitalize blocking call to accept-process-output quit inhibited

[Stefan Monnier]

>> While the two may be related, the "blocking call to
>> accept-process-output" is just a warning message
>> (i.e. shouldn't affect the actual behavior) used to indicate
>> that the code could potentially freeze Emacs

> Concerning auto-capitalize.el another error Iobtain is the following:

Since the accept-process-output thingy is a warning (and not an error),
this one is not "another" error.

> Debugger entered--Lisp error: (wrong-type-argument stringp nil)
>   re-search-backward(nil nil t)
>   auto-capitalize(28 29 0)
>   self-insert-command(1)
>   call-interactively(self-insert-command nil nil)
 
If you load auto-capitalize.el instead of auto-capitalize.elc, the
backtrace will give more detailed info.


        Stefan

Re: autoc-apitalize blocking call to accept-process-output quit inhibited

[Kevin Rodgers]

On 6/26/13 10:28 AM, Uwe Brauer wrote:
> Concerning auto-capitalize.el another error Iobtain is the following:
>
> Debugger entered--Lisp error: (wrong-type-argument stringp nil)
>    re-search-backward(nil nil t)
>    auto-capitalize(28 29 0)
>    self-insert-command(1)
>    call-interactively(self-insert-command nil nil)

There are 3 calls to re-search-backward in auto-capitalize.el, all with nil and
t as the optional 2nd and 3rd args respectively.

The 1st call assumes that paragraph-start is a valid non-nil regexp, the 2nd
call assumes that sentence-end is a valid non-nil regexp, and the 3rd call
passes abbrev-regexp which it has just bound to a constant string.

So either paragraph-start or sentence-end is null in the buffer where the
auto-capitalize function is running.  I suppose those variable references
should be changed to something like these to try to avoid that error:

(or paragraph-start (default-value 'paragraph-start))

(or sentence-end (and (fboundp 'sentence-end) (sentence-end))

-- 
Kevin Rodgers
Denver, Colorado, USA

Re: autoc-apitalize blocking call to accept-process-output quit inhibited

[Uwe Brauer]

Kevin Rodgers <kevin.d.rodgers@gmail.com> writes:

       > On 6/26/13 10:28 AM, Uwe Brauer wrote:
      > > Concerning auto-capitalize.el another error Iobtain is the following:
      > >

       > There are 3 calls to re-search-backward in auto-capitalize.el,
       > all with nil and t as the optional 2nd and 3rd args
       > respectively.

       > The 1st call assumes that paragraph-start is a valid non-nil
       > regexp, the 2nd call assumes that sentence-end is a valid
       > non-nil regexp, and the 3rd call passes abbrev-regexp which it
       > has just bound to a constant string.

       > So either paragraph-start or sentence-end is null in the buffer
       > where the auto-capitalize function is running.  I suppose those
       > variable references should be changed to something like these
       > to try to avoid that error:

       > (or paragraph-start (default-value 'paragraph-start))

       > (or sentence-end (and (fboundp 'sentence-end) (sentence-end)))
                                                                     ^^^
there is a ) missing, no? In any case, these two parts made the
difference!!! Auto capitalize works again, the only issue is this very
annoying message described earlier!

Thanks for the patch!


Uwe