6:32 a.m.
Hi,
Although it may be a known problem, I think it is evil that the
call-process-region function requires the point to have to be in
the beginning or the end position when the 4th arg DELETEP is t.
It simply erases the buffer's contents when that requirement is
not satisfied. For instance:
(with-temp-buffer
(insert "Hello World")
(goto-char 2)
(call-process-region (point-min) (point-max) "cat" t t)
(buffer-string))
=> ""
Such an accident does not happen in Emacs or XEmacs 21.4. For
this reason, emacs-w3m has not run on XEmacs 21.5 since
2002-05-28 (but it will run by the end of today). I attach a
patch, although it may be too late.
2003-11-21 Katsumi Yamaoka <yamaoka(a)jpl.org>
* code-process.el (call-process-region): Move point to end of
region when DELETEP is non-nil.
--- code-process.el~ 2002-11-10 22:07:29 +0000
+++ code-process.el 2003-11-21 06:30:19 +0000
＠＠ -173,6 +173,8 ＠＠
;; of the region.
(let ((s (and deletep (copy-marker start t)))
(e (and deletep (copy-marker end))))
+ (if deletep
+ (goto-char end))
(let ((retval
(apply #'call-process program (list (current-buffer) start end)
buffer displayp args)))
8:24 a.m.
Katsumi Yamaoka <yamaoka(a)jpl.org> writes:
Although it may be a known problem, I think it is evil that the
call-process-region function requires the point to have to be in
the beginning or the end position when the 4th arg DELETEP is t.
It sounds like a bug.
It simply erases the buffer's contents when that requirement is
not satisfied.
Why does that happen? The code looks like it's trying to correctly
track deletions and insertions by using markers.
It could be that your fix hides another bug, elsewhere in the code,
but I could be mistaken. What do you think?
8:50 a.m.
>>>> In [emacs-w3m : No.06092]
>>>> Hrvoje Niksic <hniksic(a)xemacs.org> wrote:
Katsumi Yamaoka <yamaoka(a)jpl.org> writes:
> Although it may be a known problem, I think it is evil that the
> call-process-region function requires the point to have to be in
> the beginning or the end position when the 4th arg DELETEP is t.
It sounds like a bug.
> It simply erases the buffer's contents when that requirement
is
> not satisfied.
Why does that happen? The code looks like it's trying to
correctly
track deletions and insertions by using markers.
It is because call-process-region of that version uses
call-process which inserts the output of a process in the point.
And it deletes the region which is pointed to by markers, after
a process is finished. Markers function rightly. Therefore,
they will be erased if the output of a process is inserted
within limits which markers point to.
It could be that your fix hides another bug, elsewhere in the code,
but I could be mistaken. What do you think?
Probably, the problem is simple and lucid.
--
Katsumi Yamaoka <yamaoka(a)jpl.org>
10:20 a.m.
>>>> In <b9yhe0yqf4i.fsf(a)jpl.org> Katsumi Yamaoka
wrote:
It is because call-process-region of that version uses
call-process which inserts the output of a process in the point.
And it deletes the region which is pointed to by markers, after
a process is finished. Markers function rightly. Therefore,
they will be erased if the output of a process is inserted
within limits which markers point to.
Oops! Using markers is needless if the points aren't changed
while processing. Here is a new patch:
--- code-process.el~ 2002-11-10 22:07:29 +0000
+++ code-process.el 2003-11-21 10:14:54 +0000
＠＠ -171,15 +171,13 ＠＠
;; in the same buffer, make sure we track the insertion, and don't get
;; any of it in the deleted region if insertion happens at either end
;; of the region.
- (let ((s (and deletep (copy-marker start t)))
- (e (and deletep (copy-marker end))))
- (let ((retval
- (apply #'call-process program (list (current-buffer) start end)
- buffer displayp args)))
- ;; If start and end were the same originally, s will be beyond e now
- (if (and deletep (> e s))
- (delete-region s e))
- retval)))
+ (if deletep
+ (goto-char end))
+ (prog1
+ (apply #'call-process program (list (current-buffer) start end)
+ buffer displayp args)
+ (if deletep
+ (delete-region start end))))
(defun start-process (name buffer program &rest program-args)
"Start a program in a subprocess. Return the process object for it.
Probably, another function should be made if Ben wanted to write
that code for the particular purpose.
--
Katsumi Yamaoka <yamaoka(a)jpl.org>
1:03 a.m.
Maintainers,
Could you please clarify whether it is a bug or specification?
If it is not a bug, it is necessary to add notes, isn't it?
>>>> In <b9y65hez0yf.fsf(a)jpl.org> Katsumi Yamaoka
wrote:
Hi,
Although it may be a known problem, I think it is evil that the
call-process-region function requires the point to have to be in
the beginning or the end position when the 4th arg DELETEP is t.
It simply erases the buffer's contents when that requirement is
not satisfied. For instance:
(with-temp-buffer
(insert "Hello World")
(goto-char 2)
(call-process-region (point-min) (point-max) "cat" t t)
(buffer-string))
=> ""
Such an accident does not happen in Emacs or XEmacs 21.4.
>>>> In <b9y7k1uqaz3.fsf(a)jpl.org> Katsumi Yamaoka
wrote:
--- code-process.el~ 2002-11-10 22:07:29 +0000
+++ code-process.el 2003-11-21 10:14:54 +0000
＠＠ -171,15 +171,13 ＠＠
;; in the same buffer, make sure we track the insertion, and don't get
;; any of it in the deleted region if insertion happens at either end
;; of the region.
- (let ((s (and deletep (copy-marker start t)))
- (e (and deletep (copy-marker end))))
- (let ((retval
- (apply #'call-process program (list (current-buffer) start end)
- buffer displayp args)))
- ;; If start and end were the same originally, s will be beyond e now
- (if (and deletep (> e s))
- (delete-region s e))
- retval)))
+ (if deletep
+ (goto-char end))
+ (prog1
+ (apply #'call-process program (list (current-buffer) start end)
+ buffer displayp args)
+ (if deletep
+ (delete-region start end))))
(defun start-process (name buffer program &rest program-args)
"Start a program in a subprocess. Return the process object for it.
2:49 a.m.
Ben?
>>>> "KY" == Katsumi Yamaoka
<yamaoka(a)jpl.org> writes:
Maintainers,
Could you please clarify whether it is a bug or specification?
If it is not a bug, it is necessary to add notes, isn't it?
>>>> In <b9y65hez0yf.fsf(a)jpl.org> Katsumi Yamaoka
wrote:
Hi,
Although it may be a known problem, I think it is evil that the
call-process-region function requires the point to have to be in
the beginning or the end position when the 4th arg DELETEP is t.
It simply erases the buffer's contents when that requirement is
not satisfied. For instance:
(with-temp-buffer
(insert "Hello World")
(goto-char 2)
(call-process-region (point-min) (point-max) "cat" t t)
(buffer-string))
=> ""
Such an accident does not happen in Emacs or XEmacs 21.4.
>>>> In <b9y7k1uqaz3.fsf(a)jpl.org> Katsumi Yamaoka
wrote:
--- code-process.el~ 2002-11-10 22:07:29 +0000
+++ code-process.el 2003-11-21 10:14:54 +0000
＠＠ -171,15 +171,13 ＠＠
;; in the same buffer, make sure we track the insertion, and don't get
;; any of it in the deleted region if insertion happens at either end
;; of the region.
- (let ((s (and deletep (copy-marker start t)))
- (e (and deletep (copy-marker end))))
- (let ((retval
- (apply #'call-process program (list (current-buffer) start end)
- buffer displayp args)))
- ;; If start and end were the same originally, s will be beyond e now
- (if (and deletep (> e s))
- (delete-region s e))
- retval)))
+ (if deletep
+ (goto-char end))
+ (prog1
+ (apply #'call-process program (list (current-buffer) start end)
+ buffer displayp args)
+ (if deletep
+ (delete-region start end))))
(defun start-process (name buffer program &rest program-args)
"Start a program in a subprocess. Return the process object for it.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
7470
days inactive
7475
days old
5 comments
3 participants
participants (3)
-
Hrvoje Niksic -
Katsumi Yamaoka -
Stephen J. Turnbull