rename-file

rename-file

[Sam Steingold]

Why isn't the buffer, which is visiting a file, renamed
when I rename the underlying file with rename-file?

Re: rename-file

[Stephen Berman]

On Wed, 26 Aug 2009 12:05:06 -0400 Sam Steingold <sds@gnu.org> wrote:

> Why isn't the buffer, which is visiting a file, renamed
> when I rename the underlying file with rename-file?

Not only that, but subsequently doing `C-h v buffer-file-name' on that
buffer returns the previous name, though that file(name) no longer
exists.

Steve Berman

Re: rename-file

[Stefan Monnier]

>> Why isn't the buffer, which is visiting a file, renamed
>> when I rename the underlying file with rename-file?

> Not only that, but subsequently doing `C-h v buffer-file-name' on that
> buffer returns the previous name, though that file(name) no longer
> exists.

That's perfectly OK.  It's exactly as if someone had done "mv" from
outside Emacs (or from a M-x shell buffer).  There's nothing wrong with
such a situation.  C-x C-s will (re)create the file.


        Stefan

Re: rename-file

[Stephen Berman]

On Wed, 26 Aug 2009 23:16:08 -0400 Stefan Monnier <monnier@iro.umontreal.ca> wrote:

>>> Why isn't the buffer, which is visiting a file, renamed
>>> when I rename the underlying file with rename-file?
>
>> Not only that, but subsequently doing `C-h v buffer-file-name' on that
>> buffer returns the previous name, though that file(name) no longer
>> exists.
>
> That's perfectly OK.  It's exactly as if someone had done "mv" from
> outside Emacs (or from a M-x shell buffer).  There's nothing wrong with
> such a situation.  C-x C-s will (re)create the file.

I should have left out the though-clause above, it was a misguided
afterthought.  What's problematic is that returning the previous name
contradicts the doc string of buffer-file-name: "Name of file visited in
current buffer, or nil if not visiting a file."

Steve Berman

Re: rename-file

[Andreas Schwab]

Stephen Berman <stephen.berman@gmx.net> writes:

> What's problematic is that returning the previous name contradicts the
> doc string of buffer-file-name: "Name of file visited in current
> buffer, or nil if not visiting a file."

That's not a contradiction.  All that buffer-file-name defines is a
relationship between a buffer and a file name, independent of the
current meaning of that name.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: rename-file

[Stephen Berman]

On Thu, 27 Aug 2009 11:09:45 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote:

> Stephen Berman <stephen.berman@gmx.net> writes:
>
>> What's problematic is that returning the previous name contradicts the
>> doc string of buffer-file-name: "Name of file visited in current
>> buffer, or nil if not visiting a file."
>
> That's not a contradiction.  All that buffer-file-name defines is
> a relationship between a buffer and a file name, independent of the
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> current meaning of that name.

If that is so, then the doc string should be changed to "A name of the
file visited in the current buffer...", to avoid confusion.  As it is, I
understand it to be equivalent to "The name of the file visited in the
current buffer..." (and indeed, "the name" is used in the comment above
the filename field of the buffer struct declaration in buffer.h), and
the name shown (after changing it with rename-file), i.e. the value of
buffer-file-name, is incorrect.  More explicitly:

1. C-x C-f bla
2. C-h v buffer-file-name => bla
3. M-x rename-file RET blabla
4. C-h v buffer-file-name => bla

Surely the return value in step 4 is unwanted, isn't it?

Steve Berman

Re: rename-file

[Andreas Schwab]

Stephen Berman <stephen.berman@gmx.net> writes:

> If that is so, then the doc string should be changed to "A name of the
> file visited in the current buffer...", to avoid confusion.

No, that's very wrong.  There can be only at most one name associated
with any buffer.

> 1. C-x C-f bla
> 2. C-h v buffer-file-name => bla
> 3. M-x rename-file RET blabla
> 4. C-h v buffer-file-name => bla
>
> Surely the return value in step 4 is unwanted, isn't it?

The visited name of the buffer didn't change in any way whatsoever.
Only set-visited-file-name can do that.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: rename-file

[Stephen Berman]

On Thu, 27 Aug 2009 12:21:36 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote:

>                         There can be only at most one name associated
> with any buffer.

Isn't it rather "exactly one file name"?  Or is it possible to have
nameless buffers?

>
>> 1. C-x C-f bla
>> 2. C-h v buffer-file-name => bla
>> 3. M-x rename-file RET blabla
>> 4. C-h v buffer-file-name => bla
>>
>> Surely the return value in step 4 is unwanted, isn't it?
>
> The visited name of the buffer didn't change in any way whatsoever.

But that's precisely the problem!

> Only set-visited-file-name can do that.

Then it should be made to take effect between steps 3 and 4 above.

I don't understand how anyone can deny that evaluating the variable
whose value is documented as the "Name of file visited in current
buffer, or nil if not visiting a file" should return the correct current
name of the visited file.

Steve Berman

Re: rename-file

[Miles Bader]

Stephen Berman <stephen.berman@gmx.net> writes:
> I don't understand how anyone can deny that evaluating the variable
> whose value is documented as the "Name of file visited in current
> buffer, or nil if not visiting a file" should return the correct current
> name of the visited file.

Because "rename-file" is a relatively low-level command, which just
tells the OS to rename the file.

If you want "smart" renaming, use dired to rename files ("R" command).
It will update buffer filenames (and buffer names) accordingly.

While the latter behavior is convenient, there's certainly nothing
_wrong_ with the former -- there's no guarantee that a buffer's filename
variable refers an existing file at all (and really there can't be,
since in the general case, emacs can't keep track of what goes on in the
filesystem behind its back).

-Miles

-- 
"An atheist doesn't have to be someone who thinks he has a proof that
there can't be a god.  He only has to be someone who believes that the
evidence on the God question is at a similar level to the evidence on
the werewolf question."  [John McCarthy]

Re: rename-file

[Stephen Berman]

On Thu, 27 Aug 2009 20:50:24 +0900 Miles Bader <miles@gnu.org> wrote:

> Stephen Berman <stephen.berman@gmx.net> writes:
>> I don't understand how anyone can deny that evaluating the variable
>> whose value is documented as the "Name of file visited in current
>> buffer, or nil if not visiting a file" should return the correct current
>> name of the visited file.
>
> Because "rename-file" is a relatively low-level command, which just
> tells the OS to rename the file.
>
> If you want "smart" renaming, use dired to rename files ("R" command).
> It will update buffer filenames (and buffer names) accordingly.
>
> While the latter behavior is convenient, there's certainly nothing
> _wrong_ with the former -- there's no guarantee that a buffer's filename
> variable refers an existing file at all (and really there can't be,
> since in the general case, emacs can't keep track of what goes on in the
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> filesystem behind its back).
  ^^^^^^^^^^^^^^^^^^^^^^^^^^

I understand the Emacs can't be expected to track filesystem changes
that happen outside its purview (but sometimes it does anyway,
e.g. auto-revert-mode), but my concern is about changes due to a
sequence of Emacs commands (steps 1-4 below), which result in a value
for buffer-file-name that to my understanding is not consistent with its
doc string.  I wouldn't have complained if the doc string had noted this
possibility, e.g.:

"Name of file visited in current buffer, or nil if not visiting a file.
Note that a non-nil value of this variable does not change if the file
name changed after visiting the file."

On Thu, 27 Aug 2009 14:18:03 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote:

> Stephen Berman <stephen.berman@gmx.net> writes:
>
>>>> 1. C-x C-f bla
>>>> 2. C-h v buffer-file-name => bla
>>>> 3. M-x rename-file RET blabla
>>>> 4. C-h v buffer-file-name => bla
>>>>
>>>> Surely the return value in step 4 is unwanted, isn't it?
>>>
>>> The visited name of the buffer didn't change in any way whatsoever.
>>
>> But that's precisely the problem!
>
> There is no assotiation between a buffer and a file, only between a
> buffer and file name.
>
>>> Only set-visited-file-name can do that.
>>
>> Then it should be made to take effect between steps 3 and 4 above.
>
> No.  That the filesystem changes does not in any way change the
       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> association of a buffer to a file name, just like deleting a file does
  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> not magically kill a buffer.

That appears to depend on the command or function that brings about the
filesystem change, e.g. write-file does change the association of a
buffer to a file name:

5. C-h v buffer-file-name => blip
6. C-x C-w blap (i.e. at the prompt type "blap" after default-directory)
7. C-h v buffer-file-name => "/<default-directory>/blap"
8. There is no longer a buffer visiting "/<default-directory>/blip"

If write-file does it, there's no reason in principle why rename-file
cannot do it, so if (or since) it (currently) does not change the
association, the doc string should note this.  Or is there a convention
that an Emacs command does not track file system changes unless its doc
string explicitly says it does?  If so, is this convention documented or
assumed as common lore?

Steve Berman

Re: rename-file

[Andreas Schwab]

Stephen Berman <stephen.berman@gmx.net> writes:

> On Thu, 27 Aug 2009 12:21:36 +0200 Andreas Schwab <schwab@linux-m68k.org> wrote:
>
>>                         There can be only at most one name associated
>> with any buffer.
>
> Isn't it rather "exactly one file name"?  Or is it possible to have
> nameless buffers?

Sure.  See the docstring of buffer-file-name.

>>> 1. C-x C-f bla
>>> 2. C-h v buffer-file-name => bla
>>> 3. M-x rename-file RET blabla
>>> 4. C-h v buffer-file-name => bla
>>>
>>> Surely the return value in step 4 is unwanted, isn't it?
>>
>> The visited name of the buffer didn't change in any way whatsoever.
>
> But that's precisely the problem!

There is no assotiation between a buffer and a file, only between a
buffer and file name.

>> Only set-visited-file-name can do that.
>
> Then it should be made to take effect between steps 3 and 4 above.

No.  That the filesystem changes does not in any way change the
association of a buffer to a file name, just like deleting a file does
not magically kill a buffer.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: rename-file

[Stefan Monnier]

> I should have left out the though-clause above, it was a misguided
> afterthought.  What's problematic is that returning the previous name
> contradicts the doc string of buffer-file-name: "Name of file visited in
> current buffer, or nil if not visiting a file."

The docstring sounds correct to me.  If you rename the underlying file,
it means the buffer's file name will refer to some other file (or no
file at all).


        Stefan

Re: rename-file

[Stefan Monnier]

> Why isn't the buffer, which is visiting a file, renamed
> when I rename the underlying file with rename-file?

Because rename-file is a low-level function.  Maybe we should introduce
a new command rename-file-buffer which automatically applies to the
current buffer's file.


        Stefan

Re: rename-file

[Lennart Borgman]

On Wed, Aug 26, 2009 at 9:21 PM, Stefan Monnier<monnier@iro.umontreal.ca> wrote:
>> Why isn't the buffer, which is visiting a file, renamed
>> when I rename the underlying file with rename-file?
>
> Because rename-file is a low-level function.  Maybe we should introduce
> a new command rename-file-buffer which automatically applies to the
> current buffer's file.

Yes, that would be good because it would be less confusing.

Re: rename-file

[Sam Steingold]

On Wed, Aug 26, 2009 at 3:21 PM, Stefan Monnier<monnier@iro.umontreal.ca> wrote:
>> Why isn't the buffer, which is visiting a file, renamed
>> when I rename the underlying file with rename-file?
>
> Because rename-file is a low-level function.  Maybe we should introduce
> a new command rename-file-buffer which automatically applies to the
> current buffer's file.

OK to commit?

(defun rename-buffer-file (new-name)
  "Rename the current buffer and the file it is visiting to NEW-NAME."
  (interactive "FRename current file and buffer to: ")
  (let ((filename (or buffer-file-name
                      (error "This buffer is not visiting a file")))
        (oldbuf (find-buffer-visiting new-name)))
    (when (or (null oldbuf)
              (y-or-n-p "Buffer %s is already visiting %s, proceed? "
                        oldbuf new-name))
      (rename-file filename new-name 1)
      (set-visited-file-name new-name nil t))))

-- 
Sam Steingold <http://sds.podval.org>

Re: rename-file

[Eli Zaretskii]

> Date: Wed, 26 Aug 2009 16:35:31 -0400
> From: Sam Steingold <sds@gnu.org>
> Cc: emacs-devel@gnu.org
> 
> > Because rename-file is a low-level function.  Maybe we should introduce
> > a new command rename-file-buffer which automatically applies to the
> > current buffer's file.
> 
> OK to commit?
> 
> (defun rename-buffer-file (new-name)
>   "Rename the current buffer and the file it is visiting to NEW-NAME."
>   (interactive "FRename current file and buffer to: ")
>   (let ((filename (or buffer-file-name
>                       (error "This buffer is not visiting a file")))
>         (oldbuf (find-buffer-visiting new-name)))
>     (when (or (null oldbuf)
>               (y-or-n-p "Buffer %s is already visiting %s, proceed? "
>                         oldbuf new-name))
>       (rename-file filename new-name 1)
>       (set-visited-file-name new-name nil t))))

Will this work in a Dired buffer, especially if there are files in
that directory visited by other buffers?

What about other buffers visiting the same file?

Re: rename-file

[Richard Stallman]

    (defun rename-buffer-file (new-name)
      "Rename the current buffer and the file it is visiting to NEW-NAME."
      (interactive "FRename current file and buffer to: ")

Given the existence of the command `rename-buffer',
this name will cause confusion.

Does it rename the buffer, or does it just set the visited file name?
The code seems to do the latter.  If so, the doc string is inaccurate.

Re: rename-file

[Sam Steingold]

On Wed, Aug 26, 2009 at 11:25 PM, Richard Stallman<rms@gnu.org> wrote:
>    (defun rename-buffer-file (new-name)
>      "Rename the current buffer and the file it is visiting to NEW-NAME."
>      (interactive "FRename current file and buffer to: ")
>
> Given the existence of the command `rename-buffer',
> this name will cause confusion.

please suggest a better name.

> Does it rename the buffer, or does it just set the visited file name?
> The code seems to do the latter.  If so, the doc string is inaccurate.

doc string says:

rename the _BUFFER_ and the _FILE_.
it renames the file and keeps the buffer associated with it.

-- 
Sam Steingold <http://sds.podval.org>

Re: rename-file

[Stefan Monnier]

> doc string says:

> rename the _BUFFER_ and the _FILE_.
> it renames the file and keeps the buffer associated with it.

BTW, the issue is not whether it renames the buffer, but wether it
renames the buffer's buffer-file-name.


        Stefan


PS: I hear dired's R does such a thing, so we should consolidate
that code.

Re: rename-file

[Richard Stallman]

    > Given the existence of the command `rename-buffer',
    > this name will cause confusion.

    please suggest a better name.

I can't suggest one, since I am not sure what this function really
does.

    > Does it rename the buffer, or does it just set the visited file name?
    > The code seems to do the latter.  If so, the doc string is inaccurate.

    doc string says:

    rename the _BUFFER_ and the _FILE_.
    it renames the file and keeps the buffer associated with it.

Then the code could use some more comments.  Where does the buffer
name get changed?

Re: rename-file

[Sam Steingold]

Richard Stallman wrote:
> 
>     rename the _BUFFER_ and the _FILE_.
>     it renames the file and keeps the buffer associated with it.
> 
> Where does the buffer
> name get changed?

in set-visited-file-name

Re: rename-file

[Sam Steingold]

Sam Steingold wrote:
> On Wed, Aug 26, 2009 at 3:21 PM, Stefan Monnier<monnier@iro.umontreal.ca> wrote:
>>> Why isn't the buffer, which is visiting a file, renamed
>>> when I rename the underlying file with rename-file?
>> Because rename-file is a low-level function.  Maybe we should introduce
>> a new command rename-file-buffer which automatically applies to the
>> current buffer's file.
> 
> OK to commit?
> 
> (defun rename-buffer-file (new-name)
>   "Rename the current buffer and the file it is visiting to NEW-NAME."
>   (interactive "FRename current file and buffer to: ")
>   (let ((filename (or buffer-file-name
>                       (error "This buffer is not visiting a file")))
>         (oldbuf (find-buffer-visiting new-name)))
>     (when (or (null oldbuf)
>               (y-or-n-p "Buffer %s is already visiting %s, proceed? "
>                         oldbuf new-name))
>       (rename-file filename new-name 1)
>       (set-visited-file-name new-name nil t))))
> 

on a second thought I decided that R in dired + M-x vc-rename-file are already
sufficient. sorry about the noise.

Re: rename-file

[martin rudalics]

 > Why isn't the buffer, which is visiting a file, renamed
 > when I rename the underlying file with rename-file?

Personally, I do the following when renaming a file or a directory:
First I do a `bookmark-set-filename' on every involved buffer.  Next I
do (the precise reasons I don't remember) for any BUFFER whose filename
is set to NEW

         (with-current-buffer buffer
           (let ((buffer-modified-p (buffer-modified-p)))
             ;; make buffer visit new
             (set-visited-file-name new nil t)
             (clear-visited-file-modtime)
             (set-buffer-modified-p buffer-modified-p)))))

`dired-rename-file' does something similar, but I never looked into
that.

Finally, I do a redisplay to assure that modelines and frame titles get
updated correctly.  When renaming directories I also set the default
directory, if necessary.

All these are sufficient for my personal needs and have not caused any
problems over the past five years.  But I'm afraid there's a number of
more hairy issues involved when a package has stored away some buffers'
names in a local list.

martin

Re: rename-file

[Daniel Colascione]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Aug 26, 2009, at 12:05 PM, Sam Steingold wrote:
> Why isn't the buffer, which is visiting a file, renamed
> when I rename the underlying file with rename-file?

Here's what I use. It's DTRT for years. (I have it bound to C-x m.)
Something like this, IMHO, should be the user-level rename-file.


(defun qtmstr-rename-file (new)
  "Rename file OLD to NEW, renaming the buffer as well.

If the file is under version control, act like vc-rename-file."

  (interactive "FRename to: ")

  (let ((oldbuf (current-buffer))
        (oldbufname (buffer-name))
        (old (buffer-file-name))
        vc)

    (if (and oldbuf (buffer-modified-p oldbuf))
	(error "Please save files before moving them"))
    (if (get-file-buffer new)
	(error "Already editing new file name"))
    (if (file-exists-p new)
	(error "New file already exists"))

    (let ((state (vc-state old)))
      (cond ((memq state '(up-to-date edited))
             (setq vc t)
             (vc-call rename-file old new)
             (vc-file-clearprops old))

            (state
             (error "Please %s files before moving them"
                    (if (stringp state) "check in" "update")))))

    (if (file-exists-p old) (rename-file old new))

    (with-current-buffer oldbuf
      (let ((buffer-read-only buffer-read-only))
        (set-visited-file-name new))
      (when vc
        (vc-backend new)
        (vc-mode-line new))
      (set-buffer-modified-p nil))
    (message "Renamed %s to %s%s"
             oldbufname new (if vc " [with VC]" ""))))
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Darwin)

iEYEARECAAYFAkqWjswACgkQ17c2LVA10VtOywCgxhjjfavlB0qDDKMKgR1D1llN
o3IAoIICnOzZbDqP+RBpT3m2tPyXg935
=4l7w
-----END PGP SIGNATURE-----