M-x dired and read-file-name-completion-ignore-case

M-x dired and read-file-name-completion-ignore-case

[Stephen Berman]

Since a fairly recent change, `C-x d' (i.e. `M-x dired') no longer
ignores case when read-file-name-completion-ignore-case is non-nil.  I
suspect this change:

2008-04-29  Stefan Monnier  <monnier@iro.umontreal.ca>
        [...]
	* dired.el (dired-read-dir-and-switches): Replace last change with
	a new approach that mixes read-file-name and read-directory-name.

(This change eliminates the call to read-directory-name (unless
next-read-file-uses-dialog-p is non-nil), which eliminates the call to
read-file-name, which checks read-file-name-completion-ignore-case.)  Is
this result intended?  If so, please mention it in NEWS, since it is
incompatible with earlier behavior.  (But I would prefer to have the
earlier behavior restored.)

Steve Berman

Re: M-x dired and read-file-name-completion-ignore-case

[Stefan Monnier]

> Since a fairly recent change, `C-x d' (i.e. `M-x dired') no longer
> ignores case when read-file-name-completion-ignore-case is non-nil.  I
> suspect this change:

> 2008-04-29  Stefan Monnier  <monnier@iro.umontreal.ca>
>         [...]
> 	* dired.el (dired-read-dir-and-switches): Replace last change with
> 	a new approach that mixes read-file-name and read-directory-name.

The above patch should fix it,


        Stefan


--- dired.el.~1.393.~	2008-05-02 13:38:03.000000000 -0400
+++ dired.el	2008-05-05 15:26:27.000000000 -0400
@@ -607,6 +607,7 @@
             (lambda ()
               (setq minibuffer-default default)
               (setq minibuffer-completing-file-name t)
+              (setq completion-ignore-case read-file-name-completion-ignore-case)
               (setq default-directory defdir))
           (substitute-in-file-name
            (completing-read

Re: M-x dired and read-file-name-completion-ignore-case

[John Paul Wallington]

On 5 May 2008, at 20:27, Stefan Monnier wrote:

> The above patch should fix it,
>
>
>        Stefan
>
>
> --- dired.el.~1.393.~	2008-05-02 13:38:03.000000000 -0400
> +++ dired.el	2008-05-05 15:26:27.000000000 -0400
> @@ -607,6 +607,7 @@
>             (lambda ()
>               (setq minibuffer-default default)
>               (setq minibuffer-completing-file-name t)
> +              (setq completion-ignore-case read-file-name- 
> completion-ignore-case)
>               (setq default-directory defdir))
>           (substitute-in-file-name
>            (completing-read


I don't think that change is good because it globally clobbers  
`completion-ignore-case' rather then let-binding it. How about  
installing the following fix?

(diff -uw output, to ignore changing indentation):

--- dired.el.~1.394.~   2008-05-05 23:36:20.000000000 +0100
+++ dired.el    2008-05-06 00:04:08.000000000 +0100
@@ -607,9 +607,8 @@
              (lambda ()
                (setq minibuffer-default default)
                (setq minibuffer-completing-file-name t)
-              (setq completion-ignore-case
-                    read-file-name-completion-ignore-case)
                (setq default-directory defdir))
+         (let ((completion-ignore-case read-file-name-completion- 
ignore-case))
            (substitute-in-file-name
             (completing-read
              (format "Dired %s(directory): " str)
@@ -628,7 +627,7 @@
                   (complete-with-action
                    action 'read-file-name-internal str nil)))
               'read-file-name-internal)
-            nil nil (abbreviate-file-name defdir) 'file-name- 
history))))))))
+             nil nil (abbreviate-file-name defdir) 'file-name- 
history)))))))))

  ;;;###autoload (define-key ctl-x-map "d" 'dired)
  ;;;###autoload 

Re: M-x dired and read-file-name-completion-ignore-case

[Stefan Monnier]

> I don't think that change is good because it globally clobbers
> completion-ignore-case' rather then let-binding it. How about  installing
> the following fix?

Looks like you're right.  I believe I've fixed it now, tho differently.


        Stefan

Re: M-x dired and read-file-name-completion-ignore-case

[Stephen Berman]

On Mon, 05 May 2008 20:58:52 -0400 Stefan Monnier <monnier@iro.umontreal.ca> wrote:

>> I don't think that change is good because it globally clobbers
>> completion-ignore-case' rather then let-binding it. How about  installing
>> the following fix?
>
> Looks like you're right.  I believe I've fixed it now, tho differently.

Thanks, I confirm it's fixed.

Steve Berman