bug#67953: 30.0.50; ls-lisp messes up columns

bug#67953: 30.0.50; ls-lisp messes up columns

[Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors]

Package: Emacs
Version: 30.0.50


After loading `ls-lisp`, I notice that with (setq
ls-lisp-use-insert-directory-program t) I get the nice:

    C-x d /u*/s* RET  ==>

      /:
      wildcard u*/s*/
      drwxr-xr-x   2 root root 20480 10 déc 11:24 usr/sbin/
      drwxr-xr-x 404 root root 16384  2 déc 15:05 usr/share/
      drwxr-xr-x   4 root root  4096 22 fév  2023 usr/src/

But with (setq ls-lisp-use-insert-directory-program nil) I get:

    C-x d /u*/s* RET  ==>

      /:
      wildcard u*/s*
      drwxr-xr-x  2 root root 20480 12-10 11:24 usr/sbin
      drwxr-xr-x404 root root 16384 12-02 15:05 usr/share
       drwxr-xr-x  4 root root 4096 2023-02-22  usr/src

Notice the weird extra space in front of the last line and the missing
space between "x" and "404" in the penultimate line.


        Stefan

bug#67953: 30.0.50; ls-lisp messes up columns

[Eli Zaretskii]

> Date: Thu, 21 Dec 2023 09:35:55 -0500
> From:  Stefan Monnier via "Bug reports for GNU Emacs,
>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
> 
> Package: Emacs
> Version: 30.0.50
> 
> 
> After loading `ls-lisp`, I notice that with (setq
> ls-lisp-use-insert-directory-program t) I get the nice:
> 
>     C-x d /u*/s* RET  ==>
> 
>       /:
>       wildcard u*/s*/
>       drwxr-xr-x   2 root root 20480 10 déc 11:24 usr/sbin/
>       drwxr-xr-x 404 root root 16384  2 déc 15:05 usr/share/
>       drwxr-xr-x   4 root root  4096 22 fév  2023 usr/src/
> 
> But with (setq ls-lisp-use-insert-directory-program nil) I get:
> 
>     C-x d /u*/s* RET  ==>
> 
>       /:
>       wildcard u*/s*
>       drwxr-xr-x  2 root root 20480 12-10 11:24 usr/sbin
>       drwxr-xr-x404 root root 16384 12-02 15:05 usr/share
>        drwxr-xr-x  4 root root 4096 2023-02-22  usr/src
> 
> Notice the weird extra space in front of the last line and the missing
> space between "x" and "404" in the penultimate line.

It's a bug in dired-align-file (it sounds like it was never tested
with a version of 'ls' that doesn't support the --dired switch).  Its
algorithm breaks down if the entry of a file does not start with
spaces.  --dired guarantees that, but ls-lisp doesn't, and I wonder
what other versions of 'ls' do.

The simple kludge below, which simply prevents it from realigning the
first column of data, seems to fix it here.  WDYT?

diff --git a/lisp/dired.el b/lisp/dired.el
index cc548ba..3838368 100644
--- a/lisp/dired.el
+++ b/lisp/dired.el
@@ -1572,14 +1572,16 @@ dired-align-file
 		 ;; the beginning or the end of the next field, depending on
 		 ;; whether this field is left or right aligned).
 		 (align-pt-offset
-		  (save-excursion
-		    (goto-char other)
-		    (move-to-column curcol)
-		    (when (looking-at
-			   (concat
-			    (if (eq (char-before) ?\s) " *" "[^ ]* *")
-			    (if num-align "[0-9][^ ]*")))
-		      (- (match-end 0) (match-beginning 0)))))
+                  (if (zerop curcol)
+                      0
+		    (save-excursion
+		      (goto-char other)
+		      (move-to-column curcol)
+		      (when (looking-at
+			     (concat
+			      (if (eq (char-before) ?\s) " *" "[^ ]* *")
+			      (if num-align "[0-9][^ ]*")))
+		        (- (match-end 0) (match-beginning 0))))))
 		 ;; Now, the number of spaces to insert is align-pt-offset
 		 ;; minus the distance to the equivalent point on the
 		 ;; current line.

bug#67953: 30.0.50; ls-lisp messes up columns

[Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors]

> The simple kludge below, which simply prevents it from realigning the
> first column of data, seems to fix it here.  WDYT?

Looks OK to me.  I'd recommend we add a comment about this (zerop
curcol) test explaining why the other branch mishandles this
case (or mentioning that we don't know why).


        Stefan


> diff --git a/lisp/dired.el b/lisp/dired.el
> index cc548ba..3838368 100644
> --- a/lisp/dired.el
> +++ b/lisp/dired.el
> @@ -1572,14 +1572,16 @@ dired-align-file
>  		 ;; the beginning or the end of the next field, depending on
>  		 ;; whether this field is left or right aligned).
>  		 (align-pt-offset
> -		  (save-excursion
> -		    (goto-char other)
> -		    (move-to-column curcol)
> -		    (when (looking-at
> -			   (concat
> -			    (if (eq (char-before) ?\s) " *" "[^ ]* *")
> -			    (if num-align "[0-9][^ ]*")))
> -		      (- (match-end 0) (match-beginning 0)))))
> +                  (if (zerop curcol)
> +                      0
> +		    (save-excursion
> +		      (goto-char other)
> +		      (move-to-column curcol)
> +		      (when (looking-at
> +			     (concat
> +			      (if (eq (char-before) ?\s) " *" "[^ ]* *")
> +			      (if num-align "[0-9][^ ]*")))
> +		        (- (match-end 0) (match-beginning 0))))))
>  		 ;; Now, the number of spaces to insert is align-pt-offset
>  		 ;; minus the distance to the equivalent point on the
>  		 ;; current line.

bug#67953: 30.0.50; ls-lisp messes up columns

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: 67953@debbugs.gnu.org
> Date: Fri, 22 Dec 2023 10:31:07 -0500
> 
> > The simple kludge below, which simply prevents it from realigning the
> > first column of data, seems to fix it here.  WDYT?
> 
> Looks OK to me.  I'd recommend we add a comment about this (zerop
> curcol) test explaining why the other branch mishandles this
> case (or mentioning that we don't know why).

Done.  I decided to install this on master, since this is a very old
problem (I can see it in Emacs 26), and the situations where it rears
its ugly head are quite rare.

Closing.