Toggle EOL

Toggle EOL

[Stefan Monnier]

I think that the EOL information in the mode-line should be
better separated from the coding-system info.  The patch below
gives it its own help-echo and its own keymap (mouse-3 switches
between the three different possible EOL).

Any comment/objection ?


	Stefan


Index: bindings.el
===================================================================
RCS file: /cvsroot/emacs/emacs/lisp/bindings.el,v
retrieving revision 1.100
diff -u -u -b -r1.100 bindings.el
--- bindings.el	16 Apr 2002 14:55:13 -0000	1.100
+++ bindings.el	28 Jun 2002 17:01:15 -0000
@@ -135,6 +135,36 @@
   "Local keymap for the coding-system part of the mode line.")
 
 
+(defvar mode-line-eol-map
+  (let ((map (make-sparse-keymap)))
+    (define-key map [mode-line mouse-3]
+      (lambda ()
+	(interactive)
+	(let ((eol (coding-system-eol-type buffer-file-coding-system)))
+	  (set-buffer-file-coding-system
+	   (cond ((eq eol 0) 'dos) ((eq eol 1) 'mac) (t 'unix))))))
+    (purecopy map))
+  "Local keymap for the `coding-system-eol-type' part of the mode line.")
+
+(defvar mode-line-eol-desc-cache nil)
+
+(defun mode-line-eol-desc ()
+  (let* ((eol (coding-system-eol-type buffer-file-coding-system))
+	 (mnemonic (coding-system-eol-type-mnemonic buffer-file-coding-system))
+	 (desc (assq eol mode-line-eol-desc-cache)))
+    (if (and desc (eq (cadr desc) mnemonic))
+	(cddr desc)
+      (setq desc
+	    (propertize
+	     mnemonic
+	     'help-echo (format "%s end-of-line; mouse-3 to change"
+				(if (eq eol 0) "Unix-style LF"
+				  (if (eq eol 1) "Dos-style CRLF"
+				    (if (eq eol 2) "Mac-style CR" "Undecided"))))
+	       'keymap mode-line-eol-map))
+      (push (cons eol (cons mnemonic desc)) mode-line-eol-desc-cache)
+      desc)))
+
 (defvar mode-line-mule-info
   `(""
     (current-input-method
@@ -145,7 +175,7 @@
 			     ".  mouse-2: disable, mouse-3: describe")
 		  local-map ,mode-line-input-method-map))
     ,(propertize
-      "%Z"
+      "%z"
       'help-echo
       #'(lambda (window object point)
 	  (with-current-buffer (window-buffer window)
@@ -157,16 +187,17 @@
 			  " buffer; mouse-3: describe coding system")
 		(concat "Unibyte " (symbol-name buffer-file-coding-system)
 			" buffer")))))
-      'local-map mode-line-coding-system-map))
+      'local-map mode-line-coding-system-map)
+    (:eval (mode-line-eol-desc))))
   "Mode-line control for displaying information of multilingual environment.
 Normally it displays current input method (if any activated) and
 mnemonics of the following coding systems:

Re: Toggle EOL

[Jan D.]

> 
> I think that the EOL information in the mode-line should be
> better separated from the coding-system info.  The patch below
> gives it its own help-echo and its own keymap (mouse-3 switches
> between the three different possible EOL).
> 
> Any comment/objection ?

I like it.  But is the patch correct?  My patch program complained
that the last hunk was corrupted.  And I think this line:
+    (:eval (mode-line-eol-desc))))

should end in 3 ), not 4.

But the concept is great, I've had some problems explaining coding
systems at work, this makes things easier.

	Jan D.

Re: Toggle EOL

[Stefan Monnier]

> > 
> > I think that the EOL information in the mode-line should be
> > better separated from the coding-system info.  The patch below
> > gives it its own help-echo and its own keymap (mouse-3 switches
> > between the three different possible EOL).
> > 
> > Any comment/objection ?
> 
> I like it.  But is the patch correct?  My patch program complained
> that the last hunk was corrupted.  And I think this line:
> +    (:eval (mode-line-eol-desc))))

Yes, there was a typo and I fiddle with the patch by hand :-(
This one should work better,


	Stefan

diff -u -u -b -r1.100 bindings.el
--- bindings.el	16 Apr 2002 14:55:13 -0000	1.100
+++ bindings.el	28 Jun 2002 20:41:28 -0000
@@ -135,6 +135,38 @@
   "Local keymap for the coding-system part of the mode line.")
 
 
+(defvar mode-line-eol-map
+  (let ((map (make-sparse-keymap)))
+    (define-key map [mode-line mouse-3]
+      (lambda ()
+	(interactive)
+	(let ((eol (coding-system-eol-type buffer-file-coding-system)))
+	  (set-buffer-file-coding-system
+	   (cond ((eq eol 0) 'dos) ((eq eol 1) 'mac) (t 'unix))))))
+    (purecopy map))
+  "Local keymap for the `coding-system-eol-type' part of the mode line.")
+
+(defvar mode-line-eol-desc-cache nil)
+
+(defun mode-line-eol-desc ()
+  (let* ((eol (coding-system-eol-type buffer-file-coding-system))
+	 (mnemonic (coding-system-eol-type-mnemonic buffer-file-coding-system))
+	 (desc (assq eol mode-line-eol-desc-cache)))
+    (if (and desc (eq (cadr desc) mnemonic))
+	(cddr desc)
+      (if desc (setq mode-line-eol-desc-cache nil)) ;Flush the cache if stale.
+      (setq desc
+	    (propertize
+	     mnemonic
+	     'help-echo (format "%s end-of-line; mouse-3 to cycle"
+				(if (eq eol 0) "Unix-style LF"
+				  (if (eq eol 1) "Dos-style CRLF"
+				    (if (eq eol 2) "Mac-style CR"
+				      "Undecided"))))
+	     'keymap mode-line-eol-map))
+      (push (cons eol (cons mnemonic desc)) mode-line-eol-desc-cache)
+      desc)))
+
 (defvar mode-line-mule-info
   `(""
     (current-input-method
@@ -145,7 +177,7 @@
 			     ".  mouse-2: disable, mouse-3: describe")
 		  local-map ,mode-line-input-method-map))
     ,(propertize
-      "%Z"
+      "%z"
       'help-echo
       #'(lambda (window object point)
 	  (with-current-buffer (window-buffer window)
@@ -157,7 +189,8 @@
 			  " buffer; mouse-3: describe coding system")
 		(concat "Unibyte " (symbol-name buffer-file-coding-system)
 			" buffer")))))
-      'local-map mode-line-coding-system-map))
+      'local-map mode-line-coding-system-map)
+    (:eval (mode-line-eol-desc)))
   "Mode-line control for displaying information of multilingual environment.
 Normally it displays current input method (if any activated) and
 mnemonics of the following coding systems:

Re: Toggle EOL

[Richard Stallman]

The general idea looks good.

Re: Profiling font-lock in xemacs

[Richard Stallman]

It would be better not to have a delay of even .25 second in handling
C-g.  It ought to respond crisply.

If the delay were made smaller, eventually it would be insignificant.
I don't know at what point that is reached.

Currently on some systems C-g handling can involve a delay of up to 1
second, the wait for "polling for input".  That's because when it was
implemented there was no better facility to use than `alarm'.
Nowadays I gather there is one; we should change the polling interval
to something shorter than a second.