bug#20790: Add more S-SPC key bindings

bug#20790: Add more S-SPC key bindings

[Juri Linkov]

After adding S-SPC to some modes in bug#2145
it's natural to expect S-SPC to be the reverse of SPC
to move in opposite direction, but in modes where SPC
moves point by one line, S-SPC unexpectedly moves back
a full screen (because it uses the default binding
scroll-down-command in special-mode).  This patch
binds S-SPC to line-oriented commands.

diff --git a/lisp/dired.el b/lisp/dired.el
index 02946e0..babdf35 100644
--- a/lisp/dired.el
+++ b/lisp/dired.el
@@ -1541,7 +1541,8 @@ (defvar dired-mode-map
     (define-key map "<" 'dired-prev-dirline)
     (define-key map ">" 'dired-next-dirline)
     (define-key map "^" 'dired-up-directory)
-    (define-key map " "  'dired-next-line)
+    (define-key map " " 'dired-next-line)
+    (define-key map [?\S-\ ] 'dired-previous-line)
     (define-key map [remap next-line] 'dired-next-line)
     (define-key map [remap previous-line] 'dired-previous-line)
     ;; hiding
diff --git a/lisp/vc/vc-dir.el b/lisp/vc/vc-dir.el
index eb03a8b..87f1e05 100644
--- a/lisp/vc/vc-dir.el
+++ b/lisp/vc/vc-dir.el
@@ -271,6 +271,7 @@ (defvar vc-dir-mode-map
     (define-key map " " 'vc-dir-next-line)
     (define-key map "\t" 'vc-dir-next-directory)
     (define-key map "p" 'vc-dir-previous-line)
+    (define-key map [?\S-\ ] 'vc-dir-previous-line)
     (define-key map [backtab] 'vc-dir-previous-directory)
     ;;; Rebind paragraph-movement commands.
     (define-key map "\M-}" 'vc-dir-next-directory)
diff --git a/lisp/arc-mode.el b/lisp/arc-mode.el
index 4df41b5..44d9a21 100644
--- a/lisp/arc-mode.el
+++ b/lisp/arc-mode.el
@@ -395,6 +395,7 @@ (defvar archive-mode-map
     (define-key map "o" 'archive-extract-other-window)
     (define-key map "p" 'archive-previous-line)
     (define-key map "\C-p" 'archive-previous-line)
+    (define-key map [?\S-\ ] 'archive-previous-line)
     (define-key map [up] 'archive-previous-line)
     (define-key map "r" 'archive-rename-entry)
     (define-key map "u" 'archive-unflag)
diff --git a/lisp/proced.el b/lisp/proced.el
index 69355ab..65bd243 100644
--- a/lisp/proced.el
+++ b/lisp/proced.el
@@ -463,6 +463,7 @@ (defvar proced-mode-map
     (define-key km "\C-n" 'next-line)
     (define-key km "\C-p" 'previous-line)
     (define-key km "\C-?" 'previous-line)
+    (define-key km [?\S-\ ] 'previous-line)
     (define-key km [down] 'next-line)
     (define-key km [up] 'previous-line)
     ;; marking

bug#20790: Add more S-SPC key bindings

[Juri Linkov]

> After adding S-SPC to some modes in bug#2145
> it's natural to expect S-SPC to be the reverse of SPC
> to move in opposite direction, but in modes where SPC
> moves point by one line, S-SPC unexpectedly moves back
> a full screen (because it uses the default binding
> scroll-down-command in special-mode).  This patch
> binds S-SPC to line-oriented commands.

Pushed.

Re: bug#20790: Add more S-SPC key bindings

[Random832]

Juri Linkov <juri@linkov.net> writes:
> After adding S-SPC to some modes in bug#2145
> it's natural to expect S-SPC to be the reverse of SPC
> to move in opposite direction, but in modes where SPC
> moves point by one line, S-SPC unexpectedly moves back
> a full screen (because it uses the default binding
> scroll-down-command in special-mode).  This patch
> binds S-SPC to line-oriented commands.

On the subject of S-SPC - I configured my terminal emulator to emit
"\e[27;2;32~" (the same escape sequence that xterm emits in some mode),
and added this to my init.el:

(define-key input-decode-map "^[[27;2;32~" (kbd "S-SPC"))

And it "works" (I don't get escape sequence junk when I type it), but
"doesn't work" in that all it ever seems to do is insert space, and
C-h k thinks I've typed SPC with no hint of it being translated from
anything else.

Anyone know what I'm doing wrong?

Re: bug#20790: Add more S-SPC key bindings

[Emanuel Berg]

Random832 <random832@fastmail.com> writes:

> On the subject of S-SPC - I configured my terminal
> emulator to emit "\e[27;2;32~" (the same escape
> sequence that xterm emits in some mode), and added
> this to my init.el:
>
> (define-key input-decode-map "^[[27;2;32~" (kbd
> "S-SPC"))
>
> And it "works" (I don't get escape sequence junk
> when I type it), but "doesn't work" in that all it
> ever seems to do is insert space, and C-h k thinks
> I've typed SPC with no hint of it being translated
> from anything else.
>
> Anyone know what I'm doing wrong?

Mine look like this:

    (define-key input-decode-map [?\u0110] [C-0])

and yes, it should say it is translated:

    <C-0> (translated from ...) is undefined

I once wrote a tutorial how to do this in the ttys:

    http://user.it.uu.se/~embe8573/tty-emacs-keys.txt

but with xterm I don't know.

Either it is a notational issue - try something else
(something simpler), just to see if the translation
stuff works - either that, or it is something else.

:)

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: bug#20790: Add more S-SPC key bindings

[Stefan Monnier]

> And it "works" (I don't get escape sequence junk when I type it), but
> "doesn't work" in that all it ever seems to do is insert space, and
> C-h k thinks I've typed SPC with no hint of it being translated from
> anything else.
> Anyone know what I'm doing wrong?

Have you bound S-SPC to a command?  If not, then Emacs will
automatically try to strip the shift modifier to find a valid command.


        Stefan

Re: bug#20790: Add more S-SPC key bindings

[Random832]

Emanuel Berg <embe8573@student.uu.se> writes:
> Mine look like this:
>
>     (define-key input-decode-map [?\u0110] [C-0])

Yes, but like you said in your tutorial, that's an arbitrary name. It's
not the real C-0, that'd be [?\C-0] == [#x4000030].

I found my issue, though. I'd been _assuming_ that Gnus sets a binding
for S-SPC to gnus-summary-prev-page/gnus-article-goto-prev-page, when in
reality it apparently doesn't. If I set a global binding on [?\S-\ ], it
works (fsvo. "works" - now I need to figure out how to add a binding to
Gnus for it.)

It is a bit annoying that I don't get the "translated from" notation in
C-h k, though. Led me down a bit of a false path in determining what the
problem was.

Re: bug#20790: Add more S-SPC key bindings

[Emanuel Berg]

Random832 <random832@fastmail.com> writes:

> Yes, but like you said in your tutorial, that's an
> arbitrary name. It's not the real C-0, that'd be
> [?\C-0] == [#x4000030].

Yeah, no, that was just an example, so it is even more
arbitrary. I mean the unicode notation what the
virtual terminal is sending (in this case [?\u0110])
looks much simpler than what you had, so perhaps there
was an issue of notation where some char got out of
place compared to how it looked with xterm.

> I found my issue, though. I'd been _assuming_ that
> Gnus sets a binding for S-SPC to
> gnus-summary-prev-page/gnus-article-goto-prev-page,
> when in reality it apparently doesn't. If I set
> a global binding on [?\S-\ ], it works (fsvo.
> "works" - now I need to figure out how to add
> a binding to Gnus for it.)

Aha, cool.

-- 
underground experts united
http://user.it.uu.se/~embe8573