Re: [PATCH]: PrintInfo Bindings
On Sun, Feb 08, 2009 at 02:44:15AM +, Thomas Adam wrote:
Hello all,
Something I've seen an increase of lately, on the fvwm list and
particularly on IRC and the forums, is the question of: What are the
default mouse/key bindings of fvwm? To which the answer varies,
since invariably the question is: Can I list all my bindings fvwm
knows about?
Well, with the attached patch, you can. I've added it to PrintInfo,
since it really is only useful in certain circumstances.
Dominik: I've deliberately not restructured this patch to sit on top
of the InitialMapCommand patches -- so if you consider applying this
patch along with that one, you might get some conflicts in NEWS,
AUTHORS and Changelog. Hopefully shouldn't be a problem though.
That's no problem.
Comment/suggestions welcome as always.
See below.
Index: AUTHORS
===
RCS file: /home/cvs/fvwm/fvwm/AUTHORS,v
retrieving revision 1.130
diff -u -r1.130 AUTHORS
--- AUTHORS 19 Oct 2008 12:04:12 - 1.130
+++ AUTHORS 8 Feb 2009 02:27:54 -
@@ -15,6 +15,7 @@
StartShaded style option.
Introduce the command expansion placeholder: $[w.visiblename]
Make style matching honour a window's visible name (c.f. $[w.visiblename])
+Added bindings option to PrintInfo command useful for debugging.
Serge (gentoosiast) Koksharov:
Documentation fixes, bug fixes.
Index: ChangeLog
===
RCS file: /home/cvs/fvwm/fvwm/ChangeLog,v
retrieving revision 1.3061
diff -u -r1.3061 ChangeLog
--- ChangeLog 30 Dec 2008 13:32:40 - 1.3061
+++ ChangeLog 8 Feb 2009 02:28:04 -
@@ -1,3 +1,16 @@
+2009-02-05 Thomas Adam thomas.ada...@gmail.com
+ * libs/charmap.c (charmap_table_to_string):
+ * libs/charmap.h:
+ Introduce charmap_to_string function which is used to build up a
+ binding string, for use with PrintInfo.
+
+ * fvwm/bindings.c (print_bindings):
+ Introduce print_bindings to print all bindings known to fvwm.
+
+ * fvwm/builtins.c (CMD_PrintInfo):
+ * fvwm/builtins.h:
+ Add support for binding as an option to PrintInfo.
+
2008-12-29 Alexandre Julliard julli...@winehq.org
* fvwm/ewmh_events.c (ewmh_WMStateMaxHoriz):
Index: NEWS
===
RCS file: /home/cvs/fvwm/fvwm/NEWS,v
retrieving revision 1.767
diff -u -r1.767 NEWS
--- NEWS 30 Dec 2008 13:32:40 - 1.767
+++ NEWS 8 Feb 2009 02:28:05 -
@@ -18,6 +18,9 @@
... is now honoured. Useful with IndexedWindowName as a style
option.
+ - New option to PrintInfo, bindings which prints out all of the Key,
+ PointerKey, Mouse and Stroke bindings which fvwm knows about.
+
* Bug fixes:
- Fixed compilation without XRender support.
Index: doc/commands/PrintInfo.xml
===
RCS file: /home/cvs/fvwm/fvwm/doc/commands/PrintInfo.xml,v
retrieving revision 1.4
diff -u -r1.4 PrintInfo.xml
--- doc/commands/PrintInfo.xml16 Jun 2007 12:38:46 - 1.4
+++ doc/commands/PrintInfo.xml8 Feb 2009 02:28:05 -
@@ -57,4 +57,10 @@
replaceableverbose/replaceable
can be 1./para
+parafvwmopt cms=PrintInfo opt=bindings/
+which prints information on all the bindings fvwm has: key, mouse and
+stroke bindings.
+replaceableverbose/replaceable
+has no effect with this option./para
+
/section
Index: fvwm/bindings.c
===
RCS file: /home/cvs/fvwm/fvwm/fvwm/bindings.c,v
retrieving revision 1.103
diff -u -r1.103 bindings.c
--- fvwm/bindings.c 23 Nov 2007 08:49:11 - 1.103
+++ fvwm/bindings.c 8 Feb 2009 02:28:10 -
@@ -639,6 +639,60 @@
return;
}
+void print_bindings(int verbose)
If the verbose flag has no effect, why is it an argument to the
function?
+{
+ Binding *b;
+
+ fprintf(stderr, Current list of bindings:\n\n);
+
+ for (b = Scr.AllBindings; b != NULL; b = b-NextBinding)
+ {
+ /* Key, PointerKey, Mouse, Stroke. */
+ char *binding_type = NULL;
+ char type[20];
+ switch (b-type)
+ {
+ case BIND_KEYPRESS:
+ binding_type = Key;
+ sprintf(type, %s, b-key_name);
+ break;
+ case BIND_PKEYPRESS:
+ binding_type = PointerKey;
+ sprintf(type, %s, b-key_name);
+ break;
+ case BIND_BUTTONPRESS:
+ case BIND_BUTTONRELEASE:
+binding_type = Mouse;
+sprintf(type, %d, b-Button_Key);
+
Re: [PATCH]: PrintInfo Bindings
2009/2/8 Dominik Vogt dominik.v...@gmx.de:
On Sun, Feb 08, 2009 at 02:44:15AM +, Thomas Adam wrote:
Comment/suggestions welcome as always.
See below.
Thanks.
+void print_bindings(int verbose)
If the verbose flag has no effect, why is it an argument to the
function?
I had thought about trying to utilise verbose at one point, but
couldn't think of the need. Changed to void instead.
+ char type[20];
+ switch (b-type)
+ {
+ case BIND_KEYPRESS:
+ binding_type = Key;
+ sprintf(type, %s, b-key_name);
+ break;
+ case BIND_PKEYPRESS:
+ binding_type = PointerKey;
+ sprintf(type, %s, b-key_name);
+ break;
+ case BIND_BUTTONPRESS:
+ case BIND_BUTTONRELEASE:
+binding_type = Mouse;
+sprintf(type, %d, b-Button_Key);
+break;
+ case BIND_STROKE:
+ binding_type = Stroke;
+ sprintf(type, %s\t%d, (char *)b-Stroke_Seq,
+ b-Button_Key);
I think we shouldn't copy arbitrary strings into a fixed length
buffer. I'm sure there are key names that are longer than 19
characters.
I checked in bindings.c:ParseBindings() -- key_string is defined as:
char key_string[201] = ;
I've mimicked this here as well for char[] type. If you think that's
still going to cause problems, I will really make it dynamic, although
from testing it locally, it's not yet caused any problems, and my
Thinkpad does have some nice XF86LongNameKeys defined, for instance.
+ break;
+ }
+
+ char *mod_string = charmap_table_to_string(
+ MaskUsedModifiers(b-Modifier),key_modifiers);
+ char *context_string = charmap_table_to_string(
+ b-Context, win_contexts);
+ char win_name[260] = ;
Uh, I know we have hard coded some length limit to the window
name, but that's just because there was or is a bug in the X
system that caused a crash.
OK. Changed to be dynamic, although I will note that again, according
to ParseBindings(), the max length of a window name is capped at 79
chars. But I've ignored that again here, and allocated win_name with
malloc().
#include config.h
#include stdio.h
#include ctype.h
+#include string.h
You can't rely on string.h being present. On some systems, it's
callend strings.h. Luckily you don't have to take care of that
yourself: config.h already includes the proper file, so you can
just delete this line.
I'm sure I had problems with that before -- but you're right. Removed
without any problems.
#include charmap.h
+#include safemalloc.h
Should be libs/safemalloc.h. Maybe I should remove libs from the
include path.
Fixed. From doing a quick cursive glance through some of the other
header files in libs/, there's a real mixture of some header files
being included as libs/foo.h and others just as foo.h. If you
don't get around to this some point next week, I will send in a patch
to clean this up on your behalf.
+
+/* Used from PrintInfo Bindings. */
+char *charmap_table_to_string(int mask, charmap_t *table)
+{
+ char *allmods = safemalloc (sizeof(table));
+ memset (allmods, '\0', sizeof(table));
Are you sure you want to reserve sizeof(table) bytes? This
would reserve space to store a pointer, not space to store a
charmap_t.
Oops. I meant: sizeof(*table) here. Feel free to replace it with:
sizeof(charmap_t) -- I know some prefer the latter over the former.
+
+ int modmask = mask;
Please don't declare variables in the middle of a block. This
does not compile with oplder compilers.
I didn't quite understand what you meant by this -- the declaration
wasn't in a block, it was just after the function had been declared.
I've tried to change it, but if it's still wrong, do say and I'll
correct it again.
Please see print-bindings2.patch attached.
Kindly,
-- Thomas Adam
Index: AUTHORS
===
RCS file: /home/cvs/fvwm/fvwm/AUTHORS,v
retrieving revision 1.130
diff -u -r1.130 AUTHORS
--- AUTHORS 19 Oct 2008 12:04:12 - 1.130
+++ AUTHORS 8 Feb 2009 17:35:08 -
@@ -15,6 +15,7 @@
StartShaded style option.
Introduce the command expansion placeholder: $[w.visiblename]
Make style matching honour a window's visible name (c.f. $[w.visiblename])
+Added bindings option to PrintInfo command useful for debugging.
Serge (gentoosiast) Koksharov:
Documentation fixes, bug fixes.
Index: NEWS
===
RCS file:
Re: [PATCH]: PrintInfo Bindings
On Sun, Feb 08, 2009 at 05:55:51PM +, Thomas Adam wrote:
2009/2/8 Dominik Vogt dominik.v...@gmx.de:
[sniiip]
+ int modmask = mask;
Please don't declare variables in the middle of a block. This
does not compile with oplder compilers.
I didn't quite understand what you meant by this -- the declaration
wasn't in a block,
All code is in blocks. Don't forget the curly braces that
delimit the function body :-)
it was just after the function had been declared.
I've tried to change it, but if it's still wrong, do say and I'll
correct it again.
In C89 you can't have
{
declaration;
statement;
declaration;
}
You have to put all the declarations at the beginning of the
block, i.e.:
Wrong (does not compile):
void foo(void)
{
int x;
memset(x, 0, sizeof(x);
int y;
}
Right:
void foo(void)
{
int x;
int y;
memset(x, 0, sizeof(x);
}
More comments later when I've got time to look at the new patch.
Ciao
Dominik ^_^ ^_^
--
Dominik Vogt
