This patch sorts keybindings in arrays by keycode or keysym to speed up look
up using binary searches. This is a preliminary work to enable more powerful
keybindings stuff, where keybindings can be cascaded or why not, attached to
specific clients.
Interstingly enough, this patch saves 100ko of initial memory (Heap) usage here.
The underlying idea is that we should be able to define keybindings_t as
trees of keybindings_t which would then define key sequences.
The OO approach kind of make sense in fact, since you create a base
keybinding (e.g. reacting on Mod4-w) and then you will probably (with
appropriate apis) be able to populate new submaps from that point more or
less dynamically.
And if you have two keybindings on Mod4-w, then adding them will replace the
previous one. This means that you can fake per-client bindings with e.g.:
k_default = keybindings.new({"Mod4"}, "w", something);
k_mplayer = keybindings.new({"Mod4"}, "w", something_else);
k_default:add()
and in your focus hook:
if /* code for testing if it's mplayer */ then
k_mplayer:add()
else
k_default:add()
end
This would not work before, it does now.
It will take way more sense with submaps of course.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
put *item to NULL on unref, as we cannot know if the pointer is valid
after an unref, so just segfault rather than hide a problem.
Also return *item on ref() it allow short versions like:
foo_list_push(&list, foo_ref(&elem));
which is kind of readable _and_ handy.
Signed-off-by: Pierre Habouzit <madcoder@debian.org>
Pierre Habouzit
Julien Danjou
Michael Gehring