> yes pleas. and tbh, I was hoping for something stronger than "I believe" :)
Well, it did build for me, so I have to guess what is different...
> do you have a good test-case for this one?
Using this:
key <FK06> {
type= "THREE_LEVEL",
symbols[Group1]= [ NoSymbol, NoSymbol, KP_3 ],
actions[Group1]= [ RedirectKey(key=<KP1>),
RedirectKey(key=<KP2>,clearMods=Shift), RedirectKey(key=<FK06>) ]
};
Regarding the (re)moved 'return 0' in particular, press and hold F6 and
then another key while F6 is still down; that other key should not be
suppressed.
Otherwise, one can try out the various different orders of
pressing/releasing Shift, ISO_Level3_Shift, and F6 and check whether the
results are correct. Apart from creating KP_1, KP_2 and KP_3, this also
means that all keys that get pressed also get released. Finally, one
can try with autorepeat, pressing and releasing Shift and
ISO_Level3_Shift while FK06 autorepeats.
For the changes to action messages:
key <FK06> {
type= "TWO_LEVEL",
symbols[Group1]= [ KP_3, KP_2 ],
actions[Group1]= [ ActionMessage(report=all,data="Base", genkeyevent),
ActionMessage(report=all, data="Shift", genkeyevent=no) ]
};
There are a few more options to test, namely, 'report=press' and
'report=release'. With a little program one can observe what
action messages get sent:
-------------------------------------------------------------------------------
#include <X11/Xlibint.h>
#include <X11/Xlib.h>
#include <X11/XKBlib.h>
#include <stdio.h>
int main(int argn, char** argv){
int xkbbasecode, errcode, statcode;
Display* dpy = XkbOpenDisplay(NULL, &xkbbasecode, &errcode, NULL, NULL,
&statcode);
XkbSelectEvents(dpy, XkbUseCoreKbd, XkbActionMessageMask,
XkbActionMessageMask);
while(1){
XkbEvent event;
XNextEvent(dpy, &event.core);
if(event.type == xkbbasecode){
if(event.any.xkb_type == XkbActionMessage){
const char* m = event.message.message;
const unsigned int kc = event.message.keycode;
printf("Keycode %d press %d message %s\n", kc,
event.message.press, m);
}
}
}
return 0;
}
-------------------------------------------------------------------------------
Regards,
Andreas
_______________________________________________
[email protected]: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: http://lists.x.org/mailman/listinfo/xorg-devel
