I was testing xcin2.5 with aterm. It does not work
properly. I can switch to the input mode correctly, but
aterm are capturing all keystrokes for building chinese
characters, xcin is not getting them at all. (I tried
kinput2 for Japanese input, it works though.)
I read rxvt code and realized that it's filtering the
XEvent before calling lookup_key. Also removed the
unnesscary XFilterEvent call within lookup_key too.
With this patch xcin will work again.
I tried this patch with kinput2 too, and it works as
well, so hopefully I didn't break anything :)
diff -ru -x CVS aterm/src/command.c
aterm-orig/src/command.c
--- aterm/src/command.c Fri Sep 7 22:14:25 2001
+++ aterm-orig/src/command.c Fri Sep 7 01:57:04
2001
@@ -1380,6 +1380,7 @@
PrivMode((!numlock_state), PrivMode_aplKP);
}
#ifndef NO_XLOCALE
+ if (!XFilterEvent(ev, *(&ev->xkey.window))) {
if (Input_Context != NULL) {
Status status_return;
@@ -1391,6 +1392,8 @@
sizeof(kbuf), &keysym,
&compose);
}
+ } else
+ len = 0;
#else /* NO_XLOCALE */
len = XLookupString(&ev->xkey, (char *) kbuf,
sizeof(kbuf), &keysym, &compo
se);
/*
@@ -1975,9 +1978,6 @@
refreshed = 0;
XNextEvent(Xdisplay, &ev);
-#ifndef NO_XLOCALE
- if(!XFilterEvent(&ev, ev.xany.window))
-#endif
process_x_event(&ev);
/* in case button actions pushed chars to
cmdbuf */
Logged In: YES
user_id=220249
Sorry my last patch was in the wrong way, I did it the other
way around. And apparently tabs and spacings were lost by
posting in the message.
Here I attached the patch as a file again.
XIM bug with xcin2.5 patch