In article <20060420075450.GA4164@xxxxxxxxxxxxx>, VÃclav Åmilauer <eudoxos@xxxxxxxx> writes:
> I tried with scim 1.4.4 and uim (1.0.0), m17n upgraded to 1.3.1 (debian
> unstable); xorg 7.0. Both of them (scim and uim) behave the same, so it
> doesn't seem to be IM engine issue; I use the following map (in
> comments, whether it is functional or not)
> ("b" "[b]") ;; yes
> ((aacute) "[aacute]") ;; no
> ((c) "[LATIN_SMALL_LETTER_C]") ;; yes
> ((C) "[LATIN_CAPITAL_LETTER_C]") ;; yes
> ((space) "[SPACE]") ;; no
> ((0) "[0]") ;; no
> ((at) "[@]") ;; no
> Apparently, letters work both as strings and as symbols ("b" and (c)).
> For non-letter characters, like @ (I tried both (at) and "@", not at the
> same time), neither works.
The only strange part about your result is that (accute)
didn't work. Did you get the actual aacute character
instead of the string "[accute]"?
> Apparently, I have a rather different behavior from you; is there a way
> you can think of to debug it? I don't mind rebuilding m17n with verbose
> debugging output and the like, but such a thing will have to wait a few
> days at least (diploma thesis deadline).
Please try this.
% killall scim-launcher
% MDEBUG_OUTPUT_FILE=/tmp/m17n-log MDEBUG_INPUT=1 GTK_IM_MODULE=scim gedit
Then all the events m17n-lib received and how m17n-lib
reacted on it is recorded in the file /tmp/m17n-lob. Finish
gedit (to be sure to flush all log) and check that file.
I tested the attached mim file with gedit as above. When I
type keys "bà 0@" in this sequence, the following characters
are input (as expected):
"b"(aacute)"SPC""0""@"
I also attach the log I got.
---
Kenichi Handa
handa@xxxxxxxx
Attachment:
test.mim
Description: Binary data
Attachment:
m17n-log
Description: Binary data