Closed GoogleCodeExporter closed 9 years ago
One more comment, the only reason that "是" appears at the first position is
because I've tried very hard to find
it and used it recently, otherwise it would appear on later pages.
Original comment by sikan.c...@gmail.com
on 4 Apr 2010 at 4:20
@sikan.chen, I'm so sorry for that, since I don't have a big-endian machine to
verify the .be files, I'd try to convert
the be files by thaikov's conversion tool, let me know if you'd like to help us
on that?
Original comment by find...@gmail.com
on 4 Apr 2010 at 5:06
Sure, that'd be my pleasure.
Original comment by sikan.c...@gmail.com
on 4 Apr 2010 at 5:18
@sikan.chen, would you please list your sha1 checksum for the lm_sc.tg3 and
pydict_sc.bin files in
/Library/Input\ Methods/SunPinyin.app/Contents/Resources? I need to know which
lm and lexicon files you are
using, thanks a lot!
Original comment by find...@gmail.com
on 5 Apr 2010 at 1:40
SHA1(/Library/Input Methods/SunPinyin.app/Contents/Resources/lm_sc.t3g)=
da1753c198d6a7a8ca1609a015b49603cf88267e
SHA1(/Library/Input Methods/SunPinyin.app/Contents/Resources/pydict_sc.bin)=
4c8ec0431983dffb6c9aa196fe64f151ddd3d8ba
Original comment by sikan.c...@gmail.com
on 5 Apr 2010 at 2:22
@sikan.chen, would you please help to try the attached pydict_sc.bin.be file?
please copy it to /Library/Input
Methods/SunPinyin.app/Contents/Resources/pydict_sc.bin, thanks a lot for your
help!
Original comment by find...@gmail.com
on 5 Apr 2010 at 6:51
Attachments:
Oops, forgot to mention, please also try to remove the history cache
(~/Library/Application\
Support/SunPinyin/history), and killall -9 SunPinyin before you copy the new
lexicon file.
Original comment by find...@gmail.com
on 5 Apr 2010 at 6:53
我一直就在说这个问题嘛⋯⋯
Original comment by leikka...@gmail.com
on 5 Apr 2010 at 7:34
@findsun, I did exactly what you've suggested (kill the process/remove the
history/replace the lexicon file) and
the candidate windows now refuses to prompt up. I even tried 1) logging out and
2) reinstalling and re-applying
the offline file replacement, same results, no candidate window...
@leikkanen, have you tried the same?
Original comment by sikan.c...@gmail.com
on 6 Apr 2010 at 2:05
@sikan.chen, I'm so sorry, looks like the big-endian lexicon file is not
properly
generated by the converter :(
Would you like to help us to build the big-endian lexicon on your own machine?
Original comment by find...@gmail.com
on 6 Apr 2010 at 2:08
Sure :) Need some basic instructions and hope my 4.5 year old iBook is capable
of doing that.
Original comment by sikan.c...@gmail.com
on 6 Apr 2010 at 2:12
I assume "BuildSunpinyin2onMac" is the one I'm supposed to take a look. Need
some time to prep the env.
Original comment by sikan.c...@gmail.com
on 6 Apr 2010 at 2:19
@sikan.chen, refer to
http://code.google.com/p/sunpinyin/wiki/BuildSunpinyin2onMac,if you don't
want to use
the open-gram language model, please download
http://sunpinyin.googlecode.com/files/lm_sc.t3g.be, and save
to data/lm_sc.t3g, and run 'make genpyt; make lexicon' under build folder.
Let me know if you meet any issue.
Original comment by find...@gmail.com
on 6 Apr 2010 at 2:21
sorry for the inconvenience, =(
i will test the converter/generator more. adding issue#86 to track it.
Original comment by tchai...@gmail.com
on 6 Apr 2010 at 2:26
@sikan.chen, @leikkanen,
我们最近重新build了大端的词库,并上传到了
http://sunpinyin.googlecode.com/files/pydict_sc.bin.be,
有时间请帮忙测试一下,多谢了!
Original comment by find...@gmail.com
on 9 Apr 2010 at 5:46
Original comment by find...@gmail.com
on 9 Apr 2010 at 5:47
You great guys, sunpinyin team! I'll give it a try no later than next Monday
(my ppc
mac is not with me at the moment).
Original comment by sikan.c...@gmail.com
on 10 Apr 2010 at 3:24
Unfortunately, the candidate window still refuses to show up with your newly generated lexicon file...
Original comment by sikan.c...@gmail.com
on 12 Apr 2010 at 1:38
we built that file on a solaris sparc machine (on unix-center.net), it should
be able to loaded into sunpinyin's ppc
version :(
Original comment by find...@gmail.com
on 12 Apr 2010 at 1:49
Please keep this one open as I am still trying... will get back to you guys
soon. Thanks.
Original comment by sikan.c...@gmail.com
on 13 Apr 2010 at 2:20
I just did a clean install (but skipped the downloading step and manually
copied two big endian files to the
right location ) and now the candidate window failing to show up problem seems
gone. So that is supposed to
be an irrelevant issue.
So once again I'm able type Chinese, but neither the one you generated on that
SPARC nor the one I generated
on my own PPC is able to mitigate the problem. Though the ordering of the
characters did change a little bit,
it changed to an equally weird way...
Here I attach the candidate window the first time I type "shi" with your SPARC
file.
Original comment by sikan.c...@gmail.com
on 13 Apr 2010 at 2:48
Attachments:
@sikan.chen, would you please try to kill sunpinyin process, and remove your
history cache, under
~/Library/Application\ Support/SunPinyin/history, then re-activate it?
Original comment by find...@gmail.com
on 13 Apr 2010 at 2:51
That's exactly what I did, but to no avail...
Original comment by sikan.c...@gmail.com
on 14 Apr 2010 at 12:49
@sikan.chen, I'm sorry for the late response, just heard that fedora has the
powerpc RPMs for sunpinyin, and I
extracted the data files, uploaded to
http://sunpinyin.googlecode.com/files/sunpinyin-data-be.tar.bz2, would
you please have a try?
Please remove the history file after you deployed the new lm_sc.t3g and
pydict_sc.bin files.
Original comment by find...@gmail.com
on 15 Apr 2010 at 3:08
Sorry for the late reply. I'm sure I've done strictly what you ask for, but the
candidate characters are exactly the
same.
Original comment by sikan.c...@gmail.com
on 19 Apr 2010 at 1:31
Original comment by find...@gmail.com
on 19 May 2010 at 2:22
Original issue reported on code.google.com by
sikan.c...@gmail.com
on 4 Apr 2010 at 4:16Attachments: