Tatsuya Kinoshita <t...@vega.ocn.ne.jp> writes: > On June 20, 2010 at 2:24AM +0900, > hirofumi (at mail.parknet.co.jp) wrote: > >> Package: eblook >> Version: 1:1.6.1-6 >> Severity: normal > [...] >> I recently saw the eblook's wrong behavior with "-e euc-jp" on >> LANG=ja_JP.UTF-8. > [...] >> With some debugging, the cause of this behavior looks like readline() >> returned the 'search "ใในใ'. Because, perhaps locale is "UTF-8", but >> input text was EUC-JP, I'm not sure though. >> >> This setup (-e euc-jp on UTF-8) is the default of lookup-el on emacs. >> >> Any idea to fix this? > > Hmm, with or without readline(), eblook on UTF-8 seems buggy.
BTW, what problem? I didn't have any problem with previous eblook version on debian/testing until now (sorry, I forgot actual version). > Does LC_ALL=C prevent your problem? > > I'm thinking about updating the lookup-el package as follows. I'm not using lookup-el package actually, because I'm using emacs on bzr with custom lookup. So, I tested it with the following, $ cat > ~/bin/eblook #!/bin/sh export LANGUAGE=C export LC_ALL=C export LANG=C exec /usr/bin/eblook "$@" $ and with this hack, it seems to work for me (tested a few word only though). However, if eblook is really buggy on UTF-8, why don't we set C to locale in eblook? Thanks. -- OGAWA Hirofumi <hirof...@mail.parknet.co.jp> -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org