ピアス エリック writes: [...]

The actual line looks like this: href="info.html">こっち</A></P></DIV>

This brings up nothing: $ grep こっち * $

I guess the Japanese in your file is iso-2022-jp encoded or SJIS encoded. 'grep' will interpret the expression you enter in the encoding of the locale you are using, which is probably EUC-JP. If you are running in ja_JP.eucJP locale and want to grep for Japanese in files which may have EUC-JP, ISO-2022-JP or SJIS encoding, you can use lgrep: mfabian@gregory:/tmp$ locale charmap EUC-JP mfabian@gregory:/tmp$ lgrep "こ+っち" ttt*html ttt-euc-jp.html:test ここっち test ttt-iso-2022-jp.html:test ここっち test ttt-sjis.html:test ここっち test mfabian@gregory:/tmp$ lgrep will check for all these Japanese encodings. lgrep is a hard link to lv, i.e. you need to have lv.rpm installed. -- Mike Fabian http://www.suse.de/~mfabian 睡眠不足はいい仕事の敵だ。

At Tue, 21 May 2002 11:49:21 +0900 (JST), ピアス エリック wrote:

That works great. Too bad there isn't a '-r' (recursive) option for lgrep.

you can combine with find (yeah it's linux :) % find . | xargs lgrep pattern -- Takashi Iwai SuSE GmbH - www.suse.de ALSA Developer ALSA Project - www.alsa-project.org