Hi,

I have so many lines such as follows. It seems that I got Japanese character crash in my solaris10. How to fix it if possible?

"<2004ǯ04��13�� (��) 21��46ʬ05�� JST"


I want to use sed to take all lines out and change it to another one, But I do not know how to do that? any idea?

Maybe I'm missing something or maybe you don't give sufficient details...

If you're suffering an information loss, it may be impossible to restore the file at the original state. By the way, are you sure about what happened? What's a "Japanese character crash"? A file in a locale whose support is not installed in your system may appear unreadable, but that does not mean your file is broken.

Take all lines? With sed you can for example look for a regular expression to match some lines, but if you're changing all of them, why don't you replace the entire file? I don't understand how you'll be filtering lines.

And...where are these lines? e.g. in a file?

Do you get stuff like this when opening any file?

If you simply open a terminal an enter "ls -l", is the output normal?

What would you like to do exactly?

Just looking for ascii info try:
$ strings <file>

You might like to read
$man tr

This is really handy illustration of tr AND if you're poking at binary files often your terminal gets messed up:
$ alias reset='echo "X[mX(BX)0OX[?5lX7X[rX8" | tr "XO" "\033\017"'

tr is used so the whole command can be in ascii (and emailed/skyped/cut&pasted without problems)
and tr translates X to \033 (ESC) (and O to \017 (SI))


from man tr:

Example 1 Creating a list of words

The following example creates a list of all words in file1,
one per line in file2, where a word is taken to be a maximal
string of letters.

tr -cs "[:alpha:]" "[\n*]" <file1 >file2