Hi, anybody knows that


how to replace a token(string or number) in a text file with a certain token(string/number) by a single line of command (might be combinations, pipelines....etc.)?? i want to do it in my bash shell scripts.


thanks!!


Jimmy

sed would be a good choice.

sed -i.bak s/oldword/newword/ filename

a backup file is created if you give the extension. The example above would create a backup named filename.bak

One of the great things about linux (& unix) is that their are often several diffrent ways to apprach just about any problem that you are faces with. This is one of those cases. For basic substitution using sed is one of my favorite choices. For instance if I want to replace 'NOT' with 'not' in a file called data I could do something like....

sed 's/NOT/not/' data > new_data; cat new_data > data; rm new_data

as you can see I dump the info out to a temporary holding file called new_data and when it is complete I overwrite the original with the modified version.

Other solutions presented here on linuxquestions include....

http://www.linuxquestions.org/questi...?postid=932751

http://www.linuxquestions.org/questi...hreadid=137374

If you want to do something a bit more complex it might be worth your time to brush up on your perl, which has some very good tools for editing text.

I hope that this helps.

For strings more complicated than letters or numbers out might want to check out some sed documentation especially sed and regular expressions.

http://www.hk8.org/old_web/unix/sedawk/index.htm