hi all,

anyone of you help me what command to be used in unix to delete all characters before pattern starts in every line of a text file

for example, the pattern is "user".

the file has line like,

1234|needhelp234|user234

from the above line, onle user234 to be retained remaining characters to be removed.

thanks
balaji

sed 's/.*\(user*\)/\1/'

thanks..it is done...

..and the solution you found was?

..so that others in your position can find out what to do...

I think pdreyer had the right solution mark though he may of missed a dot before the * ie

Code:

sed 's/.*\(user.*\)/\1/'

Those brackets remember everything from user onwards and put it into a "variable" called \1 then the whole string is replaced by this variable.

There are many ways to skin a cat..

As an alternative, you can also use the "cut" command for the same result :

echo "1234|needhelp234|user234" | cut -f 3 -d "|"
or
cat /my/file | cut -f 3 -d "|"

The cut -f 3 -d "|" says to retain only field number 3 (-f3) of the input, and assume that all fields are delimited with the "|" character (-d"|").