Test if pattern in file exists, evaluate if

Question

I want to test if a pattern exists in a file.

If the pattern exists (In a file among a list), do a replacement and update the corresponded file.
If the pattern doesn't exists, do nothing.

So far I have done:

    if [ -f $((`awk '/Pattern1/' $FILE`))];then
        sed 's/Pattern1/\nWORD/g' $FILE > a
        mv a $FILE
    fi

But, when I test it, I get the following error:

syntax error: invalid arithmetic operator

And I don't want to evaluate the numeric part, just if the pttern~string is there or not.

Input of file 1

Rho =  1.1955904623E+02     Rho at Nucleus =  1.1955904594E+02

Input of file2

Rho =  1.1955904623E+02

If you are using awk to look for the pattern why not use awk to do the entire job? Seems strange to switch to sed. — David Hoelzer, Jun 06 '15 at 21:51
one idea, you can use wildcards for filenames, like *.txt to catch all txt files in folder, then no need for using bash, only sed or awk will be enough — josifoski, Jun 06 '15 at 22:08
@josifoski good idea and yes, you were right, the but was only a space. Thank you for helping out and have a good day : ) — Another.Chemist, Jun 06 '15 at 22:11

score 2 · Accepted Answer · edited May 23 '17 at 12:22

2

You could simply go ahead with the substitution, if the pattern exist something will be changed, if not, nothing will be done to the file.

awk '{sub(/Pattern1/, "\nWORD"); print}' $FILE > tmp && mv tmp $FILE

edited May 23 '17 at 12:22

Community

answered Jun 06 '15 at 22:10

Tarc

1 Answers1