Search & remove text in files from a source file

Question

I want to search for and remove a block text from several files.

The block of text to be matched against is in file, say /home/user/myblock.txt I want to parse the directory /home/user/rep and remove the content of mybloc.txt from all the files in the directory.

Answer 1

If the block of text to remove spans multiple lines, and if each individual file fits comfortably in RAM, the easiest approach is to load each file entirely into RAM and perform a global string replacement. Perl makes this easy, with -i -p to modify files in place record by record and -0777 to not set any record separator (the default record separator is a newline).

perl -0777 -i -pe '
    BEGIN {
        open BLOCK, "<", "/home/user/myblock.txt";
        $block = <BLOCK>;
        die unless length($block);
        close BLOCK;
    }
    s/\Q$block//g
' /home/user/rep/*

Answer 2

You can try this:

    var=`cat /home/user/myblock.txt`;
    sed -e s/"$var"//g /home/user/rep/somefile;

Of course, this will just print the somefile without your block of text, but you can easily save it with output redirection.

If you want to edit the file "in place", use the -i argument to sed, but be careful. You may lose all your data. It's a better idea to save to some temporary file and rename it.

This is not working if myblock.txt has multiple lines.