Question

I have a text file with this pattern:

\t (hello world)

I need to replace the text in parenthesis with a unique value (e.g. ob1, obj2, etc), such that

\t (obj)
\t (obj)
\t (obj)

becomes…

\t (obj1)
\t (obj2)
\t (obj3)

Or anything else that is unique. A solution that uses any tool that works in cygwin will work. My attempt using bash and sed to do this failed:

#!/bin/bash
x=1
for fl in myfile; do
    cp $fl $fl.old
    sed 's/\\t \(.*\)/\\t \("${x}"\)/g' $fl.old > $fl.new
    x=$((x+1))
    echo $x
done

Answer 1

The best way I know how to do this is with a perl in-place edit:

E.g. myfile.txt contains:

\t (obj)
\t (obj)
\t (obj)

Run the in-place edit:

perl -i.bak -ne '$a=int(rand()*2000); s/\((.*?)\)/{$1$a}/g; print' myfile.txt

myfile.txt now contains:

\t (obj1869)
\t (obj665)
\t (obj1459)

Obviously adjust 2000 to your requirements.

EDIT: If you wish to use incrementing identifiers use:

perl -i.bak -ne '$a++; s/\((.*?)\)/{$1$a}/g; print' myfile.txt