shell parsing a line to look for a certain tag

Question

I am planning to create a simple script to edit a file based on values stored within a properties file. So essentially I am planning to loop through each line in the original fi

Answer 1

For instance you can search for strings and manipulate them in one step using sed the stream editor.

echo $line | sed -rn 's:^.*/#(certs.+):\1:p'

This will print only the relevant parts after the /# of the relevant lines.

Answer 2

This is portable to Bourne shell and thus, of course, ksh and Bash.

case $line in
    '/#'* ) tag="${line#/\#}" ;;
esac

To put it into some sort of context, here is a more realistic example of how you might use it:

while read line; do
    case $line in
        '/#'* ) tag="${line#/\#}" ;;
        *) continue ;; # skip remainder of loop for lines without a tag
    esac
    echo "$tag"
    # Or maybe do something more complex, such as
    case $tag in
        cert)
           echo 'We have a cert!' >&2 ;;
        bingo)
           echo 'You are the winner.' >&2
           break # terminate loop
           ;;
        esac
done <$OLD_FILE >$NEW_FILE

Answer 3

You can use BASH regex capabilities:

while read line
do
if [[ "$line" =~ ^.*/#certs(.*)$ ]]; then
    # do your processing here 
    # echo ${BASH_REMATCH[1]} is the part after /#certs
    echo echo ${BASH_REMATCH[1]} >> ${NEW_FILE}
fi
done < ${OLD_FILE}