[SOLVED] How to parse every match of sed command

Issue

I have a string [u'SOMEVALUE1', u'SOMEVALUE2', u'SOMEVALUE3'], I would like to parse every element matched by my sed command. The element matched are in the single quote. Here is my script

#!/bin/bash
ARR="[u'SOMEVALUE1', u'SOMEVALUE1', u'SOMEVALUE1']"
for id in $(sed -n "s/^.*'\(.*\)'.*$/\1/ p" <<< ${ARR});
do
    echo "$id"
done

I have only the first value returned.

Solution

The wildcard .* will match the longest leftmost possible string. If your intention is to match the individual substrings which are in single quotes, try

grep -o "'[^']*'" <<<"$ARR"

To remove the single quotes around the values, simply pipe to sed "s/'//g" and to loop over the lines printed by a pipe, do

... commands ... |
while read -r id; do
    : things with "$id"
done

Answered By - tripleee