I have a value stored in a variable like
input="###key1,value1###key2,value2###key3,value3###"
I wanted to have output like
key1 has value1 values
key2 has value2 values
key3 has value3 values
Could anybody help me out how to achieve it using loops in bash scripting?
Thanks,:)
and one more doubt, what if the
input="###key1, value1###key2, value2###key3,value3###
that is with whitespace in front of values? How to trim that?
Here is a way by using loop in bash.
lee:~/StackOverFlow $ cat test.sh
input="$1"
input="$(echo "$input" | sed -E 's/^#+//; s/#+$//; s/#+/\\n/g;')"
echo -e "$input" | while read item; do
k="$(echo $item | awk -F\, '{print $1}' | sed -E 's/^[[:space:]]+//')"
v="$(echo $item | awk -F\, '{print $2}' | sed -E 's/^[[:space:]]+//')"
echo "$k has $v values"
done
Output
lee:~/StackOverFlow $ bash test.sh "###key1,value1###key2,value2###key3,value3###"
key1 has value1 values
key2 has value2 values
key3 has value3 values
lee:~/StackOverFlow $ bash test.sh "###key1, value1###key2, value2###key3,value3###"
key1 has value1 values
key2 has value2 values
key3 has value3 values
# the start and end of the line (-E Interpret regular expressions as extended)# with \n
You could use the following approach (only for grep versions with P flag):
#!/usr/bin/env bash
input="###key1,value1###key2,value2###key3,value3###"
eval $(grep -oP '[^#]*(?=###)' <<< $(echo "$input") | sed "s/,/=/")
echo $key1
echo $key2
echo $key3
Output:
value1
value2
value3
For your second case the spaces around = must be removed, so instead of sed "s/,/=/" you need sed "s/,/=/;s/ *//"
With gnu sed
sed 's/##*/\n/g;s/^\n\|\n$//g;s/,/ has /g;s/\n\|$/ values&/g' <<<"$input"
