I am currently working with a data file that has 8 columns delimited by pipes. The last column contains an unpredictable number of key-value pairs, each separated by the equals sign and spaces between each pair. However, the challenge is that the values within each pair can also include spaces.
For instance, here is a sample record from the input data file:
Val1|Val2|Val3|Val4|Val5|Val6|Val7| key1=70 key2=11\=12 key3=Some other value key4=Another value
The desired final output should present all key-value pairs in the 8th column in JSON format like this:
Val1|Val2|Val3|Val4|Val5|Val6|Val7|{"key1":"70", "key2":"11\=12", "key3":"Some other value", "key4":"Another value"}
This is a snippet of the current data format I have been able to achieve (assume this is stored in a variable named mydata):
Val1|Val2|Val3|Val4|Val5|Val6|Val7|{"key1":"70 key2":"11\=12 key3":"Some other value key4":"Another value"}
However, what is missing are the necessary ", " characters that would properly separate the previous value from the next key name.
Currently, I need to perform a search and replace operation within the loop of an awk command so that during each iteration, it correctly formats the key-value pairs as specified.
I have formulated the final awk command outline, but it lacks the crucial search and replace operation required for each iteration within the loop:
echo $mydata | awk -F '\":\"' '{ str1 = $1; for (i = 2; i <= NF; i++) str1 = str1 "\":\"" $i; print str1 }'
Thus, the question remains: how exactly do I integrate the necessary search and replace logic into the for loop of the awk command?
An additional piece of code that I have executed up to this point is as follows:
test="Val1|Val2|Val3|Val4|Val5|Val6|Val7| key1=70 key2=11\=12 key3=Some other value key4=Another value";
echo $test | awk -F '|' '{print $1 "|" $2 "\x1c" $3 "|" $4 "|" $5 "|" $6 "|" $7 "|" "{\"" $8 "\"}" }' \
| sed -e 's/{\" /{\"/g' -e 's/\\=/\x07/g' -e 's/[[:space:]]*=/\":\"/g' -e 's/\x07/\\=/g'