Convert properties to JSON using jq. #properties $json #jq

README-convert-properties-to-json-using-jq.md

README

Say we have a properties file at foo.properties.

First, clean-up whitespace and empty lines, store in foo-1.properties:

cat foo.properties | \
    sed 's/[[:space:]]*=[[:space:]]*/=/' | \
    sed 's/[[:space:]]*$//' | \
    sed '/^$/d' > foo-1.properties

Pipe to jq using raw mode (-R, to treat input as raw text lines) and slurp mode (-s to merge stream of objects into a single object)

cat foo-1.properties | \
    jq -R -s 'split("\n") | map(split("=")) | map({(.[0]): .[1]}) | add' > foo.json

See also: https://jqplay.org/s/v3fqcUGzvx

You could use only jq for that

cat foo.properties | jq -R -s 'split("\n") | map(select(length > 0)) | map(select(startswith("#") | not)) | map(split("=")) | map({(.[0]): .[1]}) | add'

cat foo.properties | jq -R -s 'split("\n") | map(select(length > 0)) | map(select(startswith("#") | not)) | map(split("=")) | map({(.[0]): .[1]}) | add'

thanks, this almost worked for me with exception to strings containing = breaking it.
The slightly tweaked version I'm using is:

cat foo.properties | jq -R -s 'split("\n") | map(select(length > 0)) | map(select(startswith("#") | not)) | map(split("=")) | map({(.[0]): .[1:] | join("=")}) | add'

You don't need cat

Usage:  jq [options] <jq filter> [file...]

Building on from above

• add whitespace capture around = split
• remove cat
• combine comment and empty line filtering into a single map
• remove unnecessary formatting and join in the final map call

jq -R -s 'split("\n") | map(select(startswith("#") or (length == 0) | not)) | map(split("\\s*=\\s*"; "")) | map({(.[0]): (.[1])}) | add' foo.properties