I was trying to follow the output of a service running on my machine. Each log line was a JSON string. The nesting was shallow, but there were a large number of key/value pairs in each log line, and it was pretty hard to find the values I was looking for in the sea of white text.
In want of a better way to handle JSON output on the command line, my own research initially turned up bat. After I shared that in the #dev_tip_of_the_day channel on Slack, several of my colleagues chimed in with more excellent suggestions.
- bat
- Pros: Extremely flexible. Handles many more formats than just JSON.
- Cons: Finding a theme that highlighted JSON the way I wanted it to was a little tough, as many themes displayed object keys and string values in the same colour.
- jq
- Pros: Super powerful. Lets you do cool stuff like displaying and searching at the same time.
- Cons: Just JSON. Very arguably not a flaw, if you’re an adherent of the Unix philosophy.
- jless
- Pros: The coolest of the bunch. Lets you interactively whip through JSON data. Very cute mascot. Handles YAML, too.
- Cons: Not well suited to streaming data, e.g. following log output.
python -m json.tool- Pros: Pretty-prints output and doesn’t require installing a third party tool (unless you count Python.)
- Cons: Not well suited to streaming data. Pretty-prints but doesn’t add colour.
Of the bunch, I think jless will become my go-to. Such a cute mascot.
