It's not specific to JSON/XML either, it works on any text that follows the usual conventions for structuring data in a textual way.
You can filter by level as well, but, usually I just press e/E to jump to the next/previous error from the current location.
There are, of course, many other features as well.
- jq
- Google Sheets (after conversion to CSV)
- Beekeeper Studio (after conversion to SQLite)
This won't scale to large projects, but it works well for the logs from my little CLI tools I enjoy writing in my spare time.
I wrote about this (with some screenshots) at https://github.com/bbkane/logos#analyze-as-json
- It isn´t finished
- The code is a mess (and not available online)
- The UX clearly needs work
- etc. etc. :)
...but I started a thing with basically exactly the same gripes as you two. The only difference: Actually, I don't think collapsing JSON is nearly enough. I want pretty printers (like in GDB) I can rewrite on the fly.So, if you're willing to plunge head-first into cold water without proper documentation, have a look at: https://run.pickaroon.app
Keep in mind that I wrote this to scratch my own itch and will change it further as time allows. On the upside, I've been using it for my own dives through logs quite successfully already.
I also started a manual-ish readme that also isn´t public yet. It's written in markdown, so I'll have to look into how easy I can throw it onto my server or not. I might just make a separate GitHub repo for the time being.
Edit: I went with the GitHub route. You can find the current iteration of the readme here: https://pickaroon.github.io/readme/
This hits very close! Will have a look at this for sure, thanks.