Show HN: JSON Query (jsonquerylang.org)

154 points by wofo ↗ HN
I'm working on a tool that will probably involve querying JSON documents and I'm asking myself how to expose that functionality to my users.

I like the power of `jq` and the fact that LLMs are proficient at it, but I find it right out impossible to come up with the right `jq` incantations myself. Has anyone here been in a similar situation? Which tool / language did you end up exposing to your users?

29 comments

[ 3.9 ms ] story [ 60.0 ms ] thread
not to be confused with jq for querying json?
Doesn't the command-line utility `jq` already define a protocol for this? How do the syntaxes compare?

(LLMs are already very adept at using `jq` so I would think it was preferable to be able to prompt a system that implements querying inside of source code as "this command uses the same format as `jq`")

Nice. I work on something similar but for .net.
(comment deleted)
Nice work with a jq-esque feel. Website is cut on mobile devices though
There are a ridiculous number of JSON query/path languages. Wish all the authors got together and harmonized on a standard.
In the k8s world there's a random collection of json path, json query, some random expression language.

Just use jq. None of the other ones are as flexible or widespread and you just end up with frustrated users.

I have a similar use case in the app I'm working on. Initially I went with JSONata, which worked, but resulted in queries that indeed felt more like incantations and were difficult even for me to understand (let alone my users).

I then switched to JavaScript / TypeScript, which I found much better overall: it's understandable to basically every developer, and LLMs are very good at it. So now in my app I have a button wherever a TypeScript snippet is required that asks the LLM for its implementation, and even "weak" models one-shot it correctly 99% of the times.

It's definitely more difficult to set up, though, as it requires a sandbox where you can run the code without fears. In my app I use QuickJS, which works very well for my use case, but might not be performant enough in other contexts.

Most alternatives being talked about are working on query strings (like `$.phoneNumbers[:1].type`) which is fine but can not be easily modeled / modified by code.

Things like https://jsonlogic.com/ works better if you wish to expose a rest api with a defined query schema or something like that. Instead of accepting a query `string`. This seems better as in you have a string format and a concrete JSON format. Also APIs to convert between them.

Also, if you are building a filter interface, having a structured representation helps:

https://react-querybuilder.js.org/demo?outputMode=export&exp...

.friends | filter(.city == "New York") | sort(.age) | pick(.name, .age)

mapValues(mapKeys(substring(get(), 0, 10)))

This is all too cute. Why not just use JavaScript syntax? You can limit it to the exact amount of functionality you want for whatever reason it is you want to limit it.

What do your users know? If they’re quite familiar with SQL for querying, I would look at duckdb.
I can't help myself and surely someone else has already done the same. But the query

  obj.friends.filter(x=>{ return x.city=='New York'})
  .sort((a, b) => a.age - b.age)
  .map(item => ({ name: item.name, age: item.age }));
does exactly the same without any plugin.

am I missing something?

You just have to wrap your mind around jq. It's a) functional, b) has pervasive generators and backtracking. So when you write `.a[].b`, which is a lot like `(.a | .[] | .b)` what you get is three generators strung together in an `and_then` fashion: `.a`, then `.[]`, and then `.b`. And here `.a` generates exactly one value, as does `.b`, but `.[]` generates as many values as are in the value produced by `.a`. And obviously `.b` won't run at all if `.a` has no values, and `.b` will run for _each_ value of `.a[]`. Once you begin to see the generators and the backtracking then everything begins to make sense.
I hate jq as much as the next guy but it’s ubiquitous and great for this sort of thing. If you want a single path style query language I’d highly recommend JsonPath. It’s so much nicer than jq for “I need every student’s gpa”.
I read the man page of `jq` and learned how to use it. It's quite well-written and contains a good introduction.

I've observed that too many users of jq aren't willing to take a few minutes to understand how stream programming works. That investment pays off in spades.

"JSON Query" is kind of a long name. You should find a way to shorten it. Maybe "jQuery" or something along those lines :P