feat: json traversal

[dmfay]

this needs more time than I have for it at the moment but I don't think it'll be too hard to get basic support for #166! element indexing (possibly some common ground with the array[1,2,3] syntax?) and jsonpath are bigger considerations though.

[matthias-Q]

I added the other json operators and updated the test. Actually it looks like this is it. All other json operations are probably handled by invocation

[Gelio]

Drive-by appreciation comment:

I am honestly amazed by your aptitude and the promptness at fixing the issues I've reported. Stellar project maintenance! I certainly wasn't expecting to see all 3 issues addressed the same day I reported them

[dmfay]

Drive-by appreciation comment:

I am honestly amazed by your aptitude and the promptness at fixing the issues I've reported. Stellar project maintenance! I certainly wasn't expecting to see all 3 issues addressed the same day I reported them

the JSON issue has secretly been bugging me for a while (along with array literals and subscripting, but that's for later) and I've always been too busy with other things & other projects to take a look, and it turns out it just fits right into the expression node now

[matthias-Q]

LGTM. Thanks for fixing the tests.

[DerekStride]

This subtree represents the term "user"->'address' ->> 'city'. It only references a single literal node which means the other two have been hidden.

I've puts together #171 to show an alternative using the binary_expression node that will surface all 3 literals and treat the navigation nodes as anonymous operators.

[matthias-Q]

We should add the new operators to the highlights.scm (i forgot that with filter yesterday as well ...)

[matthias-Q]

This looks good to me. I am merging it. I am so surprised that this went so smoothly.

(I was hoping that after the recent refactor we could also tackle array and not in, but it seems like they are still difficult to solve)

[dmfay]

This looks good to me. I am merging it. I am so surprised that this went so smoothly.

(I was hoping that after the recent refactor we could also tackle array and not in, but it seems like they are still difficult to solve)

I went looking and we actually got arrays at some point!

[matthias-Q]

@dmfay yes, but not as a function. See #115 Besides, I think #128 is more important :-/