Clamping of array indices in Proposal A about json-path-comparison

[remorhaz]

Python behaviour does not help guide us as it will give an array out of bounds error

I interpreted Python errors as guide for returning an empty result.

[cburgmer]

We were indeed missing a query for this case, I added one: https://cburgmer.github.io/json-path-comparison/results/bracket_notation_with_negative_number_on_short_array.html

Turns out that Proposal A is not aligned with the consensus, and I would say the proposal is also plain wrong

IMHO this should also correlate with the positive number case which shows a clear consensus: https://cburgmer.github.io/json-path-comparison/results/bracket_notation_with_number_on_short_array.html

Should we change Proposal A's implementation in this case?

[glyn]

Should we change Proposal A's implementation in this case?

Yes, I think so.

[cburgmer]

Pushed a change, happy to see this closed if you agree.

[glyn]

Yes, the fix seems good. Thanks.