[anandijain]

@quinnj

This PR making me realize I'm bad at type parameters, I think we still want the jsontable(x::JSON3.Array) dispatch that I commented out.

Should I make it jsontable(x::JSON3.Array{Any}) or something?

Thanks

[quinnj]

I think just uncommenting it should be fine? Were you seeing issues not specifying a type paramter?

[codecov]

Codecov Report

Base: 91.57% // Head: 92.55% // Increases project coverage by +0.97% 🎉

Coverage data is based on head (c6fac42) compared to base (239f4f2).
Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #23      +/-   ##
==========================================
+ Coverage   91.57%   92.55%   +0.97%     
==========================================
  Files           1        1              
  Lines          95       94       -1     
==========================================
  Hits           87       87              
+ Misses          8        7       -1     

Impacted Files	Coverage Δ
src/JSONTables.jl	92.55% <100.00%> (+0.97%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[quinnj]

[P2] Retain the JSON-array validation fallback (src/JSONTables.jl:34).

The bounded object-vector method is the right direction, but removing the generic JSON3.Array method changes invalid-input behavior. On this exact head (c6fac42d4a2b5d177407297cc198f612fa237d7e), four public checks fail: jsontable(JSON3.read("[]")) now succeeds, while parsed [1,2], [{},1], and [null] throw MethodError instead of the existing ArgumentError. The remaining generic wrapper attempts to parse those already parsed arrays again.

Keep the validation fallback and add a more specific JSON3.Array{JSON3.Object} bridge to the bounded vector method. This avoids the dispatch ambiguity while retaining the invalid-array contract. I implemented that approach, with initial-empty-row handling and public regressions, in #26; its six current-head checks are green and it awaits independent approval.

AI disclosure: This work was prepared with assistance from OpenAI Codex.