Summary
Follow-up to #6339 (drop json5 dep, handle non-finite floats in socket parse fallback) to make the user-visible behavior explicit in docs/tests.
Problem
The parser fallback behavior for non-finite floats is subtle and easy to regress:
Infinity / -Infinity should round-trip to frontend JS numbers.
NaN is preserved via sentinel + reviver behavior in the fallback path.
Without a clear documentation note and explicit compatibility framing in tests, future refactors may accidentally change rendering/round-trip behavior.
Proposed Follow-up
- Add a short docs note in
docs/vars/base_vars.md describing expected handling of NaN, Infinity, and -Infinity.
- Add an explicit compatibility comment around the
special_floats integration assertion to codify expected UI representation.
Acceptance Criteria
- User-facing docs mention non-finite float behavior and compatibility expectations.
- Integration test intent is explicit for maintainers reviewing future parser changes.
Summary
Follow-up to #6339 (
drop json5 dep, handle non-finite floats in socket parse fallback) to make the user-visible behavior explicit in docs/tests.Problem
The parser fallback behavior for non-finite floats is subtle and easy to regress:
Infinity/-Infinityshould round-trip to frontend JS numbers.NaNis preserved via sentinel + reviver behavior in the fallback path.Without a clear documentation note and explicit compatibility framing in tests, future refactors may accidentally change rendering/round-trip behavior.
Proposed Follow-up
docs/vars/base_vars.mddescribing expected handling ofNaN,Infinity, and-Infinity.special_floatsintegration assertion to codify expected UI representation.Acceptance Criteria