Repository navigation
fix(validation): reject ambiguous paired result JSON - #468
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Paired recomputation currently accepts a results file containing conflicting duplicate metric names when the last value matches the observations. Parse the expected JSON with the existing duplicate-key hook and return a scoped rejection before granting numerical acceptance.
This also rejects nested keys and names that collide after JSON escape decoding. Arithmetic, sampling, malformed-JSON handling and neighboring check results stay unchanged. The
numeric_comparechange in #466 remains separate.Fixes #467
Validation: 95 affected handoff and numeric-precision checks passed, including paired sampling and source-byte preservation. Against the exact original parser, six duplicate-key regressions fail while four ordinary paired controls pass. The baseline used actual files and real NumPy, without an arithmetic shim.