Repository navigation
Retain conflicting voucher evidence and reject nonfinite amounts - #6
Merged
Merged
Conversation
dev-belly
changed the title
dev-belly
marked this pull request as ready for review
October 24, 2026 05:49
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Exact-duplicate removal compared only eight columns, so records with the same ID/amount but different currency, FX, invoice, approval or other evidence could be silently discarded. Numeric infinities and overflowing FX products could also survive cleaning.
Compare every normalized source column for exact duplicates, keeping conflicting records visible to downstream key validation. Convert nonfinite numeric inputs and translated products to missing values so the existing exclusion/fallback rules apply. Preserve raw extracts and describe the input-quality boundary.
Complete this existing draft against current main, including the merged full-precision selection fingerprint and review CSV repairs. Validation: 104 cleaning tests and the full 519-test suite pass; full 30,128-voucher pipeline, Ruff and diff checks pass. Every report, including the 300-row queue and model/rule diagnostics, remains byte-identical to the same-environment pre-cleaning baseline. README test counts are checked by the suite, and every uploaded blob hash matches the local file.