Unicode label cleanup with a

Unicode label cleanup with a collision report

AI-assisted, self-initiated Python demo using seven invented labels. Applies Unicode NFC and folds whitespace while preserving case, record order and string identifiers.

Verified result: four labels changed, one collision group retained for review, and one empty result flagged. Nonbreaking spaces and tabs become ordinary single spaces. 0010 and 10 stay distinct. Running the cleaned labels again produces no further changes.

Outputs: normalized labels, before/after log, collision list and empty-record list. Part of a three-tool suite with six passing automated tests. Source and synthetic inputs available on request. No client work or production deployment is claimed.