A catalog
worth checking.
A local CSV workflow that cleans agreed formatting and keeps uncertain records visible for human review.
The result.
A fictional 16-row product catalog became nine usable records, six records for review and one exact duplicate kept in a separate file. All source rows are accounted for. The original source remains unchanged.
| Input issue | Result |
|---|---|
| SKU 0007 and stock 008 | SKU kept as text 0007; stock normalised to 8 |
| Extra spaces around Desk Lamp | Agreed whitespace cleanup |
| INR price ₹1,299.00 | 1299.00 in the agreed INR format |
| Two identical notebook records | One retained; one recorded as an exact duplicate |
| Two prices for the same travel-mug SKU | Both held for review |
| Missing price or ambiguous date | Held for review; no guessed replacement |
| Formula-like product text | Held for review and protected as text in exports |
Inspect the delivery.
The workbook includes Raw, Cleaned, Review, Duplicates, Changes and Read me sheets. The CSV outputs can be inspected independently. These are sample data, not client records or a testimonial.
A small brief first.
Agree the columns, locale, duplicate rules and output before starting. Share only a small, non-sensitive sample for initial scoping. Access to private or client data needs explicit permission and appropriate handling.
Discuss a cleanup task ↗