How the Unique Key Conflict Finder Works
AnyTool finds duplicate keys entirely in the browser. The CSV/TSV is parsed with PapaParse (RFC-4180, so quoted commas survive), then a key is built per row from the column(s) you select — joined with a control character so composite keys (e.g. email + tenant) cannot collide. Keys are compared with optional case-folding and whitespace trimming, and blank keys are tracked as a distinct category rather than dropped. The report is precise: total rows, distinct keys, duplicate keys (keys appearing more than once), rows-in-conflict (all rows sharing a duplicated key), excess rows (Σ count−1), and empty-key rows. Conflicts are grouped by key with the real 1-based row numbers and sample rows, sorted by frequency. You can export just the conflicting rows or a keep-first de-duplicated file. Nothing is uploaded.
- Runs client-side — the dataset is never uploaded
- RFC-4180 parsing; composite keys joined with a non-colliding separator
- Case-fold, trim, and blank-key options
- Accurate distinct / duplicate / rows-in-conflict / empty-key stats
- Export conflicting rows or a keep-first deduped CSV
- No CDN, no backend, no row cap or signup
