Synthetic CSV Import Validation: Dates, IDs and Exceptions

CSV import validation — synthetic demonstration

Fictional data, not client work. AI-assisted preparation with executed validation checks. This demonstrates a small import-validation method, not production throughput or customer results.

Contract: keep IDs as text; dates must be real calendar dates in YYYY-MM-DD format; quantities must be integers. Reject unresolved rows into an exception report rather than guessing. No source system is accessed.

Five test rows ID 001 | 2026-09-01 | quantity 10 | Accept ID 0007 | 2026-09-30 | quantity 2 | Accept; preserve ID text ID 008 | 09/10/2026 | quantity 5 | Exception: ambiguous date without a locale rule ID 009 | 2026-02-30 | quantity 3 | Exception: invalid calendar date ID 010 | 2026-10-01 | quantity oops | Exception: non-integer quantity

Checked results Five input rows become two accepted rows and three exceptions. Accepted quantities sum to 12. Numeric quantities in rejected rows sum to 8. All numeric inputs total 20 = 12 + 8; the one nonnumeric quantity is explicitly excluded, not replaced with zero. ID 0007 stays 0007.

Delivery format An accepted-row file, a row-level exception register, a validation contract and control totals. This example performs no live import. Rejected rows need corrected source data or a newly agreed rule before resubmission. Original rows remain unchanged.