Code Craft: Validate Cart

Phone-screen Code Craft prompt. Given a cart with item entries, validate whether the cart is orderable under quantity, availability, and min/max-limit rules. The prompt is intentionally open-ended; candidates are expected to negotiate exact rule semantics before coding. Requirements Input: a cart containing item lines. Each line should at minimum expose an item id / SKU and requested quantity; the availability and min/max limits may be supplied in the item metadata or by a mock inventory / catalog service. Validate at least three rule families: Requested quantity must be valid for the item. Item must be available / in stock. Item-specific minimum and maximum quantity limits must be respected. Output: either a boolean validity result or a structured validation result with per-item errors; clarify which shape the interviewer wants before implementing. The prompt is open-ended: the interviewer expects discussion on rule shape, data ownership, and how errors should be surfaced. Notes Keep the implementation rule-driven rather than hard-coding checks inline. A clean shape is CartValidator.validate(cart) -> ValidationResult, with small helper methods or rule objects for quantity, availability, and min/max limits. Prefer returning structured errors such as {item_id, code, message} so the caller can render actionable feedback. A plain boolean is too thin unless the interviewer explicitly requests it. Edge cases to test: empty cart, duplicate item lines, quantity zero / negative, item unavailable, requested quantity below min, requested quantity above max, and multiple validation errors in one cart. If availability comes from a service, separate fetching item metadata from applying validation rules. This keeps the validator testable and makes failure handling easier to discuss. Preparation Practice implementing a small rule-engine-style validator in 25 minutes. Prepare a table-driven test suite covering the edge cases above. Have a short answer ready for catalog / inventory-service failure: fail closed for checkout, show retryable validation error, and log the failed dependency.