Test Datasets
Test Datasets are reusable parameter tables for manual and automated test cases. A dataset separates test logic from input data so one test case can cover many scenarios, while revisions preserve changes over time.
What Datasets Solve
Use datasets when one scenario needs multiple input combinations:
- user roles and permissions;
- currencies, countries, and tax profiles;
- payment card types and 3DS outcomes;
- product SKUs and discount rules;
- feature flags and configuration variants;
- imported parameters from pytest or external CSV/JSON files.
Instead of keeping 20 nearly identical test cases, the team keeps one scenario and 20 data rows.
Core Concepts
| Concept | Description |
|---|---|
| Dataset | Project-scoped set of columns and rows. |
| Source Type | Origin of the data: Manual, Pytest Parametrize, or Imported. |
| Source Ref | Link to a file, test, fixture, document, or external reference. |
| Column | Data field such as country, currency, or coupon_code. |
| Row | Concrete scenario such as bg_eur_valid_coupon. |
| Revision | Version of the dataset structure and rows. |
| Binding | Link between a dataset and a test case. |