Bdd

Document Test Cases In BDD

1. Store And Organize Files

1. Store feature files under docs/features/ or docs/tests/features/.
2. Create one .feature file per API capability, feature, or resource group.
3. Group scenarios by requirement or business rule, not by raw HTTP method alone.
4. Keep aggregate index files under docs/testing/ optional; treat the .feature files themselves as the canonical narrative artifacts.
5. For non-trivial requirements, document at least three scenarios:
   1. MSS for the main success path.
   2. EXT for meaningful valid variations.
   3. ERR for validation, auth, not-found, conflict, or drift behavior.
6. For trivial low-risk reads, one MSS scenario is sufficient.

2. Use Standard Gherkin Structure

1. Start with a Feature: statement that names the API capability.
2. Use Background: only for shared setup such as base environment, auth state, seeded data, or contract version.
3. Use Scenario: for single cases and Scenario Outline: only when the same behavior really varies by data.
4. Use Examples: tables for compact data variations.
5. Add tags above scenarios for requirement ids, operation ids, and scope such as @smoke, @regression, @contract, @negative, or @US-123.

3. Write API-Focused Steps

1. Put environment, auth state, or seeded data in Given.
2. Put request actions in When.
3. Put status, content type, critical headers, body semantics, and side effects in Then.
4. Use And and But only to extend the current clause clearly.
5. Keep wording business-readable and API-specific.
6. Do not turn the scenario into a low-value transcript of raw HTTP syntax.

4. Start From The Template

1. Start from feature-template.feature for new capability files.
2. Keep scenario titles stable because traceability reports link to them.
3. Keep one feature file per capability or resource group; do not collapse the whole service into one giant file.

5. Depth Rules

1. Always include a standard MSS scenario for the primary valid request.
2. Add EXT coverage for optional fields, alternate filters, paging, sorting, or valid role-specific variants.
3. Add ERR coverage for invalid input, unauthorized access, missing resources, duplicate requests, or contract drift.
4. Add a dedicated drift scenario when the runtime behavior does not match the documented contract.

6. Example

@SPC-OWN-003 @addOwner @workflow
Feature: Owner management

  Scenario: Create an owner with required fields
    Given the Spring Petclinic API is available
    And no owner exists for the generated last name
    When a client submits a valid owner payload to POST /api/owners
    Then the response status is 201
    And the response body contains the created owner id
    And the created owner can be retrieved by id

7. Examples

1. Input: Write BDD for GET /orders/{id} unauthorized access. Output: A tagged ERR scenario with Given, When, and Then steps for 401 behavior.
2. Input: Document the create-order requirement in BDD. Output: Separate MSS, EXT, and ERR scenarios unless the requirement is trivial.

8. Troubleshooting

1. Problem: The scenario reads like a curl command transcript. Fix: Keep steps behavioral and move raw payload detail into data tables or examples only when needed.
2. Problem: One feature file mixes unrelated capabilities. Fix: Split by resource or business capability.
3. Problem: A complex requirement has only one scenario. Fix: Add the missing EXT and ERR scenarios before calling the documentation complete.