Add lockInfo tests#612
Conversation
Signed-off-by: PatersonProjects <keldonhoff@gmail.com>
…ks fields through use of fsynclock Signed-off-by: PatersonProjects <keldonhoff@gmail.com>
documentdb-triage-tool
Bot
added
compatibility test
|
🤖 Auto-triaged by documentdb-triage-tool. Applied: Reasoningcomponent from path globs (test-coverage); effort from diff stats (226+0 LOC, 5 files); LLM: Adds new test coverage for the $lockInfo diagnostic command, a functional enhancement to the test suite referencing tracked issues. If a label is wrong, remove it manually and ping |
Signed-off-by: PatersonProjects <keldonhoff@gmail.com>
| pytestmark = pytest.mark.admin | ||
|
|
||
|
|
||
| CORE_BEHAVIOR_TESTS: list[DiagnosticTestCase] = [ |
There was a problem hiding this comment.
The parametrized cases in test_lockInfo_core_behavior (returns_ok and returns_lockInfo_array) are identical to test_lockInfo_response_structure (ok_field and lockInfo_is_array). Consider removing the duplicates from one file — they belong in response_structure since they validate response shape.
| pytestmark = pytest.mark.admin | ||
|
|
||
|
|
||
| ERROR_TESTS: list[DiagnosticTestCase] = [ |
There was a problem hiding this comment.
No test for case-mismatched command name (e.g., {"LockInfo": 1} or {"LOCKINFO": 1}). The hostInfo PR includes such a test. Consider adding one for consistency.
| assertProperties(result, {"ok": Eq(1.0)}, msg=spec.msg, raw_res=True) | ||
|
|
||
|
|
||
| INT_EDGE_TESTS: list[DiagnosticTestCase] = [ |
There was a problem hiding this comment.
For consistency with the sibling ping PR (#616) which uses NUMERIC_EDGE_CASES covering -1, 0, and FLOAT_INFINITY, consider adding FLOAT_INFINITY here as well. Both commands ignore their value, and infinity is a special double value beyond what the BSON type acceptance test covers (3.14).
2 participants
This PR adds the tests for the lockInfo diagnostic command. Most of the tests are very straightforward, but response validation requires a lock in place to ensure fields can be checked and so is marked as no-parallel.
Ref: Issue #43, #443