v0.28.2

[0.28.2] - 2025-05-07

• engines: Enhance graceful shutdown by canceling ongoing queries when shutting down postgres connections. This improves the behavior when receiving a Ctrl-C signal by ensuring that any long-running queries are properly canceled before the connection is closed.
• bin: In parallel mode, the output of a test case will be printed promptly before shutting down the database connection.

v0.28.1

[0.28.1] - 2025-04-18

• bin: Support graceful shutdown. When receiving a Ctrl-C, the program will cancel all running test cases, log cancelled and skipped test cases, drop temporary databases (if in parallel mode), close database connections, and finally exit with a non-zero code.

v0.28.0

[0.28.0] - 2025-03-06

• runner: Add Partitioner and Runner::with_partitioner to enable partitioning of test cases, allowing only a subset of the glob result to be executed. This can be helpful for running tests in parallel in CI.
• bin: Add --partition-id and --partition-count to set a hash partitioning for the test cases. If users are running in Buildkite CI with parallelism: .. specified in the workflow file, this will be automatically configured.

v0.27.2

[0.27.2] - 2025-02-18

• engines/bin: fix stdin to be closed properly to avoid hangs in the external engine.

v0.27.1

[0.27.1] - 2025-02-17

• runner: Add Runner::set_var method to allow adding runner-local variables for substitution.
• bin: Add __DATABASE__ variable for accessing current database name from SLT files.

v0.27.0

[0.27.0] - 2025-02-11

• runner: add shutdown method to DB and AsyncDB trait to allow for graceful shutdown of the database connection. Users are encouraged to call Runner::shutdown or Runner::shutdown_async after running tests to ensure that the database connections are properly closed.

v0.26.4

[0.26.4] - 2025-01-27

• runner: add random string in path generation to avoid conflict when using include.
• bin: detect connection refused error and exit early to make error message clearer.

v0.26.3

[0.26.3] - 2025-01-14

• bin: when --fail-fast is enabled, abort all remaining connections before dropping temporary databases.

v0.26.2

[0.26.2] - 2025-01-08

• bin: support --fail-fast, and add env vars SLT_FAIL_FAST and SLT_KEEP_DB_ON_FAILURE

v0.26.1

[0.26.1] - 2025-01-08

• parser/runner: support system ok retry