Building

uv sync
uv run pytest            # starts Neo4j, Memgraph, FalkorDB and Postgres in Docker
ORACLE_TESTS=true uv run pytest -k oracle   # and Oracle, for its tests alone
uv run ruff check . && uv run pyright

drivine/sync and tests/sync_integration are written from the async source: change that, and run uv run python scripts/unasync.py. A unit test fails while they are out of date. drivine/sync/concurrency.py, the counterpart of drivine/concurrency.py, is the one sync module written by hand, and a test marked async_only (one that runs tasks concurrently) is left out of the sync tests.

uv run python scripts/compile_bench.py times import drivine, each SQL store’s first compile and its warm compiles over the Cypher spec in tests/cypher. CI’s compile-bench job runs it on Python 3.13, puts the table in the run’s summary, and fails where the least of at least 7 fresh import drivines is worse by more than 10% than main’s, timed in turn on the same runner (--import-against, a checkout of main beside the tree: a runner’s CPU type moved import by a quarter where the code didn’t, #496), or where a store’s first compile’s median, or its uncached compile’s median or p99, is worse than the baseline by more than 10% (--against tests/bench/compile-baseline.json): the baseline is the compiler Drivine had before the shared one, measured by CI before the switch (#467). The baseline’s import is printed for reference only.

The integration tests run on Neo4j, Memgraph, FalkorDB, Postgres, Oracle and SQLite. A test marked written_in_cypher seeds and reads back in Cypher with statements of no language, and on the SQL stores those are sent as Cypher, which is compiled to SQL (tests/integration/sql_cypher_fixtures.py, as drivine4j’s SqlCypherFixtures). A test marked graph_engine_only runs only on the graph engines, for the reason it gives, and one marked fails_on_sql_stores is expected to fail on the SQL stores until the issue it names is fixed.