You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Surfaced by the multi-language coverage fan-out while conformance-testing these SPEC-IDs against databricks/databricks-sql-nodejs. Each finding is committed as an expected-failure (xfail) test in the coverage PR — the test asserts the CORRECT (post-fix) behavior and stays red until THIS driver (databricks/databricks-sql-nodejs) is fixed, then flips green as a tripwire.
failing test: getTables — tableTypes=[] (empty, non-null) matches ALL table types [xfail: driver bug] [thrift], getTables — tableTypes=[] (empty, non-null) matches ALL table types [xfail: driver bug] [sea] (see the coverage PR diff under tests/)
METADATA-036 [sea]: the SEA metadata path sends a redundant CloseStatement DELETE even when the metadata ExecuteStatement response reported status.state=CLOSED; the driver's regular-query path honors the terminal state correctly, so this is metadata-specific
failing test: METADATA-036 — metadata statement close honors ExecuteStatement status.state [sea] (see the coverage PR diff under tests/)
Context
The behavior was first fixed in a DIFFERENT driver — reference PR: https://github.com/databricks/databricks-sql-kernel/pull/183 — which seeded the shared language-neutral spec. This issue tracks the same conformance gap in databricks/databricks-sql-nodejs; the reference PR is for cross-referencing the intended behavior, NOT a change to this repo.
METADATA-036 (an extra CloseStatement after a metadata statement already reported CLOSED): on the kernel backend the driver only calls the native binding's close(), so any extra request comes from the Rust kernel. If it still reproduces, it needs a fix in databricks-sql-kernel.
Closing, since nothing is left to change in this driver.
Activity
vuanhphung commentedon Nov 13, 2026
Re-checked on
main(7be705e) against a live warehouse, on both the Thrift and kernel backends:getTablesreturns 972 rows with notableTypesfilter and 0 rows withtableTypes: []. That is what the shared spec now requires: fix(csharp): empty tableTypes filter matches none, not all (JDBC parity) adbc-drivers/hiveserver2#83 reversed this case, so an empty, non-null filter matches no table types, the same as databricks-jdbc.close(), so any extra request comes from the Rust kernel. If it still reproduces, it needs a fix in databricks-sql-kernel.Closing, since nothing is left to change in this driver.