bug: [Athena] does not list tables

[anjakefala]

What happened?

When I try to list the tables in a database, the result is empty.

What version of ibis are you using?

main branch commit 3d10def

What backend(s) are you using, if any?

Athena

Relevant log output

In [6]: con.drop_database('mydatabase', force=True)
---------------------------------------------------------------------------
OperationalError                          Traceback (most recent call last)
Cell In[6], line 1
----> 1 con.drop_database('mydatabase', force=True)

File ~/git/ibis/ibis/backends/athena/__init__.py:461, in Backend.drop_database(self, name, catalog, force)
    459 name = sg.table(name, catalog=catalog, quoted=self.compiler.quoted)
    460 sql = sge.Drop(this=name, kind="SCHEMA", replace=force)
--> 461 with self._safe_raw_sql(sql, unload=False):
    462     pass

File ~/anaconda3/envs/ibis-dev/lib/python3.11/contextlib.py:137, in _GeneratorContextManager.__enter__(self)
    135 del self.args, self.kwds, self.func
    136 try:
--> 137     return next(self.gen)
    138 except StopIteration:
    139     raise RuntimeError("generator didn't yield") from None

File ~/git/ibis/ibis/backends/athena/__init__.py:291, in Backend._safe_raw_sql(self, query, unload, *args, **kwargs)
    289     query = query.sql(self.dialect)
    290 with self.con.cursor(unload=unload) as cur:
--> 291     yield cur.execute(query, *args, **kwargs)

File ~/anaconda3/envs/ibis-dev/lib/python3.11/site-packages/pyathena/arrow/cursor.py:148, in ArrowCursor.execute(self, operation, parameters, work_group, s3_staging_dir, cache_size, cache_expiration_time, result_reuse_enable, result_reuse_minutes, paramstyle, **kwargs)
    137     self.result_set = AthenaArrowResultSet(
    138         connection=self._connection,
    139         converter=self._converter,
   (...)
    145         **kwargs,
    146     )
    147 else:
--> 148     raise OperationalError(query_execution.state_change_reason)
    149 return self

OperationalError: FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. MetaException(message:Unable to drop database: InvalidOperationException(message:Database mydatabase is not empty.))

In [7]: con.list_tables('mydatabase')
Out[7]: []

Code of Conduct

• I agree to follow this project's Code of Conduct

[cpcloud]

This is expected behavior: the first argument to list_tables is like, not database:

    def list_tables(
        self, like: str | None = None, database: tuple[str, str] | str | None = None
    ) -> list[str]:

[cpcloud]

FYI in 10.0, these arguments will be keyword-only, so it's a bit less error prone.

You'll no longer be able to write con.list_tables without specifying either like=... or database=... or both, all as keyword arguments.

[anjakefala]

Oh, sorry for mixing it up, but that's a helpful API change! Thanks!