chore(deps): bump flakes

.pre-commit-config.yaml

@@ -105,6 +105,7 @@ repos:
         name: shellcheck
         language: system
         entry: shellcheck
+        args: ["--shell", "bash"]
         types_or:
           - sh
           - shell

flake.nix

@@ -23,6 +23,15 @@
         nixpkgs.follows = "nixpkgs";
       };
     };
+
+    pyproject-build-systems = {
+      url = "github:pyproject-nix/build-system-pkgs";
+      inputs = {
+        pyproject-nix.follows = "pyproject-nix";
+        uv2nix.follows = "uv2nix";
+        nixpkgs.follows = "nixpkgs";
+      };
+    };
   };
 
   outputs =
@@ -32,11 +41,12 @@
     , nixpkgs
     , pyproject-nix
     , uv2nix
+    , pyproject-build-systems
     , ...
     }: {
       overlays.default = nixpkgs.lib.composeManyExtensions [
         gitignore.overlay
-        (import ./nix/overlay.nix { inherit uv2nix pyproject-nix; })
+        (import ./nix/overlay.nix { inherit uv2nix pyproject-nix pyproject-build-systems; })
       ];
     } // flake-utils.lib.eachDefaultSystem (
       localSystem:

ibis/backends/impala/tests/test_exprs.py

@@ -132,7 +132,7 @@ def test_builtins(con, alltypes):
         s.repeat(i1),
     ]
 
-    proj_exprs = [expr.name("e%d" % i) for i, expr in enumerate(exprs)]
+    proj_exprs = [expr.name(f"e{i:d}") for i, expr in enumerate(exprs)]
 
     projection = table.select(proj_exprs)
     projection.limit(10).execute()
@@ -418,7 +418,7 @@ def test_decimal_timestamp_builtins(con):
         exprs.append(ts + offset)
         exprs.append(ts - offset)
 
-    proj_exprs = [expr.name("e%d" % i) for i, expr in enumerate(exprs)]
+    proj_exprs = [expr.name(f"e{i:d}") for i, expr in enumerate(exprs)]
 
     projection = table.select(proj_exprs).limit(10)
     projection.execute()
@@ -474,7 +474,7 @@ def test_aggregations(alltypes):
         d.var(where=cond),
     ]
 
-    metrics = [expr.name("e%d" % i) for i, expr in enumerate(exprs)]
+    metrics = [expr.name(f"e{i:d}") for i, expr in enumerate(exprs)]
 
     agged_table = table.aggregate(metrics)
     agged_table.execute()
@@ -511,7 +511,7 @@ def test_analytic_functions(alltypes):
         f.max(),
     ]
 
-    proj_exprs = [expr.name("e%d" % i) for i, expr in enumerate(exprs)]
+    proj_exprs = [expr.name(f"e{i:d}") for i, expr in enumerate(exprs)]
 
     proj_table = g.mutate(proj_exprs)
     proj_table.execute()

ibis/backends/impala/udf.py

@@ -21,7 +21,7 @@
 import ibis.expr.operations as ops
 from ibis import util
 
-__all__ = ["scalar_function", "aggregate_function", "wrap_udf", "wrap_uda"]
+__all__ = ["aggregate_function", "scalar_function", "wrap_uda", "wrap_udf"]
 
 
 class Function(abc.ABC):

ibis/backends/oracle/tests/test_client.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from datetime import date  # noqa: TCH003
+from datetime import date  # noqa: TC003
 
 import oracledb
 import pandas as pd

ibis/backends/polars/rewrites.py

@@ -7,7 +7,7 @@
 from ibis.common.annotations import attribute
 from ibis.common.collections import FrozenDict
 from ibis.common.patterns import replace
-from ibis.common.typing import VarTuple  # noqa: TCH001
+from ibis.common.typing import VarTuple  # noqa: TC001
 from ibis.expr.schema import Schema
 
 

ibis/backends/sql/compilers/__init__.py

@@ -3,8 +3,8 @@
 __all__ = [
     "BigQueryCompiler",
     "ClickHouseCompiler",
-    "DatabricksCompiler",
     "DataFusionCompiler",
+    "DatabricksCompiler",
     "DruidCompiler",
     "DuckDBCompiler",
     "ExasolCompiler",
@@ -16,8 +16,8 @@
     "PostgresCompiler",
     "PySparkCompiler",
     "RisingWaveCompiler",
-    "SnowflakeCompiler",
     "SQLiteCompiler",
+    "SnowflakeCompiler",
     "TrinoCompiler",
 ]
 

ibis/backends/sql/compilers/base.py

@@ -80,7 +80,7 @@ class AggGen:
     class _Accessor:
         """An internal type to handle getattr/getitem access."""
 
-        __slots__ = ("handler", "compiler")
+        __slots__ = ("compiler", "handler")
 
         def __init__(self, handler: Callable, compiler: SQLGlotCompiler):
             self.handler = handler
@@ -175,7 +175,7 @@ def __getitem__(self, key: str) -> Callable[..., sge.Anonymous]:
 
 
 class FuncGen:
-    __slots__ = ("dialect", "namespace", "anon", "copy")
+    __slots__ = ("anon", "copy", "dialect", "namespace")
 
     def __init__(
         self, *, dialect: sg.Dialect, namespace: str | None = None, copy: bool = False

ibis/backends/sql/rewrites.py

@@ -14,11 +14,11 @@
 import ibis.expr.datatypes as dt
 import ibis.expr.operations as ops
 from ibis.common.annotations import attribute
-from ibis.common.collections import FrozenDict  # noqa: TCH001
+from ibis.common.collections import FrozenDict  # noqa: TC001
 from ibis.common.deferred import var
 from ibis.common.graph import Graph
 from ibis.common.patterns import InstanceOf, Object, Pattern, replace
-from ibis.common.typing import VarTuple  # noqa: TCH001
+from ibis.common.typing import VarTuple  # noqa: TC001
 from ibis.expr.rewrites import d, p, replace_parameter
 from ibis.expr.schema import Schema
 

ibis/backends/tests/test_aggregation.py

@@ -470,12 +470,19 @@ def mean_and_std(v):
             lambda t, where: t.string_col.approx_nunique(where=where),
             lambda t, where: t.string_col[where].nunique(),
             id="approx_nunique",
-            marks=pytest.mark.xfail_version(
-                duckdb=["duckdb>=1.1"],
-                raises=AssertionError,
-                reason="not exact, even at this tiny scale",
-                strict=False,
-            ),
+            marks=[
+                pytest.mark.xfail_version(
+                    duckdb=["duckdb>=1.1"],
+                    raises=AssertionError,
+                    reason="not exact, even at this tiny scale",
+                    strict=False,
+                ),
+                pytest.mark.notimpl(
+                    ["datafusion"],
+                    reason="data type is not supported",
+                    raises=Exception,
+                ),
+            ],
         ),
         param(
             lambda t, where: t.bigint_col.bit_and(where=where),

ibis/backends/tests/test_examples.py

@@ -16,7 +16,9 @@
     reason="nix on linux cannot download duckdb extensions or data due to sandboxing",
 )
 @pytest.mark.notimpl(["pyspark", "exasol", "databricks"])
-@pytest.mark.notyet(["clickhouse", "druid", "impala", "mssql", "trino", "risingwave"])
+@pytest.mark.notyet(
+    ["clickhouse", "druid", "impala", "mssql", "trino", "risingwave", "datafusion"]
+)
 @pytest.mark.parametrize(
     ("example", "columns"),
     [

ibis/backends/tests/test_expr_caching.py

@@ -55,7 +55,7 @@ def test_persist_expression_contextmanager(backend, con, alltypes):
     assert non_cached_table.op() not in con._cache_op_to_entry
 
 
-@mark.notimpl(["datafusion", "flink", "impala", "trino", "druid"])
+@mark.notimpl(["flink", "impala", "trino", "druid"])
 @pytest.mark.never(
     ["risingwave"],
     raises=com.UnsupportedOperationError,
@@ -83,6 +83,7 @@ def test_persist_expression_multiple_refs(backend, con, alltypes):
     backend.assert_frame_equal(
         non_cached_table.order_by("id").to_pandas(),
         cached_table.order_by("id").to_pandas(),
+        check_dtype=False,
     )
 
     name = cached_table.op().name
@@ -105,7 +106,7 @@ def test_persist_expression_multiple_refs(backend, con, alltypes):
     assert name not in con.list_tables()
 
 
-@mark.notimpl(["datafusion", "flink", "impala", "trino", "druid"])
+@mark.notimpl(["flink", "impala", "trino", "druid"])
 @mark.notimpl(["exasol"], reason="Exasol does not support temporary tables")
 @pytest.mark.never(
     ["risingwave"],
@@ -127,7 +128,7 @@ def test_persist_expression_repeated_cache(alltypes, con):
     assert name not in con.list_tables()
 
 
-@mark.notimpl(["datafusion", "flink", "impala", "trino", "druid"])
+@mark.notimpl(["flink", "impala", "trino", "druid"])
 @mark.notimpl(["exasol"], reason="Exasol does not support temporary tables")
 @pytest.mark.never(
     ["risingwave"],

ibis/backends/tests/test_temporal.py

@@ -781,11 +781,6 @@ def convert_to_offset(x):
                     raises=AssertionError,
                     reason="duckdb returns dateoffsets",
                 ),
-                pytest.mark.notimpl(
-                    ["trino"],
-                    raises=AssertionError,
-                    reason="doesn't match pandas results, unclear what the issue is, perhaps timezones",
-                ),
                 pytest.mark.notimpl(
                     ["flink"],
                     raises=Py4JJavaError,

ibis/backends/tests/tpc/ds/test_queries.py

@@ -2072,11 +2072,6 @@ def test_40(catalog_sales, catalog_returns, warehouse, item, date_dim):
 
 
 @tpc_test("ds")
-@pytest.mark.notyet(
-    ["datafusion"],
-    raises=Exception,
-    reason="Error during planning: Correlated column is not allowed in predicate",
-)
 @pytest.mark.notyet(
     ["clickhouse"],
     raises=ClickHouseDatabaseError,
@@ -2230,9 +2225,6 @@ def test_44(store_sales, item):
 
 
 @tpc_test("ds")
-@pytest.mark.notyet(
-    ["datafusion"], raises=Exception, reason="Unsupported feature in DataFusion SQL"
-)
 def test_45(web_sales, customer, customer_address, date_dim, item):
     return (
         web_sales.join(customer, [("ws_bill_customer_sk", "c_customer_sk")])
@@ -4647,6 +4639,7 @@ def items(returns, *, prefix):
 
 
 @tpc_test("ds")
+@pytest.mark.notyet(["datafusion"], reason="coercion failure", raises=Exception)
 def test_84(
     customer,
     customer_address,

ibis/backends/trino/converter.py

@@ -9,16 +9,18 @@
     @classmethod
     def convert_Interval(cls, s, dtype, pandas_dtype):
         def parse_trino_timedelta(value):
-            # format is 'days hour:minute:second.millisecond'
-            days, rest = value.split(" ", 1)
-            hms, millis = rest.split(".", 1)
-            hours, minutes, seconds = hms.split(":")
-            return datetime.timedelta(
-                days=int(days),
-                hours=int(hours),
-                minutes=int(minutes),
-                seconds=int(seconds),
-                milliseconds=int(millis),
-            )
+            if isinstance(value, str):
+                # format is 'days hour:minute:second.millisecond'
+                days, rest = value.split(" ", 1)
+                hms, millis = rest.split(".", 1)
+                hours, minutes, seconds = hms.split(":")
+                return datetime.timedelta(
+                    days=int(days),
+                    hours=int(hours),
+                    minutes=int(minutes),
+                    seconds=int(seconds),
+                    milliseconds=int(millis),
+                )
+            return value
 
         return s.map(parse_trino_timedelta, na_action="ignore")