Lower minimum pyarrow dependency back to v10

.github/workflows/build_test.yml

@@ -241,7 +241,7 @@ jobs:
           chrome-version: stable
       - name: Install ChromeDriver
         run: |
-          pip install chromedriver-binary-auto==0.2.6
+          pip install chromedriver-binary-auto
       - name: Download vegafusion wheel
         uses: actions/download-artifact@cbed621e49e4c01b044d60f6c80ea4ed6328b281 # pin@v2
         with:

python/vegafusion-jupyter/dev-environment-3.8.yml

@@ -22,6 +22,6 @@ dependencies:
   - flaky
   - click
   - toml
-  - pyarrow
+  - pyarrow=10.0.1
   - python-duckdb >=0.7.1
   - polars

python/vegafusion/tests/test_transformed_data.py

@@ -1,14 +1,15 @@
 from pathlib import Path
 
 import pandas as pd
-import pytz
+import pyarrow as pa
 import pytest
 from altair.utils.execeval import eval_block
 import vegafusion as vf
 from vega_datasets import data
 import polars as pl
 import altair as alt
 
+pa_major_minor = tuple((int(v) for v in pa.__version__.split(".")[:2]))
 
 here = Path(__file__).parent
 altair_mocks_dir = here / "altair_mocks"
@@ -220,6 +221,7 @@ def test_transformed_data_exclude():
     assert "mean_wheat" in datasets[1]
 
 
+@pytest.mark.skipif(pa_major_minor < (11, 0), reason="pyarrow 11+ required")
 @pytest.mark.parametrize("connection", get_connections())
 def test_gh_286(connection):
     # https://github.com/hex-inc/vegafusion/issues/286

python/vegafusion/vegafusion/dataset/dataframe.py

@@ -1,9 +1,8 @@
 from abc import ABC, abstractmethod
 import pyarrow as pa
 from vegafusion.proto.datafusion_pb2 import LogicalExprNode
-from pyarrow.interchange.dataframe import _PyArrowDataFrame
 
-from typing import Optional, List, Literal, Union
+from typing import Optional, List, Literal, Union, Any
 
 
 class DataFrameOperationNotSupportedError(RuntimeError):
@@ -175,7 +174,7 @@ def impute(
 
     def __dataframe__(
         self, nan_as_null: bool = False, allow_copy: bool = True, **kwargs
-    ) -> _PyArrowDataFrame:
+    ) -> Any:
         """DataFrame interchange protocol support"""
         table = self.collect()
         return table.__dataframe__(nan_as_null=nan_as_null, allow_copy=allow_copy)

python/vegafusion/vegafusion/dataset/duckdb.py

@@ -3,7 +3,7 @@
 from duckdb import DuckDBPyRelation
 from .sql import SqlDataset
 from ..connection.duckdb import duckdb_relation_to_schema
-from pyarrow.interchange.dataframe import _PyArrowDataFrame
+from typing import Any
 
 
 class DuckDbDataset(SqlDataset):
@@ -34,7 +34,7 @@ def fetch_query(self, query: str, schema: pa.Schema) -> pa.Table:
 
     def __dataframe__(
         self, nan_as_null: bool = False, allow_copy: bool = True, **kwargs
-    ) -> _PyArrowDataFrame:
+    ) -> Any:
         return self._relation.to_arrow_table().__dataframe__(
             nan_as_null=nan_as_null, allow_copy=allow_copy
         )

python/vegafusion/vegafusion/dataset/sql.py

@@ -1,6 +1,6 @@
 from abc import ABC, abstractmethod
 import pyarrow as pa
-from pyarrow.interchange.dataframe import _PyArrowDataFrame
+from typing import Any
 
 
 class SqlDataset(ABC):
@@ -60,6 +60,6 @@ def fallback(self) -> bool:
 
     def __dataframe__(
         self, nan_as_null: bool = False, allow_copy: bool = True, **kwargs
-    ) -> _PyArrowDataFrame:
+    ) -> Any:
         """DataFrame interchange protocol support"""
         raise NotImplementedError()