INTPYTHON-497 Extend rather than replace TypeRegistry in write function (#346)

Author: blink1073

.github/workflows/test-python.yml

@@ -143,3 +143,33 @@ jobs:
         run: just docs
       - name: Run linkcheck
         run: just docs-linkcheck
+  test_minimum:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          persist-credentials: false
+      - name: Setup Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: 3.9
+          cache: 'pip'
+          cache-dependency-path: '**/pyproject.toml'
+      - name: Install Deps
+        run: |
+          python -m pip install uv rust-just
+      - name: Set up env
+        run: |
+           echo "LIBBSON_INSTALL_DIR=$PWD/libbson" >> $GITHUB_ENV
+           echo "LD_LIBRARY_PATH=$PWD/libbson/lib" >> $GITHUB_ENV
+      - name: Install libbson
+        run: just build-libbson
+      - name: Run tests
+        shell: bash
+        run: |
+          uv venv
+          source .venv/bin/activate
+          uv pip install -e ".[test]" --resolution=lowest-direct
+          pytest

bindings/python/pymongoarrow/api.py

@@ -49,6 +49,7 @@
     is_uint32,
     is_uint64,
 )
+from pymongo.collection import Collection
 from pymongo.common import MAX_WRITE_BATCH_SIZE
 
 from pymongoarrow.context import PyMongoArrowContext
@@ -488,7 +489,9 @@ def transform_python(self, value):
         return Decimal128(value)
 
 
-def write(collection, tabular, *, exclude_none: bool = False, auto_convert: bool = True):
+def write(
+    collection: Collection, tabular, *, exclude_none: bool = False, auto_convert: bool = True
+):
     """Write data from `tabular` into the given MongoDB `collection`.
 
     :Parameters:
@@ -559,10 +562,13 @@ def write(collection, tabular, *, exclude_none: bool = False, auto_convert: bool
 
     # Add handling for special case types.
     codec_options = collection.codec_options
+    base_codecs = []
+    if hasattr(codec_options.type_registry, "codecs"):
+        base_codecs = codec_options.type_registry.codecs
     if pd is not None:
-        type_registry = TypeRegistry([_PandasNACodec(), _DecimalCodec()])
+        type_registry = TypeRegistry([*base_codecs, _PandasNACodec(), _DecimalCodec()])
     else:
-        type_registry = TypeRegistry([_DecimalCodec()])
+        type_registry = TypeRegistry([*base_codecs, _DecimalCodec()])
     codec_options = codec_options.with_options(type_registry=type_registry)
 
     while cur_offset < tab_size:

bindings/python/pyproject.toml

@@ -51,8 +51,8 @@ Source = "https://github.com/mongodb-labs/mongo-arrow/tree/main/bindings/python"
 Tracker = "https://jira.mongodb.org/projects/INTPYTHON/issues"
 
 [project.optional-dependencies]
-test = ["pytz", "pytest"]
-test-polars = ["polars"]
+test = ["pytz>=2025.2", "pytest>=8.0"]
+test-polars = ["polars>=1.10"]
 test-pandas = ["pandas>=1.3.5,<3"]
 
 [tool.setuptools]
@@ -106,6 +106,7 @@ xfail_strict = true
 filterwarnings = [
   "error",
   "module:The global interpreter lock:RuntimeWarning", # from pandas
+  "module:matching against an empty string will:pytest.PytestWarning", # from pandas
 ]
 
 [tool.ruff]

bindings/python/test/pandas_types/test_binary.py

@@ -91,6 +91,10 @@ def test_contains(self):
         # We cannot compare a Binary object to an array.
         pass
 
+    def test_array_interface_copy(self):
+        # We cannot avoid copying with our extension arrays.
+        pass
+
 
 class TestConstructors(base.BaseConstructorsTests):
     def test_array_from_scalars(self):

bindings/python/test/pandas_types/test_code.py

@@ -87,6 +87,10 @@ def test_contains(self):
         # We cannot compare a Code object to an array.
         pass
 
+    def test_array_interface_copy(self):
+        # We cannot avoid copying with our extension arrays.
+        pass
+
 
 class TestConstructors(base.BaseConstructorsTests):
     pass

bindings/python/test/pandas_types/test_decimal128.py

@@ -84,7 +84,9 @@ def test_is_not_object_type(self, data):
 
 
 class TestInterface(base.BaseInterfaceTests):
-    pass
+    def test_array_interface_copy(self):
+        # We cannot avoid copying with our extension arrays.
+        pass
 
 
 class TestConstructors(base.BaseConstructorsTests):

bindings/python/test/pandas_types/test_objectid.py

@@ -83,7 +83,9 @@ def test_is_not_object_type(self, data):
 
 
 class TestInterface(base.BaseInterfaceTests):
-    pass
+    def test_array_interface_copy(self):
+        # We cannot avoid copying with our extension arrays.
+        pass
 
 
 class TestConstructors(base.BaseConstructorsTests):