Add import check for optional dependency on pyiceberg_core (apache#2221)

Added a NotInstalledException check when pyiceberg_core is imported for
pyarrow_transforms This will raise a helpful error message when endusers
try to use methods that depend on pyiceberg_core but haven't installed
the optional dependency

<!--
Thanks for opening a pull request!
-->

Closes apache#1987 

# Rationale for this change
If an enduser hasn't installed the `pyiceberg_core` optional dependency,
using pyarrow transforms will crash with an unhelpful error. This PR
gives the enduser a nicer error message that informs them to install the
optional dependency

# Are these changes tested?
Yes, a test was added to verify the behaviour

# Are there any user-facing changes?

<!-- In the case of user-facing changes, please add the changelog label.
-->

---------

Co-authored-by: Anders Bogsnes <[email protected]>

Author: 2 people

pyiceberg/transforms.py

@@ -17,7 +17,9 @@
 
 import base64
 import datetime as py_datetime
+import importlib
 import struct
+import types
 from abc import ABC, abstractmethod
 from enum import IntEnum
 from functools import singledispatch
@@ -28,6 +30,7 @@
 import mmh3
 from pydantic import Field, PositiveInt, PrivateAttr
 
+from pyiceberg.exceptions import NotInstalledError
 from pyiceberg.expressions import (
     BoundEqualTo,
     BoundGreaterThan,
@@ -106,6 +109,17 @@
 TRUNCATE_PARSER = ParseNumberFromBrackets(TRUNCATE)
 
 
+def _try_import(module_name: str, extras_name: Optional[str] = None) -> types.ModuleType:
+    try:
+        return importlib.import_module(module_name)
+    except ImportError:
+        if extras_name:
+            msg = f'{module_name} needs to be installed. pip install "pyiceberg[{extras_name}]"'
+        else:
+            msg = f"{module_name} needs to be installed."
+        raise NotInstalledError(msg) from None
+
+
 def _transform_literal(func: Callable[[L], L], lit: Literal[L]) -> Literal[L]:
     """Small helper to upwrap the value from the literal, and wrap it again."""
     return literal(func(lit.value))
@@ -382,8 +396,7 @@ def __repr__(self) -> str:
         return f"BucketTransform(num_buckets={self._num_buckets})"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        from pyiceberg_core import transform as pyiceberg_core_transform
-
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.bucket, self._num_buckets)
 
     @property
@@ -509,9 +522,8 @@ def __repr__(self) -> str:
         return "YearTransform()"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        import pyarrow as pa
-        from pyiceberg_core import transform as pyiceberg_core_transform
-
+        pa = _try_import("pyarrow")
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.year, expected_type=pa.int32())
 
 
@@ -570,8 +582,8 @@ def __repr__(self) -> str:
         return "MonthTransform()"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        import pyarrow as pa
-        from pyiceberg_core import transform as pyiceberg_core_transform
+        pa = _try_import("pyarrow")
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
 
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.month, expected_type=pa.int32())
 
@@ -639,8 +651,8 @@ def __repr__(self) -> str:
         return "DayTransform()"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        import pyarrow as pa
-        from pyiceberg_core import transform as pyiceberg_core_transform
+        pa = _try_import("pyarrow", extras_name="pyarrow")
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
 
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.day, expected_type=pa.int32())
 
@@ -692,7 +704,7 @@ def __repr__(self) -> str:
         return "HourTransform()"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        from pyiceberg_core import transform as pyiceberg_core_transform
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
 
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.hour)
 
@@ -915,7 +927,7 @@ def __repr__(self) -> str:
         return f"TruncateTransform(width={self._width})"
 
     def pyarrow_transform(self, source: IcebergType) -> "Callable[[pa.Array], pa.Array]":
-        from pyiceberg_core import transform as pyiceberg_core_transform
+        pyiceberg_core_transform = _try_import("pyiceberg_core", extras_name="pyiceberg-core").transform
 
         return _pyiceberg_transform_wrapper(pyiceberg_core_transform.truncate, self._width)
 

tests/test_transforms.py

@@ -30,7 +30,9 @@
     RootModel,
     WithJsonSchema,
 )
+from pytest_mock import MockFixture
 
+from pyiceberg.exceptions import NotInstalledError
 from pyiceberg.expressions import (
     AlwaysFalse,
     BooleanExpression,
@@ -1668,3 +1670,15 @@ def test_truncate_pyarrow_transforms(
 ) -> None:
     transform: Transform[Any, Any] = TruncateTransform(width=width)
     assert expected == transform.pyarrow_transform(source_type)(input_arr)
+
+
+@pytest.mark.parametrize(
+    "transform", [BucketTransform(num_buckets=5), TruncateTransform(width=5), YearTransform(), MonthTransform(), DayTransform()]
+)
+def test_calling_pyarrow_transform_without_pyiceberg_core_installed_correctly_raises_not_imported_error(
+    transform, mocker: MockFixture
+) -> None:
+    mocker.patch.dict("sys.modules", {"pyiceberg_core": None})
+
+    with pytest.raises(NotInstalledError):
+        transform.pyarrow_transform(StringType())