portable date python changes

Author: ahmedabu98

sdks/python/apache_beam/io/jdbc.py

@@ -360,6 +360,10 @@ def __init__(
                    of the output PCollection elements. This bypasses automatic
                    schema inference during pipeline construction.
     """
+    # override new portable Date type with the current Jdbc type
+    # TODO: switch JdbcIO to return portable Date type
+    LogicalType.register_logical_type(JdbcDateType)
+
     classpath = classpath or DEFAULT_JDBC_CLASSPATH
 
     dataSchema = None

sdks/python/apache_beam/portability/common_urns.py

@@ -92,3 +92,4 @@
 var_bytes = LogicalTypes.Enum.VAR_BYTES
 fixed_char = LogicalTypes.Enum.FIXED_CHAR
 var_char = LogicalTypes.Enum.VAR_CHAR
+date = LogicalTypes.Enum.DATE

sdks/python/apache_beam/transforms/managed_iceberg_it_test.py

@@ -15,6 +15,7 @@
 # limitations under the License.
 #
 
+import datetime
 import os
 import unittest
 import uuid
@@ -49,7 +50,8 @@ def _create_row(self, num: int):
         bytes_=bytes(num),
         bool_=(num % 2 == 0),
         float_=(num + float(num) / 100),
-        arr_=[num, num, num])
+        arr_=[num, num, num],
+        date_=datetime.date.today() - datetime.timedelta(days=num)))
 
   def test_write_read_pipeline(self):
     iceberg_config = {

sdks/python/apache_beam/typehints/schemas.py

@@ -34,6 +34,7 @@
   bytes       <-----> BYTES
   ByteString  ------> BYTES
   Timestamp   <-----> LogicalType(urn="beam:logical_type:micros_instant:v1")
+  datetime.date <---> LogicalType(urn="beam:logical_type:date:v1")
   Decimal     <-----> LogicalType(urn="beam:logical_type:fixed_decimal:v1")
   Mapping     <-----> MapType
   Sequence    <-----> ArrayType
@@ -1004,6 +1005,33 @@ def to_language_type(self, value):
     return Timestamp(seconds=int(value.seconds), micros=int(value.micros))
 
 
+@LogicalType._register_internal
+class Date(NoArgumentLogicalType[datetime.date, np.int64]):
+  """Date logical type that handles ``datetime.date``, days since epoch."""
+  EPOCH = datetime.date(1970, 1, 1)
+
+  @classmethod
+  def urn(cls):
+    return common_urns.date.urn
+
+  @classmethod
+  def representation_type(cls):
+    # type: () -> type
+    return np.int64
+
+  @classmethod
+  def language_type(cls):
+    return datetime.date
+
+  def to_representation_type(self, value):
+    # type: (datetime.date) -> np.int64
+    return (value - self.EPOCH).days
+
+  def to_language_type(self, value):
+    # type: (np.int64) -> datetime.date
+    return self.EPOCH + datetime.timedelta(days=value)
+
+
 @LogicalType._register_internal
 class PythonCallable(NoArgumentLogicalType[PythonCallableWithSource, str]):
   """A logical type for PythonCallableSource objects."""
@@ -1244,7 +1272,6 @@ def argument(self):
 # TODO: A temporary fix for missing jdbc logical types.
 # See the discussion in https://github.com/apache/beam/issues/35738 for
 # more detail.
-@LogicalType._register_internal
 class JdbcDateType(LogicalType[datetime.date, MillisInstant, str]):
   """
   For internal use only; no backwards-compatibility guarantees.

sdks/python/apache_beam/typehints/schemas_test.py

@@ -20,6 +20,7 @@
 # pytype: skip-file
 
 import dataclasses
+import datetime
 import itertools
 import pickle
 import unittest
@@ -105,6 +106,7 @@ class ComplexSchema(NamedTuple):
   optional_array: Optional[Sequence[np.float32]]
   array_optional: Sequence[Optional[bool]]
   timestamp: Timestamp
+  date: datetime.date
 
 
 def get_test_beam_fieldtype_protos():