Release ArrayRecord 0.4.1.

PiperOrigin-RevId: 554817684

Author: Marvin182

.github/workflows/python-tests.yml

@@ -9,7 +9,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: ['3.8', '3.9', '3.10']
+        python-version: ['3.9', '3.10']
     env:
       DOCKER_BUILDKIT: 1
       TMP_FOLDER: /tmp/array_record

python/BUILD

@@ -34,9 +34,7 @@ py_library(
     srcs = ["array_record_data_source.py"],
     data = [":array_record_module.so"],
     deps = [
-        "//pyglib:gfile",
         # Implicit etils (/epath) dependency.
-        "//third_party/py/grain/google/ntk/array_record/python:fast_array_record",
     ],
 )
 
@@ -51,8 +49,6 @@ py_test(
     ],
     deps = [
         ":array_record_data_source",
-        "//third_party/py/grain/google/ntk/array_record/python:fast_array_record",
-        "//third_party/py/numpy",
         "@com_google_absl_py//absl/testing:absltest",
         "@com_google_absl_py//absl/testing:flagsaver",
         "@com_google_absl_py//absl/testing:parameterized",

python/array_record_data_source.py

@@ -46,6 +46,8 @@ def __getitem__(self, record_keys: Sequence[int]) -> Sequence[T]:
 from absl import logging
 from etils import epath
 
+from . import array_record_module
+
 # TODO(jolesiak): Decide what to do with these flags, e.g., remove them (could
 # be appropriate if we decide to use asyncio) or move them somewhere else and
 # pass the number of threads as an argument. For now, since we experiment, it's
@@ -156,7 +158,7 @@ def get_read_instruction(path: PathLikeOrFileInstruction) -> _ReadInstruction:
       end = int(m.group(3))
     else:
       path = os.fspath(path)
-      reader = array_record.ArrayRecordReader(path)
+      reader = array_record_module.ArrayRecordReader(path)
       start = 0  # Using whole file.
       end = reader.num_records()
       reader.close()
@@ -173,15 +175,15 @@ def get_read_instruction(path: PathLikeOrFileInstruction) -> _ReadInstruction:
 
 def _create_reader(filename: epath.PathLike):
   """Returns an ArrayRecordReader for the given filename."""
-  return array_record.ArrayRecordReader(
+  return array_record_module.ArrayRecordReader(
       filename,
       options="readahead_buffer_size:0",
       file_reader_buffer_size=32768,
   )
 
 
 def _check_group_size(
-    filename: epath.PathLike, reader: array_record.ArrayRecordReader
+    filename: epath.PathLike, reader: array_record_module.ArrayRecordReader
 ) -> None:
   """Logs an error if the group size of the underlying file is not 1."""
   options = reader.writer_options_string()

python/array_record_data_source_test.py

@@ -23,7 +23,6 @@
 from absl.testing import absltest
 from absl.testing import flagsaver
 from absl.testing import parameterized
-import numpy as np
 
 from array_record.python import array_record_data_source
 from array_record.python import array_record_module
@@ -138,15 +137,6 @@ def test_array_record_data_source_random_order(self):
     self.assertEqual(expected_data, actual_data)
     self.assertTrue(all(reader is None for reader in ar._readers))
 
-  def test_array_record_data_source_numpy_index(self):
-    # some random permutation
-    with array_record_data_source.ArrayRecordDataSource([
-        self.testdata_dir / "digits.array_record-00000-of-00002",
-        self.testdata_dir / "digits.array_record-00001-of-00002",
-    ]) as ar:
-      actual_data = ar[np.int64(3)]
-    self.assertEqual(b"3", actual_data)
-
   def test_array_record_data_source_random_order_batched(self):
     # some random permutation
     indices_to_read = [3, 0, 5, 9, 2, 1, 4, 7, 8, 6]

setup.py

@@ -10,21 +10,21 @@
 
 setup(
     name='array_record',
-    version='0.4.0',
+    version='0.4.1',
     description='A file format that achieves a new frontier of IO efficiency',
     author='ArrayRecord team',
     author_email='no-reply@google.com',
     packages=find_packages(),
     include_package_data=True,
     package_data={'': ['*.so']},
-    python_requires='>=3.8',
+    python_requires='>=3.9',
     install_requires=REQUIRED_PACKAGES,
     url='https://github.com/google/array_record',
     license='Apache-2.0',
     classifiers=[
-        'Programming Language :: Python :: 3.8',
         'Programming Language :: Python :: 3.9',
         'Programming Language :: Python :: 3.10',
+        'Programming Language :: Python :: 3.11',
     ],
     zip_safe=False,
 )