Update proto files.

PiperOrigin-RevId: 523948721

Author: fineguy

tensorflow_datasets/core/proto/dataset_info_generated_pb2.py

@@ -94,6 +94,7 @@
     DESCRIPTOR, 'dataset_info_pb2', globals()
 )
 if _descriptor._USE_C_DESCRIPTORS == False:
+
   DESCRIPTOR._options = None
   DESCRIPTOR._serialized_options = b'\370\001\001'
   _SUPERVISEDKEYS_DICT_DICTENTRY._options = None

tensorflow_datasets/core/proto/feature_generated_pb2.py

@@ -81,6 +81,7 @@
 _builder.BuildMessageAndEnumDescriptors(DESCRIPTOR, globals())
 _builder.BuildTopDescriptorsAndMessages(DESCRIPTOR, 'feature_pb2', globals())
 if _descriptor._USE_C_DESCRIPTORS == False:
+
   DESCRIPTOR._options = None
   DESCRIPTOR._serialized_options = b'\370\001\001'
   _FEATURESDICT_FEATURESENTRY._options = None

tensorflow_datasets/core/proto/generate_py_proto.sh

@@ -6,10 +6,10 @@
 set -e # Fail on errors
 
 # Ensure we have the desired protoc version.
-MIN_VERSION="libprotoc 3.11.2"
+MIN_VERSION="libprotoc 3.20.0"
 CURR_VERSION="$(protoc --version)"
 if ["$(printf '%s\n' "$MIN_VERSION" "$CURR_VERSION" | sort -V | head -n1)" != "$MIN_VERSION"]; then
-  echo 'Please use version 3.11.2 or above.'
+  echo 'Please use version 3.20.0 or above.'
   echo 'Please run install_protoc.sh to install it.'
   exit
 else

tensorflow_datasets/proto/README.md

@@ -1,6 +1,11 @@
 This is the directory with 'proto' files that are used by the datasets.
 
+## Protoc compiler
 Proto files have to be compiled with protoc compiler -
-for installation instructions, please see core/proto/install_protoc.sh
+for installation instructions, please see `core/proto/install_protoc.sh`
+
+## Proto generation
+- `tf_example.proto` and `tf_feature.proto` can be generated by running `build_tf_proto.py`;
+- `waymo_dataset.proto` can be generated by manually running protoc compiler.
 
 Please always add both the original .proto and the compiled python version.

tensorflow_datasets/proto/build_tf_proto.py

@@ -121,7 +121,12 @@ def main(_) -> None:
 
     # Launch protoc
     subprocess.run(
-        ['protoc', os.fspath(tfds_proto_path), '--python_out=.'],
+        [
+            'protoc',
+            os.fspath(tfds_proto_path),
+            f'--proto_path={PATH_TO_TFDS.parent}',
+            f'--python_out={PATH_TO_TFDS.parent}',
+        ],
         check=True,
     )
     name_before = PATH_TO_PROTO / f'tf_{proto_name.value}_pb2.py'

tensorflow_datasets/proto/tf_example_generated_pb2.py

@@ -18,10 +18,10 @@
 # -*- coding: utf-8 -*-
 # Generated by the protocol buffer compiler.  DO NOT EDIT!
 # source: tensorflow_datasets/proto/tf_example.proto
-
+"""Generated protocol buffer code."""
+from google.protobuf.internal import builder as _builder
 from google.protobuf import descriptor as _descriptor
-from google.protobuf import message as _message
-from google.protobuf import reflection as _reflection
+from google.protobuf import descriptor_pool as _descriptor_pool
 from google.protobuf import symbol_database as _symbol_database
 # @@protoc_insertion_point(imports)
 
@@ -31,153 +31,22 @@
 from tensorflow_datasets.proto import tf_feature_generated_pb2 as tensorflow__datasets_dot_proto_dot_tf__feature__pb2
 
 
-DESCRIPTOR = _descriptor.FileDescriptor(
-    name='tensorflow_datasets/proto/tf_example.proto',
-    package='tensorflow_copy',
-    syntax='proto3',
-    serialized_options=b'\n\026org.tensorflow.exampleB\rExampleProtosP\001ZSgithub.com/tensorflow/tensorflow/tensorflow/go/core/example/example_protos_go_proto\370\001\001',
-    serialized_pb=(
-        b'\n*tensorflow_datasets/proto/tf_example.proto\x12\x0ftensorflow_copy\x1a*tensorflow_datasets/proto/tf_feature.proto"6\n\x07\x45xample\x12+\n\x08\x66\x65\x61tures\x18\x01'
-        b' \x01(\x0b\x32\x19.tensorflow_copy.Features"s\n\x0fSequenceExample\x12*\n\x07\x63ontext\x18\x01'
-        b' \x01(\x0b\x32\x19.tensorflow_copy.Features\x12\x34\n\rfeature_lists\x18\x02'
-        b' \x01(\x0b\x32\x1d.tensorflow_copy.FeatureListsB\x81\x01\n\x16org.tensorflow.exampleB\rExampleProtosP\x01ZSgithub.com/tensorflow/tensorflow/tensorflow/go/core/example/example_protos_go_proto\xf8\x01\x01\x62\x06proto3'
-    ),
-    dependencies=[
-        tensorflow__datasets_dot_proto_dot_tf__feature__pb2.DESCRIPTOR,
-    ],
-)
-
-
-_EXAMPLE = _descriptor.Descriptor(
-    name='Example',
-    full_name='tensorflow_copy.Example',
-    filename=None,
-    file=DESCRIPTOR,
-    containing_type=None,
-    fields=[
-        _descriptor.FieldDescriptor(
-            name='features',
-            full_name='tensorflow_copy.Example.features',
-            index=0,
-            number=1,
-            type=11,
-            cpp_type=10,
-            label=1,
-            has_default_value=False,
-            default_value=None,
-            message_type=None,
-            enum_type=None,
-            containing_type=None,
-            is_extension=False,
-            extension_scope=None,
-            serialized_options=None,
-            file=DESCRIPTOR,
-        ),
-    ],
-    extensions=[],
-    nested_types=[],
-    enum_types=[],
-    serialized_options=None,
-    is_extendable=False,
-    syntax='proto3',
-    extension_ranges=[],
-    oneofs=[],
-    serialized_start=107,
-    serialized_end=161,
+DESCRIPTOR = _descriptor_pool.Default().AddSerializedFile(
+    b'\n*tensorflow_datasets/proto/tf_example.proto\x12\x0ftensorflow_copy\x1a*tensorflow_datasets/proto/tf_feature.proto"6\n\x07\x45xample\x12+\n\x08\x66\x65\x61tures\x18\x01'
+    b' \x01(\x0b\x32\x19.tensorflow_copy.Features"s\n\x0fSequenceExample\x12*\n\x07\x63ontext\x18\x01'
+    b' \x01(\x0b\x32\x19.tensorflow_copy.Features\x12\x34\n\rfeature_lists\x18\x02'
+    b' \x01(\x0b\x32\x1d.tensorflow_copy.FeatureListsB\x81\x01\n\x16org.tensorflow.exampleB\rExampleProtosP\x01ZSgithub.com/tensorflow/tensorflow/tensorflow/go/core/example/example_protos_go_proto\xf8\x01\x01\x62\x06proto3'
 )
 
-
-_SEQUENCEEXAMPLE = _descriptor.Descriptor(
-    name='SequenceExample',
-    full_name='tensorflow_copy.SequenceExample',
-    filename=None,
-    file=DESCRIPTOR,
-    containing_type=None,
-    fields=[
-        _descriptor.FieldDescriptor(
-            name='context',
-            full_name='tensorflow_copy.SequenceExample.context',
-            index=0,
-            number=1,
-            type=11,
-            cpp_type=10,
-            label=1,
-            has_default_value=False,
-            default_value=None,
-            message_type=None,
-            enum_type=None,
-            containing_type=None,
-            is_extension=False,
-            extension_scope=None,
-            serialized_options=None,
-            file=DESCRIPTOR,
-        ),
-        _descriptor.FieldDescriptor(
-            name='feature_lists',
-            full_name='tensorflow_copy.SequenceExample.feature_lists',
-            index=1,
-            number=2,
-            type=11,
-            cpp_type=10,
-            label=1,
-            has_default_value=False,
-            default_value=None,
-            message_type=None,
-            enum_type=None,
-            containing_type=None,
-            is_extension=False,
-            extension_scope=None,
-            serialized_options=None,
-            file=DESCRIPTOR,
-        ),
-    ],
-    extensions=[],
-    nested_types=[],
-    enum_types=[],
-    serialized_options=None,
-    is_extendable=False,
-    syntax='proto3',
-    extension_ranges=[],
-    oneofs=[],
-    serialized_start=163,
-    serialized_end=278,
-)
-
-_EXAMPLE.fields_by_name['features'].message_type = (
-    tensorflow__datasets_dot_proto_dot_tf__feature__pb2._FEATURES
-)
-_SEQUENCEEXAMPLE.fields_by_name['context'].message_type = (
-    tensorflow__datasets_dot_proto_dot_tf__feature__pb2._FEATURES
+_builder.BuildMessageAndEnumDescriptors(DESCRIPTOR, globals())
+_builder.BuildTopDescriptorsAndMessages(
+    DESCRIPTOR, 'tensorflow_datasets.proto.tf_example_pb2', globals()
 )
-_SEQUENCEEXAMPLE.fields_by_name['feature_lists'].message_type = (
-    tensorflow__datasets_dot_proto_dot_tf__feature__pb2._FEATURELISTS
-)
-DESCRIPTOR.message_types_by_name['Example'] = _EXAMPLE
-DESCRIPTOR.message_types_by_name['SequenceExample'] = _SEQUENCEEXAMPLE
-_sym_db.RegisterFileDescriptor(DESCRIPTOR)
-
-Example = _reflection.GeneratedProtocolMessageType(
-    'Example',
-    (_message.Message,),
-    {
-        'DESCRIPTOR': _EXAMPLE,
-        '__module__': 'tensorflow_datasets.proto.tf_example_pb2',
-        # @@protoc_insertion_point(class_scope:tensorflow_copy.Example)
-    },
-)
-_sym_db.RegisterMessage(Example)
-
-SequenceExample = _reflection.GeneratedProtocolMessageType(
-    'SequenceExample',
-    (_message.Message,),
-    {
-        'DESCRIPTOR': _SEQUENCEEXAMPLE,
-        '__module__': 'tensorflow_datasets.proto.tf_example_pb2',
-        # @@protoc_insertion_point(class_scope:tensorflow_copy.SequenceExample)
-    },
-)
-_sym_db.RegisterMessage(SequenceExample)
-
-
-DESCRIPTOR._options = None
+if _descriptor._USE_C_DESCRIPTORS == False:
+  DESCRIPTOR._options = None
+  DESCRIPTOR._serialized_options = b'\n\026org.tensorflow.exampleB\rExampleProtosP\001ZSgithub.com/tensorflow/tensorflow/tensorflow/go/core/example/example_protos_go_proto\370\001\001'
+  _EXAMPLE._serialized_start = 107
+  _EXAMPLE._serialized_end = 161
+  _SEQUENCEEXAMPLE._serialized_start = 163
+  _SEQUENCEEXAMPLE._serialized_end = 278
 # @@protoc_insertion_point(module_scope)