Merge pull request #2544 from iamsdas:desc

PiperOrigin-RevId: 336117107

Author: copybara-github

tensorflow_datasets/core/dataset_info.py

@@ -121,13 +121,19 @@ def __init__(self,
     """
     self._builder = builder
 
+    if builder.builder_config:
+      config_name = builder.builder_config.name
+      config_description = builder.builder_config.description
+    else:
+      config_name = None
+      config_description = None
+
     self._info_proto = dataset_info_pb2.DatasetInfo(
         name=builder.name,
         description=utils.dedent(description),
-        version=str(builder._version),  # pylint: disable=protected-access
-        config_name=(
-            builder.builder_config.name if builder.builder_config else None
-        ),
+        version=str(builder.version),
+        config_name=config_name,
+        config_description=config_description,
         citation=utils.dedent(citation),
         redistribution_info=dataset_info_pb2.RedistributionInfo(
             license=utils.dedent(redistribution_info.pop("license")),

tensorflow_datasets/core/proto/dataset_info.proto

@@ -61,6 +61,9 @@ message DatasetInfo {
   // Config name of the dataset
   string config_name = 13;
 
+  // Config description of the dataset
+  string config_description = 14;
+
   // A citation string if one exists for this dataset.
   string citation = 3;
 
@@ -90,5 +93,5 @@ message DatasetInfo {
 
   RedistributionInfo redistribution_info = 11;
 
-  // Next available: 14
+  // Next available: 15
 }

tensorflow_datasets/core/proto/dataset_info_generated_pb2.py

@@ -18,7 +18,6 @@
 # -*- coding: utf-8 -*-
 # Generated by the protocol buffer compiler.  DO NOT EDIT!
 # source: dataset_info.proto
-"""Generated protocol buffer code."""
 
 from google.protobuf import descriptor as _descriptor
 from google.protobuf import message as _message
@@ -38,7 +37,7 @@
   package='tensorflow_datasets',
   syntax='proto3',
   serialized_options=b'\370\001\001',
-  serialized_pb=b'\n\x12\x64\x61taset_info.proto\x12\x13tensorflow_datasets\x1a-tensorflow_metadata/proto/v0/statistics.proto\x1a)tensorflow_metadata/proto/v0/schema.proto\"\x1f\n\x0f\x44\x61tasetLocation\x12\x0c\n\x04urls\x18\x01 \x03(\t\"\x9d\x01\n\tSplitInfo\x12\x0c\n\x04name\x18\x01 \x01(\t\x12\x12\n\nnum_shards\x18\x02 \x01(\x03\x12\x15\n\rshard_lengths\x18\x04 \x03(\x03\x12\x11\n\tnum_bytes\x18\x05 \x01(\x03\x12\x44\n\nstatistics\x18\x03 \x01(\x0b\x32\x30.tensorflow.metadata.v0.DatasetFeatureStatistics\"/\n\x0eSupervisedKeys\x12\r\n\x05input\x18\x01 \x01(\t\x12\x0e\n\x06output\x18\x02 \x01(\t\"%\n\x12RedistributionInfo\x12\x0f\n\x07license\x18\x01 \x01(\t\"\xc9\x04\n\x0b\x44\x61tasetInfo\x12\x0c\n\x04name\x18\x01 \x01(\t\x12\x13\n\x0b\x64\x65scription\x18\x02 \x01(\t\x12\x0f\n\x07version\x18\t \x01(\t\x12\x13\n\x0b\x63onfig_name\x18\r \x01(\t\x12\x10\n\x08\x63itation\x18\x03 \x01(\t\x12\x19\n\rsize_in_bytes\x18\x04 \x01(\x03\x42\x02\x18\x01\x12\x15\n\rdownload_size\x18\x0c \x01(\x03\x12\x36\n\x08location\x18\x05 \x01(\x0b\x32$.tensorflow_datasets.DatasetLocation\x12W\n\x12\x64ownload_checksums\x18\n \x03(\x0b\x32\x37.tensorflow_datasets.DatasetInfo.DownloadChecksumsEntryB\x02\x18\x01\x12.\n\x06schema\x18\x06 \x01(\x0b\x32\x1e.tensorflow.metadata.v0.Schema\x12.\n\x06splits\x18\x07 \x03(\x0b\x32\x1e.tensorflow_datasets.SplitInfo\x12<\n\x0fsupervised_keys\x18\x08 \x01(\x0b\x32#.tensorflow_datasets.SupervisedKeys\x12\x44\n\x13redistribution_info\x18\x0b \x01(\x0b\x32\'.tensorflow_datasets.RedistributionInfo\x1a\x38\n\x16\x44ownloadChecksumsEntry\x12\x0b\n\x03key\x18\x01 \x01(\t\x12\r\n\x05value\x18\x02 \x01(\t:\x02\x38\x01\x42\x03\xf8\x01\x01\x62\x06proto3'
+  serialized_pb=b'\n\x12\x64\x61taset_info.proto\x12\x13tensorflow_datasets\x1a-tensorflow_metadata/proto/v0/statistics.proto\x1a)tensorflow_metadata/proto/v0/schema.proto\"\x1f\n\x0f\x44\x61tasetLocation\x12\x0c\n\x04urls\x18\x01 \x03(\t\"\x9d\x01\n\tSplitInfo\x12\x0c\n\x04name\x18\x01 \x01(\t\x12\x12\n\nnum_shards\x18\x02 \x01(\x03\x12\x15\n\rshard_lengths\x18\x04 \x03(\x03\x12\x11\n\tnum_bytes\x18\x05 \x01(\x03\x12\x44\n\nstatistics\x18\x03 \x01(\x0b\x32\x30.tensorflow.metadata.v0.DatasetFeatureStatistics\"/\n\x0eSupervisedKeys\x12\r\n\x05input\x18\x01 \x01(\t\x12\x0e\n\x06output\x18\x02 \x01(\t\"%\n\x12RedistributionInfo\x12\x0f\n\x07license\x18\x01 \x01(\t\"\xe5\x04\n\x0b\x44\x61tasetInfo\x12\x0c\n\x04name\x18\x01 \x01(\t\x12\x13\n\x0b\x64\x65scription\x18\x02 \x01(\t\x12\x0f\n\x07version\x18\t \x01(\t\x12\x13\n\x0b\x63onfig_name\x18\r \x01(\t\x12\x1a\n\x12\x63onfig_description\x18\x0e \x01(\t\x12\x10\n\x08\x63itation\x18\x03 \x01(\t\x12\x19\n\rsize_in_bytes\x18\x04 \x01(\x03\x42\x02\x18\x01\x12\x15\n\rdownload_size\x18\x0c \x01(\x03\x12\x36\n\x08location\x18\x05 \x01(\x0b\x32$.tensorflow_datasets.DatasetLocation\x12W\n\x12\x64ownload_checksums\x18\n \x03(\x0b\x32\x37.tensorflow_datasets.DatasetInfo.DownloadChecksumsEntryB\x02\x18\x01\x12.\n\x06schema\x18\x06 \x01(\x0b\x32\x1e.tensorflow.metadata.v0.Schema\x12.\n\x06splits\x18\x07 \x03(\x0b\x32\x1e.tensorflow_datasets.SplitInfo\x12<\n\x0fsupervised_keys\x18\x08 \x01(\x0b\x32#.tensorflow_datasets.SupervisedKeys\x12\x44\n\x13redistribution_info\x18\x0b \x01(\x0b\x32\'.tensorflow_datasets.RedistributionInfo\x1a\x38\n\x16\x44ownloadChecksumsEntry\x12\x0b\n\x03key\x18\x01 \x01(\t\x12\r\n\x05value\x18\x02 \x01(\t:\x02\x38\x01\x42\x03\xf8\x01\x01\x62\x06proto3'
   ,
   dependencies=[tensorflow__metadata_dot_proto_dot_v0_dot_statistics__pb2.DESCRIPTOR,tensorflow__metadata_dot_proto_dot_v0_dot_schema__pb2.DESCRIPTOR,])
 
@@ -237,8 +236,8 @@
   extension_ranges=[],
   oneofs=[
   ],
-  serialized_start=944,
-  serialized_end=1000,
+  serialized_start=972,
+  serialized_end=1028,
 )
 
 _DATASETINFO = _descriptor.Descriptor(
@@ -277,63 +276,70 @@
       is_extension=False, extension_scope=None,
       serialized_options=None, file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='citation', full_name='tensorflow_datasets.DatasetInfo.citation', index=4,
+      name='config_description', full_name='tensorflow_datasets.DatasetInfo.config_description', index=4,
+      number=14, type=9, cpp_type=9, label=1,
+      has_default_value=False, default_value=b"".decode('utf-8'),
+      message_type=None, enum_type=None, containing_type=None,
+      is_extension=False, extension_scope=None,
+      serialized_options=None, file=DESCRIPTOR),
+    _descriptor.FieldDescriptor(
+      name='citation', full_name='tensorflow_datasets.DatasetInfo.citation', index=5,
       number=3, type=9, cpp_type=9, label=1,
       has_default_value=False, default_value=b"".decode('utf-8'),
       message_type=None, enum_type=None, containing_type=None,
       is_extension=False, extension_scope=None,
       serialized_options=None, file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='size_in_bytes', full_name='tensorflow_datasets.DatasetInfo.size_in_bytes', index=5,
+      name='size_in_bytes', full_name='tensorflow_datasets.DatasetInfo.size_in_bytes', index=6,
       number=4, type=3, cpp_type=2, label=1,
       has_default_value=False, default_value=0,
       message_type=None, enum_type=None, containing_type=None,
       is_extension=False, extension_scope=None,
       serialized_options=b'\030\001', file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='download_size', full_name='tensorflow_datasets.DatasetInfo.download_size', index=6,
+      name='download_size', full_name='tensorflow_datasets.DatasetInfo.download_size', index=7,
       number=12, type=3, cpp_type=2, label=1,
       has_default_value=False, default_value=0,
       message_type=None, enum_type=None, containing_type=None,
       is_extension=False, extension_scope=None,
       serialized_options=None, file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='location', full_name='tensorflow_datasets.DatasetInfo.location', index=7,
+      name='location', full_name='tensorflow_datasets.DatasetInfo.location', index=8,
       number=5, 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='download_checksums', full_name='tensorflow_datasets.DatasetInfo.download_checksums', index=8,
+      name='download_checksums', full_name='tensorflow_datasets.DatasetInfo.download_checksums', index=9,
       number=10, type=11, cpp_type=10, label=3,
       has_default_value=False, default_value=[],
       message_type=None, enum_type=None, containing_type=None,
       is_extension=False, extension_scope=None,
       serialized_options=b'\030\001', file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='schema', full_name='tensorflow_datasets.DatasetInfo.schema', index=9,
+      name='schema', full_name='tensorflow_datasets.DatasetInfo.schema', index=10,
       number=6, 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='splits', full_name='tensorflow_datasets.DatasetInfo.splits', index=10,
+      name='splits', full_name='tensorflow_datasets.DatasetInfo.splits', index=11,
       number=7, type=11, cpp_type=10, label=3,
       has_default_value=False, default_value=[],
       message_type=None, enum_type=None, containing_type=None,
       is_extension=False, extension_scope=None,
       serialized_options=None, file=DESCRIPTOR),
     _descriptor.FieldDescriptor(
-      name='supervised_keys', full_name='tensorflow_datasets.DatasetInfo.supervised_keys', index=11,
+      name='supervised_keys', full_name='tensorflow_datasets.DatasetInfo.supervised_keys', index=12,
       number=8, 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='redistribution_info', full_name='tensorflow_datasets.DatasetInfo.redistribution_info', index=12,
+      name='redistribution_info', full_name='tensorflow_datasets.DatasetInfo.redistribution_info', index=13,
       number=11, type=11, cpp_type=10, label=1,
       has_default_value=False, default_value=None,
       message_type=None, enum_type=None, containing_type=None,
@@ -352,7 +358,7 @@
   oneofs=[
   ],
   serialized_start=415,
-  serialized_end=1000,
+  serialized_end=1028,
 )
 
 _SPLITINFO.fields_by_name['statistics'].message_type = tensorflow__metadata_dot_proto_dot_v0_dot_statistics__pb2._DATASETFEATURESTATISTICS
@@ -418,4 +424,4 @@
 _DATASETINFO_DOWNLOADCHECKSUMSENTRY._options = None
 _DATASETINFO.fields_by_name['size_in_bytes']._options = None
 _DATASETINFO.fields_by_name['download_checksums']._options = None
-# @@protoc_insertion_point(module_scope)
+# @@protoc_insertion_point(module_scope)