BigQuery: Remove dataset property from TableReference and add project…

…/dataset_id properties (#4011)

* adds dataset_id and project properties to TableReference

* Remove dataset property from Table and TableReference

bigquery/google/cloud/bigquery/job.py

@@ -1119,8 +1119,8 @@ def _build_resource(self):
         """Generate a resource for :meth:`begin`."""
 
         source_ref = {
-            'projectId': self.source.dataset.project,
-            'datasetId': self.source.dataset.dataset_id,
+            'projectId': self.source.project,
+            'datasetId': self.source.dataset_id,
             'tableId': self.source.table_id,
         }
 

bigquery/google/cloud/bigquery/table.py

@@ -64,17 +64,27 @@ class TableReference(object):
     """
 
     def __init__(self, dataset_ref, table_id):
-        self._dataset_ref = dataset_ref
+        self._project = dataset_ref.project
+        self._dataset_id = dataset_ref.dataset_id
         self._table_id = table_id
 
     @property
-    def dataset(self):
-        """Pointer to the dataset.
+    def project(self):
+        """Project bound to the table.
 
-        :rtype: :class:`google.cloud.bigquery.dataset.DatasetReference`
-        :returns: a pointer to the dataset.
+        :rtype: str
+        :returns: the project (derived from the dataset reference).
         """
-        return self._dataset_ref
+        return self._project
+
+    @property
+    def dataset_id(self):
+        """ID of dataset containing the table.
+
+        :rtype: str
+        :returns: the ID (derived from the dataset reference).
+        """
+        return self._dataset_id
 
     @property
     def table_id(self):
@@ -92,7 +102,8 @@ def path(self):
         :rtype: str
         :returns: the path based on project, dataset and table IDs.
         """
-        return '%s/tables/%s' % (self._dataset_ref.path, self._table_id)
+        return '/projects/%s/datasets/%s/tables/%s' % (
+            self._project, self._dataset_id, self._table_id)
 
 
 class Table(object):
@@ -111,8 +122,9 @@ class Table(object):
     _schema = None
 
     def __init__(self, table_ref, schema=(), client=None):
+        self._project = table_ref.project
         self._table_id = table_ref.table_id
-        self._dataset = table_ref.dataset
+        self._dataset_id = table_ref.dataset_id
         self._properties = {}
         # Let the @property do validation.
         self.schema = schema
@@ -125,7 +137,7 @@ def project(self):
         :rtype: str
         :returns: the project (derived from the dataset).
         """
-        return self._dataset.project
+        return self._project
 
     @property
     def dataset_id(self):
@@ -134,7 +146,7 @@ def dataset_id(self):
         :rtype: str
         :returns: the ID (derived from the dataset).
         """
-        return self._dataset.dataset_id
+        return self._dataset_id
 
     @property
     def table_id(self):
@@ -152,7 +164,8 @@ def path(self):
         :rtype: str
         :returns: the path based on project, dataset and table IDs.
         """
-        return '%s/tables/%s' % (self._dataset.path, self.table_id)
+        return '/projects/%s/datasets/%s/tables/%s' % (
+            self._project, self._dataset_id, self._table_id)
 
     @property
     def schema(self):
@@ -550,8 +563,8 @@ def _build_resource(self):
         """Generate a resource for ``create`` or ``update``."""
         resource = {
             'tableReference': {
-                'projectId': self._dataset.project,
-                'datasetId': self._dataset.dataset_id,
+                'projectId': self._project,
+                'datasetId': self._dataset_id,
                 'tableId': self.table_id},
         }
         if self.description is not None:
@@ -596,7 +609,7 @@ def create(self, client=None):
         """
         client = self._require_client(client)
         path = '/projects/%s/datasets/%s/tables' % (
-            self._dataset.project, self._dataset.dataset_id)
+            self._project, self._dataset_id)
         api_response = client._connection.api_request(
             method='POST', path=path, data=self._build_resource())
         self._set_properties(api_response)
@@ -1188,7 +1201,8 @@ def upload_from_file(self,
         _maybe_rewind(file_obj, rewind=rewind)
         _check_mode(file_obj)
         metadata = _get_upload_metadata(
-            source_format, self._schema, self._dataset, self.table_id)
+            source_format, self._schema, self._project,
+            self._dataset_id, self.table_id)
         _configure_job_metadata(metadata, allow_jagged_rows,
                                 allow_quoted_newlines, create_disposition,
                                 encoding, field_delimiter,
@@ -1353,7 +1367,7 @@ def _get_upload_headers(user_agent):
     }
 
 
-def _get_upload_metadata(source_format, schema, dataset, table_id):
+def _get_upload_metadata(source_format, schema, project, dataset_id, table_id):
     """Get base metadata for creating a table.
 
     :type source_format: str
@@ -1363,8 +1377,11 @@ def _get_upload_metadata(source_format, schema, dataset, table_id):
     :type schema: list
     :param schema: List of :class:`SchemaField` associated with a table.
 
-    :type dataset: :class:`~google.cloud.bigquery.dataset.Dataset`
-    :param dataset: A dataset which contains a table.
+    :type project: str
+    :param table_id: The project bound to the table.
+
+    :type dataset_id: str
+    :param table_id: The dataset_id of the dataset.
 
     :type table_id: str
     :param table_id: The table_id of the table.
@@ -1375,8 +1392,8 @@ def _get_upload_metadata(source_format, schema, dataset, table_id):
     load_config = {
         'sourceFormat': source_format,
         'destinationTable': {
-            'projectId': dataset.project,
-            'datasetId': dataset.dataset_id,
+            'projectId': project,
+            'datasetId': dataset_id,
             'tableId': table_id,
         },
     }

bigquery/tests/system.py

@@ -568,7 +568,7 @@ def _load_table_for_extract_table(
         self.to_delete.insert(0, blob)
 
         dataset = retry_403(Config.CLIENT.create_dataset)(
-            Dataset(table.dataset.dataset_id))
+            Dataset(table.dataset_id))
         self.to_delete.append(dataset)
         table = dataset.table(table.table_id)
         self.to_delete.insert(0, table)

bigquery/tests/unit/test_dataset.py

@@ -104,7 +104,8 @@ def test_ctor_defaults(self):
     def test_table(self):
         dataset_ref = self._make_one('some-project-1', 'dataset_1')
         table_ref = dataset_ref.table('table_1')
-        self.assertIs(table_ref.dataset, dataset_ref)
+        self.assertEqual(table_ref.dataset_id, 'dataset_1')
+        self.assertEqual(table_ref.project, 'some-project-1')
         self.assertEqual(table_ref.table_id, 'table_1')
 
 
@@ -543,7 +544,8 @@ def test_table_wo_schema(self):
         table = dataset.table('table_id')
         self.assertIsInstance(table, Table)
         self.assertEqual(table.table_id, 'table_id')
-        self.assertIs(table._dataset, dataset)
+        self.assertEqual(table.dataset_id, self.DS_ID)
+        self.assertEqual(table.project, self.PROJECT)
         self.assertEqual(table.schema, [])
 
     def test_table_w_schema(self):
@@ -558,7 +560,8 @@ def test_table_w_schema(self):
         table = dataset.table('table_id', schema=[full_name, age])
         self.assertIsInstance(table, Table)
         self.assertEqual(table.table_id, 'table_id')
-        self.assertIs(table._dataset, dataset)
+        self.assertEqual(table.dataset_id, self.DS_ID)
+        self.assertEqual(table.project, self.PROJECT)
         self.assertEqual(table.schema, [full_name, age])
 
 

bigquery/tests/unit/test_job.py

@@ -1220,8 +1220,8 @@ def _verifyResourceProperties(self, job, resource):
         self.assertEqual(job.destination_uris, config['destinationUris'])
 
         table_ref = config['sourceTable']
-        self.assertEqual(job.source.dataset.project, table_ref['projectId'])
-        self.assertEqual(job.source.dataset.dataset_id, table_ref['datasetId'])
+        self.assertEqual(job.source.project, table_ref['projectId'])
+        self.assertEqual(job.source.dataset_id, table_ref['datasetId'])
         self.assertEqual(job.source.table_id, table_ref['tableId'])
 
         if 'compression' in config:
@@ -1908,7 +1908,6 @@ def test_statement_type(self):
         self.assertEqual(job.statement_type, statement_type)
 
     def test_referenced_tables(self):
-        from google.cloud.bigquery.dataset import Dataset
         from google.cloud.bigquery.table import Table
 
         ref_tables_resource = [{
@@ -1942,24 +1941,21 @@ def test_referenced_tables(self):
 
         self.assertIsInstance(local1, Table)
         self.assertEqual(local1.table_id, 'local1')
-        self.assertIsInstance(local1._dataset, Dataset)
         self.assertEqual(local1.dataset_id, 'dataset')
         self.assertEqual(local1.project, self.PROJECT)
-        self.assertIs(local1._dataset._client, client)
+        self.assertIs(local1._client, client)
 
         self.assertIsInstance(local2, Table)
         self.assertEqual(local2.table_id, 'local2')
-        self.assertIsInstance(local2._dataset, Dataset)
         self.assertEqual(local2.dataset_id, 'dataset')
         self.assertEqual(local2.project, self.PROJECT)
-        self.assertIs(local2._dataset._client, client)
+        self.assertIs(local2._client, client)
 
         self.assertIsInstance(remote, Table)
         self.assertEqual(remote.table_id, 'other-table')
-        self.assertIsInstance(remote._dataset, Dataset)
         self.assertEqual(remote.dataset_id, 'other-dataset')
         self.assertEqual(remote.project, 'other-project-123')
-        self.assertIs(remote._dataset._client, client)
+        self.assertIs(remote._client, client)
 
     def test_undeclared_query_paramters(self):
         from google.cloud.bigquery._helpers import ArrayQueryParameter

bigquery/tests/unit/test_table.py

@@ -55,7 +55,7 @@ def test_ctor_defaults(self):
         dataset_ref = DatasetReference('project_1', 'dataset_1')
 
         table_ref = self._make_one(dataset_ref, 'table_1')
-        self.assertIs(table_ref.dataset, dataset_ref)
+        self.assertEqual(table_ref.dataset_id, dataset_ref.dataset_id)
         self.assertEqual(table_ref.table_id, 'table_1')
 
 
@@ -173,7 +173,6 @@ def test_ctor(self):
         table = self._make_one(table_ref, client=client)
 
         self.assertEqual(table.table_id, self.TABLE_NAME)
-        self.assertIs(table._dataset, dataset)
         self.assertEqual(table.project, self.PROJECT)
         self.assertEqual(table.dataset_id, self.DS_ID)
         self.assertEqual(
@@ -1710,7 +1709,8 @@ def _initiate_resumable_upload_helper(self, num_retries=None):
         data = b'goodbye gudbi gootbee'
         stream = io.BytesIO(data)
         metadata = _get_upload_metadata(
-            'CSV', table._schema, table._dataset, table.table_id)
+            'CSV', table._schema, table.project,
+            table.dataset_id, table.table_id)
         upload, transport = table._initiate_resumable_upload(
             client, stream, metadata, num_retries)
 
@@ -1777,7 +1777,8 @@ def _do_multipart_upload_success_helper(
         data = b'Bzzzz-zap \x00\x01\xf4'
         stream = io.BytesIO(data)
         metadata = _get_upload_metadata(
-            'CSV', table._schema, table._dataset, table.table_id)
+            'CSV', table._schema, table.project,
+            table.dataset_id, table.table_id)
         size = len(data)
         response = table._do_multipart_upload(
             client, stream, metadata, size, num_retries)
@@ -1928,7 +1929,7 @@ def test_upload_file_resumable_metadata(self):
                 'load': {
                     'sourceFormat': config_args['source_format'],
                     'destinationTable': {
-                        'projectId': table._dataset.project,
+                        'projectId': table.project,
                         'datasetId': table.dataset_id,
                         'tableId': table.table_id,
                     },
@@ -2254,18 +2255,20 @@ def test_w_subfields(self):
 class Test__get_upload_metadata(unittest.TestCase):
 
     @staticmethod
-    def _call_fut(source_format, schema, dataset, name):
+    def _call_fut(source_format, schema, project, dataset_id, name):
         from google.cloud.bigquery.table import _get_upload_metadata
 
-        return _get_upload_metadata(source_format, schema, dataset, name)
+        return _get_upload_metadata(
+            source_format, schema, project, dataset_id, name)
 
     def test_empty_schema(self):
         source_format = 'AVRO'
         dataset = mock.Mock(project='prediction',
                             spec=['dataset_id', 'project'])
         dataset.dataset_id = 'market'  # mock.Mock() treats `name` specially.
         table_name = 'chairs'
-        metadata = self._call_fut(source_format, [], dataset, table_name)
+        metadata = self._call_fut(source_format, [], dataset.project,
+                                  dataset.dataset_id, table_name)
 
         expected = {
             'configuration': {
@@ -2290,7 +2293,8 @@ def test_with_schema(self):
         dataset.dataset_id = 'movie'  # mock.Mock() treats `name` specially.
         table_name = 'teebull-neem'
         metadata = self._call_fut(
-            source_format, [full_name], dataset, table_name)
+            source_format, [full_name], dataset.project,
+            dataset.dataset_id, table_name)
 
         expected = {
             'configuration': {