Init GcsClient only when Pandas Dataframe is passed + Reformat import…

… statements

automl/google/cloud/automl_v1beta1/tables/gcs_client.py

@@ -16,9 +16,10 @@
 
 """Wraps the Google Cloud Storage client library for use in tables helper."""
 
-import google
 import time
 
+from google.api_core import exceptions
+
 try:
     import pandas
 except ImportError:  # pragma: NO COVER
@@ -31,7 +32,7 @@
 
 _PANDAS_REQUIRED = "pandas is required to verify type DataFrame."
 _STORAGE_REQUIRED = (
-    "google.cloud.storage is required to create Google Cloud Storage client."
+    "google-cloud-storage is required to create Google Cloud Storage client."
 )
 
 
@@ -66,8 +67,8 @@ def ensure_bucket_exists(self, project, region):
         Creates this bucket if it doesn't exist.
 
         Args:
-            project (string): The project that stores the bucket.
-            region (string): The region of the bucket.
+            project (str): The project that stores the bucket.
+            region (str): The region of the bucket.
 
         Returns:
             A string representing the created bucket name.
@@ -76,18 +77,18 @@ def ensure_bucket_exists(self, project, region):
 
         try:
             self.client.get_bucket(bucket_name)
-        except google.cloud.exceptions.NotFound:
-            bucket = self.client.Bucket(bucket_name)
+        except exceptions.NotFound:
+            bucket = self.client.bucket(bucket_name)
             bucket.create(project=project, location=region)
         return bucket_name
 
     def upload_pandas_dataframe(self, bucket_name, dataframe, uploaded_csv_name=None):
         """Uploads a Pandas DataFrame as CSV to the bucket.
 
         Args:
-            bucket_name (string): The bucket name to upload the CSV to.
+            bucket_name (str): The bucket name to upload the CSV to.
             dataframe (pandas.DataFrame): The Pandas Dataframe to be uploaded.
-            uploaded_csv_name (Optional[string]): The name for the uploaded CSV.
+            uploaded_csv_name (Optional[str]): The name for the uploaded CSV.
 
         Returns:
             A string representing the GCS URI of the uploaded CSV.

automl/google/cloud/automl_v1beta1/tables/tables_client.py

@@ -755,9 +755,8 @@ def import_data(
 
         request = {}
 
-        self.__ensure_gcs_client_is_initialized(credentials)
-
         if pandas_dataframe is not None:
+            self.__ensure_gcs_client_is_initialized(credentials)
             bucket_name = self.gcs_client.ensure_bucket_exists(project, region)
             gcs_input_uri = self.gcs_client.upload_pandas_dataframe(
                 bucket_name, pandas_dataframe
@@ -2749,9 +2748,8 @@ def batch_predict(
 
         input_request = None
 
-        self.__ensure_gcs_client_is_initialized(credentials)
-
         if pandas_dataframe is not None:
+            self.__ensure_gcs_client_is_initialized(credentials)
             bucket_name = self.gcs_client.ensure_bucket_exists(project, region)
             gcs_input_uri = self.gcs_client.upload_pandas_dataframe(
                 bucket_name, pandas_dataframe

automl/noxfile.py

@@ -70,7 +70,7 @@ def default(session):
     for local_dep in LOCAL_DEPS:
         session.install("-e", local_dep)
     session.install("-e", ".")
-    session.install("-e", ".[all]")
+    session.install("-e", ".[pandas,storage]")
 
     # Run py.test against the unit tests.
     session.run(
@@ -118,6 +118,7 @@ def system(session):
         session.install("-e", local_dep)
     session.install("-e", "../test_utils/")
     session.install("-e", ".")
+    session.install("-e", ".[pandas,storage]")
 
     # Run py.test against the system tests.
     if system_test_exists:

automl/setup.py

@@ -30,13 +30,6 @@
     "storage": ["google-cloud-storage >= 1.18.0, < 2.0.0dev"],
 }
 
-all_extras = []
-
-for extra in extras:
-    all_extras.extend(extras[extra])
-
-extras["all"] = all_extras
-
 package_root = os.path.abspath(os.path.dirname(__file__))
 
 readme_filename = os.path.join(package_root, "README.rst")

automl/tests/system/gapic/v1beta1/test_system_tables_client_v1.py

@@ -102,7 +102,7 @@ def test_import_pandas_dataframe(self):
         client = automl_v1beta1.TablesClient(project=PROJECT, region=REGION)
         display_name = _id("t_import_pandas")
         dataset = client.create_dataset(display_name)
-        dataframe = pandas.DataFrame({})
+        dataframe = pandas.DataFrame({"test-col1": [1, 2], "test-col2": [3, 4]})
         op = client.import_data(
             project=PROJECT, dataset=dataset, pandas_dataframe=dataframe
         )

automl/tests/unit/gapic/v1beta1/test_gcs_client_v1beta1.py

@@ -16,7 +16,6 @@
 
 """Unit tests."""
 
-import google
 import mock
 import pandas
 import pytest
@@ -35,8 +34,8 @@ def test_ensure_bucket_exists(self):
         mock_bucket = mock.Mock()
         gcs_client = self.gcs_client(
             {
-                "get_bucket.side_effect": google.cloud.exceptions.NotFound("err"),
-                "Bucket.return_value": mock_bucket,
+                "get_bucket.side_effect": exceptions.NotFound("err"),
+                "bucket.return_value": mock_bucket,
             }
         )
         returned_bucket_name = gcs_client.ensure_bucket_exists(
@@ -45,7 +44,7 @@ def test_ensure_bucket_exists(self):
         gcs_client.client.get_bucket.assert_called_with(
             "my-project-automl-tables-staging"
         )
-        gcs_client.client.Bucket.assert_called_with("my-project-automl-tables-staging")
+        gcs_client.client.bucket.assert_called_with("my-project-automl-tables-staging")
         mock_bucket.create.assert_called_with(
             project="my-project", location="us-central1"
         )
@@ -60,7 +59,7 @@ def test_ensure_bucket_exists_bucket_already_exists(self):
         gcs_client.client.get_bucket.assert_called_with(
             "my-project-automl-tables-staging"
         )
-        gcs_client.client.Bucket.assert_not_called()
+        gcs_client.client.bucket.assert_not_called()
         assert returned_bucket_name == "my-project-automl-tables-staging"
 
     def test_upload_pandas_dataframe(self):