Add default LoadJobConfig to Client

chelsea-lin · chelsea-lin · commit 1cb30f1f4ca6 · 2023-03-14T22:57:43.000Z
diff --git a/google/cloud/bigquery/client.py b/google/cloud/bigquery/client.py
@@ -210,6 +210,9 @@ class Client(ClientWithProject):
         default_query_job_config (Optional[google.cloud.bigquery.job.QueryJobConfig]):
             Default ``QueryJobConfig``.
             Will be merged into job configs passed into the ``query`` method.
+        default_load_job_config (Optional[google.cloud.bigquery.job.LoadJobConfig]):
+            Default ``LoadJobConfig``.
+            Will be merged into job configs passed into the ``load_table_*`` methods.
         client_info (Optional[google.api_core.client_info.ClientInfo]):
             The client info used to send a user-agent string along with API
             requests. If ``None``, then default info will be used. Generally,
@@ -235,6 +238,7 @@ def __init__(
         _http=None,
         location=None,
         default_query_job_config=None,
+        default_load_job_config=None,
         client_info=None,
         client_options=None,
     ) -> None:
@@ -260,6 +264,7 @@ def __init__(
         self._connection = Connection(self, **kw_args)
         self._location = location
         self._default_query_job_config = copy.deepcopy(default_query_job_config)
+        self._default_load_job_config = copy.deepcopy(default_load_job_config)
 
     @property
     def location(self):
@@ -277,6 +282,17 @@ def default_query_job_config(self):
     def default_query_job_config(self, value: QueryJobConfig):
         self._default_query_job_config = copy.deepcopy(value)
 
+    @property
+    def default_load_job_config(self):
+        """Default ``LoadJobConfig``.
+        Will be merged into job configs passed into the ``load_table_*`` methods.
+        """
+        return self._default_load_job_config
+
+    @default_load_job_config.setter
+    def default_load_job_config(self, value: LoadJobConfig):
+        self._default_load_job_config = copy.deepcopy(value)
+
     def close(self):
         """Close the underlying transport objects, releasing system resources.
 
@@ -2348,9 +2364,19 @@ def load_table_from_uri(
 
         destination = _table_arg_to_table_ref(destination, default_project=self.project)
 
+        # Make a copy so that the job config isn't modified in-place.
         if job_config:
-            job_config = copy.deepcopy(job_config)
             _verify_job_config_type(job_config, google.cloud.bigquery.job.LoadJobConfig)
+            job_config = copy.deepcopy(job_config)
+        else:
+            job_config = job.LoadJobConfig()
+
+        # Merge this job config with a default job config
+        if self._default_load_job_config:
+            _verify_job_config_type(
+                self._default_load_job_config, google.cloud.bigquery.job.LoadJobConfig
+            )
+            job_config = job_config._fill_from_default(self._default_load_job_config)
 
         load_job = job.LoadJob(job_ref, source_uris, destination, self, job_config)
         load_job._begin(retry=retry, timeout=timeout)
@@ -2437,9 +2463,21 @@ def load_table_from_file(
 
         destination = _table_arg_to_table_ref(destination, default_project=self.project)
         job_ref = job._JobReference(job_id, project=project, location=location)
+
+        # Make a copy so that the job config isn't modified in-place.
         if job_config:
-            job_config = copy.deepcopy(job_config)
             _verify_job_config_type(job_config, google.cloud.bigquery.job.LoadJobConfig)
+            job_config = copy.deepcopy(job_config)
+        else:
+            job_config = job.LoadJobConfig()
+
+        # Merge this job config with a default job config
+        if self._default_load_job_config:
+            _verify_job_config_type(
+                self._default_load_job_config, google.cloud.bigquery.job.LoadJobConfig
+            )
+            job_config = job_config._fill_from_default(self._default_load_job_config)
+
         load_job = job.LoadJob(job_ref, None, destination, self, job_config)
         job_resource = load_job.to_api_repr()
 
@@ -2569,13 +2607,10 @@ def load_table_from_dataframe(
         """
         job_id = _make_job_id(job_id, job_id_prefix)
 
+        # Make a copy so that the job config isn't modified in-place.
         if job_config:
             _verify_job_config_type(job_config, google.cloud.bigquery.job.LoadJobConfig)
-            # Make a copy so that the job config isn't modified in-place.
-            job_config_properties = copy.deepcopy(job_config._properties)
-            job_config = job.LoadJobConfig()
-            job_config._properties = job_config_properties
-
+            job_config = copy.deepcopy(job_config)
         else:
             job_config = job.LoadJobConfig()
 
diff --git a/tests/unit/job/test_base.py b/tests/unit/job/test_base.py
@@ -1104,7 +1104,7 @@ def test_ctor_with_unknown_property_raises_error(self):
             config = self._make_one()
             config.wrong_name = None
 
-    def test_fill_from_default(self):
+    def test_fill_query_job_config_from_default(self):
         from google.cloud.bigquery import QueryJobConfig
 
         job_config = QueryJobConfig()
@@ -1120,6 +1120,22 @@ def test_fill_from_default(self):
         self.assertTrue(final_job_config.use_query_cache)
         self.assertEqual(final_job_config.maximum_bytes_billed, 1000)
 
+    def test_fill_load_job_from_default(self):
+        from google.cloud.bigquery import LoadJobConfig
+
+        job_config = LoadJobConfig()
+        job_config.create_session = True
+        job_config.encoding = "UTF-8"
+
+        default_job_config = LoadJobConfig()
+        default_job_config.ignore_unknown_values = True
+        default_job_config.encoding = "ISO-8859-1"
+
+        final_job_config = job_config._fill_from_default(default_job_config)
+        self.assertTrue(final_job_config.create_session)
+        self.assertTrue(final_job_config.ignore_unknown_values)
+        self.assertEqual(final_job_config.encoding, "UTF-8")
+
     def test_fill_from_default_conflict(self):
         from google.cloud.bigquery import QueryJobConfig
 
diff --git a/tests/unit/test_client.py b/tests/unit/test_client.py
