fix(snippetgen): fix client streaming samples

.github/workflows/tests.yaml

@@ -359,4 +359,4 @@ jobs:
             python -m pip install autopep8
       - name: Check diff
         run: |
-            find gapic tests -name "*.py" -not -path 'tests/integration/goldens/*' | xargs autopep8 --diff --exit-code
+            find gapic tests -name "*.py" -not -path 'tests/**/goldens/*' | xargs autopep8 --diff --exit-code

DEVELOPMENT.md

@@ -26,13 +26,13 @@ Execute unit tests by running one of the sessions prefixed with `unit-`.
 -   Lint sources by running `autopep8`. The specific command is the following.
 
     ```
-    find gapic tests -name "*.py" -not -path 'tests/integration/goldens/*' | xargs autopep8 --diff --exit-code
+    find gapic tests -name "*.py" -not -path 'tests/**/goldens/*' | xargs autopep8 --diff --exit-code
     ```
 
 -  Format sources in place:
 
     ```
-    find gapic tests -name "*.py" -not -path 'tests/integration/goldens/*' | xargs autopep8 --in-place
+    find gapic tests -name "*.py" -not -path 'tests/**/goldens/*' | xargs autopep8 --in-place
     ```
 
 ## Integration Tests

gapic/samplegen/samplegen.py

@@ -466,9 +466,7 @@ def validate_and_transform_request(
 
         Raises:
             InvalidRequestSetup: If a dict in the request lacks a "field" key,
-                                 a "value" key, if there is an unexpected keyword,
-                                 or if more than one base parameter is given for
-                                 a client-side streaming calling form.
+                                 a "value" key or if there is an unexpected keyword.
             BadAttributeLookup: If a request field refers to a non-existent field
                                 in the request message type.
             ResourceRequestMismatch: If a request attempts to describe both
@@ -548,16 +546,6 @@ def validate_and_transform_request(
                     AttributeRequestSetup(**r_dup)  # type: ignore
                 )
 
-        client_streaming_forms = {
-            types.CallingForm.RequestStreamingClient,
-            types.CallingForm.RequestStreamingBidi,
-        }
-
-        if len(base_param_to_attrs) > 1 and calling_form in client_streaming_forms:
-            raise types.InvalidRequestSetup(
-                "Too many base parameters for client side streaming form"
-            )
-
         # We can only flatten a collection of request parameters if they're a
         # subset of the flattened fields of the method.
         flattenable = self.flattenable_fields >= set(base_param_to_attrs)

gapic/templates/examples/feature_fragments.j2

@@ -153,7 +153,7 @@ with open({{ attr.input_parameter }}, "rb") as f:
 client = {{ module_name }}.{{ client_name }}()
 {% endmacro %}
 
-{% macro render_request_setup(full_request, module_name, request_type) %}
+{% macro render_request_setup(full_request, module_name, request_type, calling_form, calling_form_enum) %}
 # Initialize request argument(s)
 {% for parameter_block in full_request.request_list if parameter_block.body %}
 {% if parameter_block.pattern %}
@@ -179,6 +179,21 @@ request = {{ module_name }}.{{ request_type.ident.name }}(
     {{ parameter.base }}={{ parameter.base if parameter.body else parameter.single.value }},
 {% endfor %}
 )
+{# Note: This template assumes only one request needs to be sent. When samples accept
+configs the client streaming logic should be modified to allow 2+ request objects. #}
+{# If client streaming, wrap the single request in a generator that produces 'requests' #}
+{% if calling_form in [calling_form_enum.RequestStreamingBidi,
+                       calling_form_enum.RequestStreamingClient] %}
+
+# This method expects an iterator which contains
+# '{{module_name}}.{{ request_type.ident.name }}' objects
+# Here we create a generator that yields a single `request` for 
+# demonstrative purposes.
+requests = [request]
+def request_generator():
+    for request in requests:
+        yield request
+{% endif %}
 {% endif %}
 {% endmacro %}
 
@@ -219,7 +234,7 @@ await{{ " "}}
 {%- endif -%}
 {% if calling_form in [calling_form_enum.RequestStreamingBidi,
                        calling_form_enum.RequestStreamingClient] %}
-client.{{ sample.rpc|snake_case }}([{{ render_request_params(sample.request.request_list)|trim }}])
+client.{{ sample.rpc|snake_case }}(requests=request_generator())
 {% else %}{# TODO: deal with flattening #}
 {# TODO: set up client streaming once some questions are answered #}
 client.{{ sample.rpc|snake_case }}({{ render_request_params_unary(sample.request)|trim }})

gapic/templates/examples/sample.py.j2

@@ -35,7 +35,7 @@ from {{ sample.module_namespace|join(".") }} import {{ sample.module_name }}
     """{{ sample.description }}"""
 
     {{ frags.render_client_setup(sample.module_name, sample.client_name)|indent }}
-    {{ frags.render_request_setup(sample.request, sample.module_name, sample.request_type)|indent }}
+    {{ frags.render_request_setup(sample.request, sample.module_name, sample.request_type, calling_form, calling_form_enum)|indent }}
     {% with method_call = frags.render_method_call(sample, calling_form, calling_form_enum, sample.transport) %}
     {{ frags.render_calling_form(method_call, calling_form, calling_form_enum, sample.transport, sample.response)|indent -}}
     {% endwith %}

tests/integration/goldens/logging/samples/generated_samples/logging_generated_logging_v2_logging_service_v2_tail_log_entries_async.py

@@ -38,8 +38,17 @@ async def sample_tail_log_entries():
         resource_names=['resource_names_value_1', 'resource_names_value_2'],
     )
 
+    # This method expects an iterator which contains
+    # 'logging_v2.TailLogEntriesRequest' objects
+    # Here we create a generator that yields a single `request` for
+    # demonstrative purposes.
+    requests = [request]
+    def request_generator():
+        for request in requests:
+            yield request
+
     # Make the request
-    stream = await client.tail_log_entries([resource_names=['resource_names_value_1', 'resource_names_value_2']])
+    stream = await client.tail_log_entries(requests=request_generator())
     async for response in stream:
         print(response)
 

tests/integration/goldens/logging/samples/generated_samples/logging_generated_logging_v2_logging_service_v2_tail_log_entries_sync.py

@@ -38,8 +38,17 @@ def sample_tail_log_entries():
         resource_names=['resource_names_value_1', 'resource_names_value_2'],
     )
 
+    # This method expects an iterator which contains
+    # 'logging_v2.TailLogEntriesRequest' objects
+    # Here we create a generator that yields a single `request` for
+    # demonstrative purposes.
+    requests = [request]
+    def request_generator():
+        for request in requests:
+            yield request
+
     # Make the request
-    stream = client.tail_log_entries([resource_names=['resource_names_value_1', 'resource_names_value_2']])
+    stream = client.tail_log_entries(requests=request_generator())
     for response in stream:
         print(response)
 

tests/snippetgen/goldens/mollusca_generated_mollusca_v1_snippets_method_bidi_streaming_async.py

@@ -38,8 +38,17 @@ async def sample_method_bidi_streaming():
         my_string="my_string_value",
     )
 
+    # This method expects an iterator which contains
+    # 'mollusca_v1.SignatureRequestOneRequiredField' objects
+    # Here we create a generator that yields a single `request` for
+    # demonstrative purposes.
+    requests = [request]
+    def request_generator():
+        for request in requests:
+            yield request
+
     # Make the request
-    stream = await client.method_bidi_streaming([my_string="my_string_value"])
+    stream = await client.method_bidi_streaming(requests=request_generator())
     async for response in stream:
         print(response)
 

tests/snippetgen/goldens/mollusca_generated_mollusca_v1_snippets_method_bidi_streaming_sync.py

@@ -38,8 +38,17 @@ def sample_method_bidi_streaming():
         my_string="my_string_value",
     )
 
+    # This method expects an iterator which contains
+    # 'mollusca_v1.SignatureRequestOneRequiredField' objects
+    # Here we create a generator that yields a single `request` for
+    # demonstrative purposes.
+    requests = [request]
+    def request_generator():
+        for request in requests:
+            yield request
+
     # Make the request
-    stream = client.method_bidi_streaming([my_string="my_string_value"])
+    stream = client.method_bidi_streaming(requests=request_generator())
     for response in stream:
         print(response)
 

tests/unit/samplegen/test_samplegen.py

@@ -1350,57 +1350,6 @@ def test_validate_request_reserved_input_param(dummy_api_schema):
         )
 
 
-def test_single_request_client_streaming(dummy_api_schema,
-        calling_form=types.CallingForm.RequestStreamingClient):
-    # Each API client method really only takes one parameter:
-    # either a single protobuf message or an iterable of protobuf messages.
-    # With unary request methods, python lets us describe attributes as positional
-    # and keyword parameters, which simplifies request construction.
-    # The 'base' in the transformed request refers to an attribute, and the
-    # 'field's refer to sub-attributes.
-    # Client streaming and bidirectional streaming methods can't use this notation,
-    # and generate an exception if there is more than one 'base'.
-    input_type = DummyMessage(
-        fields={
-            "cephalopod": DummyField(
-                message=DummyMessage(
-                    fields={
-                        "order": DummyField(
-                            message=DummyMessage(type="ORDER_TYPE")
-                        )
-                    },
-                    type="CEPHALOPOD_TYPE"
-                )
-            ),
-            "gastropod": DummyField(
-                message=DummyMessage(
-                    fields={
-                        "order": DummyField(
-                            message=DummyMessage(type="ORDER_TYPE")
-                        )
-                    },
-                    type="GASTROPOD_TYPE"
-                )
-            )
-        },
-        type="MOLLUSC_TYPE"
-    )
-    v = samplegen.Validator(DummyMethod(input=input_type), dummy_api_schema)
-    with pytest.raises(types.InvalidRequestSetup):
-        v.validate_and_transform_request(
-            types.CallingForm.RequestStreamingClient,
-            [
-                {"field": "cephalopod.order", "value": "cephalopoda"},
-                {"field": "gastropod.order", "value": "pulmonata"},
-            ],
-        )
-
-
-def test_single_request_bidi_streaming():
-    test_single_request_client_streaming(
-        types.CallingForm.RequestStreamingBidi)
-
-
 def test_validate_request_calling_form():
     assert (
         types.CallingForm.method_default(DummyMethod(lro=True))

tests/unit/samplegen/test_template.py

@@ -200,7 +200,7 @@ def test_render_request_unflattened():
     check_template(
         '''
         {% import "feature_fragments.j2" as frags %}
-        {{ frags.render_request_setup(request, module_name, request_type) }}
+        {{ frags.render_request_setup(request, module_name, request_type, calling_form, calling_form_enum) }}
         ''',
         '''
         # Initialize request argument(s)
@@ -289,7 +289,9 @@ def test_render_request_unflattened():
                 )
             },
             ident=common_types.DummyIdent(name="CreateMolluscRequest")
-        )
+        ),
+        calling_form_enum=CallingForm,
+        calling_form=CallingForm.Request,
     )
 
 
@@ -1015,7 +1017,7 @@ def test_render_method_call_bidi():
                                   calling_form, calling_form_enum, transport) }}
         ''',
         '''
-        client.categorize_mollusc([video])
+        client.categorize_mollusc(requests=request_generator())
         ''',
         request=samplegen.FullRequest(
             request_list=[
@@ -1040,7 +1042,7 @@ def test_render_method_call_bidi_async():
                                   calling_form, calling_form_enum, transport) }}
         ''',
         '''
-        await client.categorize_mollusc([video])
+        await client.categorize_mollusc(requests=request_generator())
         ''',
         request=samplegen.FullRequest(
             request_list=[
@@ -1065,7 +1067,7 @@ def test_render_method_call_client():
         calling_form, calling_form_enum, transport) }}
         ''',
         '''
-        client.categorize_mollusc([video])
+        client.categorize_mollusc(requests=request_generator())
         ''',
         request=samplegen.FullRequest(
             request_list=[
@@ -1090,7 +1092,7 @@ def test_render_method_call_client_async():
         calling_form, calling_form_enum, transport) }}
         ''',
         '''
-        await client.categorize_mollusc([video])
+        await client.categorize_mollusc(requests=request_generator())
         ''',
         request=samplegen.FullRequest(
             request_list=[