feat: add overwrite mode and deepdiff

gooddata-sdk/test-requirements.txt

@@ -8,3 +8,4 @@ urllib3==1.26.9
 python-dotenv~=1.0.0
 attrs>=21.4.0,<=24.2.0
 cattrs>=22.1.0,<=24.1.1
+deepdiff~=8.5.0

gooddata-sdk/tests/catalog/test_catalog_workspace.py

@@ -34,6 +34,7 @@
     CatalogWebhook,
 )
 from gooddata_sdk.utils import recreate_directory
+from tests_support.compare_utils import deep_eq
 from tests_support.vcrpy_utils import get_vcr
 
 from tests.catalog.utils import _refresh_workspaces
@@ -165,8 +166,8 @@ def test_get_declarative_workspaces(test_config):
 
     expected_o = CatalogDeclarativeWorkspaces.from_dict(data)
 
-    assert workspaces_o == expected_o
-    assert workspaces_o.to_api().to_dict(camel_case=True) == data
+    assert deep_eq(expected_o, workspaces_o)
+    assert deep_eq(data, workspaces_o.to_api().to_dict(camel_case=True))
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_declarative_workspaces.yaml"))
@@ -610,10 +611,10 @@ def test_put_declarative_workspace(test_config):
         workspace_o = sdk.catalog_workspace.get_declarative_workspace(
             test_config["workspace_test"], exclude=["ACTIVITY_INFO"]
         )
-        assert workspace_e != workspace_o
+        assert not deep_eq(workspace_e, workspace_o)
         workspace_e.remove_wdf_refs()
-        assert workspace_e == workspace_o
-        assert workspace_e.to_dict() == workspace_o.to_dict()
+        assert deep_eq(workspace_e, workspace_o)
+        assert deep_eq(workspace_e.to_dict(), workspace_o.to_dict())
     finally:
         _refresh_workspaces(sdk)
 
@@ -634,8 +635,8 @@ def test_store_declarative_workspace(test_config):
         workspace_id=test_config["workspace"], layout_root_path=path
     )
 
-    assert workspaces_e == workspaces_o
-    assert workspaces_e.to_dict(camel_case=True) == workspaces_o.to_dict(camel_case=True)
+    assert deep_eq(workspaces_e, workspaces_o)
+    assert deep_eq(workspaces_e.to_dict(camel_case=True), workspaces_o.to_dict(camel_case=True))
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_load_and_put_declarative_workspace.yaml"))
@@ -655,8 +656,8 @@ def test_load_and_put_declarative_workspace(test_config):
         workspace_o = sdk.catalog_workspace.get_declarative_workspace(
             workspace_id=test_config["workspace"], exclude=["ACTIVITY_INFO"]
         )
-        assert workspace_e == workspace_o
-        assert workspace_e.to_dict(camel_case=True) == workspace_o.to_dict(camel_case=True)
+        assert deep_eq(workspace_e, workspace_o)
+        assert deep_eq(workspace_e.to_dict(camel_case=True), workspace_o.to_dict(camel_case=True))
     finally:
         _refresh_workspaces(sdk)
 

gooddata-sdk/tests/catalog/test_catalog_workspace_content.py

@@ -26,6 +26,7 @@
 )
 from gooddata_sdk.compute.model.filter import AbsoluteDateFilter, RelativeDateFilter
 from gooddata_sdk.utils import recreate_directory
+from tests_support.compare_utils import deep_eq
 from tests_support.vcrpy_utils import get_vcr
 
 from tests.catalog.test_catalog_workspace import _refresh_workspaces
@@ -91,8 +92,8 @@ def test_store_declarative_ldm(test_config):
     sdk.catalog_workspace_content.store_declarative_ldm(workspace_id, path)
     ldm_o = sdk.catalog_workspace_content.load_declarative_ldm(workspace_id, path)
 
-    assert ldm_e == ldm_o
-    assert ldm_e.to_api().to_dict() == ldm_o.to_api().to_dict()
+    assert deep_eq(ldm_e, ldm_o)
+    assert deep_eq(ldm_e.to_api().to_dict(), ldm_o.to_api().to_dict())
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_load_and_put_declarative_ldm.yaml"))
@@ -109,10 +110,10 @@ def test_load_and_put_declarative_ldm(test_config):
 
         sdk.catalog_workspace_content.load_and_put_declarative_ldm(identifier, path, standalone_copy=True)
         ldm_o = sdk.catalog_workspace_content.get_declarative_ldm(identifier)
-        assert ldm_e != ldm_o
+        assert not deep_eq(ldm_e, ldm_o)
         ldm_e.remove_wdf_refs()
-        assert ldm_e == ldm_o
-        assert ldm_e.to_api().to_dict() == ldm_o.to_api().to_dict()
+        assert deep_eq(ldm_e, ldm_o)
+        assert deep_eq(ldm_e.to_api().to_dict(), ldm_o.to_api().to_dict())
     finally:
         _refresh_workspaces(sdk)
 
@@ -193,8 +194,8 @@ def test_store_declarative_analytics_model(test_config):
     sdk.catalog_workspace_content.store_declarative_analytics_model(workspace_id, path)
     analytics_model_o = sdk.catalog_workspace_content.load_declarative_analytics_model(workspace_id, path)
 
-    assert analytics_model_e == analytics_model_o
-    assert analytics_model_e.to_api().to_dict() == analytics_model_o.to_api().to_dict()
+    assert deep_eq(analytics_model_e, analytics_model_o)
+    assert deep_eq(analytics_model_e.to_api().to_dict(), analytics_model_o.to_api().to_dict())
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_load_and_put_declarative_analytics_model.yaml"))
@@ -214,8 +215,8 @@ def test_load_and_put_declarative_analytics_model(test_config):
         analytics_model_o = sdk.catalog_workspace_content.get_declarative_analytics_model(
             identifier, exclude=["ACTIVITY_INFO"]
         )
-        assert analytics_model_e == analytics_model_o
-        assert analytics_model_e.to_api().to_dict() == analytics_model_o.to_api().to_dict()
+        assert deep_eq(analytics_model_e, analytics_model_o)
+        assert deep_eq(analytics_model_e.to_api().to_dict(), analytics_model_o.to_api().to_dict())
     finally:
         _refresh_workspaces(sdk)
 
@@ -231,8 +232,8 @@ def test_put_declarative_analytics_model(test_config):
 
         sdk.catalog_workspace_content.put_declarative_analytics_model(identifier, analytics_model_e)
         analytics_model_o = sdk.catalog_workspace_content.get_declarative_analytics_model(identifier)
-        assert analytics_model_e == analytics_model_o
-        assert analytics_model_e.to_api().to_dict() == analytics_model_o.to_api().to_dict()
+        assert deep_eq(analytics_model_e, analytics_model_o)
+        assert deep_eq(analytics_model_e.to_api().to_dict(), analytics_model_o.to_api().to_dict())
     finally:
         _refresh_workspaces(sdk)
 
@@ -248,10 +249,10 @@ def test_put_declarative_ldm(test_config):
     try:
         sdk.catalog_workspace_content.put_declarative_ldm(identifier, ldm_e, standalone_copy=True)
         ldm_o = sdk.catalog_workspace_content.get_declarative_ldm(identifier)
-        assert ldm_e != ldm_o
+        assert not deep_eq(ldm_e, ldm_o)
         ldm_e.remove_wdf_refs()
-        assert ldm_e == ldm_o
-        assert ldm_e.to_api().to_dict() == ldm_o.to_api().to_dict()
+        assert deep_eq(ldm_e, ldm_o)
+        assert deep_eq(ldm_e.to_api().to_dict(), ldm_o.to_api().to_dict())
     finally:
         _refresh_workspaces(sdk)
 
@@ -269,8 +270,8 @@ def test_get_declarative_analytics_model(test_config):
 
     expected_o = CatalogDeclarativeAnalytics.from_dict(data)
 
-    assert analytics_model_o == expected_o
-    assert analytics_model_o.to_api().to_dict(camel_case=True) == data
+    assert deep_eq(expected_o, analytics_model_o)
+    assert deep_eq(data, analytics_model_o.to_api().to_dict(camel_case=True))
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_get_declarative_analytics_model_child.yaml"))
@@ -286,8 +287,8 @@ def test_get_declarative_analytics_model_child(test_config):
 
     expected_o = CatalogDeclarativeAnalytics.from_dict(data)
 
-    assert analytics_model_o == expected_o
-    assert analytics_model_o.to_api().to_dict(camel_case=True) == data
+    assert deep_eq(expected_o, analytics_model_o)
+    assert deep_eq(data, analytics_model_o.to_api().to_dict(camel_case=True))
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_get_declarative_ldm.yaml"))
@@ -301,8 +302,8 @@ def test_get_declarative_ldm(test_config):
 
     expected_o = CatalogDeclarativeModel.from_dict(data)
 
-    assert ldm_o == expected_o
-    assert ldm_o.to_api().to_dict(camel_case=True) == data
+    assert deep_eq(expected_o, ldm_o)
+    assert deep_eq(data, ldm_o.to_api().to_dict(camel_case=True))
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "demo_catalog.yaml"))
@@ -367,7 +368,7 @@ def test_ldm_store_load(test_config):
 
     sdk.catalog_workspace_content.store_ldm_to_disk(test_config["workspace"], path)
     loaded_ldm = sdk.catalog_workspace_content.load_ldm_from_disk(path)
-    assert loaded_ldm == ldm
+    assert deep_eq(ldm, loaded_ldm)
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "analytics_store_load.yaml"))
@@ -378,7 +379,7 @@ def test_analytics_store_load(test_config):
 
     sdk.catalog_workspace_content.store_analytics_model_to_disk(test_config["workspace"], path)
     loaded_analytics_model = sdk.catalog_workspace_content.load_analytics_model_from_disk(path)
-    assert loaded_analytics_model == analytics_model
+    assert deep_eq(analytics_model, loaded_analytics_model)
 
 
 @gd_vcr.use_cassette(str(_fixtures_dir / "label_elements.yaml"))
@@ -468,7 +469,7 @@ def test_explicit_workspace_data_filter(test_config):
 
         updated_ldm = sdk.catalog_workspace_content.get_declarative_ldm(workspace_id=test_config["workspace"])
 
-        assert model_cpy == updated_ldm
+        assert deep_eq(model_cpy, updated_ldm)
 
         dataset = sdk.catalog_workspace_content.get_full_catalog(workspace_id=test_config["workspace"]).get_dataset(
             dataset_id
@@ -496,6 +497,6 @@ def test_export_definition_analytics_layout(test_config):
         analytics_e = sdk.catalog_workspace_content.get_declarative_analytics_model(
             test_config["workspace"], exclude=["ACTIVITY_INFO"]
         )
-        assert analytics_o.analytics.export_definitions == analytics_e.analytics.export_definitions
+        assert deep_eq(analytics_o.analytics.export_definitions, analytics_e.analytics.export_definitions)
     finally:
         _refresh_workspaces(sdk)

gooddata-sdk/tests/overview.md

@@ -64,3 +64,12 @@ def test_something():
 ```
 
 **NOTE**: you do not have to set token value in config file unless you are going to create new recordings.
+
+You can set the envvar `OVERWRITE=1` when you want to rerecord existing cassettes.
+
+## Testing on bigger structures
+
+You are welcome and actually encouraged to use DeepDiff when you are comparing bigger structures.
+It can be imported from `tests_support.compare_utils` in the `deep_eq` function.
+This simplifies debugging in case of errors.
+See other tests how it is used.

tests-support/tests_support/compare_utils.py

@@ -0,0 +1,12 @@
+# (C) 2023 GoodData Corporation
+from __future__ import annotations
+
+from deepdiff import DeepDiff
+
+
+def deep_eq(expected: any, actual: any) -> bool:
+    if expected != actual:
+        print(DeepDiff(expected, actual))
+        return False
+
+    return True

tests-support/tests_support/vcrpy_utils.py

@@ -2,12 +2,14 @@
 from __future__ import annotations
 
 import json
+import os
 import typing
 from json import JSONDecodeError
 from typing import Any, Optional
 
 import vcr
 import yaml
+from vcr.record_mode import RecordMode
 
 VCR_MATCH_ON = ("method", "scheme", "host", "port", "path", "query", "body")
 NON_STATIC_HEADERS = ["DATE", "X-GDC-TRACE-ID"]
@@ -22,6 +24,7 @@ def get_vcr() -> vcr.VCR:
         before_record_request=custom_before_request,
         before_record_response=custom_before_response,
         decode_compressed_response=True,
+        record_mode=RecordMode.ALL if "OVERWRITE" in os.environ else RecordMode.ONCE,
     )
 
     gd_vcr.register_serializer("custom", CustomSerializerYaml())