Merge pull request #1167 from janmatzek/jmat-SVS-1259-forbid-extra-fields-when-validating-provisioning-input

feat(gooddata-pipelines): forbid extra fields when validating provisioning input

Author: janmatzek

gooddata-pipelines/TODO.md

@@ -10,15 +10,10 @@ A list of outstanding tasks, features, or technical debt to be addressed in this
 
 - [ ] Integrate with GoodDataApiClient
 - [ ] Consider replacing the SdkMethods wrapper with direct calls to the SDK methods
-- [ ] Consider using orjson library instead of json to load test data
 - [ ] Cleanup custom exceptions
 - [ ] Improve test coverage. Write missing unit tests for legacy code (e.g., user data filters)
 
 ## Documentation
 
 - [ ] Improve package README
-- [ ] Workspace provisioning
-- [ ] User provisioning
-- [ ] User group provisioning
-- [ ] Permission provisioning
 - [ ] User data filter provisioning

gooddata-pipelines/gooddata_pipelines/backup_and_restore/backup_input_processor.py

@@ -1,7 +1,6 @@
 # (C) 2025 GoodData Corporation
 
-from dataclasses import dataclass
-
+import attrs
 import requests
 
 from gooddata_pipelines.api import GoodDataApi
@@ -46,7 +45,7 @@ def set_endpoints(self) -> None:
         )
         self.all_workspaces_endpoint = f"{self.base_workspace_endpoint}?page=0&size={self.page_size}&sort=name,asc&metaInclude=page"
 
-    @dataclass
+    @attrs.define
     class _ProcessDataOutput:
         workspace_ids: list[str]
         sub_parents: list[str] | None = None

gooddata-pipelines/gooddata_pipelines/backup_and_restore/backup_manager.py

@@ -6,10 +6,10 @@
 import tempfile
 import time
 import traceback
-from dataclasses import dataclass
 from pathlib import Path
 from typing import Any, Type
 
+import attrs
 import requests
 import yaml
 from gooddata_sdk.utils import PROFILES_FILE_PATH, profile_content
@@ -40,7 +40,7 @@
 from gooddata_pipelines.utils.rate_limiter import RateLimiter
 
 
-@dataclass
+@attrs.define
 class BackupBatch:
     list_of_ids: list[str]
 

gooddata-pipelines/gooddata_pipelines/backup_and_restore/constants.py

@@ -1,11 +1,11 @@
 # (C) 2025 GoodData Corporation
 import datetime
-from dataclasses import dataclass
 
+import attrs
 from gooddata_sdk._version import __version__ as sdk_version
 
 
-@dataclass(frozen=True)
+@attrs.frozen
 class DirNames:
     """
     Folder names used in the SDK backup process:
@@ -21,14 +21,14 @@ class DirNames:
     UDF = "user_data_filters"
 
 
-@dataclass(frozen=True)
+@attrs.frozen
 class ApiDefaults:
     DEFAULT_PAGE_SIZE = 100
     DEFAULT_BATCH_SIZE = 100
     DEFAULT_API_CALLS_PER_SECOND = 1.0
 
 
-@dataclass(frozen=True)
+@attrs.frozen
 class BackupSettings(ApiDefaults):
     MAX_RETRIES = 3
     RETRY_DELAY = 5  # seconds

gooddata-pipelines/gooddata_pipelines/provisioning/entities/user_data_filters/models/udf_models.py

@@ -2,31 +2,29 @@
 
 """This module defines data models for user data filters in a GoodData workspace."""
 
-# TODO: consider using attrs instead of dataclasses for these models. Dataclasses
-# have different functionality per Python version (not package version).
+import attrs
+from pydantic import BaseModel, ConfigDict
 
-from dataclasses import dataclass, field
 
-
-@dataclass
+@attrs.define
 class UserDataFilterGroup:
     udf_id: str
     udf_values: list[str]
 
 
-@dataclass
+@attrs.define
 class WorkspaceUserDataFilters:
     workspace_id: str
-    user_data_filters: list["UserDataFilterGroup"] = field(default_factory=list)
+    user_data_filters: list["UserDataFilterGroup"] = attrs.field(factory=list)
+
 
+class UserDataFilterFullLoad(BaseModel):
+    model_config = ConfigDict(extra="forbid")
 
-@dataclass
-class UserDataFilterFullLoad:
     workspace_id: str
     udf_id: str
     udf_value: str
 
 
-@dataclass
 class UserDataFilterIncrementalLoad(UserDataFilterFullLoad):
     is_active: bool

gooddata-pipelines/gooddata_pipelines/provisioning/entities/user_data_filters/user_data_filters.py

@@ -50,6 +50,8 @@ class UserDataFilterProvisioner(
     ldm_column_name: str = ""
     maql_column_name: str = ""
 
+    FULL_LOAD_TYPE = UserDataFilterFullLoad
+
     def set_ldm_column_name(self, ldm_column_name: str) -> None:
         """Set the LDM column name for user data filters.
 
@@ -214,8 +216,6 @@ def _provision_full_load(self) -> None:
         )
         self._create_user_data_filters(grouped_db_user_data_filters)
 
-        self.logger.info("User data filters provisioning completed")
-
     def _provision_incremental_load(self) -> None:
         """Provision user data filters in GoodData workspaces."""
         raise NotImplementedError("Not implemented yet.")

gooddata-pipelines/gooddata_pipelines/provisioning/entities/users/models/permissions.py

@@ -9,7 +9,7 @@
     CatalogDeclarativeSingleWorkspacePermission,
     CatalogDeclarativeWorkspacePermissions,
 )
-from pydantic import BaseModel
+from pydantic import BaseModel, ConfigDict
 
 from gooddata_pipelines.provisioning.utils.exceptions import BaseUserException
 
@@ -23,6 +23,8 @@ class EntityType(str, Enum):
 
 
 class BasePermission(BaseModel):
+    model_config = ConfigDict(extra="forbid")
+
     permission: str
     workspace_id: str
     entity_id: str

gooddata-pipelines/gooddata_pipelines/provisioning/entities/users/models/user_groups.py

@@ -1,9 +1,17 @@
 # (C) 2025 GoodData Corporation
 
-from pydantic import BaseModel, Field, ValidationInfo, field_validator
+from pydantic import (
+    BaseModel,
+    ConfigDict,
+    Field,
+    ValidationInfo,
+    field_validator,
+)
 
 
 class UserGroupBase(BaseModel):
+    model_config = ConfigDict(extra="forbid")
+
     user_group_id: str
     user_group_name: str
     parent_user_groups: list[str] = Field(default_factory=list)

gooddata-pipelines/gooddata_pipelines/provisioning/entities/users/models/users.py

@@ -3,7 +3,7 @@
 from typing import Any
 
 from gooddata_sdk.catalog.user.entity_model.user import CatalogUser
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, ConfigDict, Field
 
 
 class UserProfile(BaseModel):
@@ -18,6 +18,8 @@ class UserProfile(BaseModel):
 class BaseUser(BaseModel):
     """Base class containing shared user fields and functionality."""
 
+    model_config = ConfigDict(extra="forbid")
+
     user_id: str
     firstname: str | None
     lastname: str | None

gooddata-pipelines/gooddata_pipelines/provisioning/entities/workspaces/models.py

@@ -22,7 +22,7 @@ class WorkspaceDataMaps:
 
 
 class WorkspaceBase(BaseModel):
-    model_config = ConfigDict(coerce_numbers_to_str=True)
+    model_config = ConfigDict(coerce_numbers_to_str=True, extra="forbid")
 
     parent_id: str
     workspace_id: str