chore(gooddata-pipelines): replace dataclasses with attrs

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

@@ -2,24 +2,20 @@
 
 """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).
-
-from dataclasses import dataclass, field
-
+import attrs
 from pydantic import BaseModel, ConfigDict
 
 
-@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):

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/utils/utils.py

@@ -2,6 +2,8 @@
 
 """Module for utilities used in GoodData Pipelines provisioning."""
 
+from typing import Any, cast
+
 import attrs
 from requests import Response
 
@@ -11,9 +13,8 @@ class AttributesMixin:
     Mixin class to provide a method for getting attributes of an object which may or may not exist.
     """
 
-    @staticmethod
     def get_attrs(
-        *objects: object, overrides: dict[str, str] | None = None
+        self, *objects: object, overrides: dict[str, str] | None = None
     ) -> dict[str, str]:
         """
         Returns a dictionary of attributes from the given objects.
@@ -27,11 +28,11 @@ def get_attrs(
         """
         # TODO: This might not work great with nested objects, values  which are lists of objects etc.
         # If we care about parsing the logs back from the string, we should consider some other approach
-        attrs: dict[str, str] = {}
+        attributes: dict[str, str] = {}
         for context_object in objects:
             if isinstance(context_object, Response):
                 # for request.Response objects, keys need to be renamed to match the log schema
-                attrs.update(
+                attributes.update(
                     {
                         "http_status": str(context_object.status_code),
                         "http_method": getattr(
@@ -42,23 +43,31 @@ def get_attrs(
                         ),
                     }
                 )
+            elif attrs.has(type(context_object)):
+                for key, value in attrs.asdict(
+                    cast(attrs.AttrsInstance, context_object)
+                ).items():
+                    self._add_to_dict(attributes, key, value)
             else:
                 # Generic handling for other objects
                 for key, value in context_object.__dict__.items():
-                    if value is None:
-                        continue
-
-                    if isinstance(value, list):
-                        attrs[key] = ", ".join(
-                            str(list_item) for list_item in value
-                        )
-                    else:
-                        attrs[key] = str(value)
+                    self._add_to_dict(attributes, key, value)
 
         if overrides:
-            attrs.update(overrides)
+            attributes.update(overrides)
+
+        return attributes
+
+    def _add_to_dict(
+        self, attributes: dict[str, str], key: str, value: Any
+    ) -> None:
+        if value is None:
+            return
 
-        return attrs
+        if isinstance(value, list):
+            attributes[key] = ", ".join(str(list_item) for list_item in value)
+        else:
+            attributes[key] = str(value)
 
 
 @attrs.define

gooddata-pipelines/tests/provisioning/entities/users/test_users.py

@@ -1,7 +1,7 @@
 # (C) 2025 GoodData Corporation
-from dataclasses import dataclass
 from typing import Literal, Optional
 
+import attrs
 import orjson
 import pytest
 from gooddata_api_client.exceptions import NotFoundException  # type: ignore
@@ -29,7 +29,7 @@
 TEST_DATA_SUBDIR = f"{TEST_DATA_DIR}/provisioning/entities/users"
 
 
-@dataclass
+@attrs.define
 class MockUser:
     id: str
     firstname: Optional[str]