Fixing groups in config file not considered for group migration job (#943)

Author: william-conti

src/databricks/labs/ucx/workspace_access/groups.py

@@ -394,8 +394,24 @@ def delete_original_workspace_groups(self):
             raise ManyError(errors)
 
     def _fetcher(self) -> Iterable[MigratedGroup]:
+        state = []
         for row in self._backend.fetch(f"SELECT * FROM {self._full_name}"):
-            yield MigratedGroup(*row)
+            state.append(MigratedGroup(*row))
+
+        if not self._include_group_names:
+            return state
+
+        new_state = []
+        group_name_with_state = {migrated_group.name_in_workspace: migrated_group for migrated_group in state}
+        for group_name in self._include_group_names:
+            if group_name in group_name_with_state:
+                new_state.append(group_name_with_state[group_name])
+            else:
+                logger.warning(
+                    f"Group {group_name} defined in configuration does not exist on the groups table. "
+                    "Consider checking if the group exist in the workspace or re-running the assessment."
+                )
+        return new_state
 
     def _crawler(self) -> Iterable[MigratedGroup]:
         workspace_groups_in_workspace = self._workspace_groups_in_workspace()

tests/unit/workspace_access/test_groups.py

@@ -1031,3 +1031,32 @@ def do_api_side_effect(*args, **kwargs):
         backend, wsclient, inventory_database="inv", workspace_group_regex=r"\(([1-9]+)\)", account_group_regex="[1-9]+"
     ).validate_group_membership()
     assert grp_membership == []
+
+
+def test_migration_state_with_filtered_group():
+    backend = MockBackend(
+        rows={
+            "SELECT": [
+                ("", "de", "de", "test-group-de", "", "", "", ""),
+                ("", "ds", "ds", "test-group-ds", "", "", "", ""),
+            ]
+        }
+    )
+    wsclient = MagicMock()
+    grp_membership = GroupManager(
+        backend, wsclient, inventory_database="inv", include_group_names=["ds", "irrelevant_group"]
+    ).get_migration_state()
+
+    assert len(grp_membership.groups) == 1
+    assert grp_membership.groups == [
+        MigratedGroup(
+            id_in_workspace='',
+            name_in_workspace='ds',
+            name_in_account='ds',
+            temporary_name='test-group-ds',
+            members='',
+            entitlements='',
+            external_id='',
+            roles='',
+        )
+    ]