simplify: inline namespace extraction, remove extra methods and tests

Made-with: Cursor

Author: thomasrockhu-codecov

apps/codecov-api/webhook_handlers/tests/test_gitlab.py

@@ -276,9 +276,7 @@ def test_handle_system_hook_when_not_enterprise(self):
             response = self._post_event_data(**event_data)
             assert response.status_code == status.HTTP_403_FORBIDDEN
 
-    def test_duplicate_repos_disambiguated_by_job_project_name(self):
-        """Job Hook payloads include project_name ('namespace/repo') which
-        can disambiguate when two owners share the same service_id."""
+    def test_duplicate_repos_disambiguated_by_project_name(self):
         owner2 = OwnerFactory(service="gitlab")
         RepositoryFactory(author=owner2, service_id=self.repo.service_id, active=True)
 
@@ -288,15 +286,13 @@ def test_duplicate_repos_disambiguated_by_job_project_name(self):
                 "object_kind": "build",
                 "project_id": self.repo.service_id,
                 "build_status": "pending",
-                "project_name": f"{self.repo.author.username}/my-repo",
+                "project_name": f"{self.repo.author.username} / my-repo",
             },
         )
         assert response.status_code == status.HTTP_200_OK
         assert response.data == WebhookHandlerErrorMessages.SKIP_PENDING_STATUSES
 
-    def test_duplicate_repos_disambiguated_by_push_path(self):
-        """Push events include project.path_with_namespace which can
-        disambiguate duplicate repos."""
+    def test_duplicate_repos_disambiguated_by_path_with_namespace(self):
         owner2 = OwnerFactory(service="gitlab")
         RepositoryFactory(author=owner2, service_id=self.repo.service_id, active=True)
 
@@ -313,38 +309,6 @@ def test_duplicate_repos_disambiguated_by_push_path(self):
         assert response.status_code == status.HTTP_200_OK
         assert response.data == "No yaml cached yet."
 
-    def test_duplicate_repos_disambiguated_by_job_project_name_spaced(self):
-        """Some GitLab versions use 'namespace / repo' format with spaces."""
-        owner2 = OwnerFactory(service="gitlab")
-        RepositoryFactory(author=owner2, service_id=self.repo.service_id, active=True)
-
-        response = self._post_event_data(
-            event=GitLabWebhookEvents.JOB,
-            data={
-                "object_kind": "build",
-                "project_id": self.repo.service_id,
-                "build_status": "pending",
-                "project_name": f"{self.repo.author.username} / my-repo",
-            },
-        )
-        assert response.status_code == status.HTTP_200_OK
-        assert response.data == WebhookHandlerErrorMessages.SKIP_PENDING_STATUSES
-
-    def test_duplicate_repos_without_namespace_falls_through(self):
-        """Without namespace info, duplicate repos still raise an error."""
-        owner2 = OwnerFactory(service="gitlab")
-        RepositoryFactory(author=owner2, service_id=self.repo.service_id, active=True)
-
-        response = self._post_event_data(
-            event=GitLabWebhookEvents.JOB,
-            data={
-                "object_kind": "build",
-                "project_id": self.repo.service_id,
-                "build_status": "pending",
-            },
-        )
-        assert response.status_code == status.HTTP_500_INTERNAL_SERVER_ERROR
-
     def test_secret_validation(self):
         owner = OwnerFactory(service="gitlab")
         repo = RepositoryFactory(

apps/codecov-api/webhook_handlers/views/gitlab.py

@@ -51,60 +51,6 @@ def _inc_err(self, reason: str):
             error_reason=reason,
         ).inc()
 
-    def _get_owner_namespace(self, request) -> str | None:
-        """Extract the owner namespace from the webhook payload.
-
-        Different GitLab event types place this in different fields:
-        - Push/MR/Pipeline events: project.path_with_namespace ("namespace/repo")
-        - System hooks: path_with_namespace ("namespace/repo")
-        - Job hooks: project_name ("namespace/repo" or "namespace / repo")
-
-        See: https://docs.gitlab.com/ee/user/project/integrations/webhook_events.html
-        """
-        path_with_namespace = request.data.get("project", {}).get(
-            "path_with_namespace"
-        ) or request.data.get("path_with_namespace")
-
-        if path_with_namespace:
-            parts = path_with_namespace.split("/")
-            if len(parts) >= 2:
-                return parts[0].strip()
-
-        project_name = request.data.get("project_name")
-        if project_name:
-            parts = project_name.split("/")
-            if len(parts) >= 2:
-                return parts[0].strip()
-
-        return None
-
-    def _get_repo_from_webhook(self, request, project_id) -> Repository:
-        """Look up the Repository for the given webhook, disambiguating by owner
-        namespace when multiple repos share the same service_id."""
-        base_filter = {
-            "author__service": self.service_name,
-            "service_id": project_id,
-        }
-        owner_namespace = self._get_owner_namespace(request)
-
-        if owner_namespace:
-            try:
-                return Repository.objects.get(
-                    **base_filter, author__username=owner_namespace
-                )
-            except Repository.DoesNotExist:
-                pass
-            except Repository.MultipleObjectsReturned:
-                log.warning(
-                    "Multiple repos found even with namespace filter",
-                    extra={
-                        "service_id": project_id,
-                        "namespace": owner_namespace,
-                    },
-                )
-
-        return get_object_or_404(Repository, **base_filter)
-
     def post(self, request, *args, **kwargs):
         """
         Helpful docs for working with GitLab webhooks
@@ -135,11 +81,23 @@ def post(self, request, *args, **kwargs):
                 self._inc_err("permission_denied")
                 raise PermissionDenied()
 
+        # Use the owner namespace from the payload to disambiguate when
+        # multiple owners have repos with the same GitLab project_id.
+        # See: https://docs.gitlab.com/ee/user/project/integrations/webhook_events.html
+        path_with_ns = (
+            request.data.get("project", {}).get("path_with_namespace")
+            or request.data.get("path_with_namespace")
+            or request.data.get("project_name")
+        )
+        repo_filter = {
+            "author__service": self.service_name,
+            "service_id": project_id,
+        }
+        if path_with_ns and "/" in path_with_ns:
+            repo_filter["author__username"] = path_with_ns.split("/")[0].strip()
+
         try:
-            repo = self._get_repo_from_webhook(request, project_id)
-        except Repository.MultipleObjectsReturned as e:
-            self._inc_err("repo_multiple_found")
-            raise e
+            repo = get_object_or_404(Repository, **repo_filter)
         except Exception as e:
             self._inc_err("repo_not_found")
             raise e