Fix mypy type errors in enterprise/integrations/github/ (#13191)

Co-authored-by: openhands <openhands@all-hands.dev>

enterprise/integrations/github/data_collector.py

@@ -116,10 +116,8 @@ class GitHubDataCollector:
 
         return suffix
 
-    def _get_installation_access_token(self, installation_id: str) -> str:
-        token_data = self.github_integration.get_access_token(
-            installation_id  # type: ignore[arg-type]
-        )
+    def _get_installation_access_token(self, installation_id: int) -> str:
+        token_data = self.github_integration.get_access_token(installation_id)
         return token_data.token
 
     def _check_openhands_author(self, name, login) -> bool:
@@ -134,7 +132,7 @@ class GitHubDataCollector:
         )
 
     def _get_issue_comments(
-        self, installation_id: str, repo_name: str, issue_number: int, conversation_id
+        self, installation_id: int, repo_name: str, issue_number: int, conversation_id
     ) -> list[dict[str, Any]]:
         """
         Retrieve all comments from an issue until a comment with conversation_id is found
@@ -234,7 +232,7 @@ class GitHubDataCollector:
             f'[Github]: Saved issue #{issue_number} for {github_view.full_repo_name}'
         )
 
-    def _get_pr_commits(self, installation_id: str, repo_name: str, pr_number: int):
+    def _get_pr_commits(self, installation_id: int, repo_name: str, pr_number: int):
         commits = []
         installation_token = self._get_installation_access_token(installation_id)
         with Github(auth=Auth.Token(installation_token)) as github_client:
@@ -431,7 +429,7 @@ class GitHubDataCollector:
         - Num openhands review comments
         """
         pr_number = openhands_pr.pr_number
-        installation_id = openhands_pr.installation_id
+        installation_id = int(openhands_pr.installation_id)
         repo_id = openhands_pr.repo_id
 
         # Get installation token and create Github client

enterprise/integrations/github/github_service.py

@@ -122,13 +122,37 @@ class SaaSGitHubService(GitHubService):
             raise Exception(f'No node_id found for repository {repo_id}')
         return node_id
 
+    async def _get_external_auth_id(self) -> str | None:
+        """Get or fetch external_auth_id from Keycloak token if not already set."""
+        if self.external_auth_id:
+            return self.external_auth_id
+
+        if self.external_auth_token:
+            try:
+                user_info = await self.token_manager.get_user_info(
+                    self.external_auth_token.get_secret_value()
+                )
+                self.external_auth_id = user_info.sub
+                logger.info(
+                    f'Determined external_auth_id from Keycloak token: {self.external_auth_id}'
+                )
+                return self.external_auth_id
+            except Exception as e:
+                logger.warning(
+                    f'Could not determine external_auth_id from token: {e}',
+                    exc_info=True,
+                )
+        return None
+
     async def get_paginated_repos(self, page, per_page, sort, installation_id):
         repositories = await super().get_paginated_repos(
             page, per_page, sort, installation_id
         )
-        asyncio.create_task(
-            store_repositories_in_db(repositories, self.external_auth_id)
-        )
+        external_auth_id = await self._get_external_auth_id()
+        if external_auth_id:
+            asyncio.create_task(
+                store_repositories_in_db(repositories, external_auth_id)
+            )
         return repositories
 
     async def get_all_repositories(
@@ -136,8 +160,10 @@ class SaaSGitHubService(GitHubService):
     ) -> list[Repository]:
         repositories = await super().get_all_repositories(sort, app_mode)
         # Schedule the background task without awaiting it
-        asyncio.create_task(
-            store_repositories_in_db(repositories, self.external_auth_id)
-        )
+        external_auth_id = await self._get_external_auth_id()
+        if external_auth_id:
+            asyncio.create_task(
+                store_repositories_in_db(repositories, external_auth_id)
+            )
         # Return repositories immediately
         return repositories

enterprise/integrations/github/github_view.py

@@ -733,7 +733,7 @@ class GithubFactory:
     @staticmethod
     async def create_github_view_from_payload(
         message: Message, keycloak_user_id: str
-    ) -> ResolverViewInterface:
+    ) -> GithubViewType:
         """Create the appropriate class (GithubIssue or GithubPRComment) based on the payload.
         Also return metadata about the event (e.g., action type).
         """