safeguard wherever reading s3 paths[0] (#451)

Author: longshuicy

ai_ta_backend/rabbitmq/ingest.py

@@ -305,22 +305,27 @@ def _ingest_single(ingest_method: Callable, s3_path: str, force_embeddings: bool
                             self.posthog.capture('distinct_id_of_the_user', event='ingest_failure',
                                 properties={
                                     'course_name': course_name,
-                                    's3_path': s3_paths,
+                                    's3_path': s3_path,
                                     'kwargs': kwargs,
                                     'error': err_msg
                                 })
             return success_status
         except Exception as e:
             err = f"❌❌ Error in /ingest: `{inspect.currentframe().f_code.co_name}`: {e}\nTraceback:\n", traceback.format_exc()  # type: ignore
+            # Use s3_path if available (from loop), otherwise use first item from s3_paths list or s3_paths itself if it's a string
+            try:
+                error_s3_path = s3_path if 's3_path' in locals() else (s3_paths[0] if isinstance(s3_paths, list) and len(s3_paths) > 0 else s3_paths)
+            except (NameError, TypeError, IndexError):
+                error_s3_path = str(s3_paths) if s3_paths else "unknown"
             success_status['failure_ingest'] = {
-                's3_path': s3_path,
+                's3_path': error_s3_path,
                 'error': f"MAJOR ERROR DURING INGEST: {err}"
             }
             if self.posthog:
                 self.posthog.capture('distinct_id_of_the_user', event='ingest_failure',
                                     properties={
                                         'course_name': course_name,
-                                        's3_path': s3_paths,
+                                        's3_path': error_s3_path,
                                         'kwargs': kwargs,
                                         'error': err
                                     })

ai_ta_backend/rabbitmq/ingest_canvas.py

@@ -253,7 +253,7 @@ def ingest_course_content(self,
             data = {
                 'course_name': course_name,
                 'readable_filename': readable_filename,
-                's3_paths': s3_path,
+                's3_paths': [s3_path],
                 'base_url': f"{self.canvas_url}/courses/{canvas_course_id}",
             }
             print(f"Posting readable_filename: '{readable_filename}' with S3 path: '{s3_path}'")

ai_ta_backend/rabbitmq/rmqueue.py

@@ -62,8 +62,16 @@ def addJobToIngestQueue(self, inputs, queue_name=None):
             logging.error("RabbitMQ is offline")
 
         # SQL record first
+        # Handle s3_paths: convert string to list, or get first element from list
+        s3_path_value = ''
+        if 's3_paths' in inputs:
+            s3_paths = inputs['s3_paths']
+            if isinstance(s3_paths, str):
+                s3_path_value = s3_paths
+            elif isinstance(s3_paths, list) and len(s3_paths) > 0:
+                s3_path_value = s3_paths[0]
         doc_progress_payload = models.DocumentsInProgress(
-            s3_path=inputs['s3_paths'][0] if 's3_paths' in inputs else '',
+            s3_path=s3_path_value,
             readable_filename=inputs['readable_filename'],
             course_name=inputs['course_name']
         )