refactor(csv_parser): Implement SAVEPOINT handling for activity import

- Added SAVEPOINT management to the `_import_activity` function to improve error handling during CSV imports.
- Moved the training type resolution logic to occur within a transaction block, ensuring that errors can be rolled back without affecting the entire import process.
- Enhanced error logging to capture evaluation failures, providing better insights during CSV import operations.

backend/csv_parser/executor.py

@@ -872,13 +872,14 @@ def _import_activity(
         dist = _sf_act(mapped.get("distance_km"))
 
         wtype = str(activity_type).strip()
-        training_type_id, training_category, training_subcategory = _resolve_training_type_for_activity(
-            cur, wtype, profile_id
-        )
-
         iso = date_d.isoformat()
 
+        # Pro Zeile: bei SQL-Fehler sonst „current transaction is aborted“ bis Xact-Ende.
+        cur.execute("SAVEPOINT csv_activity_row")
         try:
+            training_type_id, training_category, training_subcategory = _resolve_training_type_for_activity(
+                cur, wtype, profile_id
+            )
             cur.execute(
                 """
                 SELECT id FROM activity_log
@@ -986,7 +987,12 @@ def _import_activity(
                     _evaluate_and_save_activity(cur, aid, activity_dict, training_type_id, profile_id)
                 except Exception as eval_err:
                     logger.warning("[csv activity] Auto-Eval fehlgeschlagen: %s", eval_err)
+            cur.execute("RELEASE SAVEPOINT csv_activity_row")
         except Exception as e:
+            try:
+                cur.execute("ROLLBACK TO SAVEPOINT csv_activity_row")
+            except Exception:
+                pass
             error_details.append({"row": rows_total, "error": str(e)})
 
     return {