tests

app.py

@@ -69,6 +69,78 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
                 )
             except Exception as e:
                 print(f"Attempted access to resource at {file_url} -> error: {e}")
+
+        # Try fetching the same files from the official GAIA dataset using the HF token.
+        gaia_repo = "gaia-benchmark/GAIA"
+        gaia_files_cache = None
+        try:
+            from huggingface_hub import list_repo_files, hf_hub_download
+        except Exception as e:
+            print(f"Skipping GAIA file fetch (huggingface_hub not available): {e}")
+            gaia_repo = None
+
+        if gaia_repo:
+            # Inspect profile to locate token; also allow env var fallbacks.
+            try:
+                print("Profile keys:", list(profile.__dict__.keys()))
+            except Exception as e:
+                print(f"Could not introspect profile: {e}")
+
+            token = None
+            for attr in ("access_token", "token"):
+                token = getattr(profile, attr, None)
+                if token:
+                    print(f"Using token from profile.{attr}")
+                    break
+            if not token:
+                # Some OAuth profiles keep tokens under .tokens or .auth
+                for attr in ("tokens", "auth"):
+                    container = getattr(profile, attr, None)
+                    if isinstance(container, dict):
+                        token = container.get("access_token") or container.get("token")
+                        if token:
+                            print(f"Using token from profile.{attr}")
+                            break
+            if not token:
+                token = os.getenv("HF_TOKEN") or os.getenv("HUGGINGFACEHUB_API_TOKEN") or os.getenv("HUGGINGFACE_HUB_TOKEN")
+                if token:
+                    print("Using token from environment.")
+
+            if not token:
+                print("Skipping GAIA file fetch (no HF token found in profile or env).")
+            else:
+                for q in with_file:
+                    fname = q.get("file_name")
+                    task_id = q.get("task_id")
+                    if gaia_files_cache is None:
+                        try:
+                            gaia_files_cache = list_repo_files(
+                                gaia_repo, repo_type="dataset", token=token
+                            )
+                            print(f"GAIA repo file count: {len(gaia_files_cache)}")
+                        except Exception as e:
+                            print(f"Failed to list GAIA repo files: {e}")
+                            gaia_files_cache = []
+                    matches = []
+                    if gaia_files_cache:
+                        # First try an exact filename match, then any path containing the task_id.
+                        matches = [p for p in gaia_files_cache if p.endswith(fname)]
+                        if not matches:
+                            matches = [p for p in gaia_files_cache if task_id in p]
+                    if not matches:
+                        print(f"GAIA file not found for task {task_id} (looking for {fname}).")
+                        continue
+                    match_path = matches[0]
+                    try:
+                        local_path = hf_hub_download(
+                            gaia_repo,
+                            match_path,
+                            repo_type="dataset",
+                            token=token,
+                        )
+                        print(f"Downloaded GAIA file for task {task_id} to {local_path}")
+                    except Exception as e:
+                        print(f"Failed to download GAIA file for task {task_id} ({match_path}): {e}")
     except requests.exceptions.RequestException as e:
         print(f"Error fetching questions: {e}")
         return f"Error fetching questions: {e}", None

requirements.txt

@@ -2,4 +2,5 @@ gradio
 requests
 
 langgraph
-langchain_openai
+langchain_openai
+huggingface_hub