handle gradio file wrapping

tool.py

@@ -2,6 +2,7 @@ from __future__ import annotations
 
 import json
 import os
+from typing import Optional, Tuple
 
 import requests
 from smolagents.tools import Tool
@@ -25,7 +26,19 @@ class BrightDataScraperTool(Tool):
         }
         super().__init__()
 
-    def forward(self, url: str) -> str:
+    def forward(self, url) -> str:
+        url_str, file_path = self._coerce_url_input(url)
+
+        if file_path:
+            try:
+                with open(file_path, "r", encoding="utf-8", errors="ignore") as f:
+                    return f.read()
+            except OSError as exc:
+                return json.dumps({"error": f"Failed to read uploaded file: {exc}"})
+
+        if not url_str:
+            return json.dumps({"error": "No valid URL provided"})
+
         api_token = os.getenv("BRIGHT_DATA_API_TOKEN")
         unlocker_zone = os.getenv("BRIGHT_DATA_UNLOCKER_ZONE", "web_unlocker1")
 
@@ -39,7 +52,7 @@ class BrightDataScraperTool(Tool):
         }
 
         payload = {
-            "url": url,
+            "url": url_str,
             "zone": unlocker_zone,
             "format": "raw",
             "data_format": "markdown",
@@ -52,3 +65,20 @@ class BrightDataScraperTool(Tool):
         except requests.exceptions.RequestException as exc:
             details = exc.response.text if getattr(exc, "response", None) is not None else ""
             return json.dumps({"error": str(exc), "details": details})
+
+    def _coerce_url_input(self, raw) -> Tuple[Optional[str], Optional[str]]:
+        if isinstance(raw, str):
+            return raw, None
+
+        if isinstance(raw, dict):
+            file_path = raw.get("path") or raw.get("name")
+            if file_path and os.path.isfile(file_path):
+                return None, file_path
+
+            orig_name = raw.get("orig_name")
+            if isinstance(orig_name, str) and orig_name:
+                if orig_name.startswith(("http://", "https://")):
+                    return orig_name, None
+                return f"https://{orig_name}", None
+
+        return None, None