fix(security): path traversal sur /download et /cleanup, validation UUID session_id et chemin input_path dans reconstruct-document

Made-with: Cursor

routes/legacy_routes.py

@@ -20,6 +20,26 @@ logger = logging.getLogger(__name__)
 router = APIRouter(prefix="/api/v1", tags=["Legacy"])
 
 
+def _safe_output_path(filename: str):
+    """
+    Resolve filename to a path under config.OUTPUT_DIR. Prevents path traversal.
+    Returns (Path, True) if valid, (None, False) if invalid.
+    """
+    if not filename or ".." in filename or "/" in filename or "\\" in filename:
+        return None, False
+    safe_name = Path(filename).name
+    if not safe_name.strip():
+        return None, False
+    base = config.OUTPUT_DIR.resolve()
+    try:
+        resolved = (config.OUTPUT_DIR / safe_name).resolve()
+        if not resolved.is_relative_to(base):
+            return None, False
+        return resolved, True
+    except (ValueError, OSError):
+        return None, False
+
+
 def _resolve_model(
     cfg_model: Optional[str],
     model_env: str,
@@ -298,32 +318,30 @@ async def translate_batch_documents(
 
 @router.get("/download/{filename}")
 async def download_file(filename: str):
-    """Download a translated file by filename"""
-    file_path = config.OUTPUT_DIR / filename
-
+    """Download a translated file by filename. Filename is sanitized to prevent path traversal."""
+    file_path, ok = _safe_output_path(filename)
+    if not ok or file_path is None:
+        raise HTTPException(status_code=400, detail="Invalid filename")
     if not file_path.exists():
         raise HTTPException(status_code=404, detail="File not found")
-
     return FileResponse(
         path=file_path,
-        filename=filename,
+        filename=file_path.name,
         media_type="application/octet-stream",
     )
 
 
 @router.delete("/cleanup/{filename}")
 async def cleanup_translated_file(filename: str):
-    """Cleanup a translated file after download"""
+    """Cleanup a translated file after download. Filename is sanitized to prevent path traversal."""
+    file_path, ok = _safe_output_path(filename)
+    if not ok or file_path is None:
+        raise HTTPException(status_code=400, detail="Invalid filename")
     try:
-        file_path = config.OUTPUT_DIR / filename
-
         if not file_path.exists():
             raise HTTPException(status_code=404, detail="File not found")
-
         file_handler.cleanup_file(file_path)
-
-        return {"message": f"File {filename} deleted successfully"}
-
+        return {"message": f"File {file_path.name} deleted successfully"}
     except HTTPException:
         raise
     except Exception as e:
@@ -454,8 +472,14 @@ async def reconstruct_document(
     ),
     target_language: str = Form(..., description="Target language code"),
 ):
-    """Reconstruct a document with translated texts"""
+    """Reconstruct a document with translated texts. session_id must be a valid UUID."""
     import json
+    import uuid
+
+    try:
+        uuid.UUID(session_id)
+    except (ValueError, TypeError):
+        raise HTTPException(status_code=400, detail="Invalid session ID")
 
     try:
         session_file = config.UPLOAD_DIR / f"session_{session_id}.json"
@@ -465,7 +489,10 @@ async def reconstruct_document(
         with open(session_file, "r", encoding="utf-8") as f:
             session_data = json.load(f)
 
-        input_path = Path(session_data["input_path"])
+        input_path = Path(session_data["input_path"]).resolve()
+        upload_dir_resolved = config.UPLOAD_DIR.resolve()
+        if not input_path.is_relative_to(upload_dir_resolved):
+            raise HTTPException(status_code=400, detail="Invalid session data")
         file_extension = session_data["file_extension"]
         original_filename = session_data["original_filename"]