Merge pull request #292 from kubet/feat/attachment-rework

fix: normalize path

backend/sandbox/api.py

@@ -1,4 +1,5 @@
 import os
+import urllib.parse
 from typing import Optional
 
 from fastapi import FastAPI, UploadFile, File, HTTPException, APIRouter, Form, Depends, Request
@@ -30,6 +31,41 @@ class FileInfo(BaseModel):
     mod_time: str
     permissions: Optional[str] = None
 
+def normalize_path(path: str) -> str:
+    """
+    Normalize a path to ensure proper UTF-8 encoding and handling.
+    
+    Args:
+        path: The file path, potentially containing URL-encoded characters
+        
+    Returns:
+        Normalized path with proper UTF-8 encoding
+    """
+    try:
+        # First, ensure the path is properly URL-decoded
+        decoded_path = urllib.parse.unquote(path)
+        
+        # Handle Unicode escape sequences like \u0308
+        try:
+            # Replace Python-style Unicode escapes (\u0308) with actual characters
+            # This handles cases where the Unicode escape sequence is part of the URL
+            import re
+            unicode_pattern = re.compile(r'\\u([0-9a-fA-F]{4})')
+            
+            def replace_unicode(match):
+                hex_val = match.group(1)
+                return chr(int(hex_val, 16))
+            
+            decoded_path = unicode_pattern.sub(replace_unicode, decoded_path)
+        except Exception as unicode_err:
+            logger.warning(f"Error processing Unicode escapes in path '{path}': {str(unicode_err)}")
+        
+        logger.debug(f"Normalized path from '{path}' to '{decoded_path}'")
+        return decoded_path
+    except Exception as e:
+        logger.error(f"Error normalizing path '{path}': {str(e)}")
+        return path  # Return original path if decoding fails
+
 async def verify_sandbox_access(client, sandbox_id: str, user_id: Optional[str] = None):
     """
     Verify that a user has access to a specific sandbox based on account membership.
@@ -114,6 +150,9 @@ async def create_file(
     user_id: Optional[str] = Depends(get_optional_user_id)
 ):
     """Create a file in the sandbox using direct file upload"""
+    # Normalize the path to handle UTF-8 encoding correctly
+    path = normalize_path(path)
+    
     logger.info(f"Received file upload request for sandbox {sandbox_id}, path: {path}, user_id: {user_id}")
     client = await db.client
     
@@ -144,6 +183,9 @@ async def list_files(
     user_id: Optional[str] = Depends(get_optional_user_id)
 ):
     """List files and directories at the specified path"""
+    # Normalize the path to handle UTF-8 encoding correctly
+    path = normalize_path(path)
+    
     logger.info(f"Received list files request for sandbox {sandbox_id}, path: {path}, user_id: {user_id}")
     client = await db.client
     
@@ -186,7 +228,14 @@ async def read_file(
     user_id: Optional[str] = Depends(get_optional_user_id)
 ):
     """Read a file from the sandbox"""
+    # Normalize the path to handle UTF-8 encoding correctly
+    original_path = path
+    path = normalize_path(path)
+    
     logger.info(f"Received file read request for sandbox {sandbox_id}, path: {path}, user_id: {user_id}")
+    if original_path != path:
+        logger.info(f"Normalized path from '{original_path}' to '{path}'")
+    
     client = await db.client
     
     # Verify the user has access to this sandbox
@@ -196,17 +245,59 @@ async def read_file(
         # Get sandbox using the safer method
         sandbox = await get_sandbox_by_id_safely(client, sandbox_id)
         
+        # Verify the file exists first
+        try:
+            filename = os.path.basename(path)
+            parent_dir = os.path.dirname(path)
+            
+            # List files in the parent directory to check if the file exists
+            files_in_dir = sandbox.fs.list_files(parent_dir)
+            
+            # Look for the target file with exact name match
+            file_exists = any(file.name == filename for file in files_in_dir)
+            
+            if not file_exists:
+                logger.warning(f"File not found: {path} in sandbox {sandbox_id}")
+                
+                # Try to find similar files to help diagnose
+                close_matches = [file.name for file in files_in_dir if filename.lower() in file.name.lower()]
+                error_detail = f"File '{filename}' not found in directory '{parent_dir}'"
+                
+                if close_matches:
+                    error_detail += f". Similar files in the directory: {', '.join(close_matches)}"
+                
+                raise HTTPException(status_code=404, detail=error_detail)
+        except Exception as list_err:
+            # If we can't list files, continue with the download attempt
+            logger.warning(f"Error checking if file exists: {str(list_err)}")
+        
         # Read file
-        content = sandbox.fs.download_file(path)
+        try:
+            content = sandbox.fs.download_file(path)
+        except Exception as download_err:
+            logger.error(f"Error downloading file {path} from sandbox {sandbox_id}: {str(download_err)}")
+            raise HTTPException(
+                status_code=404, 
+                detail=f"Failed to download file: {str(download_err)}"
+            )
         
         # Return a Response object with the content directly
         filename = os.path.basename(path)
         logger.info(f"Successfully read file {filename} from sandbox {sandbox_id}")
+        
+        # Ensure proper encoding by explicitly using UTF-8 for the filename in Content-Disposition header
+        # This applies RFC 5987 encoding for the filename to support non-ASCII characters
+        encoded_filename = filename.encode('utf-8').decode('latin-1')
+        content_disposition = f"attachment; filename*=UTF-8''{encoded_filename}"
+        
         return Response(
             content=content,
             media_type="application/octet-stream",
-            headers={"Content-Disposition": f"attachment; filename={filename}"}
+            headers={"Content-Disposition": content_disposition}
         )
+    except HTTPException:
+        # Re-raise HTTP exceptions without wrapping
+        raise
     except Exception as e:
         logger.error(f"Error reading file in sandbox {sandbox_id}: {str(e)}")
         raise HTTPException(status_code=500, detail=str(e))

frontend/src/components/thread/content/ThreadContent.tsx

@@ -9,9 +9,9 @@ import { FileCache } from '@/hooks/use-cached-file';
 import { useAuth } from '@/components/AuthProvider';
 import { Project } from '@/lib/api';
 import {
-  extractPrimaryParam,
-  getToolIcon,
-  safeJsonParse,
+    extractPrimaryParam,
+    getToolIcon,
+    safeJsonParse,
 } from '@/components/thread/utils';
 
 // Define the set of tags whose raw XML should be hidden during streaming

frontend/src/components/thread/file-attachment.tsx

@@ -124,7 +124,19 @@ function getFileUrl(sandboxId: string | undefined, path: string): string {
         path = `/workspace/${path.startsWith('/') ? path.substring(1) : path}`;
     }
 
+    // Handle any potential Unicode escape sequences
+    try {
+        // Replace escaped Unicode sequences with actual characters
+        path = path.replace(/\\u([0-9a-fA-F]{4})/g, (_, hexCode) => {
+            return String.fromCharCode(parseInt(hexCode, 16));
+        });
+    } catch (e) {
+        console.error('Error processing Unicode escapes in path:', e);
+    }
+
     const url = new URL(`${process.env.NEXT_PUBLIC_BACKEND_URL}/sandboxes/${sandboxId}/files/content`);
+
+    // Properly encode the path parameter for UTF-8 support
     url.searchParams.append('path', path);
 
     return url.toString();

frontend/src/hooks/use-cached-file.ts

@@ -24,6 +24,15 @@ function normalizePath(path: string): string {
     path = `/workspace/${path.startsWith('/') ? path.substring(1) : path}`;
   }
   
+  // Handle Unicode escape sequences like \u0308
+  try {
+    path = path.replace(/\\u([0-9a-fA-F]{4})/g, (_, hexCode) => {
+      return String.fromCharCode(parseInt(hexCode, 16));
+    });
+  } catch (e) {
+    console.error('Error processing Unicode escapes in path:', e);
+  }
+  
   return path;
 }
 
@@ -111,6 +120,8 @@ export function useCachedFile<T = string>(
       const normalizedPath = normalizePath(filePath || '');
       
       const url = new URL(`${process.env.NEXT_PUBLIC_BACKEND_URL}/sandboxes/${sandboxId}/files/content`);
+      
+      // Properly encode the path parameter for UTF-8 support
       url.searchParams.append('path', normalizedPath);
       
       // Fetch with authentication
@@ -377,6 +388,11 @@ export const FileCache = {
     console.log(`[FILE CACHE] Preloading ${filePaths.length} files for sandbox ${sandboxId}`);
     
     return Promise.all(filePaths.map(async (path) => {
+      // Handle Unicode escape sequences in paths
+      path = path.replace(/\\u([0-9a-fA-F]{4})/g, (_, hexCode) => {
+        return String.fromCharCode(parseInt(hexCode, 16));
+      });
+      
       const normalizedPath = normalizePath(path);
       const key = getCacheKey(sandboxId, path);
       
@@ -389,6 +405,8 @@ export const FileCache = {
       
       try {        
         const url = new URL(`${process.env.NEXT_PUBLIC_BACKEND_URL}/sandboxes/${sandboxId}/files/content`);
+        
+        // Properly encode the path parameter for UTF-8 support
         url.searchParams.append('path', normalizedPath);
         
         const response = await fetch(url.toString(), {
@@ -491,6 +509,11 @@ export async function getCachedFile(
     ? 'blob' 
     : (options.contentType || 'text');
   
+  // First ensure the file path has any Unicode escape sequences properly handled
+  filePath = filePath.replace(/\\u([0-9a-fA-F]{4})/g, (_, hexCode) => {
+    return String.fromCharCode(parseInt(hexCode, 16));
+  });
+  
   const key = getCacheKey(sandboxId, filePath);
   const startTime = performance.now();
   
@@ -535,7 +558,10 @@ export async function getCachedFile(
   
   try {
     const url = new URL(`${process.env.NEXT_PUBLIC_BACKEND_URL}/sandboxes/${sandboxId}/files/content`);
-    url.searchParams.append('path', normalizePath(filePath));
+    const normalizedPath = normalizePath(filePath);
+    
+    // Properly encode the path parameter for UTF-8 characters
+    url.searchParams.append('path', normalizedPath);
     
     const response = await fetch(url.toString(), {
       headers: {

frontend/src/lib/api.ts

@@ -1135,6 +1135,19 @@ export const createSandboxFileJson = async (
   }
 };
 
+// Helper function to normalize file paths with Unicode characters
+function normalizePathWithUnicode(path: string): string {
+  try {
+    // Replace escaped Unicode sequences with actual characters
+    return path.replace(/\\u([0-9a-fA-F]{4})/g, (_, hexCode) => {
+      return String.fromCharCode(parseInt(hexCode, 16));
+    });
+  } catch (e) {
+    console.error('Error processing Unicode escapes in path:', e);
+    return path;
+  }
+}
+
 export const listSandboxFiles = async (
   sandboxId: string,
   path: string,
@@ -1146,7 +1159,12 @@ export const listSandboxFiles = async (
     } = await supabase.auth.getSession();
 
     const url = new URL(`${API_URL}/sandboxes/${sandboxId}/files`);
-    url.searchParams.append('path', path);
+    
+    // Normalize the path to handle Unicode escape sequences
+    const normalizedPath = normalizePathWithUnicode(path);
+    
+    // Properly encode the path parameter for UTF-8 support
+    url.searchParams.append('path', normalizedPath);
 
     const headers: Record<string, string> = {};
     if (session?.access_token) {
@@ -1189,7 +1207,12 @@ export const getSandboxFileContent = async (
     } = await supabase.auth.getSession();
 
     const url = new URL(`${API_URL}/sandboxes/${sandboxId}/files/content`);
-    url.searchParams.append('path', path);
+    
+    // Normalize the path to handle Unicode escape sequences
+    const normalizedPath = normalizePathWithUnicode(path);
+    
+    // Properly encode the path parameter for UTF-8 support
+    url.searchParams.append('path', normalizedPath);
 
     const headers: Record<string, string> = {};
     if (session?.access_token) {