fix: clean up storage adapter and pin sentence-transformers to 3.0.1

app/services/hf_storage_adapter.py

@@ -1,7 +1,6 @@
 """Hugging Face Hub storage adapter."""
 from app.ports.storage import StoragePort
 from app.config import get_settings
-from typing import Optional
 import logging
 import os
 from pathlib import Path
@@ -12,98 +11,88 @@ settings = get_settings()
 
 class HFStorageAdapter(StoragePort):
     """Hugging Face Hub implementation of StoragePort."""
-    
+
     def __init__(self):
         self.bucket_path = settings.HF_BUCKET_PATH
-        # Extract repo_id from bucket path: hf://buckets/Peterase/Ragora-doc-store -> Peterase/Ragora-doc-store
+        # Extract repo_id: hf://buckets/Peterase/Ragora-doc-store -> Peterase/Ragora-doc-store
         self.repo_id = self.bucket_path.replace("hf://buckets/", "")
         self.local_cache = Path("/tmp/ragora_cache")
         self.local_cache.mkdir(exist_ok=True)
+        # HF token for authenticated operations (set in HF Space secrets)
+        self.hf_token = os.getenv("HF_TOKEN")
         logger.info(f"Using HF bucket: {self.bucket_path} (repo: {self.repo_id})")
-    
+        if not self.hf_token:
+            logger.warning("HF_TOKEN not set - uploads may fail for private repos")
+
     async def upload(self, key: str, data: bytes, content_type: str) -> str:
-        """Upload file to HF bucket."""
+        """Upload file to HF dataset repo."""
         try:
             # Save locally first
             local_path = self.local_cache / key
             local_path.parent.mkdir(parents=True, exist_ok=True)
-            
-            with open(local_path, 'wb') as f:
+
+            with open(local_path, "wb") as f:
                 f.write(data)
-            
-            # Upload to HF bucket using huggingface_hub API
+
+            # Upload using huggingface_hub Python API
             from huggingface_hub import HfApi
-            api = HfApi()
-            
-            # Upload file to the bucket
+            api = HfApi(token=self.hf_token)
             api.upload_file(
                 path_or_fileobj=str(local_path),
                 path_in_repo=key,
                 repo_id=self.repo_id,
                 repo_type="dataset"
             )
-            
+
             logger.info(f"Uploaded file: {key}")
             return key
         except Exception as e:
             logger.error(f"Error uploading file: {e}")
             raise
-    
+
     async def download(self, key: str) -> bytes:
-        """Download file from HF bucket."""
+        """Download file from HF dataset repo."""
         try:
-            # Check local cache first
             local_path = self.local_cache / key
-            
+
             if not local_path.exists():
-                # Download from HF bucket using huggingface_hub API
                 from huggingface_hub import hf_hub_download
-                
                 local_path.parent.mkdir(parents=True, exist_ok=True)
-                
-                downloaded_path = hf_hub_download(
+                hf_hub_download(
                     repo_id=self.repo_id,
                     filename=key,
                     repo_type="dataset",
                     local_dir=str(self.local_cache),
-                    local_dir_use_symlinks=False
+                    token=self.hf_token
                 )
-                
                 logger.info(f"Downloaded file: {key}")
-            
-            with open(local_path, 'rb') as f:
-                data = f.read()
-            
-            return data
+
+            with open(local_path, "rb") as f:
+                return f.read()
         except Exception as e:
             logger.error(f"Error downloading file: {e}")
             raise
-    
+
     async def delete(self, key: str) -> None:
-        """Delete file from HF bucket."""
+        """Delete file from HF dataset repo."""
         try:
-            # Delete from HF bucket using huggingface_hub API
             from huggingface_hub import HfApi
-            api = HfApi()
-            
+            api = HfApi(token=self.hf_token)
             api.delete_file(
                 path_in_repo=key,
                 repo_id=self.repo_id,
                 repo_type="dataset"
             )
-            
-            # Delete from local cache
+
             local_path = self.local_cache / key
             if local_path.exists():
                 local_path.unlink()
-            
+
             logger.info(f"Deleted file: {key}")
         except Exception as e:
             logger.error(f"Error deleting file: {e}")
             raise
-    
+
     async def get_presigned_url(self, key: str, expires: int = 3600) -> str:
-        """Get presigned URL for downloading."""
-        # For HF datasets, return the direct HF URL
-        hf_url = f"https://huggingface.co/datasets/{self.repo_id}/resolve/main/{key}"
-        return hf_url
+        """Get direct download URL."""
+        return f"https://huggingface.co/datasets/{self.repo_id}/resolve/main/{key}"

requirements.txt

@@ -11,7 +11,7 @@ python-docx==1.1.0
 
 # Vector & embeddings
 qdrant-client==1.7.3
-sentence-transformers==2.3.1
+sentence-transformers==3.0.1
 
 # Storage
 minio==7.2.3
@@ -23,12 +23,8 @@ httpx==0.26.0
 PyJWT==2.8.0
 cryptography==42.0.0
 
-# Testing
-pytest==7.4.4
-pytest-asyncio==0.23.3
-
 # Hugging Face Hub
-huggingface-hub>=0.20.0
+huggingface-hub==0.24.0
 
 # Pinecone (new package name)
 pinecone>=5.0.0