Update app/services/vector_service.py

app/services/vector_service.py

@@ -9,31 +9,48 @@ logger = logging.getLogger(__name__)
 
 class VectorService:
     def __init__(self):
-        self.client = chromadb.PersistentClient(
-            path="./chroma_db",
-            settings=Settings(
+        try:
+            # Create absolute path for ChromaDB in HuggingFace Spaces
+            chroma_path = "/app/chroma_db"
+            os.makedirs(chroma_path, exist_ok=True)
+            
+            print(f"🔧 [CHROMA] Using database path: {chroma_path}", flush=True)
+            
+            # Use Client instead of PersistentClient for HuggingFace compatibility
+            self.client = chromadb.Client(Settings(
+                chroma_db_impl="duckdb+parquet",
+                persist_directory=chroma_path,
                 anonymized_telemetry=False,
                 allow_reset=True
-            )
-        )
-        logger.info("🗄️ ChromaDB client initialized")
+            ))
+            
+            print("✅ [CHROMA] ChromaDB client initialized successfully!", flush=True)
+            logger.info("🗄️ ChromaDB client initialized")
+            
+        except Exception as e:
+            print(f"❌ [CHROMA] Failed to initialize ChromaDB: {e}", flush=True)
+            logger.error(f"❌ Failed to initialize ChromaDB: {e}")
+            raise Exception(f"Failed to initialize ChromaDB: {e}")
     
     def create_collection(self, repository_id: int) -> chromadb.Collection:
         collection_name = f"repo_{repository_id}"
         
         try:
             collection = self.client.get_collection(collection_name)
+            print(f"📚 [CHROMA] Using existing collection: {collection_name}", flush=True)
             logger.info(f"📚 Using existing collection: {collection_name}")
         except:
             collection = self.client.create_collection(
                 name=collection_name,
                 metadata={"repository_id": repository_id}
             )
+            print(f"🆕 [CHROMA] Created new collection: {collection_name}", flush=True)
             logger.info(f"🆕 Created new collection: {collection_name}")
         
         return collection
     
     async def store_embeddings(self, repository_id: int, embedded_chunks: List[Dict]):
+        print(f"💾 [CHROMA] Storing {len(embedded_chunks)} embeddings for repository {repository_id}", flush=True)
         logger.info(f"💾 Storing {len(embedded_chunks)} embeddings for repository {repository_id}")
         
         collection = self.create_collection(repository_id)
@@ -59,16 +76,26 @@ class VectorService:
             ids.append(chunk_id)
         
         batch_size = 100
-        for i in range(0, len(documents), batch_size):
+        total_batches = (len(documents) + batch_size - 1) // batch_size
+        
+        for batch_num, i in enumerate(range(0, len(documents), batch_size), 1):
             end_idx = min(i + batch_size, len(documents))
             
-            collection.add(
-                documents=documents[i:end_idx],
-                embeddings=embeddings[i:end_idx],
-                metadatas=metadatas[i:end_idx],
-                ids=ids[i:end_idx]
-            )
+            try:
+                collection.add(
+                    documents=documents[i:end_idx],
+                    embeddings=embeddings[i:end_idx],
+                    metadatas=metadatas[i:end_idx],
+                    ids=ids[i:end_idx]
+                )
+                
+                print(f"✅ [CHROMA] Stored batch {batch_num}/{total_batches} ({end_idx-i} embeddings)", flush=True)
+                
+            except Exception as e:
+                print(f"❌ [CHROMA] Error storing batch {batch_num}: {e}", flush=True)
+                raise
         
+        print(f"🎉 [CHROMA] Successfully stored all {len(embedded_chunks)} embeddings for repository {repository_id}!", flush=True)
         logger.info(f"✅ Successfully stored all embeddings for repository {repository_id}")
     
     async def search_similar_code(self, repository_id: int, query_embedding: List[float], top_k: int = 5) -> List[Dict]:
@@ -108,11 +135,13 @@ class VectorService:
         logger.info(f"🔍 Found {len(search_results)} similar code chunks")
         return search_results
     
-    def delete_repository_data(self, repository_id: int):
+    async def delete_repository_data(self, repository_id: int):
         collection_name = f"repo_{repository_id}"
         
         try:
             self.client.delete_collection(collection_name)
+            print(f"🗑️ [CHROMA] Deleted collection: {collection_name}", flush=True)
             logger.info(f"🗑️ Deleted collection: {collection_name}")
-        except:
+        except Exception as e:
+            print(f"⚠️ [CHROMA] Collection {collection_name} not found for deletion: {e}", flush=True)
             logger.warning(f"⚠️ Collection {collection_name} not found for deletion")