Vector.

app/agents/vectorization_agent.py

@@ -1,29 +1,25 @@
 # app/agents/vectorization_agent.py
 from langchain.text_splitter import RecursiveCharacterTextSplitter
-from langchain_openai import OpenAIEmbeddings
-embedding_model = OpenAIEmbeddings(model="text-embedding-3-small")
-import chromadb
-
-client = chromadb.PersistentClient(path="./chroma_db")
-vector_store = client.get_or_create_collection(name="documents")
 
 text_splitter = RecursiveCharacterTextSplitter(
     chunk_size=1000,
     chunk_overlap=100,
 )
 
-def agent_vectorize_and_store(doc_id: str, text: str, category: str, language: str):
-    """
-    Agent 4: Vectorizes a document and stores it in ChromaDB.
-    """
+def agent_vectorize_and_store(
+        doc_id: str,
+        text: str,
+        category: str,
+        language: str,
+        embedding_model,
+        vector_store
+):
     print(f"--- [Background Task] Starting vectorization (ID: {doc_id})...")
 
     try:
-        return
-
         chunks = text_splitter.split_text(text)
         if not chunks:
-            print(f"--- [Background Task] document {doc_id} has no text to vectorize.")
+            print(f"--- [Background task] document is empty, skip vectorization. (ID: {doc_id})")
             return
 
         chunk_ids = [f"{doc_id}_{i}" for i in range(len(chunks))]
@@ -38,6 +34,6 @@ def agent_vectorize_and_store(doc_id: str, text: str, category: str, language: s
             documents=chunks,
             metadatas=metadatas
         )
-        print(f"--- [Background Task] Document {doc_id} vectorized and stored successfully.")
+        print(f"--- [Background Task] Document {doc_id} vectorized。")
     except Exception as e:
-        print(f"--- [background Task] Vectorization failed (ID: {doc_id}): {e}")
+        print(f"--- [Background Task] Document vectorization failed (ID: {doc_id}): {e}")

app/core/vector_store.py

@@ -1,47 +1,28 @@
-# app/core/vector_store.py
 import os
 import chromadb
 from dotenv import load_dotenv
 from langchain_openai import AzureOpenAIEmbeddings, OpenAIEmbeddings
 
 load_dotenv()
-
 LLM_PROVIDER = os.getenv("LLM_PROVIDER", "openai").lower()
-
 embedding_model = None
 
 print(f"--- [Core] Initializing Embeddings with provider: {LLM_PROVIDER} ---")
 
 if LLM_PROVIDER == "azure":
-    required_vars = [
-        "AZURE_OPENAI_ENDPOINT", "AZURE_OPENAI_API_KEY",
-        "OPENAI_API_VERSION", "AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME"
-    ]
-    if not all(os.getenv(var) for var in required_vars):
-        raise ValueError("One or more Azure OpenAI environment variables for embeddings are not set.")
-
     embedding_model = AzureOpenAIEmbeddings(
         azure_endpoint=os.getenv("AZURE_OPENAI_ENDPOINT"),
         api_key=os.getenv("AZURE_OPENAI_API_KEY"),
         api_version=os.getenv("OPENAI_API_VERSION"),
         azure_deployment=os.getenv("AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME"),
     )
-
 elif LLM_PROVIDER == "openai":
-    if not os.getenv("OPENAI_API_KEY"):
-        raise ValueError("OPENAI_API_KEY is not set for the 'openai' provider.")
-
     embedding_model = OpenAIEmbeddings(
         api_key=os.getenv("OPENAI_API_KEY"),
         model=os.getenv("OPENAI_EMBEDDING_MODEL_NAME", "text-embedding-3-small")
     )
-
 else:
-    raise ValueError(f"Unsupported LLM_PROVIDER: {LLM_PROVIDER}. Please use 'azure' or 'openai'.")
-
+    raise ValueError(f"Unsupported LLM_PROVIDER: {LLM_PROVIDER}.")
 
 client = chromadb.PersistentClient(path="./chroma_db")
-vector_store = client.get_or_create_collection(
-    name="documents",
-    metadata={"hnsw:space": "cosine"}
-)
+vector_store = client.get_or_create_collection(name="documents")

app/routers/documents.py

@@ -5,10 +5,10 @@ from typing import Dict, Any, List
 from fastapi.concurrency import run_in_threadpool
 from PIL import Image
 from io import BytesIO
-
 from .. import agents
 from ..core.pdf_processor import convert_pdf_to_images, image_to_base64_str
 from ..core.ocr import extract_text_from_images
+from ..core.vector_store import embedding_model, vector_store
 
 # Create an APIRouter instance
 router = APIRouter(
@@ -102,10 +102,12 @@ async def upload_and_process_document(
         full_text = await run_in_threadpool(extract_text_from_images, images)
         background_tasks.add_task(
             agents.agent_vectorize_and_store,
-            doc_id,
-            full_text,
-            category,
-            language
+            doc_id=doc_id,
+            text=full_text,
+            category=category,
+            language=language,
+            embedding_model=embedding_model,
+            vector_store=vector_store
         )
         print("--- [Main] Vectorization job added to background tasks.")
 
@@ -118,4 +120,4 @@ async def upload_and_process_document(
 async def get_result(doc_id: str):
     if doc_id in db_results:
         return db_results[doc_id]
-    raise HTTPException(status_code=404, detail="Document not found.")
+    raise HTTPException(status_code=404, detail="Document not found.")