commit

app.py

@@ -49,20 +49,33 @@ class RAGSystem:
             length_function=len
         )
 
+    # def add_documents(self, documents: List[str]):
+    #     """ドキュメントをベクトルストアに追加"""
+    #     chunks = []
+    #     for doc in documents:
+    #         chunks.extend(self.text_splitter.split_text(doc))
+        
+    #     # ChromaDBに保存
+    #     self.vectorstore.add_texts(chunks)
+
     def add_documents(self, documents: List[str]):
         """ドキュメントをベクトルストアに追加"""
         chunks = []
-        for doc in documents:
-            chunks.extend(self.text_splitter.split_text(doc))
-        
-        # ChromaDBに保存
-        self.vectorstore.add_texts(chunks)
+        metadatas = []
 
+        for i, doc in enumerate(documents):
+            split_chunks = self.text_splitter.split_text(doc)
+            chunks.extend([f"passage: {chunk}" for chunk in split_chunks])
+            metadatas.extend([{"source_doc_id": i}] * len(split_chunks))
+
+        # ChromaDB に保存（プレフィックス付き・メタデータ付き）
+        self.vectorstore.add_texts(texts=chunks, metadatas=metadatas)
+        
     def query(self, question: str, k: int = 8) -> str:  # k値をさらに増加
         """質問に対する回答を生成"""
         # 関連ドキュメントを検索
         docs = self.vectorstore.similarity_search_with_score(
-            question,
+            f"query: {question}",
             k=k
         )