'commit'

dsRag/Milvus/X4_InsertData.py

@@ -36,9 +36,12 @@ for filename in os.listdir(txt_dir):
     if filename.endswith('.txt'):
         filepath = os.path.join(txt_dir, filename)
         with open(filepath, 'r', encoding='utf-8') as f:
-            content = f.read().strip()
+            # 只读取第一行作为向量计算
+            first_line = f.readline().strip()
+            # 读取全部内容用于后续查询
+            full_content = first_line + '\n' + f.read()
             
-            if not content:
+            if not first_line:
                 print(f"跳过空文件: {filename}")
                 continue
                 
@@ -46,15 +49,15 @@ for filename in os.listdir(txt_dir):
 
             # 5. 获取当前时间和会话ID
             timestamp = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())
-            tags = {"tags": ["MATH_DATA_1", "小学数学"]} # 直接使用Python字典，Milvus会自动转换为JSON
+            tags = {"tags": ["MATH_DATA_1", "小学数学"], "full_content": full_content} # 添加完整内容
 
-            # 6. 将文本转换为嵌入向量
-            embedding = text_to_embedding(content)
+            # 6. 将第一行文本转换为嵌入向量
+            embedding = text_to_embedding(first_line)
 
             # 7. 插入数据
             entities = [
                 [tags],  # tags
-                [content],  # user_input
+                [first_line],  # user_input
                 [timestamp],  # timestamp
                 [embedding]  # embedding
             ]

dsRag/Milvus/X6_search_near_data.py

@@ -59,7 +59,7 @@ results = collection_manager.search(
     current_embedding,
     search_params,
     expr=expr,  # 使用in操作符
-    limit=10
+    limit=5
 )
 
 end_time = time.time()

dsRag/Start.py

@@ -99,7 +99,11 @@ async def generate_stream(client, milvus_pool, collection_manager, query):
                         logger.info(f"ID: {hit.id}")
                         logger.info(f"标签: {record['tags']}")
                         logger.info(f"用户问题: {record['user_input']}")
-                        context = context + record['user_input']
+                        
+                        # 获取完整内容
+                        full_content = record['tags'].get('full_content', record['user_input'])
+                        context = context + full_content
+                        
                         logger.info(f"时间: {record['timestamp']}")
                         logger.info(f"距离: {hit.distance}")
                         logger.info("-" * 40)  # 分隔线