ruxia
578 lines
22 KiB
Python
578 lines
22 KiB
Python
from langchain.chains import LLMChain
|
|
from langchain_openai import OpenAI
|
|
from langchain_experimental.utilities import PythonREPL
|
|
from project_implementation import ProjectBuilder
|
|
from prompt_templates import FUNCTION_CALL_PROMPT
|
|
import inspect
|
|
import project
|
|
import io
|
|
import sys
|
|
from parameter_rewriting import rewrite_query_parameters, KnowledgeGraphProcessor
|
|
import json
|
|
from langchain.agents import Tool, AgentExecutor, create_react_agent
|
|
from langchain.prompts import PromptTemplate
|
|
|
|
from llm import llm
|
|
|
|
|
|
# 获取ProjectTookiIt类的方法定义
|
|
def get_project_class_methods():
|
|
"""
|
|
从project模块中提取ProjectTookiIt类的方法定义
|
|
|
|
Returns:
|
|
str: 格式化后的方法定义字符串
|
|
"""
|
|
project_class_code = inspect.getsource(project.ProjectTookiIt)
|
|
|
|
lines = project_class_code.split("\n")
|
|
result_lines = []
|
|
|
|
in_class = False
|
|
skip_init = False
|
|
|
|
for line in lines:
|
|
if line.strip().startswith("class ProjectTookiIt"):
|
|
in_class = True
|
|
result_lines.append(line)
|
|
elif in_class:
|
|
if line.strip().startswith("def __init__"):
|
|
skip_init = True
|
|
elif skip_init and line.strip() and not line.startswith(" " * 8):
|
|
skip_init = False
|
|
|
|
if not skip_init:
|
|
result_lines.append(line)
|
|
|
|
return "\n".join(result_lines)
|
|
|
|
|
|
# 创建动态提示模板
|
|
project_class_methods = get_project_class_methods()
|
|
|
|
# 创建 Chain
|
|
function_call_chain = LLMChain(llm=llm, prompt=FUNCTION_CALL_PROMPT, output_key="code")
|
|
|
|
# Python 执行器
|
|
repl = PythonREPL()
|
|
|
|
# 创建知识图谱处理器实例
|
|
kg_processor = KnowledgeGraphProcessor()
|
|
|
|
|
|
# 定义搜索知识库的工具
|
|
def search_knowledge_base(query):
|
|
"""
|
|
在知识库中搜索关键词
|
|
|
|
Args:
|
|
query (str): 搜索关键词
|
|
|
|
Returns:
|
|
str: 搜索结果的JSON字符串
|
|
"""
|
|
found_data = kg_processor._search_in_kg(query)
|
|
if found_data:
|
|
return json.dumps(found_data, ensure_ascii=False, indent=2)
|
|
else:
|
|
return f"未找到与'{query}'相关的信息"
|
|
|
|
|
|
# 定义获取节点定义的工具
|
|
def get_node_definition(node_type):
|
|
"""
|
|
获取节点类型的定义
|
|
|
|
Args:
|
|
node_type (str): 节点类型名称
|
|
|
|
Returns:
|
|
str: 节点类型定义
|
|
"""
|
|
definition = kg_processor._get_node_definition(node_type)
|
|
return definition
|
|
|
|
|
|
# 创建工具列表
|
|
tools = [
|
|
Tool(
|
|
name="search_knowledge_base",
|
|
func=search_knowledge_base,
|
|
description="在知识库中搜索关键词，返回相关信息。输入应该是一个搜索关键词。",
|
|
),
|
|
Tool(
|
|
name="get_node_definition",
|
|
func=get_node_definition,
|
|
description="获取节点类型的定义，返回类型定义代码。输入应该是一个节点类型名称。",
|
|
),
|
|
]
|
|
|
|
# 创建Agent
|
|
agent = create_react_agent(llm, tools, FUNCTION_CALL_PROMPT)
|
|
|
|
# 创建Agent执行器
|
|
agent_executor = AgentExecutor(agent=agent, tools=tools, verbose=True, handle_parsing_errors=True)
|
|
|
|
|
|
def nl_query_to_function_call(input_data):
|
|
"""
|
|
将自然语言查询转换为函数调用并执行，或直接执行提供的代码
|
|
|
|
Args:
|
|
input_data (dict): 包含type和value的字典
|
|
{
|
|
"type": "query|code",
|
|
"value": "查询内容或代码"
|
|
}
|
|
|
|
Returns:
|
|
dict: 包含状态码、消息和数据的字典
|
|
"""
|
|
input_type = input_data.get("type", "query")
|
|
input_value = input_data.get("value", "")
|
|
max_retries = 1 # 设置最大重试次数
|
|
current_retry = 0
|
|
|
|
original_query = input_value # 保存原始查询用于RAG
|
|
|
|
print(f"\n====== 开始处理查询 ======")
|
|
print(f"查询类型: {input_type}")
|
|
print(f"查询内容: {input_value}")
|
|
|
|
while current_retry <= max_retries:
|
|
print(f"\n----- 尝试 #{current_retry + 1} -----")
|
|
|
|
# 如果type是query，使用LLM生成代码
|
|
if input_type == "query" and current_retry == 0:
|
|
# 从查询中提取关键部分
|
|
import re
|
|
|
|
# 提取【】中的内容
|
|
path_parts = re.findall(r"【([^】]+)】", input_value)
|
|
|
|
# 创建临时代码
|
|
temp_code = f'search("{input_value}")'
|
|
|
|
# 对于每个路径，只添加最后一个部分
|
|
for part in path_parts:
|
|
if "/" in part:
|
|
# 提取路径中的最后一个部分
|
|
last_part = part.split("/")[-1].strip()
|
|
if last_part:
|
|
temp_code += f'\nsearch("{last_part}")'
|
|
else:
|
|
# 如果没有/，直接使用整个部分
|
|
temp_code += f'\nsearch("{part}")'
|
|
|
|
# 获取知识库内容和节点定义
|
|
knowledge_base, node_definitions = kg_processor._get_relevant_knowledge(temp_code)
|
|
|
|
# 使用Agent执行查询
|
|
agent_response = agent_executor.invoke(
|
|
{
|
|
"query": input_value,
|
|
"project_class_methods": project_class_methods,
|
|
"KnowledgeBase": knowledge_base,
|
|
"NodeDefinition": node_definitions,
|
|
}
|
|
)
|
|
|
|
# 从Agent响应中提取代码
|
|
code = agent_response["output"]
|
|
print(f"\n生成的代码:\n{code}")
|
|
else:
|
|
print(f"\n使用重写后的代码:\n{input_value}")
|
|
code = input_value
|
|
|
|
# 保存原始代码用于返回
|
|
original_code = code
|
|
|
|
# 执行生成的函数并捕获输出
|
|
try:
|
|
# 创建一个新的命名空间来执行代码，包含必要的导入
|
|
namespace = {
|
|
"ProjectBuilder": ProjectBuilder, # 添加ProjectBuilder到命名空间
|
|
"project_implementation": __import__("project_implementation"),
|
|
"project": __import__("project"),
|
|
}
|
|
|
|
# 执行生成的代码，定义neo4j_find_function函数
|
|
exec(code, namespace)
|
|
|
|
# 重定向stdout来捕获print输出
|
|
old_stdout = sys.stdout
|
|
redirected_output = io.StringIO()
|
|
sys.stdout = redirected_output
|
|
|
|
try:
|
|
# 执行函数
|
|
print("\n执行代码...")
|
|
result = namespace["neo4j_find_function"]()
|
|
|
|
# 获取捕获的输出
|
|
output = redirected_output.getvalue().strip()
|
|
print(f"\n原始输出:\n{output}")
|
|
|
|
# 检查结果是否为空
|
|
is_empty_result = (
|
|
not output
|
|
or output.lower() == "none"
|
|
or output == "[]"
|
|
or "未找到" in output
|
|
or "[]" in output
|
|
or "None" in output
|
|
or result is None
|
|
)
|
|
|
|
# 如果结果为空，走重写流程
|
|
if is_empty_result:
|
|
print("\n查询未找到结果，尝试定位具体缺失节点...")
|
|
|
|
# 解析原始查询路径中的最后一个节点名
|
|
import re
|
|
|
|
match = re.search(r"【([^】]+)】\s*$", original_query)
|
|
missing_node = match.group(1) if match else "未知节点"
|
|
|
|
error_info = {
|
|
"error_type": "NodeNotFoundError",
|
|
"error_message": f"{missing_node} 未找到，请检查该节点是否存在。",
|
|
"missing_node": missing_node,
|
|
"original_query": original_query,
|
|
"executed_code": original_code,
|
|
}
|
|
|
|
print("结构化错误信息：")
|
|
print(json.dumps(error_info, ensure_ascii=False, indent=2))
|
|
|
|
if current_retry < max_retries:
|
|
print("\n尝试使用RAG重写查询...")
|
|
try:
|
|
# 使用RAG重写查询和代码，并传递错误信息
|
|
rewritten = rewrite_query_parameters(original_query, original_code, error_info)
|
|
|
|
print(f"\nRAG重写结果: {json.dumps(rewritten, ensure_ascii=False, indent=2)}")
|
|
|
|
# 更新查询和代码
|
|
if "query" in rewritten and "code" in rewritten and rewritten["code"] != original_code:
|
|
print("\nRAG重写成功，使用新代码重试...")
|
|
input_value = rewritten["code"] # 直接使用重写后的代码
|
|
input_type = "code" # 切换到代码模式
|
|
current_retry += 1
|
|
continue # 继续下一次循环
|
|
else:
|
|
print("\nRAG重写未产生新代码，返回原始错误")
|
|
except Exception as e:
|
|
print(f"\nRAG重写失败: {e}")
|
|
# 记录错误但继续执行
|
|
|
|
# RAG重写失败或未产生新代码，返回原始错误
|
|
query_status = (
|
|
"第一次查询失败，RAG重写也失败"
|
|
if current_retry == 0
|
|
else f"第{current_retry+1}次查询失败，RAG重写也失败"
|
|
)
|
|
print(f"\n{query_status}")
|
|
return {
|
|
"code": 1,
|
|
"message": f"{missing_node} 未找到，请检查该节点是否存在。",
|
|
"data": {"value": "", "code": original_code},
|
|
"error_info": error_info,
|
|
"query_status": query_status,
|
|
}
|
|
|
|
# 清理输出，只保留有用的结果部分
|
|
clean_output = output
|
|
|
|
# 如果输出包含查询结果数量和对象引用
|
|
if "查询结果数量:" in output and "<project." in output:
|
|
# 提取查询结果部分
|
|
import re
|
|
|
|
# 尝试提取节点属性
|
|
node_match = re.search(r"找到节点: <Node.*?properties=({.*?})>", output, re.DOTALL)
|
|
if node_match:
|
|
props_str = node_match.group(1).replace("'", '"')
|
|
try:
|
|
import ast
|
|
|
|
props = ast.literal_eval(props_str)
|
|
clean_output = json.dumps(props, ensure_ascii=False, indent=2)
|
|
except:
|
|
pass
|
|
|
|
# 如果有查询结果数量信息
|
|
count_match = re.search(r"查询结果数量: (\d+)", output)
|
|
if count_match:
|
|
count = count_match.group(1)
|
|
if count == "0":
|
|
clean_output = "未找到匹配的数据。"
|
|
is_empty_result = True
|
|
elif not node_match: # 如果没有提取到节点属性但有结果
|
|
clean_output = f"找到 {count} 条匹配结果"
|
|
|
|
# 检查结果对象
|
|
if result is not None:
|
|
if isinstance(result, list):
|
|
if not result: # 空列表
|
|
is_empty_result = True
|
|
else:
|
|
# 处理非空列表
|
|
formatted_items = []
|
|
for item in result:
|
|
if hasattr(item, "__dict__"):
|
|
# 提取对象的所有属性
|
|
attrs = {k: v for k, v in item.__dict__.items() if not k.startswith("_")}
|
|
formatted_items.append(attrs)
|
|
else:
|
|
formatted_items.append(str(item))
|
|
|
|
if not is_empty_result: # 只有在不是空结果时才返回成功
|
|
query_status = (
|
|
"第一次查询成功"
|
|
if current_retry == 0
|
|
else f"第{current_retry+1}次查询成功(RAG重写后)"
|
|
)
|
|
print(f"\n{query_status}")
|
|
return {
|
|
"code": 0,
|
|
"message": "成功",
|
|
"data": {
|
|
"value": json.dumps(formatted_items, ensure_ascii=False, indent=2),
|
|
"code": original_code,
|
|
},
|
|
"query_status": query_status,
|
|
}
|
|
elif hasattr(result, "__dict__"):
|
|
# 单个对象
|
|
attrs = {k: v for k, v in result.__dict__.items() if not k.startswith("_")}
|
|
|
|
if not is_empty_result: # 只有在不是空结果时才返回成功
|
|
query_status = (
|
|
"第一次查询成功" if current_retry == 0 else f"第{current_retry+1}次查询成功(RAG重写后)"
|
|
)
|
|
print(f"\n{query_status}")
|
|
return {
|
|
"code": 0,
|
|
"message": "成功",
|
|
"data": {
|
|
"value": json.dumps(attrs, ensure_ascii=False, indent=2),
|
|
"code": original_code,
|
|
},
|
|
"query_status": query_status,
|
|
}
|
|
|
|
# 如果没有对象属性但有清理后的输出，且不是空结果
|
|
if (
|
|
clean_output
|
|
and clean_output.lower() != "none"
|
|
and clean_output != "[]"
|
|
and "未找到" not in clean_output
|
|
and not is_empty_result
|
|
):
|
|
query_status = (
|
|
"第一次查询成功" if current_retry == 0 else f"第{current_retry+1}次查询成功(RAG重写后)"
|
|
)
|
|
print(f"\n{query_status}")
|
|
return {
|
|
"code": 0,
|
|
"message": "成功",
|
|
"data": {"value": clean_output, "code": original_code},
|
|
"query_status": query_status,
|
|
}
|
|
|
|
finally:
|
|
# 恢复stdout
|
|
sys.stdout = old_stdout
|
|
|
|
except Exception as e:
|
|
import traceback
|
|
|
|
error_details = traceback.format_exc()
|
|
print(f"\n执行代码时出错: {error_details}")
|
|
|
|
# 如果走到这里，说明结果为空或未找到匹配项，应该执行RAG重写流程
|
|
print("\n查询未找到结果，尝试定位具体缺失节点...")
|
|
|
|
# 解析原始查询路径中的最后一个节点名
|
|
import re
|
|
|
|
match = re.search(r"【([^】]+)】\s*$", original_query)
|
|
missing_node = match.group(1) if match else "未知节点"
|
|
|
|
error_info = {
|
|
"error_type": "NodeNotFoundError",
|
|
"error_message": f"{missing_node} 未找到，请检查该节点是否存在。",
|
|
"missing_node": missing_node,
|
|
"original_query": original_query,
|
|
"executed_code": original_code,
|
|
}
|
|
|
|
print("结构化错误信息：")
|
|
print(json.dumps(error_info, ensure_ascii=False, indent=2))
|
|
|
|
if current_retry < max_retries:
|
|
print("\n尝试使用RAG重写查询...")
|
|
try:
|
|
# 使用RAG重写查询和代码，并传递错误信息
|
|
rewritten = rewrite_query_parameters(original_query, original_code, error_info)
|
|
|
|
print(f"\nRAG重写结果: {json.dumps(rewritten, ensure_ascii=False, indent=2)}")
|
|
|
|
# 更新查询和代码
|
|
if "query" in rewritten and "code" in rewritten and rewritten["code"] != original_code:
|
|
print("\nRAG重写成功，使用新代码重试...")
|
|
input_value = rewritten["code"] # 直接使用重写后的代码
|
|
input_type = "code" # 切换到代码模式
|
|
current_retry += 1
|
|
continue # 继续下一次循环
|
|
else:
|
|
print("\nRAG重写未产生新代码，返回原始错误")
|
|
except Exception as e:
|
|
print(f"\nRAG重写失败: {e}")
|
|
# 记录错误但继续执行
|
|
|
|
# RAG重写失败或未产生新代码，返回原始错误
|
|
query_status = (
|
|
"第一次查询失败，RAG重写也失败"
|
|
if current_retry == 0
|
|
else f"第{current_retry+1}次查询失败，RAG重写也失败"
|
|
)
|
|
print(f"\n{query_status}")
|
|
return {
|
|
"code": 1,
|
|
"message": f"{missing_node} 未找到，请检查该节点是否存在。",
|
|
"data": {"value": "", "code": original_code},
|
|
"error_info": error_info,
|
|
"query_status": query_status,
|
|
}
|
|
|
|
# 如果所有重试都失败
|
|
print("\n所有重试都失败，无法找到匹配的结果")
|
|
query_status = "所有重试都失败"
|
|
return {
|
|
"code": 1,
|
|
"message": "所有重试都失败，无法找到匹配的结果",
|
|
"data": {"value": "", "code": original_code},
|
|
"query_status": query_status,
|
|
}
|
|
|
|
|
|
def format_result(result):
|
|
"""
|
|
格式化查询结果
|
|
|
|
Args:
|
|
result: 查询结果（可能为 list、dict 或其他类型）
|
|
|
|
Returns:
|
|
str: 格式化后的结果
|
|
"""
|
|
# 处理 project 对象
|
|
if hasattr(result, "__module__") and result.__module__ == "project":
|
|
# 这是一个 project 模块中的对象
|
|
attrs = {k: v for k, v in result.__dict__.items() if not k.startswith("_")}
|
|
return json.dumps(attrs, ensure_ascii=False, indent=2)
|
|
|
|
# 处理 project 对象列表
|
|
if isinstance(result, list) and all(
|
|
hasattr(item, "__module__") and item.__module__ == "project" for item in result if hasattr(item, "__module__")
|
|
):
|
|
formatted_items = []
|
|
for item in result:
|
|
if hasattr(item, "__dict__"):
|
|
attrs = {k: v for k, v in item.__dict__.items() if not k.startswith("_")}
|
|
formatted_items.append(attrs)
|
|
else:
|
|
formatted_items.append(str(item))
|
|
|
|
return json.dumps(formatted_items, ensure_ascii=False, indent=2)
|
|
|
|
# 如果结果是字符串，可能包含调试信息，需要提取有用部分
|
|
if isinstance(result, str):
|
|
# 尝试提取最终结果部分
|
|
if "[]" in result:
|
|
return "未找到匹配的数据。"
|
|
|
|
# 如果包含节点信息，提取关键部分
|
|
import re
|
|
|
|
node_match = re.search(r"找到.*?labels=.*?properties=(.*?)>", result)
|
|
if node_match:
|
|
try:
|
|
# 提取属性部分并格式化
|
|
props_str = node_match.group(1).replace("'", '"')
|
|
import ast
|
|
|
|
props = ast.literal_eval(props_str)
|
|
|
|
formatted = "找到节点:\n"
|
|
for k, v in props.items():
|
|
formatted += f" {k}: {v}\n"
|
|
return formatted
|
|
except:
|
|
pass
|
|
|
|
# 如果包含查询结果数量
|
|
count_match = re.search(r"查询结果数量: (\d+)", result)
|
|
if count_match:
|
|
count = count_match.group(1)
|
|
if count == "0":
|
|
return "未找到匹配的数据。"
|
|
|
|
# 如果是列表
|
|
if isinstance(result, list):
|
|
if not result:
|
|
return "未找到匹配的数据。"
|
|
|
|
lines = [f"找到 {len(result)} 条匹配结果:"]
|
|
for i, item in enumerate(result, 1):
|
|
lines.append(f"\n结果 {i}:")
|
|
if hasattr(item, "items"): # 检查是否有items方法（字典或类似字典的对象）
|
|
try:
|
|
for k, v in item.items():
|
|
lines.append(f" {k}: {v}")
|
|
except:
|
|
lines.append(f" {item}")
|
|
else:
|
|
lines.append(f" {item}")
|
|
return "\n".join(lines)
|
|
|
|
# 如果是字典
|
|
elif isinstance(result, dict):
|
|
lines = ["查询结果:"]
|
|
for k, v in result.items():
|
|
lines.append(f" {k}: {v}")
|
|
return "\n".join(lines)
|
|
|
|
# 其他类型
|
|
else:
|
|
return str(result)
|
|
|
|
|
|
def format_dict_or_item(item):
|
|
"""
|
|
格式化字典或其他对象
|
|
|
|
Args:
|
|
item: 字典或其他对象
|
|
|
|
Returns:
|
|
str: 格式化后的字符串
|
|
"""
|
|
if isinstance(item, dict):
|
|
formatted = ""
|
|
for key, value in item.items():
|
|
formatted += f" {key}: {value}\n"
|
|
return formatted
|
|
|
|
return str(item)
|
|
|
|
|
|
question = {
|
|
"type": "query",
|
|
"value": "查找一下【工程数据/安装工程/安装/架空输电线路本体工程/杆塔工程/杆塔组立/铁塔、钢管杆组立】的类型为【主材】的【阿巴阿巴】",
|
|
}
|
|
result = nl_query_to_function_call(question)
|
|
print(result)
|