|
|
import re
|
|
|
from typing import List, Dict, Any
|
|
|
import requests
|
|
|
from vanna.base import VannaBase
|
|
|
from Config import *
|
|
|
|
|
|
class DeepSeekVanna(VannaBase):
|
|
|
def __init__(self):
|
|
|
super().__init__()
|
|
|
self.api_key = MODEL_API_KEY
|
|
|
self.base_url = "https://dashscope.aliyuncs.com/api/v1/services/aigc/text-generation/generation" # 阿里云专用API地址
|
|
|
self.model = QWEN_MODEL_NAME # 根据实际模型名称调整
|
|
|
self.training_data = []
|
|
|
self.chat_history = []
|
|
|
|
|
|
# ---------- 必须实现的抽象方法 ----------
|
|
|
def add_ddl(self, ddl: str, **kwargs) -> None:
|
|
|
self.training_data.append({"type": "ddl", "content": ddl})
|
|
|
|
|
|
def add_documentation(self, doc: str, **kwargs) -> None:
|
|
|
self.training_data.append({"type": "documentation", "content": doc})
|
|
|
|
|
|
def add_question_sql(self, question: str, sql: str, **kwargs) -> None:
|
|
|
self.training_data.append({"type": "qa", "question": question, "sql": sql})
|
|
|
|
|
|
def get_related_ddl(self, question: str, **kwargs) -> str:
|
|
|
return "\n".join([item["content"] for item in self.training_data if item["type"] == "ddl"])
|
|
|
|
|
|
def get_related_documentation(self, question: str, **kwargs) -> str:
|
|
|
return "\n".join([item["content"] for item in self.training_data if item["type"] == "documentation"])
|
|
|
|
|
|
def get_training_data(self, **kwargs) -> List[Dict[str, Any]]:
|
|
|
return self.training_data
|
|
|
|
|
|
# ---------- 对话方法 ----------
|
|
|
def system_message(self, message: str) -> None:
|
|
|
self.chat_history = [{"role": "system", "content": message}]
|
|
|
|
|
|
def user_message(self, message: str) -> None:
|
|
|
self.chat_history.append({"role": "user", "content": message})
|
|
|
|
|
|
def assistant_message(self, message: str) -> None:
|
|
|
self.chat_history.append({"role": "assistant", "content": message})
|
|
|
|
|
|
def submit_prompt(self, prompt: str, **kwargs) -> str:
|
|
|
return self.generate_sql(question=prompt)
|
|
|
|
|
|
# ---------- 其他方法 ----------
|
|
|
def remove_training_data(self, id: str, **kwargs) -> bool:
|
|
|
return True
|
|
|
|
|
|
def generate_embedding(self, text: str, **kwargs) -> List[float]:
|
|
|
return []
|
|
|
|
|
|
def get_similar_question_sql(self, question: str, **kwargs) -> List[Dict[str, Any]]:
|
|
|
return []
|
|
|
|
|
|
|
|
|
|
|
|
def _clean_sql_output(self, raw_sql: str) -> str:
|
|
|
"""增强版清洗逻辑"""
|
|
|
# 移除所有非SQL内容
|
|
|
cleaned = re.sub(r'^.*?(?=SELECT)', '', raw_sql, flags=re.IGNORECASE|re.DOTALL)
|
|
|
# 提取第一个完整SQL语句
|
|
|
match = re.search(r'(SELECT\s.+?;)', cleaned, re.IGNORECASE|re.DOTALL)
|
|
|
if match:
|
|
|
# 标准化空格和换行
|
|
|
clean_sql = re.sub(r'\s+', ' ', match.group(1)).strip()
|
|
|
# 确保没有重复SELECT
|
|
|
clean_sql = re.sub(r'(SELECT\s+)+', 'SELECT ', clean_sql, flags=re.IGNORECASE)
|
|
|
return clean_sql
|
|
|
return raw_sql
|
|
|
|
|
|
def _build_sql_prompt(self, question: str) -> str:
|
|
|
"""强化提示词"""
|
|
|
return f"""严格按以下要求生成Postgresql查询：
|
|
|
|
|
|
表结构：
|
|
|
{self.get_related_ddl(question)}
|
|
|
|
|
|
问题：{question}
|
|
|
|
|
|
生成规则：
|
|
|
1. 只输出一个标准的SELECT语句
|
|
|
2. 绝对不要使用任何代码块标记
|
|
|
3. 语句以分号结尾
|
|
|
4. 不要包含任何解释或注释
|
|
|
5. 若需要多表查询，使用显式JOIN语法
|
|
|
6. 确保没有重复的SELECT关键字
|
|
|
"""
|
|
|
|
|
|
def generate_sql(self, question: str, **kwargs) -> str:
|
|
|
"""同步生成SQL"""
|
|
|
try:
|
|
|
headers = {
|
|
|
"Authorization": f"Bearer {self.api_key}",
|
|
|
"Content-Type": "application/json"
|
|
|
}
|
|
|
|
|
|
data = {
|
|
|
"model": self.model,
|
|
|
"input": {
|
|
|
"messages": [{
|
|
|
"role": "user",
|
|
|
"content": self._build_sql_prompt(question)
|
|
|
}]
|
|
|
},
|
|
|
"parameters": {
|
|
|
"temperature": 0.1,
|
|
|
"max_tokens": 5000,
|
|
|
"result_format": "text"
|
|
|
}
|
|
|
}
|
|
|
|
|
|
response = requests.post(self.base_url, headers=headers, json=data)
|
|
|
response.raise_for_status()
|
|
|
|
|
|
raw_sql = response.json()['output']['text']
|
|
|
return self._clean_sql_output(raw_sql)
|
|
|
|
|
|
except Exception as e:
|
|
|
print(f"\nAPI请求错误: {str(e)}")
|
|
|
return ""
|
