60 lines
1.8 KiB
Python
60 lines
1.8 KiB
Python
import json
|
|
from typing import Any, Dict, Optional
|
|
|
|
import httpx
|
|
|
|
from config import Config
|
|
|
|
|
|
class RagflowClient:
|
|
"""RAGFlow 客户端（仅检索）"""
|
|
|
|
def __init__(self):
|
|
cfg = Config.get_section("ragflow")
|
|
self._base_url = cfg.get("url", "")
|
|
self._api_key = cfg.get("api_key", "")
|
|
self._retrieval_path = cfg.get("retrieval", "/api/v1/retrieval")
|
|
self._dataset_ids = cfg.get("dataset_ids", "")
|
|
|
|
def _build_url(self) -> str:
|
|
return self._base_url.rstrip("/") + "/" + self._retrieval_path.lstrip("/")
|
|
|
|
def retrieve(self, query: str, top_k: int = 3) -> Dict[str, Any]:
|
|
"""检索匹配文档"""
|
|
if not self._base_url or not self._retrieval_path:
|
|
raise RuntimeError("未配置 ragflow.url 或 ragflow.retrieval")
|
|
url = self._build_url()
|
|
headers = {"Authorization": f"Bearer {self._api_key}"} if self._api_key else {}
|
|
payload = {
|
|
"dataset_ids": self._dataset_ids,
|
|
"query": query,
|
|
"top_k": top_k,
|
|
}
|
|
|
|
with httpx.Client(timeout=30) as client:
|
|
response = client.post(url, json=payload, headers=headers)
|
|
response.raise_for_status()
|
|
return response.json()
|
|
|
|
|
|
def extract_table_name(record: Dict[str, Any]) -> Optional[str]:
|
|
"""从检索结果中提取表名"""
|
|
if not record:
|
|
return None
|
|
|
|
metadata = record.get("metadata") or {}
|
|
for key in ("table", "table_name"):
|
|
if key in metadata:
|
|
return metadata.get(key)
|
|
|
|
for key in ("table", "table_name"):
|
|
if key in record:
|
|
return record.get(key)
|
|
|
|
content = record.get("content") or record.get("text") or ""
|
|
for line in str(content).splitlines():
|
|
if line.lower().startswith("table:"):
|
|
return line.split(":", 1)[1].strip()
|
|
|
|
return None
|