hmo
ab23dfd234
mo_bridge.py (rewrite): - get_stock_news(): DSA SearchService 7 engines → MoFin analysis context - get_market_review(): DSA run_market_review() with 24h cache - get_stock_analysis(): DSA AgentExecutor.run() with 15 strategies - enrich_analysis_context(): one-call context injection strategy_lifecycle.py: - reassess_with_context() now injects DSA market + news context - Auto-detects HK vs A-share region for market review - Graceful fallback if DSA unavailable
291 lines
10 KiB
Python
291 lines
10 KiB
Python
#!/usr/bin/env python3
|
||
"""
|
||
mo_bridge.py — MoFin ↔ DSA 全功能集成桥
|
||
|
||
真正调用 DSA 的三大功能：
|
||
1. 新闻搜索 — SearchService.search_comprehensive_intel()（7 个搜索引擎）
|
||
2. 大盘复盘 — run_market_review()（A股/港股/美股）
|
||
3. 策略问股 — AgentExecutor.run()（DSA 的 15 种策略，作为 MoFin 的第二意见）
|
||
|
||
用法：
|
||
from mo_bridge import enrich_analysis_context, get_stock_analysis
|
||
|
||
# 在 MoFin 分析前注入 DSA 上下文（大盘 + 新闻）
|
||
ctx = enrich_analysis_context("00700", "腾讯控股", region="hk")
|
||
prompt += f"\n\n{ctx}"
|
||
|
||
# 用 DSA 的策略做独立分析
|
||
opinion = get_stock_analysis("600519", "贵州茅台", skills=["ma_golden_cross"])
|
||
"""
|
||
|
||
import sys
|
||
import os
|
||
import json
|
||
import logging
|
||
from pathlib import Path
|
||
from datetime import datetime
|
||
|
||
logger = logging.getLogger(__name__)
|
||
|
||
# ── DSA 路径 ─────────────────────────────────────────────────────────
|
||
|
||
_DSA_CANDIDATES = [
|
||
"/home/hmo/daily-stock-analysis",
|
||
str(Path(__file__).resolve().parent.parent / "daily-stock-analysis" / "ZhuLinsen-daily_stock_analysis-a448886"),
|
||
]
|
||
|
||
_DSA_BASE = None
|
||
for _c in _DSA_CANDIDATES:
|
||
_p = Path(_c)
|
||
if _p.is_dir() and (_p / "data_provider" / "base.py").exists():
|
||
_DSA_BASE = _p
|
||
break
|
||
|
||
_HAS_DSA = _DSA_BASE is not None
|
||
|
||
if _HAS_DSA:
|
||
sys.path.insert(0, str(_DSA_BASE))
|
||
|
||
# ── 懒加载 DSA 模块 ──────────────────────────────────────────────────
|
||
|
||
_dsa_search_service = None
|
||
_dsa_config = None
|
||
|
||
|
||
def _ensure_dsa_search():
|
||
global _dsa_search_service
|
||
if _dsa_search_service is not None:
|
||
return _dsa_search_service
|
||
if not _HAS_DSA:
|
||
return None
|
||
try:
|
||
from src.search_service import get_search_service
|
||
_dsa_search_service = get_search_service()
|
||
except Exception as e:
|
||
logger.warning("DSA SearchService 加载失败: %s", e)
|
||
return _dsa_search_service
|
||
|
||
|
||
def _ensure_dsa_config():
|
||
global _dsa_config
|
||
if _dsa_config is not None:
|
||
return _dsa_config
|
||
if not _HAS_DSA:
|
||
return None
|
||
try:
|
||
from src.config import get_config
|
||
_dsa_config = get_config()
|
||
except Exception as e:
|
||
logger.warning("DSA Config 加载失败: %s", e)
|
||
return _dsa_config
|
||
|
||
|
||
# ── 1. 新闻搜索 ─────────────────────────────────────────────────────
|
||
|
||
def get_stock_news(stock_code: str, stock_name: str = "", max_results: int = 5) -> str:
|
||
"""通过 DSA 的 7 个搜索引擎获取股票相关新闻。
|
||
|
||
Args:
|
||
stock_code: 股票代码 (如 '600519', '00700', 'AAPL')
|
||
stock_name: 股票名称 (提高搜索精度)
|
||
max_results: 最多返回条数
|
||
|
||
Returns:
|
||
str: Markdown 格式新闻摘要，可直接注入分析 prompt。失败时返回 ''。
|
||
"""
|
||
service = _ensure_dsa_search()
|
||
if not service:
|
||
return ""
|
||
|
||
try:
|
||
intel = service.search_comprehensive_intel(
|
||
stock_code, stock_name or stock_code, max_searches=3
|
||
)
|
||
if not intel:
|
||
return ""
|
||
|
||
lines = [f"## 📰 {stock_name or stock_code} 最新情报"]
|
||
|
||
news = intel.get("latest_news")
|
||
if news and news.results:
|
||
lines.append("\n### 最新新闻")
|
||
for r in news.results[:max_results]:
|
||
date_str = f" ({r.published_date})" if r.published_date else ""
|
||
snippet = r.snippet[:150] if r.snippet else ""
|
||
lines.append(f"- **{r.title}**{date_str}: {snippet}")
|
||
|
||
risk = intel.get("risk_check")
|
||
if risk and risk.results:
|
||
lines.append("\n### ⚠️ 风险关注")
|
||
for r in risk.results[:3]:
|
||
lines.append(f"- {r.title}: {r.snippet[:100] if r.snippet else ''}")
|
||
|
||
return "\n".join(lines)
|
||
|
||
except Exception as e:
|
||
logger.warning("DSA 新闻搜索失败: %s", e)
|
||
return ""
|
||
|
||
|
||
# ── 2. 大盘复盘 ─────────────────────────────────────────────────────
|
||
|
||
def get_market_review(region: str = "cn") -> str:
|
||
"""获取 DSA 的大盘复盘报告。
|
||
优先读本地缓存（24h内），没有则调用 DSA 实时生成。
|
||
|
||
Args:
|
||
region: 'cn'=A股, 'hk'=港股, 'us'=美股, 'both'=全市场
|
||
|
||
Returns:
|
||
str: Markdown 格式大盘复盘摘要
|
||
"""
|
||
if not _HAS_DSA:
|
||
return ""
|
||
|
||
# 先读缓存
|
||
cache_dir = Path(str(_DSA_BASE)) / "data" / "market_review"
|
||
if cache_dir.exists():
|
||
try:
|
||
files = sorted(cache_dir.glob("*.md"), key=os.path.getmtime, reverse=True)
|
||
if files:
|
||
if (datetime.now().timestamp() - os.path.getmtime(str(files[0]))) < 86400:
|
||
content = files[0].read_text(encoding="utf-8")
|
||
lines = [l for l in content.split("\n")[:30] if len(l.strip()) > 3]
|
||
return "## 📈 今日大盘背景\n" + "\n".join(lines)
|
||
except Exception:
|
||
pass
|
||
|
||
# 实时调用 DSA
|
||
try:
|
||
from src.core.market_review import run_market_review
|
||
|
||
class StubNotifier:
|
||
def is_available(self): return False
|
||
def send(self, *a, **kw): return True
|
||
def save_report_to_file(self, *a, **kw): return None
|
||
|
||
config = _ensure_dsa_config()
|
||
result = run_market_review(
|
||
notifier=StubNotifier(), config=config,
|
||
override_region=region, send_notification=False,
|
||
save_report_file=False, persist_history=False,
|
||
trigger_source="mofin",
|
||
)
|
||
if result and isinstance(result, str):
|
||
lines = [l for l in result.split("\n")[:25] if len(l.strip()) > 3]
|
||
return "## 📈 今日大盘复盘\n" + "\n".join(lines)
|
||
except Exception as e:
|
||
logger.warning("DSA 大盘复盘失败: %s", e)
|
||
|
||
return ""
|
||
|
||
|
||
# ── 3. 策略问股（第二意见）───────────────────────────────────────────
|
||
|
||
def get_stock_analysis(
|
||
stock_code: str,
|
||
stock_name: str = "",
|
||
skills: list = None,
|
||
) -> dict | None:
|
||
"""用 DSA 的 15 种内置策略独立分析一只股票。
|
||
|
||
Args:
|
||
stock_code: 股票代码
|
||
stock_name: 股票名称
|
||
skills: 策略列表，默认 ['ma_golden_cross', 'bull_trend']
|
||
|
||
Returns:
|
||
dict: {source, sentiment_score, operation_advice, trend_prediction,
|
||
analysis_summary, risk_warning, strategies_used, raw}
|
||
"""
|
||
if not _HAS_DSA:
|
||
return None
|
||
|
||
if not skills:
|
||
skills = ["ma_golden_cross", "bull_trend"]
|
||
|
||
try:
|
||
from src.agent.factory import build_agent_executor
|
||
|
||
executor = build_agent_executor(skills=skills)
|
||
result = executor.run(
|
||
task=f"分析 {stock_code} {stock_name}",
|
||
context={"stock_code": stock_code, "stock_name": stock_name, "report_language": "zh"},
|
||
)
|
||
|
||
if result.success and result.dashboard:
|
||
d = result.dashboard
|
||
return {
|
||
"source": "DSA",
|
||
"sentiment_score": d.get("sentiment_score", 0),
|
||
"operation_advice": d.get("operation_advice", ""),
|
||
"trend_prediction": d.get("trend_prediction", ""),
|
||
"analysis_summary": d.get("analysis_summary", ""),
|
||
"risk_warning": d.get("risk_warning", ""),
|
||
"strategies_used": skills,
|
||
"raw": result.content[:500] if result.content else "",
|
||
}
|
||
except Exception as e:
|
||
logger.warning("DSA Agent 分析 %s 失败: %s", e)
|
||
|
||
return None
|
||
|
||
|
||
def get_strategy_opinion_text(opinion: dict) -> str:
|
||
"""将 get_stock_analysis() 的结果格式化为可读文本"""
|
||
if not opinion:
|
||
return ""
|
||
return (
|
||
f"## 🤖 DSA 策略参考\n"
|
||
f"- 评分: {opinion.get('sentiment_score', '?')}/100\n"
|
||
f"- 建议: {opinion.get('operation_advice', '?')}\n"
|
||
f"- 趋势: {opinion.get('trend_prediction', '?')}\n"
|
||
f"- 策略: {', '.join(opinion.get('strategies_used', []))}\n"
|
||
f"- 摘要: {opinion.get('analysis_summary', '')}\n"
|
||
f"- 风险: {opinion.get('risk_warning', '')}"
|
||
)
|
||
|
||
|
||
# ── 4. 综合上下文（一键调用）─────────────────────────────────────────
|
||
|
||
def enrich_analysis_context(
|
||
stock_code: str = "",
|
||
stock_name: str = "",
|
||
region: str = "cn",
|
||
include_news: bool = True,
|
||
include_market: bool = True,
|
||
) -> str:
|
||
"""一键获取 DSA 全部分析上下文，注入 MoFin 的 LLM prompt。
|
||
|
||
在 strategy_lifecycle.reassess_with_context() 或 Hermes cron job 的 prompt 前调用。
|
||
|
||
Returns:
|
||
str: 可直接拼接到 LLM prompt 的 Markdown 文本
|
||
"""
|
||
parts = []
|
||
|
||
if include_market:
|
||
market = get_market_review(region)
|
||
if market:
|
||
parts.append(market)
|
||
|
||
if include_news and stock_code:
|
||
news = get_stock_news(stock_code, stock_name)
|
||
if news:
|
||
parts.append(news)
|
||
|
||
return "\n\n".join(parts) if parts else ""
|
||
|
||
|
||
# ── 自检 ─────────────────────────────────────────────────────────────
|
||
|
||
if __name__ == "__main__":
|
||
print(f"DSA: {'available' if _HAS_DSA else 'NOT FOUND'} ({_DSA_BASE})")
|
||
if _HAS_DSA:
|
||
print("\n--- 新闻测试 (600519) ---")
|
||
n = get_stock_news("600519", "贵州茅台", max_results=2)
|
||
print(n[:300] if n else "(无结果)")
|
||
print("\n--- 大盘测试 ---")
|
||
m = get_market_review("cn")
|
||
print(m[:300] if m else "(无结果)")
|