250 lines
8.3 KiB
Python
250 lines
8.3 KiB
Python
"""新浪财经数据客户端。
|
||
|
||
作为东方财富/腾讯的兜底数据源：
|
||
- 分钟 K 线：用于盘中量能证据
|
||
- 实时行情：可按股票列表批量拉取
|
||
|
||
新浪接口属于公开网页行情接口，字段稳定性弱于正式数据服务，因此只作为 fallback。
|
||
"""
|
||
|
||
import json
|
||
import logging
|
||
import re
|
||
from collections import defaultdict
|
||
|
||
import httpx
|
||
import pandas as pd
|
||
|
||
from app.data.cache import cache
|
||
from app.data.models import StockQuote
|
||
|
||
logger = logging.getLogger(__name__)
|
||
|
||
SINA_KLINE_URL = "https://quotes.sina.cn/cn/api/json_v2.php/CN_MarketDataService.getKLineData"
|
||
SINA_QUOTE_URL = "https://hq.sinajs.cn/list="
|
||
HEADERS = {
|
||
"Referer": "https://finance.sina.com.cn",
|
||
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36",
|
||
}
|
||
|
||
|
||
def _ts_code_to_sina(ts_code: str) -> str:
|
||
code, market = ts_code.split(".")
|
||
return f"{market.lower()}{code}"
|
||
|
||
|
||
async def get_min_kline(ts_code: str, period: str = "5", count: int = 48) -> pd.DataFrame:
|
||
"""获取新浪分钟 K 线，返回与东方财富分钟线一致的列。"""
|
||
scale = period if period in {"5", "15", "30", "60"} else "5"
|
||
cache_key = f"sina_min_kline:{ts_code}:{scale}:{count}"
|
||
cached = cache.get(cache_key)
|
||
if cached is not None:
|
||
return cached
|
||
|
||
params = {
|
||
"symbol": _ts_code_to_sina(ts_code),
|
||
"scale": scale,
|
||
"ma": "no",
|
||
"datalen": str(count),
|
||
}
|
||
|
||
try:
|
||
async with httpx.AsyncClient(timeout=10, follow_redirects=True) as client:
|
||
resp = await client.get(SINA_KLINE_URL, params=params, headers=HEADERS)
|
||
resp.raise_for_status()
|
||
|
||
rows = _parse_kline_payload(resp.text)
|
||
if not rows:
|
||
return pd.DataFrame()
|
||
|
||
result = []
|
||
for item in rows[-count:]:
|
||
amount = item.get("amount")
|
||
volume = item.get("volume", 0)
|
||
result.append({
|
||
"time": item.get("day", ""),
|
||
"open": float(item.get("open", 0) or 0),
|
||
"close": float(item.get("close", 0) or 0),
|
||
"high": float(item.get("high", 0) or 0),
|
||
"low": float(item.get("low", 0) or 0),
|
||
"volume": float(volume or 0),
|
||
# 新浪分钟线常见字段只有 volume；量能分布分析需要 amount 时用 volume 兜底。
|
||
"amount": float(amount if amount not in (None, "") else volume or 0),
|
||
})
|
||
|
||
df = pd.DataFrame(result)
|
||
cache.set(cache_key, df, ttl=180)
|
||
return df
|
||
except Exception as e:
|
||
logger.warning("新浪分钟K线获取失败 %s: %s", ts_code, e)
|
||
return pd.DataFrame()
|
||
|
||
|
||
async def get_realtime_quotes_batch(ts_codes: list[str]) -> dict[str, StockQuote]:
|
||
"""批量获取新浪实时行情。"""
|
||
results: dict[str, StockQuote] = {}
|
||
if not ts_codes:
|
||
return results
|
||
|
||
batch_size = 80
|
||
for i in range(0, len(ts_codes), batch_size):
|
||
batch = ts_codes[i:i + batch_size]
|
||
symbols = ",".join(_ts_code_to_sina(code) for code in batch)
|
||
try:
|
||
async with httpx.AsyncClient(timeout=10, follow_redirects=True) as client:
|
||
resp = await client.get(f"{SINA_QUOTE_URL}{symbols}", headers=HEADERS)
|
||
resp.raise_for_status()
|
||
resp.encoding = "gbk"
|
||
|
||
quote_map = _parse_quote_payload(resp.text)
|
||
for ts_code in batch:
|
||
data = quote_map.get(_ts_code_to_sina(ts_code))
|
||
quote = _quote_from_fields(ts_code, data)
|
||
if quote:
|
||
results[ts_code] = quote
|
||
except Exception as e:
|
||
logger.warning("新浪批量行情获取失败: %s", e)
|
||
|
||
return results
|
||
|
||
|
||
async def get_sector_realtime_ranking_by_industry(limit: int = 20) -> list[dict]:
|
||
"""用新浪全市场实时行情 + Tushare 静态行业映射聚合今日行业榜。
|
||
|
||
这是东方财富板块榜不可用时的 fallback，不依赖 Tushare 当日行情。
|
||
"""
|
||
from app.data.tushare_client import tushare_client
|
||
|
||
cache_key = f"sina_sector_industry:{limit}"
|
||
cached = cache.get(cache_key)
|
||
if cached is not None:
|
||
return cached
|
||
|
||
stock_basic = tushare_client.get_stock_basic()
|
||
if stock_basic.empty or "industry" not in stock_basic.columns:
|
||
logger.warning("新浪行业榜兜底失败: 股票静态行业映射为空")
|
||
return []
|
||
|
||
stock_basic = stock_basic.dropna(subset=["industry"])
|
||
code_to_industry = {}
|
||
code_to_name = {}
|
||
for _, row in stock_basic.iterrows():
|
||
ts_code = row.get("ts_code")
|
||
industry = row.get("industry")
|
||
if not ts_code or not industry:
|
||
continue
|
||
code_to_industry[ts_code] = str(industry)
|
||
code_to_name[ts_code] = str(row.get("name") or ts_code)
|
||
|
||
quotes = await get_realtime_quotes_batch(list(code_to_industry.keys()))
|
||
if not quotes:
|
||
return []
|
||
|
||
grouped = defaultdict(lambda: {
|
||
"amount": 0.0,
|
||
"pct_sum": 0.0,
|
||
"count": 0,
|
||
"up_count": 0,
|
||
"down_count": 0,
|
||
"leading": None,
|
||
})
|
||
|
||
for ts_code, quote in quotes.items():
|
||
industry = code_to_industry.get(ts_code)
|
||
if not industry:
|
||
continue
|
||
bucket = grouped[industry]
|
||
bucket["amount"] += quote.amount or 0
|
||
bucket["pct_sum"] += quote.pct_chg
|
||
bucket["count"] += 1
|
||
if quote.pct_chg > 0:
|
||
bucket["up_count"] += 1
|
||
elif quote.pct_chg < 0:
|
||
bucket["down_count"] += 1
|
||
leading = bucket["leading"]
|
||
if leading is None or quote.pct_chg > leading["pct_chg"]:
|
||
bucket["leading"] = {
|
||
"ts_code": ts_code,
|
||
"name": quote.name or code_to_name.get(ts_code, ts_code),
|
||
"pct_chg": quote.pct_chg,
|
||
}
|
||
|
||
result = []
|
||
for idx, (industry, bucket) in enumerate(grouped.items(), start=1):
|
||
count = bucket["count"] or 1
|
||
leading = bucket["leading"] or {}
|
||
result.append({
|
||
"sector_code": f"SINA_{idx:03d}",
|
||
"sector_name": industry,
|
||
"board_type": "industry",
|
||
"pct_change": round(bucket["pct_sum"] / count, 2),
|
||
"amount": round(bucket["amount"], 2),
|
||
"turnover_rate": 0,
|
||
"up_count": int(bucket["up_count"]),
|
||
"down_count": int(bucket["down_count"]),
|
||
"leading_stock_name": leading.get("name", ""),
|
||
"leading_stock_code": leading.get("ts_code", ""),
|
||
"leading_stock_pct": float(leading.get("pct_chg", 0) or 0),
|
||
"source": "sina",
|
||
})
|
||
|
||
result.sort(key=lambda item: (item["pct_change"], item["amount"]), reverse=True)
|
||
result = result[:limit]
|
||
cache.set(cache_key, result, ttl=180)
|
||
logger.info("新浪行业实时榜兜底: 获取 %s 个行业", len(result))
|
||
return result
|
||
|
||
|
||
def _parse_kline_payload(text: str) -> list[dict]:
|
||
text = (text or "").strip()
|
||
if not text:
|
||
return []
|
||
if text.startswith("["):
|
||
return json.loads(text)
|
||
start = text.find("[")
|
||
end = text.rfind("]")
|
||
if start >= 0 and end > start:
|
||
return json.loads(text[start:end + 1])
|
||
return []
|
||
|
||
|
||
def _parse_quote_payload(text: str) -> dict[str, list[str]]:
|
||
result: dict[str, list[str]] = {}
|
||
pattern = re.compile(r'var hq_str_([^=]+)="([^"]*)";')
|
||
for symbol, payload in pattern.findall(text or ""):
|
||
if payload:
|
||
result[symbol] = payload.split(",")
|
||
return result
|
||
|
||
|
||
def _quote_from_fields(ts_code: str, fields: list[str] | None) -> StockQuote | None:
|
||
if not fields or len(fields) < 32:
|
||
return None
|
||
try:
|
||
name = fields[0]
|
||
open_price = float(fields[1] or 0)
|
||
pre_close = float(fields[2] or 0)
|
||
price = float(fields[3] or 0)
|
||
high = float(fields[4] or 0)
|
||
low = float(fields[5] or 0)
|
||
volume = float(fields[8] or 0)
|
||
amount = float(fields[9] or 0) / 10000
|
||
pct_chg = ((price - pre_close) / pre_close * 100) if pre_close else 0
|
||
if price <= 0:
|
||
return None
|
||
return StockQuote(
|
||
ts_code=ts_code,
|
||
name=name,
|
||
price=price,
|
||
pct_chg=round(pct_chg, 2),
|
||
volume=volume,
|
||
amount=amount,
|
||
turnover_rate=0,
|
||
high=high,
|
||
low=low,
|
||
open=open_price,
|
||
pre_close=pre_close,
|
||
)
|
||
except (IndexError, ValueError):
|
||
return None
|