Files
qlib/backend/tests/test_sina_provider.py
T
Simon 442999f701 feat(data): 财务/日线同步增量 + 新浪「两边一致」校验兜底 + 逐只进度
- financial 默认增量:按 A 股披露节奏判断已最新并跳过;--full 强制全量重拉
- Tushare fina_indicator 增加报告期窗口与 100 条/请求自动分页(修复老报告期静默截断)
- 新浪兜底收紧为校验兜底:两源重叠历史一致才导入缺失键,行标记 source=sina;
  财务可比字段取 eps/销售毛利率(ROE 两端口径不同不作依据),日线只比较最近重叠交易日
- CLI 输出逐只进度与导入内容描述(来源/行数/报告期与公告区间),失败股票留待重跑
- financial_indicator 增 source 列(迁移 d3f6c9a21b04);新增一致性/分页/服务测试
2026-09-08 21:48:09 +08:00

212 lines
7.3 KiB
Python

"""新浪 Provider 测试:代码转换、JSONP 解析、能力边界(不触网)。"""
from __future__ import annotations
import json
from datetime import date
import pytest
from app.infrastructure.data_sources.errors import (
DataSourceError,
DataSourceNotSupported,
)
from app.infrastructure.data_sources.sina import SinaProvider, _extract_jsonp, _to_sina_symbol
_KLINE_OK = (
'var data=[{"day":"2024-08-30","open":"1700.0","high":"1720.0","low":"1690.0",'
'"close":"1710.0","volume":"20000"},{"day":"2024-08-31","open":"1710.0",'
'"high":"1725.0","low":"1705.0","close":"1720.0","volume":"18000"}]'
)
class _FakeResp:
def __enter__(self):
return self
def __exit__(self, *exc):
return False
def read(self):
return _KLINE_OK.encode("utf-8")
class _BoomResp(_FakeResp):
def read(self):
raise OSError("socket timeout")
class TestSymbolMap:
def test_mapping(self) -> None:
assert _to_sina_symbol("600519.SH") == "sh600519"
assert _to_sina_symbol("000001.SZ") == "sz000001"
assert _to_sina_symbol("830001.BJ") == "bj830001"
class TestJsonp:
def test_extract(self) -> None:
rows = _extract_jsonp(_KLINE_OK)
assert len(rows) == 2
assert rows[0]["close"] == "1710.0"
def test_bad_payload_raises(self) -> None:
with pytest.raises(DataSourceError, match="无法解析"):
_extract_jsonp("not jsonp")
class TestGetDaily:
def test_ok_with_date_filter(self) -> None:
provider = SinaProvider(urlopen=lambda _url, **kw: _FakeResp())
bars = provider.get_daily("600519.SH", date(2024, 8, 30), date(2024, 8, 30))
assert len(bars) == 1
assert bars[0].close is not None
assert bars[0].symbol == "600519.SH"
def test_network_error_wrapped(self) -> None:
provider = SinaProvider(urlopen=lambda _url, **kw: _BoomResp())
with pytest.raises(DataSourceError, match="sina 请求失败"):
provider.get_daily("600519.SH", date(2024, 8, 1), date(2024, 8, 31))
class TestCapabilities:
def test_not_supported(self) -> None:
provider = SinaProvider()
with pytest.raises(DataSourceNotSupported):
provider.get_stock_basic()
with pytest.raises(DataSourceNotSupported):
provider.get_trade_cal(date(2024, 1, 1), date(2024, 1, 31))
with pytest.raises(DataSourceNotSupported):
provider.get_adjust_factor("600519.SH", date(2024, 1, 1), date(2024, 1, 31))
_FIN_PAYLOAD = json.dumps(
{
"result": {
"data": {
"report_date": [{"date_value": "2024-06-30"}, {"date_value": "2024-03-31"}],
"report_list": {
"2024-06-30": {
"publish_date": "2024-08-31",
"data": [
{"item_title": "基本每股收益", "item_value": "1.23"},
{"item_title": "净资产收益率(ROE)", "item_value": "15.5"},
{"item_title": "净利润", "item_value": "1230000000"},
],
},
"2024-03-31": {"publish_date": "", "data": []},
},
}
}
},
ensure_ascii=False,
)
class _FinResp:
def __enter__(self):
return self
def __exit__(self, *exc):
return False
def read(self):
return _FIN_PAYLOAD.encode("utf-8")
class TestFinancial:
def test_parse_schema_parity_with_tushare(self) -> None:
"""财务 schema 与 Tushare 对齐:report_date / announce_date / 指标。"""
provider = SinaProvider(urlopen=lambda _req, **kw: _FinResp())
rows = provider.get_financial("600519.SH")
assert len(rows) == 1 # 无 publish_date 的行被跳过(防未来函数)
row = rows[0]
assert row.symbol == "600519.SH"
assert row.report_date == date(2024, 6, 30)
assert row.announce_date == date(2024, 8, 31)
assert float(row.eps) == 1.23
assert float(row.roe) == 15.5
assert float(row.net_profit) == 1230000000
assert row.source == "sina" # 兜底行必须标记来源(与 Tushare 首选行区分)
# FinancialIndicator 校验能通过 → 格式与 Tushare 一致
assert row.model_dump()["eps"] is not None
def test_window_filter(self) -> None:
payload = json.dumps(
{
"result": {
"data": {
"report_date": [
{"date_value": "2024-06-30"},
{"date_value": "2024-03-31"},
],
"report_list": {
"2024-06-30": {
"publish_date": "2024-08-31",
"data": [{"item_title": "基本每股收益", "item_value": "1.23"}],
},
"2024-03-31": {
"publish_date": "2024-04-27",
"data": [{"item_title": "基本每股收益", "item_value": "0.5"}],
},
},
}
}
},
ensure_ascii=False,
)
class _TwoResp:
def __enter__(self):
return self
def __exit__(self, *exc):
return False
def read(self):
return payload.encode("utf-8")
provider = SinaProvider(urlopen=lambda _req, **kw: _TwoResp())
rows = provider.get_financial("600519.SH", date(2024, 6, 30), date(2024, 6, 30))
assert [r.report_date for r in rows] == [date(2024, 6, 30)]
def test_bad_structure_raises(self) -> None:
class _BadResp(_FinResp):
def read(self):
return b'{"result": {}}'
provider = SinaProvider(urlopen=lambda _req, **kw: _BadResp())
try:
provider.get_financial("600519.SH")
except DataSourceError:
return
raise AssertionError("应抛 DataSourceError")
class TestFormatParity:
"""两源格式一致:统一 DailyBar schema,新浪兜底行带 source/adjust 标记。"""
def test_sina_bar_marks_source_and_adjust(self) -> None:
provider = SinaProvider(urlopen=lambda _url, **kw: _FakeResp())
bars = provider.get_daily("600519.SH", date(2024, 8, 30), date(2024, 8, 30))
assert len(bars) == 1
bar = bars[0]
assert bar.source == "sina"
assert bar.adjust == "qfq" # 新浪为前复权,与 Tushare 不复权主口径区分
# 与 Tushare 同 schema:必备字段齐全
assert bar.symbol == "600519.SH"
assert bar.close is not None
class TestJsonpParens:
def test_real_world_paren_wrapped(self) -> None:
payload = 'var data=([{"day":"2024-08-30","close":"1710.0"}])'
rows = _extract_jsonp(payload)
assert rows[0]["close"] == "1710.0"
def test_with_leading_comment_and_trailing_semicolon(self) -> None:
payload = (
"/*<script>location.href='//sina.com';</script>*/\n"
'var data=([{"day":"2024-08-30","close":"1710.0"}]);'
)
rows = _extract_jsonp(payload)
assert rows[0]["close"] == "1710.0"