lib-x17-fintech 2.1.3__py3-none-any.whl

xfintech/data/source/tushare/stock/dayline/dayline.py

@@ -0,0 +1,177 @@
+from __future__ import annotations
+
+from typing import Any, Dict, List, Optional
+
+import pandas as pd
+
+from xfintech.data.common.cache import Cache
+from xfintech.data.common.coolant import Coolant
+from xfintech.data.common.params import Params
+from xfintech.data.common.retry import Retry
+from xfintech.data.job import JobHouse
+from xfintech.data.source.tushare.job import TushareJob
+from xfintech.data.source.tushare.session.session import Session
+from xfintech.data.source.tushare.stock.dayline.constant import (
+    KEY,
+    NAME,
+    PAGINATE,
+    SOURCE,
+    TARGET,
+)
+
+
+@JobHouse.register(KEY, alias=KEY)
+class Dayline(TushareJob):
+    """
+    描述:
+    - 获取A股日线行情数据, 每天15点-16点之间入库
+    - 本接口是未复权行情，停牌期间不提供数据
+    - 基础积分每分钟内可调取500次
+    - API文档: https://tushare.pro/document/2?doc_id=27
+    - SCALE: CrossSection & Individual
+    - TYPE: Partitioned
+    - PAGINATE: 6000 rows / 1000 pages
+
+    属性:
+    - name: str, 作业名称 'dayline'。
+    - key: str, 作业键 '/tushare/dayline'。
+    - session: Session, Tushare会话对象。
+    - source: TableInfo, 源表信息（Tushare原始格式）。
+    - target: TableInfo, 目标表信息（转换后格式）。
+    - params: Params, 查询参数。
+        - ts_code: str, 可选, 股票代码(支持多个股票，逗号分隔)
+        - trade_date: str, 可选, 交易日期（YYYYMMDD）
+        - start_date: str, 可选, 开始日期（YYYYMMDD）
+        - end_date: str, 可选, 结束日期（YYYYMMDD）
+    - coolant: Coolant, 请求冷却控制。
+    - paginate: Paginate, 分页控制（pagesize=6000, pagelimit=1000）。
+    - retry: Retry, 重试策略。
+    - cache: Cache, 缓存管理。
+
+    方法:
+    - run(): 执行作业，返回日线行情DataFrame。
+    - _run(): 内部执行逻辑，处理日期参数。
+    - transform(data): 转换数据格式，将源格式转为目标格式。
+    - list_codes(): 返回所有股票代码列表。
+    - list_dates(): 返回所有交易日期列表。
+
+    例子:
+    ```python
+        from xfintech.data.source.tushare.session import Session
+        from xfintech.data.source.tushare.stock.dayline import Dayline
+
+        session = Session(credential="your_token")
+        job = Dayline(
+            session=session,
+            params={
+                "ts_code": "000001.SZ",
+                "start_date": "20241101",
+                "end_date": "20241231"
+            }
+        )
+        df = job.run()
+    ```
+    """
+
+    def __init__(
+        self,
+        session: Session,
+        params: Optional[Params | Dict[str, Any]] = None,
+        coolant: Optional[Coolant | Dict[str, Any]] = None,
+        retry: Optional[Retry | Dict[str, Any]] = None,
+        cache: Optional[Cache | Dict[str, str] | bool] = None,
+    ) -> None:
+        super().__init__(
+            name=NAME,
+            key=KEY,
+            session=session,
+            source=SOURCE,
+            target=TARGET,
+            params=params,
+            coolant=coolant,
+            paginate=PAGINATE,
+            retry=retry,
+            cache=cache,
+        )
+
+    def _run(self) -> pd.DataFrame:
+        cached = self._load_cache()
+        if cached is not None:
+            return cached
+
+        # Prepare payload dict
+        payload = self.params.to_dict()
+        payload = self._parse_date_params(
+            payload,
+            keys=["trade_date", "start_date", "end_date"],
+        )
+        payload = self._parse_string_params(
+            payload,
+            keys=["ts_code"],
+        )
+
+        # Fetch and transform data
+        data = self._fetchall(
+            api=self.connection.daily,
+            **payload,
+        )
+        result = self.transform(data)
+        self._save_cache(result)
+        return result
+
+    # Transform logic
+    def transform(
+        self,
+        data: pd.DataFrame,
+    ) -> pd.DataFrame:
+        cols = self.target.list_column_names()
+        if data is None or data.empty:
+            return pd.DataFrame(columns=cols)
+
+        out = data.copy()
+        out["code"] = out["ts_code"].astype(str)
+        out["date"] = pd.to_datetime(
+            out["trade_date"],
+            format="%Y%m%d",
+            errors="coerce",
+        ).dt.strftime("%Y-%m-%d")
+        out["datecode"] = out["trade_date"].astype(str)
+        out["percent_change"] = pd.to_numeric(
+            out["pct_chg"],
+            errors="coerce",
+        )
+        out["volume"] = pd.to_numeric(
+            out["vol"],
+            errors="coerce",
+        )
+
+        # Convert numeric fields
+        numeric_fields = [
+            "open",
+            "high",
+            "low",
+            "close",
+            "pre_close",
+            "change",
+            "amount",
+        ]
+        for field in numeric_fields:
+            out[field] = pd.to_numeric(
+                out[field],
+                errors="coerce",
+            )
+
+        # Finalize output
+        out = out[cols].drop_duplicates()
+        out = out.sort_values(by=["code", "date"])
+        out = out.reset_index(drop=True)
+        self.markpoint("transform[OK]")
+        return out
+
+    def list_codes(self) -> List[str]:
+        df = self.run()
+        return sorted(df["code"].unique().tolist())
+
+    def list_dates(self) -> List[str]:
+        df = self.run()
+        return sorted(df["date"].unique().tolist())