tradingSystem/app/services/company_report.py

# -*- coding: utf-8 -*-
"""个股深度评析报告的代取 (2026-09-18)。

报告由数据基座每晚生成, 上游计划与逻辑状态里只带一条链接 (company_review.report_url),
链接打开是一份 markdown 纯文本。原来页面把链接开成新标签, 浏览器按纯文本显示,
用户看到的是一屏 markdown 源码。现在由持仓系统后端代取一次, 页面拿到正文后在自己的
抽屉里排版, 与整页同一套样式。

只做三件事:
一, 按代码找链接。持仓票从逻辑状态映射取, 候选票从当日计划主榜与观察档取, 与
    /api/research 的公司评析来源一致。链接只认上游给的, 不接受页面传入的任意地址。
二, 带超时拉取正文。
三, 按链接缓存十分钟。报告每晚生成一次, 盘中不会变; 点开抽屉不该每次都去打数据基座。

取不到就把原因写进 error 返回, 不抛异常 (页面按失败态显示并给重试)。
"""
from __future__ import annotations

import logging
import time

import requests

logger = logging.getLogger(__name__)

CACHE_SEC = 600          # 正文缓存秒数
TIMEOUT_SEC = 8          # 拉取超时
MAX_BYTES = 2_000_000    # 超过这个长度当异常处理, 不往页面塞
SRC_HELD, SRC_CAND, SRC_NONE = "held", "candidate", "none"

_cache: dict = {}        # url -> {"at": 时刻, "text": 正文}


def _url_of(cr) -> str | None:
    if not isinstance(cr, dict):
        return None
    u = cr.get("report_url")
    u = str(u).strip() if u else ""
    return u if u.startswith(("http://", "https://")) else None


def report_url_for(code: str, *, state_map=None, plan=None) -> tuple[str | None, str]:
    """返回 (链接, 来源)。来源: held (持仓逻辑状态) / candidate (当日计划) / none。
    state_map 与 plan 可注入 (单测用); 缺省时现取, 任一路取失败只记日志不抛。"""
    if state_map is None:
        try:
            from app.services import logic_state_service
            state_map = logic_state_service.state_map()
        except Exception as e:  # noqa: BLE001
            logger.warning("[评析报告] 取持仓逻辑状态失败 %s: %s", code, e)
            state_map = {}
    st = (state_map or {}).get(code)
    u = _url_of((st or {}).get("company_review")) if isinstance(st, dict) else None
    if u:
        return u, SRC_HELD
    if plan is None:
        try:
            from app.services import plan_feed
            plan = plan_feed.get_plan()
        except Exception as e:  # noqa: BLE001
            logger.warning("[评析报告] 取候选计划失败 %s: %s", code, e)
            plan = {}
    for r in (plan or {}).get("main") or []:
        if r.get("ts_code") == code:
            u = _url_of(r.get("company_review"))
            if u:
                return u, SRC_CAND
    for r in (plan or {}).get("observe") or []:
        if r.get("ts_code") == code:
            u = _url_of(r.get("company_review"))
            if u:
                return u, SRC_CAND
    return None, SRC_NONE


def _http_get_text(url: str, timeout: int) -> str:
    r = requests.get(url, timeout=timeout)
    r.raise_for_status()
    if len(r.content) > MAX_BYTES:
        raise ValueError(f"报告过大 ({len(r.content)} 字节)")
    r.encoding = r.encoding or "utf-8"
    return r.text


def fetch_text(url: str, *, now=None, timeout: int = TIMEOUT_SEC, getter=None) -> tuple[str, bool]:
    """取正文, 返回 (正文, 是否命中缓存)。失败抛异常, 由 get 兜成 error。"""
    now = time.time() if now is None else now
    hit = _cache.get(url)
    if hit and now - hit["at"] < CACHE_SEC:
        return hit["text"], True
    text = (getter or _http_get_text)(url, timeout)
    if not isinstance(text, str) or not text.strip():
        raise ValueError("报告内容为空")
    _cache[url] = {"at": now, "text": text}
    return text, False


def _fail_why(e: Exception, timeout: int) -> str:
    if isinstance(e, requests.exceptions.Timeout):
        return f"数据基座没有回应，等了 {timeout} 秒"
    if isinstance(e, requests.exceptions.ConnectionError):
        return "连不上数据基座"
    if isinstance(e, requests.exceptions.HTTPError):
        code = getattr(getattr(e, "response", None), "status_code", None)
        return f"数据基座回了 {code}" if code else f"数据基座回了错误: {e}"
    return f"{type(e).__name__}: {e}"


def get(code: str, *, now=None, timeout: int = TIMEOUT_SEC, getter=None, state_map=None, plan=None) -> dict:
    """页面用的一站式取法。返回固定键: ok / ts_code / url / source / markdown / cached / fetched_at / error。"""
    url, src = report_url_for(code, state_map=state_map, plan=plan)
    out = {"ok": False, "ts_code": code, "url": url, "source": src, "markdown": "", "cached": False,
           "fetched_at": None, "error": None}
    if not url:
        out["error"] = ("这只票没有个股深度评析报告的链接" +
                        ("（不在持仓也不在当日选股计划里）" if src == SRC_NONE else ""))
        return out
    try:
        text, cached = fetch_text(url, now=now, timeout=timeout, getter=getter)
    except Exception as e:  # noqa: BLE001
        out["error"] = f"评析报告读取失败: {_fail_why(e, timeout)}"
        logger.warning("[评析报告] %s %s: %s", code, url, out["error"])
        return out
    at = _cache.get(url, {}).get("at")
    out.update({"ok": True, "markdown": text, "cached": cached,
                "fetched_at": time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(at)) if at else None})
    return out


def invalidate():
    _cache.clear()