From 0d498c51f7e76fdc19b60972437e569b2c31f730 Mon Sep 17 00:00:00 2001 From: hello_strong <1145607886@qq.com> Date: Sat, 3 Oct 2026 23:42:35 +0800 Subject: [PATCH] feat(weread): multi-account quota failover MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - wx.lic 支持 weread_data.accounts: [{cookie, vid, name}, ...] 多账号数组 - 旧版单账号 weread_data.cookie 格式自动兼容 - 新增 _switch_weread_account() round-robin 切号 - 新增 _weread_http_get() 带 failover：检测 -2014/-2041/-2012/499 自动切下一个账号重试 - _weread_get / _get_mp_articles_page / _get_mp_cover / _get_mp_content 全部走新方法 - 不涉及 req/resp 接口变更，调用方零改动 --- core/wx/model/weread.py | 149 +++++++++++++++++++++++++++++++------ core/wx/model/weread_mp.py | 33 ++++---- 2 files changed, 140 insertions(+), 42 deletions(-) diff --git a/core/wx/model/weread.py b/core/wx/model/weread.py index b9219573..80ef2630 100644 --- a/core/wx/model/weread.py +++ b/core/wx/model/weread.py @@ -58,9 +58,16 @@ class MpsWeread(WxGather): self._weread_ticket: str = "" self._weread_vid: str = "" self._weread_name: str = "" + # 多账号：wx.lic 的 weread_data.accounts 数组，或包装单账号旧格式 + self._weread_accounts: list = [] + self._current_account_idx: int = 0 def _load_weread_auth(self): - """加载微信读书的 Cookie""" + """加载微信读书的 Cookie + + 支持多账号：``weread_data.accounts`` 为数组，每个元素是独立 cookie/vid/name。 + 旧版单账号 ``weread_data.cookie`` 格式自动兼容。 + """ from core.config import Config, cfg as app_cfg import os @@ -78,33 +85,130 @@ class MpsWeread(WxGather): except Exception: weread_data = {} - self._weread_cookies = app_cfg.get("weread.cookie", "") or weread_data.get("cookie", "") - self._weread_ticket = app_cfg.get("weread.ticket", "") or weread_data.get("ticket", "") - self._weread_vid = app_cfg.get("weread.vid", "") or weread_data.get("vid", "") - self._weread_name = weread_data.get("name", "") + # 构建多账号列表：兼容单账号旧格式 + accounts = weread_data.get("accounts", None) + if accounts and isinstance(accounts, list): + self._weread_accounts = accounts + else: + # 单账号旧格式 → 包装成列表 + if weread_data.get("cookie"): + self._weread_accounts = [weread_data] + else: + self._weread_accounts = [] + + # 选择当前激活账号：优先 env 变量覆盖，否则用第一个 + self._current_account_idx = 0 + self._weread_cookies = app_cfg.get("weread.cookie", "") or ( + self._weread_accounts[0].get("cookie", "") if self._weread_accounts else "" + ) + self._weread_ticket = app_cfg.get("weread.ticket", "") or ( + self._weread_accounts[0].get("ticket", "") if self._weread_accounts else "" + ) + self._weread_vid = app_cfg.get("weread.vid", "") or ( + self._weread_accounts[0].get("vid", "") if self._weread_accounts else "" + ) + self._weread_name = ( + self._weread_accounts[0].get("name", "") if self._weread_accounts else "" + ) + + def _switch_weread_account(self) -> bool: + """切换到下一个微信读书账号（round-robin）。 + + 返回 ``True`` 表示成功切到了**不同**的账号； + ``False`` 表示只有一个账号或已切回原点。 + """ + if len(self._weread_accounts) <= 1: + return False + prev_idx = self._current_account_idx + self._current_account_idx = (self._current_account_idx + 1) % len( + self._weread_accounts + ) + acc = self._weread_accounts[self._current_account_idx] + self._weread_cookies = acc.get("cookie", "") + self._weread_ticket = acc.get("ticket", "") + self._weread_vid = acc.get("vid", "") + self._weread_name = acc.get("name", "") + return True + + def _weread_http_get(self, url: str, params: dict = None, include_ticket=False, timeout=(10, 30)): + """GET 请求微信读书 API，带多账号配额 failover。 + + 检测到配额耗尽（HTTP 499 / errcode -2014 / -2041 / -2012）时， + 自动切换到 ``wx.lic`` 中下一个账号并重试。 + + 返回值：``requests.Response`` 对象（像 ``requests.get`` 一样使用）。 + """ + import requests + from core.wx.model.weread_mp import WereadMPAPIError + + tried_accounts = set() + while True: + headers = { + "Cookie": self._weread_cookies, + "User-Agent": self.user_agent, + "Accept": "application/json, text/plain, */*", + "Accept-Language": "zh-CN,zh;q=0.9,en;q=0.8", + "Origin": "https://weread.qq.com", + "Referer": "https://weread.qq.com/", + } + if include_ticket and self._weread_ticket: + headers["x-wr-ticket"] = self._weread_ticket + + try: + resp = requests.get( + url, params=params, headers=headers, + proxies=self._get_proxies(), timeout=timeout, + ) + except requests.exceptions.Timeout: + raise WereadMPAPIError("timeout", f"{url}") from None + except requests.RequestException as exc: + raise WereadMPAPIError("network_error", str(exc)) from exc + + # 检查响应中的配额耗尽信号 + is_quota_exhausted = False + if resp.status_code == 499: + is_quota_exhausted = True + elif resp.status_code == 200: + try: + payload = resp.json() + code = payload.get("errcode", payload.get("errCode", 0)) + if code in (-2014, -2041, -2012): + is_quota_exhausted = True + except (ValueError, AttributeError): + pass + + if is_quota_exhausted: + if self._current_account_idx not in tried_accounts: + tried_accounts.add(self._current_account_idx) + idx_before = self._current_account_idx + if self._switch_weread_account() and self._current_account_idx not in tried_accounts: + logger.info( + f"配额耗尽（当前账号 {idx_before}），" + f"切换到账号 {self._current_account_idx}（{self._weread_name or '未命名'}）重试" + ) + continue + # 所有账号都试过了 + raise WereadMPAPIError( + 499 if resp.status_code == 499 else (code if 'code' in locals() else -2041), + f"所有账号配额均已耗尽（{len(tried_accounts)} 个）", + retriable=False, + ) + + return resp def _weread_get(self, url: str, params: dict = None) -> Optional[dict]: - """带微信读书 Cookie 的 GET 请求（web 域，浏览器完整请求头）""" - import requests + """带微信读书 Cookie 的 GET 请求（web 域，浏览器完整请求头） - headers = { - "Cookie": self._weread_cookies, - "User-Agent": self.user_agent, - "Accept": "application/json, text/plain, */*", - "Accept-Language": "zh-CN,zh;q=0.9,en;q=0.8", - "Origin": "https://weread.qq.com", - "Referer": "https://weread.qq.com/", - } + 内部走 ``_weread_http_get``，带多账号配额 failover。 + """ + try: + resp = self._weread_http_get(url, params=params, timeout=15) + except Exception: + return None try: - proxies = self._get_proxies() - resp = requests.get( - url, params=params, headers=headers, proxies=proxies, timeout=15 - ) if resp.status_code == 200: payload = resp.json() - # 微信读书以 HTTP 200 返回业务错误（登录态失效等）， - # 必须检查 errcode，否则「登录超时」会被当成空书架（0 本书） errcode = payload.get("errCode", payload.get("errcode", 0)) if errcode in (-2012, -2010, -2041): print_warning( @@ -123,9 +227,6 @@ class MpsWeread(WxGather): else: print_warning(f"Weread API 返回 {resp.status_code}: {url}") return None - except requests.exceptions.Timeout: - print_warning(f"Weread API 超时: {url}") - return None except Exception as e: print_warning(f"Weread API 异常: {e}") return None diff --git a/core/wx/model/weread_mp.py b/core/wx/model/weread_mp.py index c6e79c38..04650d91 100644 --- a/core/wx/model/weread_mp.py +++ b/core/wx/model/weread_mp.py @@ -3,7 +3,6 @@ import time from urllib.parse import quote -import requests from bs4 import BeautifulSoup from core.log import logger @@ -164,15 +163,17 @@ class MpsWereadMP(MpsWeread): def _get_mp_articles_page(self, book_id: str, offset=0): # /web/mp/articles 曾在部分旧 Cookie 上返回 -2041，当时据此回退到 cover 方案； # 实测该列表接口可用，是增量补抓的主路径。失败时由 get_Articles 回退 cover。 + # 内部走 _weread_http_get，带多账号配额 failover。 try: - response = requests.get( + response = self._weread_http_get( f"{WEREAD_WEB_BASE}/web/mp/articles", params={"bookId": book_id, "offset": offset}, - headers=self._request_headers(include_ticket=True), - proxies=self._get_proxies(), - timeout=(10, 30), + include_ticket=True, ) - except requests.RequestException as exc: + except Exception as exc: + from core.wx.model.weread_mp import WereadMPAPIError + if isinstance(exc, WereadMPAPIError): + raise raise WereadMPAPIError("network_error", str(exc)) from exc if response.status_code != 200: raise WereadMPAPIError( @@ -194,14 +195,13 @@ class MpsWereadMP(MpsWeread): 返回: {"name", "title", "pic", "reviewId", ...}；无文章或接口异常时抛错。 """ try: - response = requests.get( + response = self._weread_http_get( f"{WEREAD_WEB_BASE}/api/mp/cover", params={"bookId": book_id}, - headers=self._request_headers(), - proxies=self._get_proxies(), - timeout=(10, 30), ) - except requests.RequestException as exc: + except Exception as exc: + if isinstance(exc, WereadMPAPIError): + raise raise WereadMPAPIError("network_error", str(exc)) from exc if response.status_code != 200: raise WereadMPAPIError( @@ -248,17 +248,14 @@ class MpsWereadMP(MpsWeread): session.close() def _get_mp_content(self, review_id: str): - headers = self._request_headers() - headers["Accept"] = "text/html,application/xhtml+xml,*/*" try: - response = requests.get( + response = self._weread_http_get( f"{WEREAD_WEB_BASE}/web/mp/content", params={"reviewId": review_id}, - headers=headers, - proxies=self._get_proxies(), - timeout=(10, 30), ) - except requests.RequestException as exc: + except Exception as exc: + if isinstance(exc, WereadMPAPIError): + raise raise WereadMPAPIError("network_error", str(exc)) from exc if response.status_code != 200: raise WereadMPAPIError(