butubb
f6ddc46d62
问题：cookie 过期导致任务失败，但没有任何通知。查下来不是代码问题 —— notify_enabled 在两个任务上都是 False，而它默认就是关的，事件（run_failed） 也确实生成了，只卡在最后一道闸门。 但那个默认值是错的。代码里的理由是「一条任务列表都推到一个群会很快变吵，所以默认静默」， 这个理由对新作品成立（可能每轮都有），对失败不成立：一次登录态失效意味着这个任务事实上 已经死了，而你不会知道，直到某天发现数据停在几周前。最该被告知的就是这种情况。 现在拆开： - notify_enabled —— 推送新作品，可能每轮都有，默认关 - notify_failures —— 推送异常（登录失效/运行失败/没抓到数据），默认开 事件按开关过滤（build_run_message）：只勾了「新作品」的任务不该因为一次失败被推消息， 反之亦然，否则拆开开关就没有意义。已有任务由 _ensure_columns 补上 notify_failures=1， 所以会自动开始收到异常推送。 列名 notify_enabled 是历史遗留（它早先是唯一的通知开关），语义已收窄为「新作品」， 用注释写明，不做列重命名 —— 那需要单独的迁移，不值为一个内部工具做。
117 lines
5.5 KiB
Python
117 lines
5.5 KiB
Python
# -*- coding: utf-8 -*-
|
|
# Copyright (c) 2025 [email protected]
|
|
#
|
|
# This file is part of MediaCrawler project.
|
|
# Repository: https://github.com/NanmiCoder/MediaCrawler/blob/main/api/schemas/monitor.py
|
|
# GitHub: https://github.com/NanmiCoder
|
|
# Licensed under NON-COMMERCIAL LEARNING LICENSE 1.1
|
|
#
|
|
# 声明：本代码仅供学习和研究目的使用。使用者应遵守以下原则：
|
|
# 1. 不得用于任何商业用途。
|
|
# 2. 使用时应遵守目标平台的使用条款和robots.txt规则。
|
|
# 3. 不得进行大规模爬取或对平台造成运营干扰。
|
|
# 4. 应合理控制请求频率，避免给目标平台带来不必要的负担。
|
|
# 5. 不得用于任何非法或不当的用途。
|
|
#
|
|
# 详细许可条款请参阅项目根目录下的LICENSE文件。
|
|
# 使用本代码即表示您同意遵守上述原则和LICENSE中的所有条款。
|
|
|
|
"""Request models for the monitoring API."""
|
|
|
|
from typing import List, Literal, Optional
|
|
|
|
from pydantic import BaseModel, Field, model_validator
|
|
|
|
# A floor on the interval is a correctness guard, not a nicety: every run
|
|
# launches a browser and hits XHS with several requests, so a short interval
|
|
# across many creators is the pattern that triggers rate limiting.
|
|
MIN_INTERVAL_MINUTES = 30
|
|
MAX_INTERVAL_MINUTES = 7 * 24 * 60
|
|
|
|
|
|
class MonitorTaskCreate(BaseModel):
|
|
name: str = Field(min_length=1, max_length=200)
|
|
platform: str = "xhs"
|
|
# One subprocess handles exactly one crawler type, so a task is either
|
|
# creator-driven or note-driven.
|
|
mode: Literal["creator", "note"]
|
|
# None means "use the value configured on the Settings page", which is what
|
|
# makes those defaults meaningful. Bounds still apply when a value is given.
|
|
interval_minutes: Optional[int] = Field(
|
|
default=None, ge=MIN_INTERVAL_MINUTES, le=MAX_INTERVAL_MINUTES
|
|
)
|
|
# --- Scheduling ---------------------------------------------------------
|
|
# All three modes are expressible with pickers; a raw cron string is
|
|
# deliberately not supported, since it is a small language to learn just to
|
|
# say "every day at nine".
|
|
schedule_mode: Literal["interval", "daily", "weekly"] = "interval"
|
|
# 0-23, e.g. [9, 12, 18]. Required for the two clock modes.
|
|
schedule_hours: List[int] = Field(default_factory=list)
|
|
# 0-6 with Monday = 0, matching Python's date.weekday(). Required for weekly.
|
|
schedule_days: List[int] = Field(default_factory=list)
|
|
# One minute for the whole schedule, so a task with three times is
|
|
# "09:30, 12:30, 18:30" rather than three separate minute choices.
|
|
schedule_minute: int = Field(default=0, ge=0, le=59)
|
|
|
|
max_notes_count: Optional[int] = Field(default=None, ge=1, le=500)
|
|
enable_comments: bool = True
|
|
# Raising this widens the comment window, which is the only lever available
|
|
# for noticing new comments -- the API has no time-sort.
|
|
max_comments_count: Optional[int] = Field(default=None, ge=1, le=500)
|
|
run_timeout_seconds: int = Field(default=3600, ge=60, le=86400)
|
|
enabled: bool = True
|
|
# 两类通知分开：新作品可能每轮都有（默认关，避免刷屏），
|
|
# 异常频率低且意味着任务已经停止工作（默认开，否则你会一直不知道）。
|
|
notify_enabled: bool = False
|
|
notify_failures: bool = True
|
|
# Raw pasted values: full URLs or bare ids, in either form.
|
|
targets: List[str] = Field(min_length=1)
|
|
|
|
@model_validator(mode="after")
|
|
def _validate_schedule(self) -> "MonitorTaskCreate":
|
|
if any(hour < 0 or hour > 23 for hour in self.schedule_hours):
|
|
raise ValueError("小时必须在 0-23 之间")
|
|
if any(day < 0 or day > 6 for day in self.schedule_days):
|
|
raise ValueError("星期必须在 0-6 之间（周一为 0）")
|
|
# A clock mode with no chosen time can never fire. Rejecting it here is
|
|
# what keeps next_occurrence()'s None branch unreachable in practice.
|
|
if self.schedule_mode in ("daily", "weekly") and not self.schedule_hours:
|
|
raise ValueError("按钟点调度至少要选一个时间")
|
|
if self.schedule_mode == "weekly" and not self.schedule_days:
|
|
raise ValueError("按周调度至少要选一个星期")
|
|
return self
|
|
|
|
|
|
class MonitorTaskUpdate(BaseModel):
|
|
name: Optional[str] = Field(default=None, min_length=1, max_length=200)
|
|
enabled: Optional[bool] = None
|
|
interval_minutes: Optional[int] = Field(
|
|
default=None, ge=MIN_INTERVAL_MINUTES, le=MAX_INTERVAL_MINUTES
|
|
)
|
|
# None means "leave alone". Cross-field validity depends on the merged state,
|
|
# so it is checked in the service rather than here.
|
|
schedule_mode: Optional[Literal["interval", "daily", "weekly"]] = None
|
|
schedule_hours: Optional[List[int]] = None
|
|
schedule_days: Optional[List[int]] = None
|
|
schedule_minute: Optional[int] = Field(default=None, ge=0, le=59)
|
|
max_notes_count: Optional[int] = Field(default=None, ge=1, le=500)
|
|
enable_comments: Optional[bool] = None
|
|
max_comments_count: Optional[int] = Field(default=None, ge=1, le=500)
|
|
run_timeout_seconds: Optional[int] = Field(default=None, ge=60, le=86400)
|
|
notify_enabled: Optional[bool] = None
|
|
notify_failures: Optional[bool] = None
|
|
# When present, replaces the whole target list.
|
|
targets: Optional[List[str]] = None
|
|
|
|
|
|
class CookiePayload(BaseModel):
|
|
cookie: str = Field(min_length=1)
|
|
|
|
|
|
class WebhookPayload(BaseModel):
|
|
url: str = Field(default="", description="企业微信机器人 Webhook 地址，留空表示停用")
|
|
|
|
|
|
class WebhookTestPayload(BaseModel):
|
|
url: Optional[str] = Field(default=None, description="不传则使用已保存的地址")
|