move files

src/plugins/nonebot_bison/__init__.py

@@ -0,0 +1,10 @@
+import nonebot
+
+from . import config_manager
+from . import config
+from . import scheduler
+from . import send
+from . import post
+from . import platform
+from . import types
+from . import utils

src/plugins/nonebot_bison/config.py

@@ -0,0 +1,136 @@
+from collections import defaultdict
+from os import path
+import os
+from typing import DefaultDict, Mapping
+
+import nonebot
+from tinydb import Query, TinyDB
+
+from .plugin_config import plugin_config
+from .types import Target, User
+from .utils import Singleton
+from .platform import platform_manager
+
+supported_target_type = platform_manager.keys()
+
+def get_config_path() -> str:
+    if plugin_config.hk_reporter_config_path:
+        data_dir = plugin_config.hk_reporter_config_path
+    else:
+        working_dir = os.getcwd()
+        data_dir = path.join(working_dir, 'data')
+    if not path.isdir(data_dir):
+        os.makedirs(data_dir)
+    return path.join(data_dir, 'hk_reporter.json')
+
+class NoSuchUserException(Exception):
+    pass
+
+class NoSuchSubscribeException(Exception):
+    pass
+
+class Config(metaclass=Singleton):
+
+    migrate_version = 2
+    
+    def __init__(self):
+        self.db = TinyDB(get_config_path(), encoding='utf-8') 
+        self.kv_config = self.db.table('kv')
+        self.user_target = self.db.table('user_target')
+        self.target_user_cache: dict[str, defaultdict[Target, list[User]]] = {}
+        self.target_user_cat_cache = {}
+        self.target_user_tag_cache = {}
+        self.target_list = {}
+        self.next_index: DefaultDict[str, int] = defaultdict(lambda: 0)
+         
+    def add_subscribe(self, user, user_type, target, target_name, target_type, cats, tags):
+        user_query = Query()
+        query = (user_query.user == user) & (user_query.user_type == user_type)
+        if (user_data := self.user_target.get(query)):
+            # update
+            subs: list = user_data.get('subs', [])
+            subs.append({"target": target, "target_type": target_type, 'target_name': target_name, 'cats': cats, 'tags': tags})
+            self.user_target.update({"subs": subs}, query)
+        else:
+            # insert
+            self.user_target.insert({
+                'user': user, 'user_type': user_type,
+                'subs': [{'target': target, 'target_type': target_type, 'target_name': target_name, 'cats': cats, 'tags': tags }]
+            })
+        self.update_send_cache()
+
+    def list_subscribe(self, user, user_type):
+        query = Query()
+        if user_sub := self.user_target.get((query.user == user) & (query.user_type ==user_type)):
+            return user_sub['subs']
+        return []
+    
+    def del_subscribe(self, user, user_type, target, target_type):
+        user_query = Query()
+        query = (user_query.user == user) & (user_query.user_type == user_type)
+        if not (query_res := self.user_target.get(query)):
+            raise NoSuchUserException()
+        subs = query_res.get('subs', [])
+        for idx, sub in enumerate(subs):
+            if sub.get('target') == target and sub.get('target_type') == target_type:
+                subs.pop(idx)
+                self.user_target.update({'subs': subs}, query)
+                self.update_send_cache()
+                return
+        raise NoSuchSubscribeException()
+
+    def update_send_cache(self):
+        res = {target_type: defaultdict(list) for target_type in supported_target_type}
+        cat_res = {target_type: defaultdict(lambda: defaultdict(list)) for target_type in supported_target_type}
+        tag_res = {target_type: defaultdict(lambda: defaultdict(list)) for target_type in supported_target_type}
+        # res = {target_type: defaultdict(lambda: defaultdict(list)) for target_type in supported_target_type}
+        for user in self.user_target.all():
+            for sub in user.get('subs', []):
+                if not sub.get('target_type') in supported_target_type:
+                    continue
+                res[sub['target_type']][sub['target']].append(User(user['user'], user['user_type']))
+                cat_res[sub['target_type']][sub['target']]['{}-{}'.format(user['user_type'], user['user'])] = sub['cats']
+                tag_res[sub['target_type']][sub['target']]['{}-{}'.format(user['user_type'], user['user'])] = sub['tags']
+        self.target_user_cache = res
+        self.target_user_cat_cache = cat_res
+        self.target_user_tag_cache = tag_res
+        for target_type in self.target_user_cache:
+            self.target_list[target_type] = list(self.target_user_cache[target_type].keys())
+
+    def get_sub_category(self, target_type, target, user_type, user):
+        return self.target_user_cat_cache[target_type][target]['{}-{}'.format(user_type, user)]
+
+    def get_sub_tags(self, target_type, target, user_type, user):
+        return self.target_user_tag_cache[target_type][target]['{}-{}'.format(user_type, user)]
+
+    def get_next_target(self, target_type):
+        # FIXME 插入或删除target后对队列的影响（但是并不是大问题
+        if not self.target_list[target_type]:
+            return None
+        self.next_index[target_type] %= len(self.target_list[target_type])
+        res = self.target_list[target_type][self.next_index[target_type]]
+        self.next_index[target_type] += 1
+        return res
+
+def start_up():
+    config = Config()
+    if not (search_res := config.kv_config.search(Query().name=="version")):
+        config.kv_config.insert({"name": "version", "value": config.migrate_version})
+    elif search_res[0].get("value") < config.migrate_version:
+        query = Query()
+        version_query = (query.name == 'version')
+        cur_version = search_res[0].get("value")
+        if cur_version == 1:
+            cur_version = 2
+            for user_conf in config.user_target.all():
+                conf_id = user_conf.doc_id
+                subs = user_conf['subs']
+                for sub in subs:
+                    sub['cats'] = []
+                    sub['tags'] = []
+                config.user_target.update({'subs': subs}, doc_ids=[conf_id])
+        config.kv_config.update({"value": config.migrate_version}, version_query)
+        # do migration
+    config.update_send_cache()
+
+nonebot.get_driver().on_startup(start_up)

src/plugins/nonebot_bison/config_manager.py

@@ -0,0 +1,186 @@
+from typing import Type
+from nonebot import logger, on_command
+from nonebot.adapters.cqhttp import Bot, Event, GroupMessageEvent
+from nonebot.adapters.cqhttp.message import Message
+from nonebot.adapters.cqhttp.permission import GROUP_ADMIN, GROUP_MEMBER, GROUP_OWNER
+from nonebot.permission import Permission, SUPERUSER
+from nonebot.rule import to_me
+from nonebot.typing import T_State
+from nonebot.matcher import Matcher
+
+from .config import Config, NoSuchSubscribeException
+from .platform import platform_manager, check_sub_target
+from .utils import parse_text
+from .types import Target
+
+def _gen_prompt_template(prompt: str):
+    if hasattr(Message, 'template'):
+        return Message.template(prompt)
+    return prompt
+
+common_platform = [p.platform_name for p in \
+        filter(lambda platform: platform.enabled and platform.is_common,
+            platform_manager.values())
+    ]
+
+help_match = on_command('help', rule=to_me(), priority=5)
+@help_match.handle()
+async def send_help(bot: Bot, event: Event, state: T_State):
+    message = '使用方法：\n@bot 添加订阅（仅管理员）\n@bot 查询订阅\n@bot 删除订阅（仅管理员）'
+    await help_match.finish(Message(await parse_text(message)))
+
+
+def do_add_sub(add_sub: Type[Matcher]):
+    @add_sub.handle()
+    async def init_promote(bot: Bot, event: Event, state: T_State):
+        state['_prompt'] = '请输入想要订阅的平台，目前支持，请输入冒号左边的名称：\n' + \
+                ''.join(['{}：{}\n'.format(platform_name, platform_manager[platform_name].name) \
+                        for platform_name in common_platform]) + \
+                '要查看全部平台请输入：“全部”'
+
+    async def parse_platform(bot: Bot, event: Event, state: T_State) -> None:
+        platform = str(event.get_message()).strip()
+        if platform == '全部':
+            message = '全部平台\n' + \
+                '\n'.join(['{}：{}'.format(platform_name, platform.name) \
+                        for platform_name, platform in platform_manager.items()])
+            await add_sub.reject(message)
+        elif platform in platform_manager:
+            state['platform'] = platform
+        else:
+            await add_sub.reject('平台输入错误')
+
+    @add_sub.got('platform', _gen_prompt_template('{_prompt}'), parse_platform)
+    @add_sub.handle()
+    async def init_id(bot: Bot, event: Event, state: T_State):
+        if platform_manager[state['platform']].has_target:
+            state['_prompt'] = '请输入订阅用户的id，详情查阅https://nonebot-hk-reporter.vercel.app/usage/#%E6%89%80%E6%94%AF%E6%8C%81%E5%B9%B3%E5%8F%B0%E7%9A%84uid'
+        else:
+            state['id'] = 'default'
+            state['name'] = await platform_manager[state['platform']].get_target_name(Target(''))
+
+    async def parse_id(bot: Bot, event: Event, state: T_State):
+        target = str(event.get_message()).strip()
+        name = await check_sub_target(state['platform'], target)
+        if not name:
+            await add_sub.reject('id输入错误')
+        state['id'] = target
+        state['name'] = name
+
+    @add_sub.got('id', _gen_prompt_template('{_prompt}'), parse_id)
+    @add_sub.handle()
+    async def init_cat(bot: Bot, event: Event, state: T_State):
+        if not platform_manager[state['platform']].categories:
+            state['cats'] = []
+            return
+        state['_prompt'] = '请输入要订阅的类别，以空格分隔，支持的类别有：{}'.format(
+                ' '.join(list(platform_manager[state['platform']].categories.values())))
+
+    async def parser_cats(bot: Bot, event: Event, state: T_State):
+        res = []
+        for cat in str(event.get_message()).strip().split():
+            if cat not in platform_manager[state['platform']].reverse_category:
+                await add_sub.reject('不支持 {}'.format(cat))
+            res.append(platform_manager[state['platform']].reverse_category[cat])
+        state['cats'] = res
+
+    @add_sub.got('cats', _gen_prompt_template('{_prompt}'), parser_cats)
+    @add_sub.handle()
+    async def init_tag(bot: Bot, event: Event, state: T_State):
+        if not platform_manager[state['platform']].enable_tag:
+            state['tags'] = []
+            return
+        state['_prompt'] = '请输入要订阅的tag，订阅所有tag输入"全部标签"'
+
+    async def parser_tags(bot: Bot, event: Event, state: T_State):
+        if str(event.get_message()).strip() == '全部标签':
+            state['tags'] = []
+        else:
+            state['tags'] = str(event.get_message()).strip().split()
+
+    @add_sub.got('tags', _gen_prompt_template('{_prompt}'), parser_tags)
+    @add_sub.handle()
+    async def add_sub_process(bot: Bot, event: Event, state: T_State):
+        config = Config()
+        config.add_subscribe(state.get('_user_id') or event.group_id, user_type='group',
+                target=state['id'],
+                target_name=state['name'], target_type=state['platform'],
+                cats=state.get('cats', []), tags=state.get('tags', []))
+        await add_sub.finish('添加 {} 成功'.format(state['name']))
+
+def do_query_sub(query_sub: Type[Matcher]):
+    @query_sub.handle()
+    async def _(bot: Bot, event: Event, state: T_State):
+        config: Config = Config()
+        sub_list = config.list_subscribe(state.get('_user_id') or event.group_id, "group")
+        res = '订阅的帐号为：\n'
+        for sub in sub_list:
+            res += '{} {} {}'.format(sub['target_type'], sub['target_name'], sub['target'])
+            platform = platform_manager[sub['target_type']]
+            if platform.categories:
+                res += ' [{}]'.format(', '.join(map(lambda x: platform.categories[x], sub['cats'])))
+            if platform.enable_tag:
+                res += ' {}'.format(', '.join(sub['tags']))
+            res += '\n'
+        await query_sub.finish(Message(await parse_text(res)))
+
+def do_del_sub(del_sub: Type[Matcher]):
+    @del_sub.handle()
+    async def send_list(bot: Bot, event: Event, state: T_State):
+        config: Config = Config()
+        sub_list = config.list_subscribe(state.get('_user_id') or event.group_id, "group")
+        res = '订阅的帐号为：\n'
+        state['sub_table'] = {}
+        for index, sub in enumerate(sub_list, 1):
+            state['sub_table'][index] = {'target_type': sub['target_type'], 'target': sub['target']}
+            res += '{} {} {} {}\n'.format(index, sub['target_type'], sub['target_name'], sub['target'])
+            platform = platform_manager[sub['target_type']]
+            if platform.categories:
+                res += ' [{}]'.format(', '.join(map(lambda x: platform.categories[x], sub['cats'])))
+            if platform.enable_tag:
+                res += ' {}'.format(', '.join(sub['tags']))
+            res += '\n'
+        res += '请输入要删除的订阅的序号'
+        await bot.send(event=event, message=Message(await parse_text(res)))
+
+    @del_sub.receive()
+    async def do_del(bot, event: Event, state: T_State):
+        try:
+            index = int(str(event.get_message()).strip())
+            config = Config()
+            config.del_subscribe(state.get('_user_id') or event.group_id, 'group', **state['sub_table'][index])
+        except Exception as e:
+            await del_sub.reject('删除错误')
+            logger.warning(e)
+        else:
+            await del_sub.finish('删除成功')
+
+async def parse_group_number(bot: Bot, event: Event, state: T_State):
+    state[state["_current_key"]] = int(str(event.get_message()))
+
+
+add_sub_matcher = on_command("添加订阅", rule=to_me(), permission=GROUP_ADMIN | GROUP_OWNER | SUPERUSER, priority=5)
+do_add_sub(add_sub_matcher)
+manage_add_sub_mather = on_command('管理-添加订阅', permission=SUPERUSER, priority=5)
+@manage_add_sub_mather.got('_user_id', "群号", parse_group_number)
+async def handle(bot: Bot, event: Event, state: T_State):
+    pass
+do_add_sub(manage_add_sub_mather)
+
+
+query_sub_macher = on_command("查询订阅", rule=to_me(), priority=5)
+do_query_sub(query_sub_macher)
+manage_query_sub_mather = on_command('管理-查询订阅', permission=SUPERUSER, priority=5)
+@manage_query_sub_mather.got('_user_id', "群号", parse_group_number)
+async def handle(bot: Bot, event: Event, state: T_State):
+    pass
+do_query_sub(manage_query_sub_mather)
+
+
+del_sub_macher = on_command("删除订阅", rule=to_me(), permission=GROUP_ADMIN | GROUP_OWNER | SUPERUSER, priority=5)
+do_del_sub(del_sub_macher)
+manage_del_sub_mather = on_command('管理-删除订阅', permission=SUPERUSER, priority=5)
+@manage_del_sub_mather.got('_user_id', "群号", parse_group_number)
+async def handle(bot: Bot, event: Event, state: T_State):
+    pass
+do_del_sub(manage_del_sub_mather)

src/plugins/nonebot_bison/platform/__init__.py

@@ -0,0 +1,28 @@
+from collections import defaultdict
+
+from .platform import Platform, NoTargetGroup
+from pkgutil import iter_modules
+from pathlib import Path
+from importlib import import_module
+
+_package_dir = str(Path(__file__).resolve().parent)
+for (_, module_name, _) in iter_modules([_package_dir]):
+    import_module(f'{__name__}.{module_name}')
+
+
+async def check_sub_target(target_type, target):
+    return await platform_manager[target_type].get_target_name(target)
+
+_platform_list = defaultdict(list)
+for _platform in Platform.registory:
+    if not _platform.enabled:
+        continue
+    _platform_list[_platform.platform_name].append(_platform)
+
+platform_manager: dict[str, Platform] = dict()
+for name, platform_list in _platform_list.items():
+    if len(platform_list) == 1:
+        platform_manager[name] = platform_list[0]()
+    else:
+        platform_manager[name] = NoTargetGroup([_platform() for _platform in platform_list])
+

src/plugins/nonebot_bison/platform/arknights.py

@@ -0,0 +1,147 @@
+import json
+from typing import Any
+
+from bs4 import BeautifulSoup as bs
+import httpx
+
+from ..post import Post
+from ..types import Category, RawPost, Target
+from ..utils import Render
+from .platform import CategoryNotSupport, NewMessage, NoTargetMixin, StatusChange
+
+
+class Arknights(NewMessage, NoTargetMixin):
+
+    categories = {1: '游戏公告'}
+    platform_name = 'arknights'
+    name = '明日方舟游戏信息'
+    enable_tag = False
+    enabled = True
+    is_common = False
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 30}
+
+    async def get_target_name(self, _: Target) -> str:
+        return '明日方舟游戏信息'
+
+    async def get_sub_list(self, _) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            raw_data = await client.get('https://ak-conf.hypergryph.com/config/prod/announce_meta/IOS/announcement.meta.json')
+            return json.loads(raw_data.text)['announceList']
+
+    def get_id(self, post: RawPost) -> Any:
+        return post['announceId']
+
+    def get_date(self, _: RawPost) -> None:
+        return None
+
+    def get_category(self, _) -> Category:
+        return Category(1)
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        announce_url = raw_post['webUrl']
+        text = ''
+        async with httpx.AsyncClient() as client:
+            raw_html = await client.get(announce_url)
+        soup = bs(raw_html, 'html.parser')
+        pics = []
+        if soup.find("div", class_="standerd-container"):
+            # 图文
+            render = Render()
+            viewport = {'width': 320, 'height': 6400, 'deviceScaleFactor': 3}
+            pic_data = await render.render(announce_url, viewport=viewport, target='div.main')
+            if pic_data:
+                pics.append(pic_data)
+            else:
+                text = '图片渲染失败'
+        elif (pic := soup.find('img', class_='banner-image')):
+            pics.append(pic['src'])
+        else:
+            raise CategoryNotSupport()
+        return Post('arknights', text=text, url='', target_name="明日方舟游戏内公告", pics=pics, compress=True, override_use_pic=False)
+
+class AkVersion(NoTargetMixin, StatusChange):
+
+    categories = {2: '更新信息'}
+    platform_name = 'arknights'
+    name = '明日方舟游戏信息'
+    enable_tag = False
+    enabled = True
+    is_common = False
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 30}
+
+    async def get_target_name(self, _: Target) -> str:
+        return '明日方舟游戏信息'
+
+    async def get_status(self, _):
+        async with httpx.AsyncClient() as client:
+            res_ver = await client.get('https://ak-conf.hypergryph.com/config/prod/official/IOS/version')
+            res_preanounce = await client.get('https://ak-conf.hypergryph.com/config/prod/announce_meta/IOS/preannouncement.meta.json')
+        res = res_ver.json()
+        res.update(res_preanounce.json())
+        return res
+
+    def compare_status(self, _, old_status, new_status):
+        res = []
+        if old_status.get('preAnnounceType') == 2 and new_status.get('preAnnounceType') == 0:
+            res.append(Post('arknights',
+                text='登录界面维护公告上线（大概是开始维护了)',
+                target_name='明日方舟更新信息'))
+        elif old_status.get('preAnnounceType') == 0 and new_status.get('preAnnounceType') == 2:
+            res.append(Post('arknights',
+                text='登录界面维护公告下线（大概是开服了，冲！）',
+                target_name='明日方舟更新信息'))
+        if old_status.get('clientVersion') != new_status.get('clientVersion'):
+            res.append(Post('arknights', text='游戏本体更新（大更新）', target_name='明日方舟更新信息'))
+        if old_status.get('resVersion') != new_status.get('resVersion'):
+            res.append(Post('arknights', text='游戏资源更新（小更新）', target_name='明日方舟更新信息'))
+        return res
+
+    def get_category(self, _):
+        return Category(2)
+
+    async def parse(self, raw_post):
+        return raw_post
+
+class MonsterSiren(NewMessage, NoTargetMixin):
+
+    categories = {3: '塞壬唱片新闻'}
+    platform_name = 'arknights'
+    name = '明日方舟游戏信息'
+    enable_tag = False
+    enabled = True
+    is_common = False
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 30}
+
+    async def get_target_name(self, _: Target) -> str:
+        return '明日方舟游戏信息'
+
+    async def get_sub_list(self, _) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            raw_data = await client.get('https://monster-siren.hypergryph.com/api/news')
+            return raw_data.json()['data']['list']
+
+    def get_id(self, post: RawPost) -> Any:
+        return post['cid']
+
+    def get_date(self, _) -> None:
+        return None
+
+    def get_category(self, _) -> Category:
+        return Category(3)
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        url = f'https://monster-siren.hypergryph.com/info/{raw_post["cid"]}'
+        async with httpx.AsyncClient() as client:
+            res = await client.get(f'https://monster-siren.hypergryph.com/api/news/{raw_post["cid"]}')
+            raw_data = res.json()
+            content = raw_data['data']['content']
+            content = content.replace('</p>', '</p>\n')
+            soup = bs(content, 'html.parser')
+            imgs = list(map(lambda x: x['src'], soup('img')))
+            text = f'{raw_post["title"]}\n{soup.text.strip()}'
+        return Post('monster-siren', text=text, pics=imgs,
+                url=url, target_name="塞壬唱片新闻", compress=True,
+                override_use_pic=False)

src/plugins/nonebot_bison/platform/bilibili.py

@@ -0,0 +1,126 @@
+import json
+from typing import Any, Optional
+
+import httpx
+
+from ..post import Post
+from ..types import Category, RawPost, Tag, Target
+from .platform import NewMessage, TargetMixin, CategoryNotSupport
+
+class Bilibili(NewMessage, TargetMixin):
+
+    categories = {
+            1: "一般动态",
+            2: "专栏文章",
+            3: "视频",
+            4: "纯文字",
+            5: "转发"
+            # 5: "短视频"
+        }
+    platform_name = 'bilibili'
+    enable_tag = True
+    enabled = True
+    is_common = True
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 10}
+    name = 'B站'
+
+    async def get_target_name(self, target: Target) -> Optional[str]:
+        async with httpx.AsyncClient() as client:
+            res = await client.get('https://api.bilibili.com/x/space/acc/info', params={'mid': target})
+            res_data = json.loads(res.text)
+            if res_data['code']:
+                return None
+            return res_data['data']['name']
+
+    async def get_sub_list(self, target: Target) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            params = {'host_uid': target, 'offset': 0, 'need_top': 0}
+            res = await client.get('https://api.vc.bilibili.com/dynamic_svr/v1/dynamic_svr/space_history', params=params, timeout=4.0)
+            res_dict = json.loads(res.text)
+            if res_dict['code'] == 0:
+                return res_dict['data']['cards']
+            else:
+                return []
+
+    def get_id(self, post: RawPost) -> Any:
+        return post['desc']['dynamic_id']
+    
+    def get_date(self, post: RawPost) -> int:
+        return post['desc']['timestamp']
+
+    def _do_get_category(self, post_type: int) -> Category:
+        if post_type == 2:
+            return Category(1)
+        elif post_type == 64:
+            return Category(2)
+        elif post_type == 8:
+            return Category(3)
+        elif post_type == 4:
+            return Category(4)
+        elif post_type == 1:
+            # 转发
+            return Category(5)
+        raise CategoryNotSupport()
+
+    def get_category(self, post: RawPost) -> Category:
+        post_type = post['desc']['type']
+        return self._do_get_category(post_type)
+
+    def get_tags(self, raw_post: RawPost) -> list[Tag]:
+        return [*map(lambda tp: tp['topic_name'], raw_post['display']['topic_info']['topic_details'])]
+
+    def _get_info(self, post_type: Category, card) -> tuple[str, list]:
+        if post_type == 1:
+            # 一般动态
+            text = card['item']['description']
+            pic = [img['img_src'] for img in card['item']['pictures']]
+        elif post_type == 2:
+            # 专栏文章
+            text = '{} {}'.format(card['title'], card['summary'])
+            pic = card['image_urls']
+        elif post_type == 3:
+            # 视频
+            text = card['dynamic']
+            pic = [card['pic']]
+        elif post_type == 4:
+            # 纯文字
+            text = card['item']['content']
+            pic = []
+        else:
+            raise CategoryNotSupport()
+        return text, pic
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        card_content = json.loads(raw_post['card'])
+        post_type = self.get_category(raw_post)
+        target_name = raw_post['desc']['user_profile']['info']['uname']
+        if post_type >= 1 and post_type < 5:
+            url = ''
+            if post_type == 1:
+                # 一般动态
+                url = 'https://t.bilibili.com/{}'.format(raw_post['desc']['dynamic_id_str'])
+            elif post_type == 2:
+                # 专栏文章
+                url = 'https://www.bilibili.com/read/cv{}'.format(raw_post['desc']['rid'])
+            elif post_type == 3:
+                # 视频
+                url = 'https://www.bilibili.com/video/{}'.format(raw_post['desc']['bvid'])
+            elif post_type == 4:
+                # 纯文字
+                url = 'https://t.bilibili.com/{}'.format(raw_post['desc']['dynamic_id_str'])
+            text, pic = self._get_info(post_type, card_content)
+        elif post_type == 5:
+            # 转发
+            url = 'https://t.bilibili.com/{}'.format(raw_post['desc']['dynamic_id_str'])
+            text = card_content['item']['content']
+            orig_type = card_content['item']['orig_type']
+            orig = json.loads(card_content['origin'])
+            orig_text, _ = self._get_info(self._do_get_category(orig_type), orig)
+            text += '\n--------------\n'
+            text += orig_text
+            pic = []
+        else:
+            raise CategoryNotSupport(post_type)
+        return Post('bilibili', text=text, url=url, pics=pic, target_name=target_name)
+

src/plugins/nonebot_bison/platform/ncm_artist.py

@@ -0,0 +1,53 @@
+from typing import Any, Optional
+
+import httpx
+from ..post import Post
+from ..types import RawPost, Target
+from .platform import TargetMixin, NewMessage
+
+class NcmArtist(TargetMixin, NewMessage):
+
+    categories = {}
+    platform_name = 'ncm-artist'
+    enable_tag = False
+    enabled = True
+    is_common = True
+    schedule_type = 'interval'
+    schedule_kw = {'minutes': 10}
+    name = "网易云-歌手"
+
+    async def get_target_name(self, target: Target) -> Optional[str]:
+        async with httpx.AsyncClient() as client:
+            res = await client.get(
+                    "https://music.163.com/api/artist/albums/{}".format(target),
+                    headers={'Referer': 'https://music.163.com/'}
+                    )
+            res_data = res.json()
+            if res_data['code'] != 200:
+                return
+            return res_data['artist']['name']
+
+    async def get_sub_list(self, target: Target) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            res = await client.get(
+                    "https://music.163.com/api/artist/albums/{}".format(target),
+                    headers={'Referer': 'https://music.163.com/'}
+                    )
+            res_data = res.json()
+            if res_data['code'] != 200:
+                return []
+            else:
+                return res_data['hotAlbums']
+
+    def get_id(self, post: RawPost) -> Any:
+        return post['id']
+
+    def get_date(self, post: RawPost) -> int:
+        return post['publishTime'] // 1000
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        text = '新专辑发布：{}'.format(raw_post['name'])
+        target_name = raw_post['artist']['name']
+        pics = [raw_post['picUrl']]
+        url = "https://music.163.com/#/album?id={}".format(raw_post['id'])
+        return Post('ncm-artist', text=text, url=url, pics=pics, target_name=target_name)

src/plugins/nonebot_bison/platform/platform.py

@@ -0,0 +1,384 @@
+from abc import abstractmethod, ABC
+from collections import defaultdict
+from dataclasses import dataclass
+from functools import reduce
+import time
+from typing import Any, Collection, Optional, Literal
+
+import httpx
+from nonebot import logger
+
+from ..plugin_config import plugin_config
+from ..post import Post
+from ..types import Category, RawPost, Tag, Target, User, UserSubInfo
+
+
+class CategoryNotSupport(Exception):
+    "raise in get_category, when post category is not supported"
+
+
+class RegistryMeta(type):
+
+    def __new__(cls, name, bases, namespace, **kwargs):
+        return super().__new__(cls, name, bases, namespace)
+
+    def __init__(cls, name, bases, namespace, **kwargs):
+        if kwargs.get('base'):
+            # this is the base class
+            cls.registory = []
+        elif not kwargs.get('abstract'):
+            # this is the subclass
+            cls.registory.append(cls)
+
+        super().__init__(name, bases, namespace, **kwargs)
+
+class RegistryABCMeta(RegistryMeta, ABC):
+    ...
+
+class StorageMixinProto(metaclass=RegistryABCMeta, abstract=True):
+    
+    has_target: bool
+
+    @abstractmethod
+    def get_stored_data(self, target: Target) -> Any:
+        ...
+
+    @abstractmethod
+    def set_stored_data(self, target: Target, data: Any):
+        ...
+
+class TargetMixin(StorageMixinProto, abstract=True):
+
+    has_target = True
+
+    def __init__(self):
+        super().__init__()
+        self.store: dict[Target, Any] = dict()
+
+    def get_stored_data(self, target: Target) -> Any:
+        return self.store.get(target)        
+
+    def set_stored_data(self, target: Target, data: Any):
+        self.store[target] = data
+
+
+class NoTargetMixin(StorageMixinProto, abstract=True):
+
+    has_target = False
+
+    def __init__(self):
+        super().__init__()
+        self.store = None
+
+    def get_stored_data(self, _: Target) -> Any:
+        return self.store
+
+    def set_stored_data(self, _: Target, data: Any):
+        self.store = data
+
+class PlatformNameMixin(metaclass=RegistryABCMeta, abstract=True):
+    platform_name: str
+    
+class CategoryMixin(metaclass=RegistryABCMeta, abstract=True):
+
+    @abstractmethod
+    def get_category(self, post: RawPost) -> Optional[Category]:
+        "Return category of given Rawpost"
+        raise NotImplementedError()
+
+class ParsePostMixin(metaclass=RegistryABCMeta, abstract=True):
+
+    @abstractmethod
+    async def parse(self, raw_post: RawPost) -> Post:
+        "parse RawPost into post"
+        ...
+
+class MessageProcessMixin(PlatformNameMixin, CategoryMixin, ParsePostMixin, abstract=True):
+    "General message process fetch, parse, filter progress"
+
+    def __init__(self):
+        super().__init__()
+        self.parse_cache: dict[Any, Post] = dict()
+
+    @abstractmethod
+    def get_id(self, post: RawPost) -> Any:
+        "Get post id of given RawPost"
+
+
+    async def _parse_with_cache(self, raw_post: RawPost) -> Post:
+        post_id = self.get_id(raw_post)
+        if post_id not in self.parse_cache:
+            retry_times = 3
+            while retry_times:
+                try:
+                    self.parse_cache[post_id] = await self.parse(raw_post)
+                    break
+                except Exception as err:
+                    retry_times -= 1
+                    if not retry_times:
+                        raise err
+        return self.parse_cache[post_id]
+
+    @abstractmethod
+    async def get_sub_list(self, target: Target) -> list[RawPost]:
+        "Get post list of the given target"
+
+    @abstractmethod
+    def get_date(self, post: RawPost) -> Optional[int]:
+        "Get post timestamp and return, return None if can't get the time"
+
+    async def filter_common(self, raw_post_list: list[RawPost]) -> list[RawPost]:
+        res = []
+        for raw_post in raw_post_list:
+            # post_id = self.get_id(raw_post)
+            # if post_id in exists_posts_set:
+            #     continue
+            if (post_time := self.get_date(raw_post)) and time.time() - post_time > 2 * 60 * 60 and \
+                    plugin_config.hk_reporter_init_filter:
+                continue
+            try:
+                self.get_category(raw_post)
+            except CategoryNotSupport:
+                continue
+            except NotImplementedError:
+                pass
+            res.append(raw_post)
+        return res
+
+class NewMessageProcessMixin(StorageMixinProto, MessageProcessMixin, abstract=True):
+    "General message process, fetch, parse, filter, and only returns the new Post"
+
+    @dataclass
+    class MessageStorage():
+        inited: bool
+        exists_posts: set[Any]
+
+    async def filter_common_with_diff(self, target: Target, raw_post_list: list[RawPost]) -> list[RawPost]:
+        filtered_post = await self.filter_common(raw_post_list)
+        store = self.get_stored_data(target) or self.MessageStorage(False, set())
+        res = []
+        if not store.inited and plugin_config.hk_reporter_init_filter:
+            # target not init
+            for raw_post in filtered_post:
+                post_id = self.get_id(raw_post)
+                store.exists_posts.add(post_id)
+            logger.info('init {}-{} with {}'.format(self.platform_name, target, store.exists_posts))
+            store.inited = True
+        else:
+            for raw_post in filtered_post:
+                post_id = self.get_id(raw_post)
+                if post_id in store.exists_posts:
+                    continue
+                res.append(raw_post)
+                store.exists_posts.add(post_id)
+        self.set_stored_data(target, store)
+        return res
+
+class UserCustomFilterMixin(CategoryMixin, ParsePostMixin, abstract=True):
+
+    categories: dict[Category, str]
+    enable_tag: bool
+
+    def __init__(self):
+        super().__init__()
+        self.reverse_category = {}
+        for key, val in self.categories.items():
+           self.reverse_category[val] = key
+
+    @abstractmethod
+    def get_tags(self, raw_post: RawPost) -> Optional[Collection[Tag]]:
+        "Return Tag list of given RawPost"
+
+    async def filter_user_custom(self, raw_post_list: list[RawPost], cats: list[Category], tags: list[Tag]) -> list[RawPost]:
+        res: list[RawPost] = []
+        for raw_post in raw_post_list:
+            if self.categories:
+                cat = self.get_category(raw_post)
+                if cats and cat not in cats:
+                    continue
+            if self.enable_tag and tags:
+                flag = False
+                post_tags = self.get_tags(raw_post)
+                for tag in post_tags or []:
+                    if tag in tags:
+                        flag = True
+                        break
+                if not flag:
+                    continue
+            res.append(raw_post)
+        return res
+
+    async def dispatch_user_post(self, target: Target, new_posts: list[RawPost], users: list[UserSubInfo]) -> list[tuple[User, list[Post]]]:
+        res: list[tuple[User, list[Post]]] = []
+        for user, category_getter, tag_getter in users:
+            required_tags = tag_getter(target) if self.enable_tag else []
+            cats = category_getter(target)
+            user_raw_post = await self.filter_user_custom(new_posts, cats, required_tags)
+            user_post: list[Post] = []
+            for raw_post in user_raw_post:
+                if isinstance(self, MessageProcessMixin):
+                    user_post.append(await self._parse_with_cache(raw_post))
+                else:
+                    user_post.append(await self.parse(raw_post))
+            res.append((user, user_post))
+        return res
+
+class Platform(PlatformNameMixin, UserCustomFilterMixin, base=True):
+    
+    # schedule_interval: int
+    schedule_type: Literal['date', 'interval', 'cron']
+    schedule_kw: dict
+    is_common: bool
+    enabled: bool
+    name: str
+
+    @abstractmethod
+    async def get_target_name(self, target: Target) -> Optional[str]:
+        ...
+
+    @abstractmethod
+    async def fetch_new_post(self, target: Target, users: list[UserSubInfo]) -> list[tuple[User, list[Post]]]:
+        ...
+
+class NewMessage(
+        Platform,
+        NewMessageProcessMixin,
+        UserCustomFilterMixin,
+        abstract=True
+        ):
+    "Fetch a list of messages, filter the new messages, dispatch it to different users"
+    
+    async def fetch_new_post(self, target: Target, users: list[UserSubInfo]) -> list[tuple[User, list[Post]]]:
+        try:
+            post_list = await self.get_sub_list(target)
+            new_posts = await self.filter_common_with_diff(target, post_list)
+            if not new_posts:
+                return []
+            else:
+                for post in new_posts:
+                    logger.info('fetch new post from {} {}: {}'.format(
+                        self.platform_name,
+                        target if self.has_target else '-',
+                        self.get_id(post)))
+            res = await self.dispatch_user_post(target, new_posts, users)
+            self.parse_cache = {}
+            return res
+        except httpx.RequestError as err:
+            logger.warning("network connection error: {}, url: {}".format(type(err), err.request.url))
+            return []
+
+class StatusChange(
+        Platform,
+        StorageMixinProto,
+        UserCustomFilterMixin,
+        abstract=True
+        ):
+    "Watch a status, and fire a post when status changes"
+
+    @abstractmethod
+    async def get_status(self, target: Target) -> Any:
+        ...
+
+    @abstractmethod
+    def compare_status(self, target: Target, old_status, new_status) -> list[RawPost]:
+        ...
+
+    @abstractmethod
+    async def parse(self, raw_post: RawPost) -> Post:
+        ...
+
+    async def fetch_new_post(self, target: Target, users: list[UserSubInfo]) -> list[tuple[User, list[Post]]]:
+        try:
+            new_status = await self.get_status(target)
+            res = []
+            if old_status := self.get_stored_data(target):
+                diff = self.compare_status(target, old_status, new_status)
+                if diff:
+                    logger.info("status changes {} {}: {} -> {}".format(
+                        self.platform_name,
+                        target if self.has_target else '-',
+                        old_status, new_status
+                        ))
+                    res = await self.dispatch_user_post(target, diff, users)
+            self.set_stored_data(target, new_status)
+            return res
+        except httpx.RequestError as err:
+            logger.warning("network connection error: {}, url: {}".format(type(err), err.request.url))
+            return []
+
+class SimplePost(
+        Platform,
+        MessageProcessMixin,
+        UserCustomFilterMixin,
+        StorageMixinProto,
+        abstract=True
+        ):
+    "Fetch a list of messages, dispatch it to different users"
+
+    async def fetch_new_post(self, target: Target, users: list[UserSubInfo]) -> list[tuple[User, list[Post]]]:
+        try:
+            new_posts = await self.get_sub_list(target)
+            if not new_posts:
+                return []
+            else:
+                for post in new_posts:
+                    logger.info('fetch new post from {} {}: {}'.format(
+                        self.platform_name,
+                        target if self.has_target else '-',
+                        self.get_id(post)))
+            res = await self.dispatch_user_post(target, new_posts, users)
+            self.parse_cache = {}
+            return res
+        except httpx.RequestError as err:
+            logger.warning("network connection error: {}, url: {}".format(type(err), err.request.url))
+            return []
+
+class NoTargetGroup(
+        Platform,
+        NoTargetMixin,
+        UserCustomFilterMixin,
+        abstract=True
+        ):
+    enable_tag = False
+    DUMMY_STR = '_DUMMY'
+    enabled = True
+
+    class PlatformProto(Platform, NoTargetMixin, UserCustomFilterMixin, abstract=True):
+        ...
+
+    def __init__(self, platform_list: list[PlatformProto]):
+        self.platform_list = platform_list
+        name = self.DUMMY_STR
+        self.categories = {}
+        categories_keys = set()
+        self.schedule_type = platform_list[0].schedule_type
+        self.schedule_kw = platform_list[0].schedule_kw
+        for platform in platform_list:
+            if name == self.DUMMY_STR:
+                name = platform.name
+            elif name != platform.name:
+                raise RuntimeError('Platform name for {} not fit'.format(self.platform_name))
+            platform_category_key_set = set(platform.categories.keys())
+            if platform_category_key_set & categories_keys:
+                raise RuntimeError('Platform categories for {} duplicate'.format(self.platform_name))
+            categories_keys |= platform_category_key_set
+            self.categories.update(platform.categories)
+            if platform.schedule_kw != self.schedule_kw or platform.schedule_type != self.schedule_type:
+                raise RuntimeError('Platform scheduler for {} not fit'.format(self.platform_name))
+        self.name = name
+        self.is_common = platform_list[0].is_common
+        super().__init__()
+
+    def __str__(self):
+        return '[' + ' '.join(map(lambda x: x.name, self.platform_list)) + ']'
+
+    async def get_target_name(self, _):
+        return await self.platform_list[0].get_target_name(_)
+
+    async def fetch_new_post(self, target, users):
+        res = defaultdict(list)
+        for platform in self.platform_list:
+            platform_res = await platform.fetch_new_post(target=target, users=users)
+            for user, posts in platform_res:
+                res[user].extend(posts)
+        return [[key, val] for key, val in res.items()]

src/plugins/nonebot_bison/platform/rss.py

@@ -0,0 +1,49 @@
+import calendar
+from typing import Any, Optional
+
+from bs4 import BeautifulSoup as bs
+import feedparser
+import httpx
+
+from ..post import Post
+from ..types import RawPost, Target
+from .platform import NewMessage, TargetMixin
+
+class Rss(NewMessage, TargetMixin):
+
+    categories = {}
+    enable_tag = False
+    platform_name = 'rss'
+    name = "Rss"
+    enabled = True
+    is_common = True
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 30}
+
+    async def get_target_name(self, target: Target) -> Optional[str]:
+        async with httpx.AsyncClient() as client:
+            res = await client.get(target, timeout=10.0)
+            feed = feedparser.parse(res.text)
+            return feed['feed']['title']
+
+    def get_date(self, post: RawPost) -> int:
+        return calendar.timegm(post.published_parsed)
+
+    def get_id(self, post: RawPost) -> Any:
+        return post.id
+
+    async def get_sub_list(self, target: Target) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            res = await client.get(target, timeout=10.0)
+            feed = feedparser.parse(res)
+            entries = feed.entries
+            for entry in entries:
+                entry['_target_name'] = feed.feed.title
+            return feed.entries
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        text = raw_post.get('title', '') + '\n' if raw_post.get('title') else ''
+        soup = bs(raw_post.description, 'html.parser')
+        text += soup.text.strip()
+        pics = list(map(lambda x: x.attrs['src'], soup('img')))
+        return Post('rss', text=text, url=raw_post.link, pics=pics, target_name=raw_post['_target_name'])

src/plugins/nonebot_bison/platform/wechat.py

@@ -0,0 +1,78 @@
+from datetime import datetime
+import hashlib
+import json
+import re
+from typing import Any, Optional
+
+from bs4 import BeautifulSoup as bs
+import httpx
+
+from ..post import Post
+from ..types import *
+# from .platform import Platform
+
+
+# class Wechat(Platform):
+
+#     categories = {}
+#     enable_tag = False
+#     platform_name = 'wechat'
+#     enabled = False
+#     is_common = False
+#     name = '微信公众号'
+
+#     @classmethod
+#     def _get_query_url(cls, target: Target):
+#         return 'https://weixin.sogou.com/weixin?type=1&s_from=input&query={}&ie=utf8&_sug_=n&_sug_type_='.format(target)
+
+#     @classmethod
+#     async def _get_target_soup(cls, target: Target) -> Optional[bs]:
+#         target_url = cls._get_query_url(target)
+#         async with httpx.AsyncClient() as client:
+#             res = await client.get(target_url)
+#         soup = bs(res.text, 'html.parser')
+#         blocks = soup.find(class_='news-list2').find_all('li',recursive=False)
+#         for block in blocks:
+#             if block.find(string=[target]):
+#                 return block
+
+#     @classmethod
+#     async def get_account_name(cls, target: Target) -> Optional[str]:
+#         if not (block := await cls._get_target_soup(target)):
+#             return None
+#         return block.find('p', class_='tit').find('a').text
+
+#     async def get_sub_list(self, target: Target) -> list[RawPost]:
+#         block = await self._get_target_soup(target)
+#         if (last_post_dt := block.find('dt', string='最近文章：')):
+#             post = {
+#                     'title': last_post_dt.find_parent().find('a').text,
+#                     'target': target,
+#                     'page_url': self._get_query_url(target),
+#                     'name': block.find('p', class_='tit').find('a').text
+#                     }
+#             return [post]
+#         else:
+#             return []
+
+#     def get_id(self, post: RawPost) -> Any:
+#         return post['title']
+
+#     def get_date(self, post: RawPost):
+#         return None
+
+#     def get_tags(self, post: RawPost):
+#         return None
+
+#     def get_category(self, post: RawPost):
+#         return None
+
+#     async def parse(self, raw_post: RawPost) -> Post:
+#         # TODO get content of post
+#         return Post(target_type='wechat',
+#                 text='{}\n详细内容请自行查看公众号'.format(raw_post['title']),
+#                 target_name=raw_post['name'],
+#                 pics=[],
+#                 url=''
+#                 )
+

src/plugins/nonebot_bison/platform/weibo.py

@@ -0,0 +1,120 @@
+from datetime import datetime
+import json
+import re
+from typing import Any, Optional
+
+from bs4 import BeautifulSoup as bs
+import httpx
+from nonebot import logger
+
+from ..post import Post
+from ..types import *
+from .platform import NewMessage, TargetMixin
+
+class Weibo(NewMessage, TargetMixin):
+
+    categories = {
+            1: '转发',
+            2: '视频',
+            3: '图文',
+            4: '文字',
+        }
+    enable_tag = True
+    platform_name = 'weibo'
+    name = '新浪微博'
+    enabled = True
+    is_common = True
+    schedule_type = 'interval'
+    schedule_kw = {'seconds': 3}
+
+    async def get_target_name(self, target: Target) -> Optional[str]:
+        async with httpx.AsyncClient() as client:
+            param = {'containerid': '100505' + target}
+            res = await client.get('https://m.weibo.cn/api/container/getIndex', params=param)
+            res_dict = json.loads(res.text)
+            if res_dict.get('ok') == 1:
+                return res_dict['data']['userInfo']['screen_name']
+            else:
+                return None
+
+    async def get_sub_list(self, target: Target) -> list[RawPost]:
+        async with httpx.AsyncClient() as client:
+            params = { 'containerid': '107603' + target}
+            res = await client.get('https://m.weibo.cn/api/container/getIndex?', params=params, timeout=4.0)
+            res_data = json.loads(res.text)
+            if not res_data['ok']:
+                return []
+            custom_filter: Callable[[RawPost], bool] = lambda d: d['card_type'] == 9
+            return list(filter(custom_filter, res_data['data']['cards']))
+
+    def get_id(self, post: RawPost) -> Any:
+        return post['mblog']['id']
+
+    def filter_platform_custom(self, raw_post: RawPost) -> bool:
+        return raw_post['card_type'] == 9
+
+    def get_date(self, raw_post: RawPost) -> float:
+        created_time = datetime.strptime(raw_post['mblog']['created_at'], '%a %b %d %H:%M:%S %z %Y')
+        return created_time.timestamp()
+
+    def get_tags(self, raw_post: RawPost) -> Optional[list[Tag]]:
+        "Return Tag list of given RawPost"
+        text = raw_post['mblog']['text']
+        soup = bs(text, 'html.parser')
+        res = list(map(
+            lambda x: x[1:-1],
+            filter(
+                lambda s: s[0] == '#' and s[-1] == '#',
+                map(lambda x:x.text, soup.find_all('span', class_='surl-text'))
+                )
+            ))
+        super_topic_img = soup.find('img', src=re.compile(r'timeline_card_small_super_default')) 
+        if super_topic_img:
+            try:
+                res.append(super_topic_img.parent.parent.find('span', class_='surl-text').text + '超话')
+            except:
+                logger.info('super_topic extract error: {}'.format(text))
+        return res
+
+    def get_category(self, raw_post: RawPost) -> Category:
+        if raw_post['mblog'].get('retweeted_status'):
+            return Category(1)
+        elif raw_post['mblog'].get('page_info') and raw_post['mblog']['page_info'].get('type') == 'video':
+            return Category(2)
+        elif raw_post['mblog'].get('pics'):
+            return Category(3)
+        else:
+            return Category(4)
+
+    def _get_text(self, raw_text: str) -> str:
+        text = raw_text.replace('<br />', '\n')
+        return bs(text, 'html.parser').text
+
+    async def parse(self, raw_post: RawPost) -> Post:
+        header = {
+                'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9',
+                'accept-language': 'zh-CN,zh;q=0.9',
+                'authority': 'm.weibo.cn',
+                'cache-control': 'max-age=0',
+                'sec-fetch-dest': 'empty',
+                'sec-fetch-mode': 'same-origin',
+                'sec-fetch-site': 'same-origin',
+                'upgrade-insecure-requests': '1',
+                'user-agent': 'Mozilla/5.0 (Linux; Android 6.0; Nexus 5 Build/MRA58N) '
+                               'AppleWebKit/537.36 (KHTML, like Gecko) Chrome/89.0.4389.72 '
+                               'Mobile Safari/537.36'}
+        info = raw_post['mblog']
+        if info['isLongText'] or info['pic_num'] > 9:
+            async with httpx.AsyncClient() as client:
+                res = await client.get('https://m.weibo.cn/detail/{}'.format(info['mid']), headers=header)
+            try:
+                full_json_text = re.search(r'"status": ([\s\S]+),\s+"hotScheme"', res.text).group(1)
+                info = json.loads(full_json_text)
+            except:
+                logger.info('detail message error: https://m.weibo.cn/detail/{}'.format(info['mid']))
+        parsed_text = self._get_text(info['text'])
+        pic_urls = [img['large']['url'] for img in info.get('pics', [])]
+        detail_url = 'https://weibo.com/{}/{}'.format(info['user']['id'], info['bid'])
+        # return parsed_text, detail_url, pic_urls
+        return Post('weibo', text=parsed_text, url=detail_url, pics=pic_urls, target_name=info['user']['screen_name'])
+

src/plugins/nonebot_bison/plugin_config.py

@@ -0,0 +1,21 @@
+from pydantic import BaseSettings
+
+import warnings
+import nonebot
+
+class PlugConfig(BaseSettings):
+
+    hk_reporter_config_path: str = ""
+    hk_reporter_use_pic: bool = False
+    hk_reporter_use_local: bool = False
+    hk_reporter_browser: str = ''
+    hk_reporter_init_filter: bool = True
+    hk_reporter_use_queue: bool = True
+
+    class Config:
+        extra = 'ignore'
+
+global_config = nonebot.get_driver().config
+plugin_config = PlugConfig(**global_config.dict())
+if plugin_config.hk_reporter_use_local:
+    warnings.warn('HK_REPORTER_USE_LOCAL is deprecated, please use HK_REPORTER_BROWSER')

src/plugins/nonebot_bison/post.py

@@ -0,0 +1,146 @@
+import base64
+from dataclasses import dataclass, field
+from functools import reduce
+from io import BytesIO
+from typing import Optional, Union
+
+from PIL import Image
+import httpx
+from nonebot import logger
+from nonebot.adapters.cqhttp.message import Message, MessageSegment
+
+from .plugin_config import plugin_config
+from .utils import parse_text
+
+@dataclass
+class Post:
+
+    target_type: str
+    text: str
+    url: Optional[str] = None
+    target_name: Optional[str] = None
+    compress: bool = False
+    override_use_pic: Optional[bool] = None
+    pics: Union[list[Union[str,bytes]], list[str], list[bytes]] = field(default_factory=list)
+    extra_msg: list[Message] = field(default_factory=list)
+
+    _message: Optional[list] = None
+
+    def _use_pic(self):
+        if not self.override_use_pic is None:
+            return self.override_use_pic
+        return plugin_config.hk_reporter_use_pic
+
+    async def _pic_url_to_image(self, data: Union[str, bytes]) -> Image.Image:
+        pic_buffer = BytesIO()
+        if isinstance(data, str):
+            async with httpx.AsyncClient() as client:
+                res = await client.get(data)
+            pic_buffer.write(res.content)
+        else:
+            pic_buffer.write(data)
+        return Image.open(pic_buffer)
+
+    def _check_image_square(self, size: tuple[int, int]) -> bool:
+        return abs(size[0] - size[1]) / size[0] < 0.01
+
+    async def _pic_merge(self) -> None:
+        if len(self.pics) < 3:
+            return
+        first_image = await self._pic_url_to_image(self.pics[0])
+        if not self._check_image_square(first_image.size):
+            return
+        images: list[Image.Image] = [first_image]
+        # first row
+        for i in range(1, 3):
+            cur_img = await self._pic_url_to_image(self.pics[i])
+            if not self._check_image_square(cur_img.size):
+                return
+            if cur_img.size[1] != images[0].size[1]: # height not equal
+                return
+            images.append(cur_img)
+        _tmp = 0
+        x_coord = [0]
+        for i in range(3):
+            _tmp += images[i].size[0]
+            x_coord.append(_tmp)
+        y_coord = [0, first_image.size[1]]
+        async def process_row(row: int) -> bool:
+            if len(self.pics) < (row + 1) * 3:
+                return False
+            row_first_img = await self._pic_url_to_image(self.pics[row * 3])
+            if not self._check_image_square(row_first_img.size):
+                return False
+            if row_first_img.size[0] != images[0].size[0]:
+                return False
+            image_row: list[Image.Image] = [row_first_img]
+            for i in range(row * 3 + 1, row * 3 + 3):
+                cur_img = await self._pic_url_to_image(self.pics[i])
+                if not self._check_image_square(cur_img.size):
+                    return False
+                if cur_img.size[1] != row_first_img.size[1]:
+                    return False
+                if cur_img.size[0] != images[i % 3].size[0]:
+                    return False
+                image_row.append(cur_img)
+            images.extend(image_row)
+            y_coord.append(y_coord[-1] + row_first_img.size[1])
+            return True
+        if await process_row(1):
+            matrix = (3,2)
+        else:
+            matrix = (3,1)
+        if await process_row(2):
+            matrix = (3,3)
+        logger.info('trigger merge image')
+        target = Image.new('RGB', (x_coord[-1], y_coord[-1]))
+        for y in range(matrix[1]):
+            for x in range(matrix[0]):
+                target.paste(images[y * matrix[0] + x], (
+                    x_coord[x], y_coord[y], x_coord[x+1], y_coord[y+1]
+                    ))
+        target_io = BytesIO()
+        target.save(target_io, 'JPEG')
+        self.pics = self.pics[matrix[0] * matrix[1]: ]
+        self.pics.insert(0, target_io.getvalue())
+
+    async def generate_messages(self):
+        if self._message is None:
+            await self._pic_merge()
+            msgs = []
+            text = ''
+            if self.text:
+                if self._use_pic():
+                    text += '{}'.format(self.text)
+                else:
+                    text += '{}'.format(self.text if len(self.text) < 500 else self.text[:500] + '...')
+            text += '\n来源: {}'.format(self.target_type)
+            if self.target_name:
+                text += ' {}'.format(self.target_name)
+            if self._use_pic():
+                msgs.append(await parse_text(text))
+                if not self.target_type == 'rss' and self.url:
+                    msgs.append(MessageSegment.text(self.url))
+            else:
+                if self.url:
+                    text += ' \n详情: {}'.format(self.url)
+                msgs.append(MessageSegment.text(text))
+            for pic in self.pics:
+                # if isinstance(pic, bytes):
+                #     pic = 'base64://' + base64.b64encode(pic).decode()
+                # msgs.append(Message("[CQ:image,file={url}]".format(url=pic)))
+                msgs.append(MessageSegment.image(pic))
+            if self.compress:
+                msgs = [reduce(lambda x, y: x.append(y), msgs, Message())]
+            msgs.extend(self.extra_msg)
+            self._message = msgs
+        return self._message
+
+    def __str__(self):
+        return 'type: {}\nfrom: {}\ntext: {}\nurl: {}\npic: {}'.format(
+                self.target_type,
+                self.target_name,
+                self.text if len(self.text) < 500 else self.text[:500] + '...',
+                self.url,
+                ', '.join(map(lambda x: 'b64img' if isinstance(x, bytes) or x.startswith('base64') else x, self.pics))
+            )

src/plugins/nonebot_bison/scheduler.py

@@ -0,0 +1,69 @@
+import logging
+
+from apscheduler.schedulers.asyncio import AsyncIOScheduler
+import logging
+import nonebot
+from nonebot import get_driver, logger
+from nonebot.log import LoguruHandler
+
+from .config import Config
+from .platform import platform_manager
+from .send import do_send_msgs
+from .send import send_msgs
+from .types import UserSubInfo
+from .plugin_config import plugin_config
+
+scheduler = AsyncIOScheduler()
+
+@get_driver().on_startup
+async def _start():
+    scheduler.configure({"apscheduler.timezone": "Asia/Shanghai"})
+    scheduler.start()
+
+# get_driver().on_startup(_start)
+
+async def fetch_and_send(target_type: str):
+    config = Config()
+    target = config.get_next_target(target_type)
+    if not target:
+        return
+    logger.debug('try to fecth new posts from {}, target: {}'.format(target_type, target))
+    send_user_list = config.target_user_cache[target_type][target]
+    send_userinfo_list = list(map(
+        lambda user: UserSubInfo(
+            user,
+            lambda target: config.get_sub_category(target_type, target, user.user_type, user.user),
+            lambda target: config.get_sub_tags(target_type, target, user.user_type, user.user)
+        ), send_user_list))
+    bot_list = list(nonebot.get_bots().values())
+    bot = bot_list[0] if bot_list else None
+    to_send = await platform_manager[target_type].fetch_new_post(target, send_userinfo_list)
+    for user, send_list in to_send:
+        for send_post in send_list:
+            logger.info('send to {}: {}'.format(user, send_post))
+            if not bot:
+                logger.warning('no bot connected')
+            else:
+                await send_msgs(bot, user.user, user.user_type, await send_post.generate_messages())
+
+for platform_name, platform in platform_manager.items():
+    if platform.schedule_type in ['cron', 'interval', 'date']:
+        logger.info(f'start scheduler for {platform_name} with {platform.schedule_type} {platform.schedule_kw}')
+        scheduler.add_job(
+                fetch_and_send, platform.schedule_type, **platform.schedule_kw,
+                args=(platform_name,))
+
+if plugin_config.hk_reporter_use_queue:
+    scheduler.add_job(do_send_msgs, 'interval', seconds=0.3, coalesce=True)
+
+    class SchedulerLogFilter(logging.Filter):
+
+        def filter(self, record: logging.LogRecord) -> bool:
+            logger.debug("logRecord", record, record.getMessage())
+            return not (record.name == "apscheduler" and 'skipped: maximum number of running instances reached' in record.getMessage())
+
+    aps_logger = logging.getLogger("apscheduler")
+    aps_logger.setLevel(30)
+    aps_logger.addFilter(SchedulerLogFilter())
+    aps_logger.handlers.clear()
+    aps_logger.addHandler(LoguruHandler())

src/plugins/nonebot_bison/send.py

@@ -0,0 +1,43 @@
+import time
+
+from nonebot import logger
+from nonebot.adapters.cqhttp.bot import Bot
+
+from .plugin_config import plugin_config
+
+QUEUE = []
+LAST_SEND_TIME = time.time()
+
+async def _do_send(bot: 'Bot', user: str, user_type: str, msg):
+    if user_type == 'group':
+        await bot.call_api('send_group_msg', group_id=user, message=msg)
+    elif user_type == 'private':
+        await bot.call_api('send_private_msg', user_id=user, message=msg)
+
+async def do_send_msgs():
+    global LAST_SEND_TIME
+    if time.time() - LAST_SEND_TIME < 1.5:
+        return
+    if QUEUE:
+        bot, user, user_type, msg, retry_time = QUEUE.pop(0)
+        try:
+            await _do_send(bot, user, user_type, msg)
+        except Exception as e:
+            if retry_time > 0:
+                QUEUE.insert(0, (bot, user, user_type, msg, retry_time - 1))
+            else:
+                msg_str = str(msg)
+                if len(msg_str) > 50:
+                    msg_str = msg_str[:50] + '...'
+                logger.warning(f'send msg err {e} {msg_str}')
+        LAST_SEND_TIME = time.time()
+
+async def send_msgs(bot, user, user_type, msgs):
+    if plugin_config.hk_reporter_use_queue:
+        for msg in msgs:
+            QUEUE.append((bot, user, user_type, msg, 2))
+    else:
+        for msg in msgs:
+            await _do_send(bot, user, user_type, msg)
+
+

src/plugins/nonebot_bison/types.py

@@ -0,0 +1,17 @@
+from typing import Any, Callable, NamedTuple, NewType
+from dataclasses import dataclass
+
+RawPost = NewType('RawPost', Any)
+Target = NewType('Target', str)
+Category = NewType('Category', int)
+Tag = NewType('Tag', str)
+
+@dataclass(eq=True, frozen=True)
+class User:
+    user: str
+    user_type: str
+
+class UserSubInfo(NamedTuple):
+    user: User
+    category_getter: Callable[[Target], list[Category]]
+    tag_getter: Callable[[Target], list[Tag]]

src/plugins/nonebot_bison/utils.py

@@ -0,0 +1,138 @@
+import asyncio
+import base64
+from html import escape
+import os
+from time import asctime
+import re
+from typing import Awaitable, Callable, Optional
+
+from nonebot.adapters.cqhttp.message import MessageSegment
+from nonebot.log import logger
+from pyppeteer import connect, launch
+from pyppeteer.browser import Browser
+from pyppeteer.chromium_downloader import check_chromium, download_chromium
+from pyppeteer.page import Page
+
+from bs4 import BeautifulSoup as bs
+
+from .plugin_config import plugin_config
+
+class Singleton(type):
+    _instances = {}
+    def __call__(cls, *args, **kwargs):
+        if cls not in cls._instances:
+            cls._instances[cls] = super(Singleton, cls).__call__(*args, **kwargs)
+        return cls._instances[cls]
+
+if not plugin_config.hk_reporter_browser and not plugin_config.hk_reporter_use_local \
+        and not check_chromium():
+    os.environ['PYPPETEER_DOWNLOAD_HOST'] = 'http://npm.taobao.org/mirrors'
+    download_chromium()
+
+class Render(metaclass=Singleton):
+
+    def __init__(self):
+        self.lock = asyncio.Lock()
+        self.browser: Browser
+        self.interval_log = ''
+        self.remote_browser = False
+
+    async def get_browser(self) -> Browser:
+        if plugin_config.hk_reporter_browser:
+            if plugin_config.hk_reporter_browser.startswith('local:'):
+                path = plugin_config.hk_reporter_browser.split('local:', 1)[1]
+                return await launch(executablePath=path, args=['--no-sandbox'])
+            if plugin_config.hk_reporter_browser.startswith('ws:'):
+                self.remote_browser = True
+                return await connect(browserWSEndpoint=plugin_config.hk_reporter_browser)
+            raise RuntimeError('HK_REPORTER_BROWSER error')
+        if plugin_config.hk_reporter_use_local:
+            return await launch(executablePath='/usr/bin/chromium', args=['--no-sandbox'])
+        return await launch(args=['--no-sandbox'])
+
+    async def close_browser(self):
+        if not self.remote_browser:
+            await self.browser.close()
+
+    async def render(self, url: str, viewport: Optional[dict] = None, target: Optional[str] = None,
+            operation: Optional[Callable[[Page], Awaitable[None]]] = None) -> Optional[bytes]:
+        retry_times = 0
+        while retry_times < 3:
+            try:
+                return await asyncio.wait_for(self.do_render(url, viewport, target, operation), 20)
+            except asyncio.TimeoutError:
+                retry_times += 1
+                logger.warning("render error {}\n".format(retry_times) + self.interval_log)
+                self.interval_log = ''
+                # if self.browser:
+                #     await self.browser.close()
+                #     self.lock.release()
+
+    def _inter_log(self, message: str) -> None:
+        self.interval_log += asctime() + '' + message + '\n'
+
+    async def do_render(self, url: str, viewport: Optional[dict] = None, target: Optional[str] = None,
+            operation: Optional[Callable[[Page], Awaitable[None]]] = None) -> Optional[bytes]:
+        async with self.lock:
+            self.browser = await self.get_browser()
+            self._inter_log('open browser')
+            page = await self.browser.newPage()
+            if operation:
+                await operation(page)
+            else:
+                await page.goto(url)
+            self._inter_log('open page')
+            if viewport:
+                await page.setViewport(viewport)
+                self._inter_log('set viewport')
+            if target:
+                target_ele = await page.querySelector(target)
+                if not target_ele:
+                    return None
+                data = await target_ele.screenshot(type='jpeg', encoding='binary')
+            else:
+                data = await page.screenshot(type='jpeg', encoding='binary')
+            self._inter_log('screenshot')
+            await page.close()
+            self._inter_log('close page')
+            await self.close_browser()
+            self._inter_log('close browser')
+            assert(isinstance(data, bytes))
+            return data
+
+    async def text_to_pic(self, text: str) -> Optional[bytes]:
+        lines = text.split('\n')
+        parsed_lines = list(map(lambda x: '<p>{}</p>'.format(escape(x)), lines))
+        html_text = '<div style="width:17em;padding:1em">{}</div>'.format(''.join(parsed_lines))
+        url = 'data:text/html;charset=UTF-8;base64,{}'.format(base64.b64encode(html_text.encode()).decode())
+        data = await self.render(url, target='div')
+        return data
+
+    async def text_to_pic_cqcode(self, text:str) -> MessageSegment:
+        data = await self.text_to_pic(text)
+        # logger.debug('file size: {}'.format(len(data)))
+        if data:
+            # logger.debug(code)
+            return MessageSegment.image(data)
+        else:
+            return MessageSegment.text('生成图片错误')
+
+async def parse_text(text: str) -> MessageSegment:
+    'return raw text if don\'t use pic, otherwise return rendered opcode'
+    if plugin_config.hk_reporter_use_pic:
+        render = Render()
+        return await render.text_to_pic_cqcode(text)
+    else:
+        return MessageSegment.text(text)
+
+def html_to_text(html: str, query_dict: dict = {}) -> str:
+    html = re.sub(r'<br\s*/?>', '<br>\n', html)
+    html = html.replace('</p>', '</p>\n')
+    soup = bs(html, 'html.parser')
+    if query_dict:
+        node = soup.find(**query_dict)
+    else:
+        node = soup
+    assert node is not None
+    return node.text.strip()
+