feat: 新增 on_message 处理消息方式

refactor: 优化正则匹配，修改消息获取方式

README.md

@@ -54,6 +54,7 @@ analysis_display_image = true # 是否显示封面 true/false
 # 哪种类型需要显示封面，与上一项相冲突，上一项为true则全开 List[str]
 analysis_display_image_list = ["video", "bangumi", "live", "article", "dynamic"]
 analysis_enable_search = false # 是否开启搜视频功能 true/false  example: "搜视频 123456"
+analysis_use_on_message = false # 是否使用 on_message 处理消息，并且处理合并转发的信息，如果遇到 cpu 占用拉满的问题 #29，可以考虑打开此项
 
 
 analysis_trust_env = false # 是否使用环境变量或者当前系统正在使用中的代理设置 true/false

nonebot_plugin_analysis_bilibili/__init__.py

@@ -1,15 +1,14 @@
 import re
 from typing import List, Union
 from aiohttp import ClientSession
-from nonebot import on_regex, logger, require
+from nonebot import on_regex, logger, require, on_message
 from nonebot.adapters import Event
 from nonebot.rule import Rule
 from nonebot.plugin import PluginMetadata
-from nonebot.params import RegexStr
 from .analysis_bilibili import config, b23_extract, bili_keyword, search_bili_by_title
 
 require("nonebot_plugin_saa")
-from nonebot_plugin_saa import (
+from nonebot_plugin_saa import (  # noqa: E402
     MessageFactory,
     MessageSegmentFactory,
     Text,
@@ -35,6 +34,7 @@
 desc_blacklist = [str(i) for i in getattr(config, "analysis_desc_blacklist", [])]
 trust_env = getattr(config, "analysis_trust_env", False)
 enable_search = getattr(config, "analysis_enable_search", False)
+use_on_message = getattr(config, "analysis_use_on_message", False)
 
 
 async def is_enable_search() -> bool:
@@ -63,15 +63,22 @@ async def is_normal(event: Event) -> bool:
     return True
 
 
+pattern = (
+    r"^(?:(?:av|cv)\d+|BV[a-zA-Z0-9]{10})|"
+    r"(?:b23\.tv|bili(?:22|23|33|2233)\.cn|\.bilibili\.com|QQ小程序(?:]|]|\])哔哩哔哩).{0,500}"
+)
+
 analysis_bili = on_regex(
-    r"^(?:(av|cv)\d+|BV[a-zA-Z0-9]{10})|"
-    r".*(?:b23\.tv|bili(?:22|23|33|2233)\.cn|\.bilibili\.com|QQ小程序(?:]|]|\])哔哩哔哩).*",
-    flags=re.I,
+    pattern,
     rule=is_normal,
     block=False,
     priority=11,
 )
 
+if use_on_message:
+    analysis_bili = on_message(rule=is_normal, block=False, priority=11)
+
+
 rule = Rule(is_enable_search, is_normal)
 search_bili = on_regex(r"^搜视频.*", rule=rule)
 
@@ -163,12 +170,23 @@ async def get_msg(
 
 
 @analysis_bili.handle()
-async def handle_analysis(event: Event, message=RegexStr()) -> None:
-    msg = await get_msg(event, message)
+async def handle_analysis(event: Event) -> None:
+    message = event.get_message()
+    # on_message
+    if use_on_message:
+        # 不解析转发消息，可能会过长导致匹配时间过长
+        for segment in message:
+            if segment.type == "forward":
+                logger.debug("analysis_bilibili 忽略转发消息")
+                return
+        if not re.search(pattern, str(message)):
+            return
+    # on_regex
+    msg = await get_msg(event, str(message))
     await send_msg(msg)
 
 
 @search_bili.handle()
-async def handle_search(event: Event, message=RegexStr()) -> None:
-    msg = await get_msg(event, message[3:].strip(), search=True)
+async def handle_search(event: Event) -> None:
+    msg = await get_msg(event, str(event.get_message())[3:].strip(), search=True)
     await send_msg(msg)

nonebot_plugin_analysis_bilibili/analysis_bilibili.py

@@ -1,7 +1,7 @@
 import re
 import nonebot
 
-from time import time, localtime, strftime
+from time import localtime, strftime
 from typing import Dict, List, Optional, Tuple, Union
 from aiohttp import ClientSession
 
@@ -73,7 +73,7 @@ async def bili_keyword(
 
 
 async def b23_extract(text: str, session: ClientSession) -> str:
-    b23 = re.compile(r"b23.tv/(\w+)|(bili(22|23|33|2233).cn)/(\w+)", re.I).search(
+    b23 = re.compile(r"b23.tv/(\w+)|(bili(22|23|33|2233).cn)/(\w+)").search(
         text.replace("\\", "")
     )
     url = f"https://{b23[0]}"
@@ -90,27 +90,25 @@ def extract(text: str) -> Tuple[str, Optional[str], Optional[str]]:
         # 视频播放定位时间
         time = re.compile(r"([?&]|&)t=\d+").search(text)
         # 主站视频 av 号
-        aid = re.compile(r"av\d+", re.I).search(text)
+        aid = re.compile(r"av\d+").search(text)
         # 主站视频 bv 号
-        bvid = re.compile(r"BV([A-Za-z0-9]{10})+", re.I).search(text)
+        bvid = re.compile(r"BV([A-Za-z0-9]{10})+").search(text)
         # 番剧视频页
-        epid = re.compile(r"ep\d+", re.I).search(text)
+        epid = re.compile(r"ep\d+").search(text)
         # 番剧剧集ssid(season_id)
-        ssid = re.compile(r"ss\d+", re.I).search(text)
+        ssid = re.compile(r"ss\d+").search(text)
         # 番剧详细页
-        mdid = re.compile(r"md\d+", re.I).search(text)
+        mdid = re.compile(r"md\d+").search(text)
         # 直播间
-        room_id = re.compile(r"live.bilibili.com/(blanc/|h5/)?(\d+)", re.I).search(text)
+        room_id = re.compile(r"live.bilibili.com/(blanc/|h5/)?(\d+)").search(text)
         # 文章
-        cvid = re.compile(
-            r"(/read/(cv|mobile|native)(/|\?id=)?|^cv)(\d+)", re.I
-        ).search(text)
+        cvid = re.compile(r"(/read/(cv|mobile|native)(/|\?id=)?|^cv)(\d+)").search(text)
         # 动态
         dynamic_id_type2 = re.compile(
-            r"(t|m).bilibili.com/(\d+)\?(.*?)(&|&)type=2", re.I
+            r"(t|m).bilibili.com/(\d+)\?(.*?)(&|&)type=2"
         ).search(text)
         # 动态
-        dynamic_id = re.compile(r"(t|m).bilibili.com/(opus/)?(\d+)", re.I).search(text)
+        dynamic_id = re.compile(r"(t|m).bilibili.com/(opus/)?(\d+)").search(text)
         if bvid:
             url = f"https://api.bilibili.com/x/web-interface/view?bvid={bvid[0]}"
         elif aid:

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "nonebot-plugin-analysis-bilibili"
-version = "2.7.7"
+version = "2.8.0"
 description = "nonebot2解析bilibili插件"
 authors = ["mengshouer"]
 license = "MIT"