LiteyukiStudio/nonebot-plugin-marshoai
4
4
Fork 0
mirror of https://github.com/LiteyukiStudio/nonebot-plugin-marshoai.git synced 2025-02-12 15:39:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

修改, 但搜索功能完全无法使用了

Browse Source
This commit is contained in:
Twisuki 2024-12-11 18:09:08 +08:00
parent ccb12f0f63
commit 086f21272f
1 changed files with 80 additions and 68 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
nonebot_plugin_marshoai/tools/marshoai-meogirl/mg_Search.py
View File

@ -2,24 +2,34 @@ from nonebot.log import logger
import re import re
import httpx import httpx
import urllib.parse
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from watchfiles import awatch
async def get_async_data (url):
async with httpx.AsyncClient() as client:
return await client.get(url)
async def search(msg : str, num : int): async def search(msg : str, num : int):
logger.info(f"搜索 : \"{msg}\"") logger.info(f"搜索 : \"{msg}\"")
result = "" result = ""
headers = { # headers = {
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.127 Safari/537.36' # 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/100.0.4896.127 Safari/537.36'
} # }
url = "https://mzh.moegirl.org.cn/index.php?search=" + msg url = "https://mzh.moegirl.org.cn/index.php?search=" + urllib.parse.quote_plus(msg)
async with httpx.AsyncClient() as client: # client = httpx.AsyncClient()
response = await client.get(url, headers = headers) # async with client:
logger.info(response.headers.get('Location')) # response = await client.post(url, headers = headers)
logger.info(f"连接萌娘百科中, 状态码 : {response.status_code}") response = await get_async_data(url)
logger.info(f"连接\"{response.headers.get('location')}\"中, 状态码 : {response.status_code}")
# 正常搜索
if response.status_code == 200:
""" """
萌娘百科搜索页面结构 萌娘百科搜索页面结构
div.searchresults # 若无, 证明页面已重定向 div.searchresults
└── p ... └── p ...
└── ul.mw-search-results # 若无, 证明无搜索结果 └── ul.mw-search-results # 若无, 证明无搜索结果
└── li # 一个搜索结果 └── li # 一个搜索结果
@ -29,11 +39,8 @@ async def search(msg : str, num : int):
└── li ... └── li ...
└── li ... └── li ...
""" """
if response.status_code == 200:
soup = BeautifulSoup(response.text, 'html.parser') soup = BeautifulSoup(response.text, 'html.parser')
# 检测div.searchresults, 是否已重定向
if soup.find('div', class_='searchresults'):
# 检测ul.mw-search-results, 是否有结果 # 检测ul.mw-search-results, 是否有结果
if soup.find('ul', class_='mw-search-results'): if soup.find('ul', class_='mw-search-results'):
ul_tag = soup.select('ul.mw-search-results')[0] ul_tag = soup.select('ul.mw-search-results')[0]
@ -62,9 +69,13 @@ async def search(msg : str, num : int):
logger.info("无结果") logger.info("无结果")
return "无结果" return "无结果"
# 无div.searchresults, 重定向 # 重定向
else: elif response.status_code == 302:
logger.info(f"\"{msg}\"已被重定向") logger.info(f"\"{msg}\"已被重定向")
# 读取重定向结果
soup = BeautifulSoup(response.text, 'html.parser')
logger.success("重定向成功")
logger.info(response)
num = 0 num = 0
""" """
@ -83,14 +94,15 @@ async def search(msg : str, num : int):
p = str(p_tag) p = str(p_tag)
p = re.sub(r'<.*?>', '', p) p = re.sub(r'<.*?>', '', p)
if p != '': if p != '':
result += str(p) + "/n" result += str(p) + "\n"
num += 1 num += 1
if num >= 5: if num >= 5:
break break
return result return result
# 状态码非200 # 状态码非200或302
else: else:
logger.error(f"网络错误, 状态码 : {response.status_code}") logger.error(f"网络错误, 状态码 : {response.status_code}")
return f"网络错误, 状态码 : {response.status_code}" return f"网络错误, 状态码 : {response.status_code}"