代码拉取完成,页面将自动刷新
# 搜索Ubuntu包
import requests
from bs4 import BeautifulSoup
from tables import *
pkgUrl = "https://mirrors.163.com/ubuntu/pool/main/"
def searchPackage(url,pkgName):
r = requests.get(url)
soup = BeautifulSoup(r.content,"lxml")
alist = soup.findAll("a")
for aitem in alist:
href = aitem.get("href")
if "../" != href:
full_url = url + href
if not isFileUrl(full_url):
searchPackage(full_url,pkgName)
else:
if href.startswith(pkgName):
print(full_url)
def celloctPackage(url):
r = requests.get(url)
soup = BeautifulSoup(r.content,"lxml")
alist = soup.findAll("a")
for aitem in alist:
href = aitem.get("href")
if "../" != href:
full_url = url + href
if not isFileUrl(full_url):
celloctPackage(full_url)
else:
with session_scope() as se:
pkgs = PkgsTable()
pkgs.pkgName = href
pkgs.pkgUrl = full_url
se.add(pkgs)
def isFileUrl(url):
if url.endswith(".deb"):
return True
return False
if __name__ == '__main__':
celloctPackage(pkgUrl)
# searchPackage(pkgUrl,"libarchive13")
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。