parent
aae3fc3457
commit
73fecf6de8
@ -0,0 +1,34 @@
|
|||||||
|
# @Time : 2021/11/1 22:09
|
||||||
|
# @Author :wenkaic
|
||||||
|
# @File : 001python大作业
|
||||||
|
# @Project : python爬虫
|
||||||
|
|
||||||
|
|
||||||
|
import urllib.request
|
||||||
|
from bs4 import BeautifulSoup
|
||||||
|
|
||||||
|
url = 'http://www.ddxs.com/doupocangqiong/'
|
||||||
|
|
||||||
|
headers = {
|
||||||
|
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/95.0.4638.54 Safari/537.36 Edg/95.0.1020.40",
|
||||||
|
}
|
||||||
|
|
||||||
|
# 请求对象定制
|
||||||
|
request = urllib.request.Request(url=url, headers=headers)
|
||||||
|
|
||||||
|
# 获取响应数据
|
||||||
|
response = urllib.request.urlopen(request)
|
||||||
|
soup = BeautifulSoup(response.read().decode('utf-8'),'lxml')
|
||||||
|
|
||||||
|
# 查询数据,返回列表
|
||||||
|
bbs = soup.select('th,tr')
|
||||||
|
|
||||||
|
#遍历列表
|
||||||
|
for i in range(0,len(bbs)):
|
||||||
|
obj = bbs[i]
|
||||||
|
print(obj.get_text().strip())
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
|
|
Loading…
Reference in new issue