使用requests，lxml爬取百思不得姐

# -*- coding:utf-8 -*-

import requests

import smtplib

from email.mime.text import MIMEText

from email.header import Header

from lxml import etree

#不得姐

url = "http://www.budejie.com/text/"

page = requests.get(url)

html = page.text

selector = etree.HTML(html)

# 第三方 SMTP 服务

mail_host="smtp.163.com"  #设置服务器

mail_user="########"    #用户名

mail_pass="######"  #口令

#邮箱

sender = '########@163.com'

receivers = ['#########@qq.com']

#头像

uimg = selector.xpath('//div[@class="u-img"]/a/img/@src')

#昵称

uname = selector.xpath('//a[@class="u-user-name"]/text()')

#内容

ucontent = selector.xpath('//div[@class="j-r-list-c-desc"]/a/text()')

#发邮件

message = MIMEText('不得姐更新提示', 'plain', 'utf-8')

message['From'] = Header("不得姐", 'utf-8')

message['To'] =  Header("我的APP", 'utf-8')

subject = '不得姐更新'

message['Subject'] = Header(subject, 'utf-8')

#一旦从不得姐获取不到信息，立即发送邮件

if not uimg or uname or ucontent:

try:

smtpObj = smtplib.SMTP()

smtpObj.connect(mail_host, 25)    # 25 为 SMTP 端口号

smtpObj.login(mail_user,mail_pass)

smtpObj.sendmail(sender, '#########@163.com', message.as_string())               ##这里本来应该填写receviers，但是发现发送失败，可能是因为QQ邮箱的smtp没有启动。所以还是让邮箱自己发给自己吧。

except smtplib.SMTPException:

print "error"