业务需要，抓取页面排名数据保存，月末发邮件

data.php

<?php
foreach (glob("./data/*.html") as $filename) {
    echo "<a href='$filename'>$filename</a><br/>";
}
?>

python_curl_bs4.py

#!/usr/bin/python2.7
# -*- coding: gbk -*-
from bs4 import BeautifulSoup
import urllib2
import smtplib
from email.mime.text import MIMEText
import time
import calendar as cal

import sys,os

def cur_file_dir():
    path = sys.path[0]
    if os.path.isdir(path):
        return path
    elif os.path.isfile(path):
        return os.path.dirname(path)

if __name__ == "__main__":
    #f = open('index.html','rb')
    #s = f.read()
    #f.close()
    #u = s.decode('gbk')
    
    path = cur_file_dir()+'/data'
    if not os.path.exists(path):
        os.makedirs(path)

    url = 'http://huodong2.4399.com/girl/girlblg/'
    t = time.strftime("%Y-%m-%d %H:%M:%S")
    d = time.strftime("%Y-%m-%d")
    l = cal.monthrange(int(time.strftime("%Y")), int(time.strftime("%-m")))
    u = urllib2.urlopen(url)

    soup = BeautifulSoup(u, "html5lib")
    title = soup.title.string.encode('utf-8')
    data = '<table>'
    li_a = soup.find_all("a", class_="side5-tit")
    li_fen = soup.find_all("span", class_="side5-fen")
    for i in range(0, len(li_a)):
      data=data+'<tr><td>'+str(i+1).encode('utf-8')+'</td><td>'+li_fen[i].string.encode('utf-8')+'</td><td>'+li_a[i]['href'].encode('utf-8')+'</td><td>'+li_a[i].string.encode('utf-8')+"</td></tr>"

    data+='</table>'

    host = "smtp.exmail.qq.com"
    port = 465
    user = "x@4399inc.com"
    pwd = "password"
    sender=user
    receiver = 'y@4399inc.com'
    cc = 'z@4399inc.com,z@4399inc.com'
    bcc = 's@qq.com'
    
    body  = "<h1>"+title+" rank</h1>"
    body += data
    body += "<p>time "+t+" from "+sender+".</p>"
    body += "<p>other day data <a href='http://t.huodong.4399.com/girl/2016girlblg/data.php'>http://t.huodong.4399.com/girl/2016girlblg/data.php</a>.</p>"
    
    f=file(path+"/"+t+".html","w+")
    f.write(body)
    f.close()
    
    if l[1]==int(time.strftime("%-d")):
        msg = MIMEText(body, "html")
        msg["subject"] = ''+title+" rank "+d
        msg["from"] = sender
        msg["to"] = receiver
        msg['Cc'] = cc
        receivers = cc.split(",") + bcc.split(",") + [receiver]
        
        s = smtplib.SMTP_SSL(host, port);
        #s.set_debuglevel(1)
        s.login(user, pwd)
        s.sendmail(sender, receivers, msg.as_string())

shell_month_end.sh

#!/usr/bin/sh
#this script is used to start the xxx in the last day of every month
ym=`date +%m" "%Y`
if [ `date +%d` = `cal $ym|xargs|awk '{print $NF}'` ]
then
#/usr/local/python-2.7.11/bin/python2.7 /workspace/work/girlblg/work.py
fi