python-spider 第一关

Posted 2020-12-19 如海

import datetime
import time
import json
import hashlib
import base64
import requests
import pandas
import browsercookie
from collections import Counter
import pandas



l = []

def get_token():
    a = ‘9622‘

    dt = datetime.datetime.now()
    timestamp = str(dt.timestamp()).split(‘.‘)[0]
    base_code = base64.b64encode((a + timestamp).encode("utf-8"))
    data = hashlib.md5(base_code)
    return data.hexdigest(), timestamp


safe, timestamp = get_token()

headers = {
    ‘User-Agent‘: ‘Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.132 Safari/537.36 QIHU 360SE‘,
    ‘Referer‘: ‘http://www.python-spider.com/challenge/1‘,
    ‘safe‘: safe,
    ‘timestamp‘: timestamp
}
cj = browsercookie.chrome()
c = Counter(‘a‘)
for i in range(1,86):
    url = ‘http://www.python-spider.com/challenge/api/json?page=%s&count=14‘%str(i)
    res = requests.get(url, headers=headers, cookies=cj)
    try:
        t = res.json()[‘infos‘]
        l.append(t)
        
    except Exception as e:
        pass


p = []
for i in l:
    for x in i:
        data = x[‘message‘]
        p.append({‘m‘:data})

df = pandas.DataFrame(p)

df[df.m.str.contains(‘招‘)]