代码实现百度翻译爬虫

解析思路点击跳转

import requests
import re
import execjs

class BaiduSpider(object):
    def __init__(self):
        self.token_url = 'https://fanyi.baidu.com/?aldtype=16047'
        self.post_url = 'https://fanyi.baidu.com/v2transapi'
        self.headers = {
            'accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3',
            'accept-language': 'zh-CN,zh;q=0.9',
            'cache-control': 'no-cache',
            'cookie': '由于过长，请自行粘贴',
            'pragma': 'no-cache',
            'upgrade-insecure-requests': '1',
            'user-agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.142 Safari/537.36',
        }

    # 获取token和gtk
    def get_token(self):
        token_url = 'https://fanyi.baidu.com/?aldtype=16047'
        # 定义请求头
        r = requests.get(self.token_url,headers=self.headers)
        token = re.findall(r"token: '(.*?)'",r.text)
        window_gtk = re.findall(r"window.*?gtk = '(.*?)';",r.text)
        if token:
            return token[0],window_gtk[0]

    # 获取sign
    def get_sign(self,word,gtk):
        with open('baidu.js','r') as f:
            js_data = f.read()

        exec_object = execjs.compile(js_data)
        sign = exec_object.eval('e("{}","{}")'.format(word,gtk))

        return sign

    # 主函数
    def main(self,word,fro,to):
        token,gtk = self.get_token()
        sign = self.get_sign(word,gtk)
        # 找到form表单数据如下,sign和token需要想办法获取
        form_data = {
            'from': fro,
            'to': to,
            'query': word,
            'transtype': 'realtime',
            'simple_means_flag': '3',
            'sign': sign,
            'token': token
        }
        r = requests.post(self.post_url,data=form_data,headers=self.headers)
        print(r.json()['trans_result']['data'][0]['dst'])

if __name__ == '__main__':
    spider = BaiduSpider()
    choice = input('1. 英译汉 2. 汉译英 : ')
    word = input('请输入要翻译的单词:')
    if choice == '1':
        fro,to = 'en','zh'
    elif choice == '2':
        fro,to = 'zh','en'

    spider.main(word,fro,to)