Python多线程豆瓣影评API接口爬虫

爬虫库

使用简单的requests库，这是一个阻塞的库，速度比较慢。
解析使用XPATH表达式
总体采用类的形式

多线程

使用concurrent.future并发模块，建立线程池，把future对象扔进去执行即可实现并发爬取效果

数据存储

使用Python ORM sqlalchemy保存到数据库，也可以使用自带的csv模块存在CSV中。

API接口

因为API接口存在数据保护情况，一个电影的每一个分类只能抓取前25页，全部评论、好评、中评、差评所有分类能爬100页，每页有20个数据，即最多为两千条数据。

因为时效性原因，不保证代码能爬到数据，只是给大家一个参考思路，上代码：

from datetime import datetime
import random
import csv
from concurrent.futures import ThreadPoolExecutor, as_completedfrom lxml import etree
import pymysql
import requestsfrom models import create_session, Comments#随机UA
USERAGENT = ['Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/14.0.835.163 Safari/535.1','Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.103 Safari/537.36','Mozilla/5.0 (Windows NT 6.1; WOW64; rv:6.0) Gecko/20100101 Firefox/6.0','Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/534.50 (KHTML, like Gecko) Version/5.1 Safari/534.50','Opera/9.80 (Windows NT 6.1; U; zh-cn) Presto/2.9.168 Version/11.50','Mozilla/5.0 (Windows; U; Windows NT 6.1; ) AppleWebKit/534.12 (KHTML, like Gecko) Maxthon/3.0 Safari/534.12'
]class CommentFetcher:headers = {'User-Agent': ''}cookie = ''cookies = {'cookie': cookie}# cookie为登录后的cookie，需要自行复制base_node = '//div[@class="comment-item"]'def __init__(self, movie_id, start, type=''):''':type: 全部评论：''， 好评：h 中评：m 差评：l:movie_id: 影片的ID号:start: 开始的记录数，0-480'''self.movie_id = movie_idself.start = startself.type = typeself.url = 'https://movie.douban.com/subject/{id}/comments?start={start}&limit=20&sort=new_score\&status=P&percent_type={type}&comments_only=1'.format(id=str(self.movie_id),start=str(self.start),type=self.type)#创建数据库连接self.session = create_session()#随机useragentdef _random_UA(self):self.headers['User-Agent'] = random.choice(USERAGENT)#获取api接口，使用get方法，返回的数据为json数据，需要提取里面的HTMLdef _get(self):self._random_UA()res = ''try:res = requests.get(self.url, cookies=self.cookies, headers=self.headers)res = res.json()['html']except Exception as e:print('IP被封，请使用代理IP')print('正在获取{} 开始的记录'.format(self.start))return resdef _parse(self):res = self._get()dom = etree.HTML(res)#id号self.id = dom.xpath(self.base_node + '/@data-cid')#用户名self.username = dom.xpath(self.base_node + '/div[@class="avatar"]/a/@title')#用户连接self.user_center = dom.xpath(self.base_node + '/div[@class="avatar"]/a/@href')#点赞数self.vote = dom.xpath(self.base_node + '//span[@class="votes"]/text()')#星级self.star = dom.xpath(self.base_node + '//span[contains(@class,"rating")]/@title')#发表时间self.time = dom.xpath(self.base_node + '//span[@class="comment-time "]/@title')#评论内容 所有span标签class名为short的节点文本self.content = dom.xpath(self.base_node + '//span[@class="short"]/text()')#保存到数据库def save_to_database(self):self._parse()for i in range(len(self.id)):try:comment = Comments(id=int(self.id[i]),username=self.username[i],user_center=self.user_center[i],vote=int(self.vote[i]),star=self.star[i],time=datetime.strptime(self.time[i], '%Y-%m-%d %H:%M:%S'),content=self.content[i])self.session.add(comment)self.session.commit()return 'finish'except pymysql.err.IntegrityError as e:print('数据重复，不做任何处理')except Exception as e:#数据添加错误，回滚self.session.rollback()finally:#关闭数据库连接self.session.close()#保存到csvdef save_to_csv(self):self._parse()f = open('comment.csv', 'w', encoding='utf-8')csv_in = csv.writer(f, dialect='excel')for i in range(len(self.id)):csv_in.writerow([int(self.id[i]),self.username[i],self.user_center[i],int(self.vote[i]),self.time[i],self.content[i]])f.close()if __name__ == '__main__':with ThreadPoolExecutor(max_workers=4) as executor:futures = []for i in ['', 'h', 'm', 'l']:for j in range(25):fetcher = CommentFetcher(movie_id=26266893, start=j * 20, type=i)futures.append(executor.submit(fetcher.save_to_csv))for f in as_completed(futures):try:res = f.done()if res:ret_data = f.result()if ret_data == 'finish':print('{} 成功保存数据'.format(str(f)))except Exception as e:f.cancel()