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_completed

from lxml import etree
import pymysql
import requests

from 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_id
        self.start = start
        self.type = type
        self.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useragent
    def _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html
    def _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 res

    def _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csv
    def 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()