利用twitter官网提供的api 及tweepy库 爬取tweets

利用twitter官网提供的api及tweepy库爬取tweets

tweepy官网文档

思路：
1.以用户为中心，爬取用户的所有推文数据
2.根据用户id寻找用户朋友的tweeter id扩展待爬用户表
3.循环1,2

几点说明：
1.爬推特数据需要，推荐用ss。代码需要http，https代理。如果是socks的话会发现浏览器能，但是代码会提示

tweepy.error.TweepError: Failed to send request: HTTPSConnectionPool(host='api.twitter.com', port=443): Max retries exceeded with url: ....

说明https连接失败。如果需要终端参考Mac命令行终端下使用s
然后在tweepy.API中加入代理信息，端口为你设定的代理端口。

api = tweepy.API(auth, proxy="127.0.0.1：1080",)

2.使用官方api需要先申请一个应用程序以获得授权，申请地址Twitter应用程序 名字描述什么的随便写好好，没有审核时间，填写后即可获得consumer_key，consumer_secret，access_token，access_token_secret这些在求取数据时需要用到。

3.官方API有速率限制具体参见[Rate limits-Twitter Development]授权用户和授权应用的请求窗口数有差异我用的。user_timeline（）状语从句：user_friends（）限制如下：

所以需要协调两个接口的调用频率。

4.当请求次数超过上限时会抛出异常，然后退出程序，解决方法时tweepy.API中将参数wait_on_rate_limit，wait_on_rate_limit_notify设置为True

到达上限时，程序将自动等待，并输出提示信息。

api = tweepy.API(auth, proxy="127.0.0.1：1080", wait_on_rate_limit=True, wait_on_rate_limit_notify=True)

5.api请求返回json格式数据如图：

6.有些用户设置不允许取数据时会提示Not authorized. 可以在异常部分处理异常，跳过改用户即可.tweepy.error信息也可以在上面的官方文档连接中查到。

代码

import tweepy
import time
import csv
import threading

consumer_key = "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
consumer_secret = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
access_token = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
access_token_secret = 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'

lock = threading.Lock()

def get_tweets():
    global user_ids
    global old_ids
    lock.acquire()
    try:
        num = 0
        while len(user_ids) > 1:
            try:
                user_id = user_ids[num]
                print('crawling user %s data...' % user_id)
                auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
                auth.set_access_token(access_token, access_token_secret)
                api = tweepy.API(auth, proxy="127.0.0.1:1080", wait_on_rate_limit=True, wait_on_rate_limit_notify=True)
                tweets = []
                new_tweets = api.user_timeline(user_id, count=200)
                tweets.extend(new_tweets)
                old = tweets[-1].id - 1
                while len(new_tweets) > 0:
                    new_tweets = api.user_timeline(user_id=user_id, count=200, max_id=old)
                    tweets.extend(new_tweets)
                    old = tweets[-1].id - 1
                    print('%s tweets downloaded' % (len(tweets)))

                out_tweets = [[tweet.id, tweet.text, tweet.created_at, tweet.lang, tweet.place, tweet.geo, tweet.source,
                               tweet.truncated, tweet.favorite_count, tweet.favorited, tweet.in_reply_to_screen_name,
                               tweet.in_reply_to_status_id, tweet.in_reply_to_user_id, tweet.is_quote_status,
                               tweet.retweet_count, tweet.retweeted, tweet.user.id, tweet.user.name, tweet.user.screen_name,
                               tweet.user.statuses_count, tweet.user.time_zone, tweet.user.url, tweet.user.notifications,
                               tweet.user.profile_background_image_url, tweet.user.profile_image_url,
                               tweet.user.profile_image_url_https, tweet.user.location, tweet.user.contributors_enabled,
                               tweet.user.created_at, tweet.user.default_profile, tweet.user.default_profile_image,
                               tweet.user.description, tweet.user.favourites_count, tweet.user.follow_request_sent,
                               tweet.user.followers_count, tweet.user.following, tweet.user.friends_count,
                               tweet.user.geo_enabled] for tweet in tweets]
                user_ids.remove(user_id)
                old_ids.append(user_id)
                with open('./data1/%s_tweets.csv' % user_id, 'w',encoding='utf-8') as file:
                    writer = csv.writer(file)
                    writer.writerows(out_tweets)
                print('saved data')
            except tweepy.TweepError as e:
                if e.reason=='Not authorized.':
                    print('this user not authorized.')
                    user_ids.remove(user_id)
                    old_ids.append(user_id)
                    continue
                else:print(e)
    finally:
        lock.release()


def get_friends():
    global user_ids
    global old_ids
    global oldest
    lock.acquire()
    try:
        print('getting user friends id...')
        auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
        auth.set_access_token(access_token, access_token_secret)
        api = tweepy.API(auth, proxy="127.0.0.1:1080", wait_on_rate_limit=True, wait_on_rate_limit_notify=True)
        ids = []
        for user in old_ids[10]:
            try:
                friends = api.friends_ids(user)
                friend = []
                for idd in friends:
                    if (idd not in old_ids) and (idd not in user_ids) and(idd not in oldest):
                        friend.append(idd)
                ids.extend(friend)

            except tweepy.TweepError as e:
                if e.reason == 'Not authorized.':
                    print('this user not authorized.')
                    old_ids.remove(user)
                    oldest.append(user)
                    continue
                else:
                    print(e)
            old_ids.remove(user)
            oldest.append(user)
        user_ids.extend(ids)
        print('done!')
        with open('crawled and expened user.txt','w',encoding='utf-8') as file:
            for x in oldest:
                file.write(str(x))
                file.write(' ')
    finally:
        lock.release()


if __name__ == '__main__':
    user_ids = [25073877,198599889]
    with open('old_ids.txt','r',encoding='utf-8') as file:
        old_ids=[x for x in file.read().split(' ')]
    while len(user_ids) > 0:
        t1=threading.Thread(target=get_tweets)
        t2=threading.Thread(target=get_friends)
        t1.start()
        t1.join()
        t2.start()
        t2.join()