メディアツイートをいいねやRT順に並べて表示

sort_media_tweet.py

# -*- coding: utf-8 -*-
"""
メディアツイートをいいねやRT順に並べて表示

必要なファイルとか:
    - twitterのConsumer Key, Consumer Secret
    - 全ツイート履歴のCSV (twitter公式の「設定 > ユーザー情報」の一番下にある「全ツイート履歴」より)

インストールが必要なパッケージ:
    - arrow
    - tweepy

"""
import csv
import datetime

import arrow
import tweepy

CONSUMER_KEY = '<Your Consumer Key>'
CONSUMER_SECRET = '<Your Consumer Secret>'
csvpath = 'path/to/tweets.csv'


def timestamp2arrow(timestamp):
    return arrow.get(timestamp, 'YYYY-MM-DD HH:mm:ss Z')


class CsvParser(object):
    def __init__(self, path, screen_name):
        self.path = path
        self.screen_name = screen_name

    def filter_images_by_since(self, since):
        with open(self.path) as csvfile:
            reader = csv.DictReader(csvfile)
            for row in reader:
                if timestamp2arrow(since) <= timestamp2arrow(row['timestamp']):
                    url = "https://twitter.com/{screen_name}/status/{tweet_id}/photo/1".format(
                        screen_name=self.screen_name,
                        tweet_id=row['tweet_id'],
                    )
                    if url == row['expanded_urls']:
                        yield row


def fetch_tweet_data(api, media_tweets, text_trancate_to=30):
    for tweet in media_tweets:
        tweet = api.get_status(id=tweet['tweet_id'])
        yield {
            'tweet_id': tweet.id,
            'likes': tweet.favorite_count,
            'retweets': tweet.retweet_count,
            'text': tweet.text.replace('\n', '')[:text_trancate_to],
            'created_at': arrow.get(tweet.created_at).to('JST').format('MM/DD HH:mm'),
        }


def sort_by(key, data_list, reverse=False):
    return sorted(data_list, key=lambda d: d[key], reverse=reverse)


def display_data(data_list, limit=10):
    for i, data in enumerate(data_list):
        print("ID:{tweet_id}, likes:{likes}, RTs:{retweets}, date:{created_at}, text:{text}".format(**data))
        if i == limit:
            break


if __name__ == '__main__':
    parser = CsvParser(csvpath, screen_name='kk6')
    media_tweets = parser.filter_images_by_since('2016-01-01 00:00:00 +0900')
    auth = tweepy.AppAuthHandler(CONSUMER_KEY, CONSUMER_SECRET)
    api = tweepy.API(auth, wait_on_rate_limit=True)
    data_list = fetch_tweet_data(api, media_tweets)
    sorted_data = sort_by('likes', data_list, reverse=True)
    display_data(sorted_data, 100)