acc_tweet.py

#!/usr/bin/env python3
# -*- coding:utf-8 -*-
from tempfile import TemporaryDirectory
import datetime

import tweepy
import requests
import dateparser as dp
from bs4 import BeautifulSoup

CONSUMER_KEY = "XXXXXXXXXXXXXXXXXXXXXXXXXXXX"
CONSUMER_SEC = "XXXXXXXXXXXXXXXXXXXXXXXXXXXX"
ACCESS_KEY = "XXXXXXXXXXXXX-XXXXXXXXXXXXXXXX"
ACCESS_SEC = "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"

SOURCE_DOMAIN = "http://www.aeoncinema.com"
SOURCE_URL = "http://www.aeoncinema.com/cinema2/chofu/movie/comingsoon.html"    # 公開予定作品

TWEET_FORMAT = "【新着】 {movie[cinema_title]} ({start_date}より上映予定)\n{movie[cinema_url]}"
TEMPORARY_DIR = TemporaryDirectory()


def tweet_new_movies(api, new_movie):
    message = TWEET_FORMAT.format(movie=new_movie, start_date=new_movie['start_date'].strftime("%Y年%m月%d日"))
    if new_movie['cinema_thumbnail_path']:
        api.update_with_media(new_movie['cinema_thumbnail_path'], status=message)
    else:
        api.update_status(message)


def get_new_items(latest_item_date):
    r = requests.get(SOURCE_URL)
    if not r.ok:
        r.status_code, r.reason
    r.encoding = 'utf-8'
    soup = BeautifulSoup(r.text, 'html.parser')

    found_cinemas = []

    newCinema = soup.find(id='conNewCinema')
    cDateBlocks = newCinema.find_all(class_='cDateBlock')

    for cdb in cDateBlocks:
        start_date = dp.parse(cdb.find(class_='startDate').text).date()
        cinemas = cdb.find_all(class_='cinemaBlock')
        if start_date <= latest_item_date:
            continue
        for cinema in cinemas:
            cinema_thumbnail_url = SOURCE_DOMAIN + cinema.img.attrs.get('src')
            if cinema_thumbnail_url.find("images/no_images") == -1:
                r = requests.get(cinema_thumbnail_url)
                with open(TEMPORARY_DIR.name + '/' + '_'.join(cinema_thumbnail_url.split('/')[-2:]), 'w+b') as f:
                    f.write(r.content)
                    cinema_thumbnail_path = f.name
            else:
                cinema_thumbnail_path = ""
            this_cinema = {
                "start_date": start_date,
                "cinema_thumbnail_url": cinema_thumbnail_url,
                "cinema_thumbnail_path": cinema_thumbnail_path,
                "cinema_title": cinema.find(class_='cbTitle').text,
                "cinema_url": SOURCE_DOMAIN + cinema.find(class_='cbTitle').a.attrs.get('href'),
            }
            found_cinemas.append(this_cinema)
    return found_cinemas


def get_latest_date():
    try:
        with open("latest.dat") as f:
            s = f.readline()
            return datetime.datetime.strptime(s, '%Y-%m-%d').date()
    except FileNotFoundError:
        return datetime.date.today()    # 初回起動時は、実行時を最終取得日とする


def update_latest_date(latest_date):
    with open("acc_latest.dat", "w") as f:
        f.write(latest_date.strftime('%Y-%m-%d'))


def main():
    auth = tweepy.OAuthHandler(CONSUMER_KEY, CONSUMER_SEC)
    auth.set_access_token(ACCESS_KEY, ACCESS_SEC)
    api = tweepy.API(auth)

    latest_item_date = get_latest_date()

    new_items = get_new_items(latest_item_date)
    if not new_items:
        return    # nothing new
    new_latest_date = max(new_items, key=lambda x: x['start_date'])['start_date']

    for new_item in new_items:
        tweet_new_movies(api, new_item)

    update_latest_date(new_latest_date)


if __name__ == "__main__":
    main()