# coding: utf-8
from __future__ import print_function
import tweepy
from .medias import VideoMedia
def parse_tweet(tweet, include_retweets, image_size, results):
"""Parse a single tweet, returning a more usful structure containing its medias."""
urls = {
'tweet_id': tweet.id_str,
'original_tweet_id': tweet.id_str,
'date': tweet.created_at,
'original_date': tweet.created_at,
'user_id': tweet.user.id_str,
'user_name': tweet.user.name,
'user_screen_name': tweet.user.screen_name,
'original_user_id': tweet.user.id_str,
'original_user_name': tweet.user.name,
'original_user_screen_name': tweet.user.screen_name,
'type': 'tweet',
'videos': [],
'images': [],
'urls': {
'periscope': [],
'instagram': [],
'others': []
},
'text': '' # tweet.full_text
}
if include_retweets and hasattr(tweet, 'retweeted_status'):
tweet = tweet.retweeted_status
urls['original_tweet_id'] = tweet.id_str
urls['original_date'] = tweet.created_at
urls['original_user_id'] = tweet.user.id_str
urls['original_user_name'] = tweet.user.name
urls['original_user_screen_name'] = tweet.user.screen_name
urls['type'] = 'retweet'
results['retweets'] += 1
else:
results['tweets'] += 1
if hasattr(tweet, 'extended_entities'):
if 'media' in tweet.extended_entities:
for media in tweet.extended_entities['media']:
# Videos
if 'video_info' in media:
variants = media['video_info']['variants']
video = max(VideoMedia(variant) for variant in variants)
urls['videos'].append(video.obj['url'])
# Images
if 'sizes' in media:
url = media['media_url_https']
if image_size in media['sizes'] or image_size == 'orig':
url += ":" + image_size
else:
print('Size `{0}` not found for image `{1}`'.format(image_size, media['media_url_https']))
urls['images'].append(url)
# Urls
if hasattr(tweet, 'entities'):
if 'urls' in tweet.entities:
for url in tweet.entities['urls']:
expanded = url['expanded_url']
if 'instagram.com' in expanded:
urls['urls']['instagram'].append(expanded)
elif 'periscope.tv' in expanded:
urls['urls']['periscope'].append(expanded)
else:
urls['urls']['others'].append(expanded)
# Remove thumbnails
if len(urls['videos']) == 1 and urls['images'] and 'video_thumb' in urls['images'][0]:
urls['images'] = urls['images'][1:]
if not urls['videos'] and not urls['images'] and not urls['urls']['periscope'] and not urls['urls']['instagram'] and hasattr(tweet, 'full_text'):
urls['text'] = tweet.full_text
results['media'].append(urls)
def get_medias(auth, user_id, include_retweets, image_size, since, since_id, until, until_id, likes):
"""Get all medias for a given Twitter user."""
auth = tweepy.OAuthHandler(auth['consumer_token'], auth['consumer_secret'])
api = tweepy.API(auth, wait_on_rate_limit=True, wait_on_rate_limit_notify=True)
results = {
'tweets': 0,
'retweets': 0,
'media': []
}
capi = api.favorites if likes else api.user_timeline
for tweet in tweepy.Cursor(capi, id=user_id, include_rts=include_retweets, include_entities=True, tweet_mode='extended', since_id=since_id, max_id=until_id).items():
if since is not None and tweet.created_at < since:
break
if until is not None and tweet.created_at > until:
continue
parse_tweet(tweet, include_retweets, image_size, results)
print('Medias for user {0}'.format(user_id))
print('- Tweets: {0}'.format(results['tweets']))
print('- Retweets: {0}'.format(results['retweets']))
print('- Parsed: {0}'.format(len(results['media'])))
return results
