Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/instagram.py

   1 from __future__ import unicode_literals
   2
   3 import re
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     get_element_by_attribute,
   8     int_or_none,
   9     limit_length,
  10     lowercase_escape,
  11     try_get,
  12 )
  13
  14
  15 class InstagramIE(InfoExtractor):
  16     _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com/p/(?P<id>[^/?#&]+))'
  17     _TESTS = [{
  18         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
  19         'md5': '0d2da106a9d2631273e192b372806516',
  20         'info_dict': {
  21             'id': 'aye83DjauH',
  22             'ext': 'mp4',
  23             'title': 'Video by naomipq',
  24             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
  25             'thumbnail': 're:^https?://.*\.jpg',
  26             'timestamp': 1371748545,
  27             'upload_date': '20130620',
  28             'uploader_id': 'naomipq',
  29             'uploader': 'Naomi Leonor Phan-Quang',
  30             'like_count': int,
  31             'comment_count': int,
  32         },
  33     }, {
  34         # missing description
  35         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
  36         'info_dict': {
  37             'id': 'BA-pQFBG8HZ',
  38             'ext': 'mp4',
  39             'title': 'Video by britneyspears',
  40             'thumbnail': 're:^https?://.*\.jpg',
  41             'timestamp': 1453760977,
  42             'upload_date': '20160125',
  43             'uploader_id': 'britneyspears',
  44             'uploader': 'Britney Spears',
  45             'like_count': int,
  46             'comment_count': int,
  47         },
  48         'params': {
  49             'skip_download': True,
  50         },
  51     }, {
  52         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  53         'only_matching': True,
  54     }, {
  55         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
  56         'only_matching': True,
  57     }]
  58
  59     @staticmethod
  60     def _extract_embed_url(webpage):
  61         mobj = re.search(
  62             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
  63             webpage)
  64         if mobj:
  65             return mobj.group('url')
  66
  67         blockquote_el = get_element_by_attribute(
  68             'class', 'instagram-media', webpage)
  69         if blockquote_el is None:
  70             return
  71
  72         mobj = re.search(
  73             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  74         if mobj:
  75             return mobj.group('link')
  76
  77     def _real_extract(self, url):
  78         mobj = re.match(self._VALID_URL, url)
  79         video_id = mobj.group('id')
  80         url = mobj.group('url')
  81
  82         webpage = self._download_webpage(url, video_id)
  83
  84         (video_url, description, thumbnail, timestamp, uploader,
  85          uploader_id, like_count, comment_count) = [None] * 8
  86
  87         shared_data = self._parse_json(
  88             self._search_regex(
  89                 r'window\._sharedData\s*=\s*({.+?});',
  90                 webpage, 'shared data', default='{}'),
  91             video_id, fatal=False)
  92         if shared_data:
  93             media = try_get(
  94                 shared_data, lambda x: x['entry_data']['PostPage'][0]['media'], dict)
  95             if media:
  96                 video_url = media.get('video_url')
  97                 description = media.get('caption')
  98                 thumbnail = media.get('display_src')
  99                 timestamp = int_or_none(media.get('date'))
 100                 uploader = media.get('owner', {}).get('full_name')
 101                 uploader_id = media.get('owner', {}).get('username')
 102                 like_count = int_or_none(media.get('likes', {}).get('count'))
 103                 comment_count = int_or_none(media.get('comments', {}).get('count'))
 104
 105         if not video_url:
 106             video_url = self._og_search_video_url(webpage, secure=False)
 107
 108         if not uploader_id:
 109             uploader_id = self._search_regex(
 110                 r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
 111                 webpage, 'uploader id', fatal=False)
 112
 113         if not description:
 114             description = self._search_regex(
 115                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 116             if description is not None:
 117                 description = lowercase_escape(description)
 118
 119         if not thumbnail:
 120             thumbnail = self._og_search_thumbnail(webpage)
 121
 122         return {
 123             'id': video_id,
 124             'url': video_url,
 125             'ext': 'mp4',
 126             'title': 'Video by %s' % uploader_id,
 127             'description': description,
 128             'thumbnail': thumbnail,
 129             'timestamp': timestamp,
 130             'uploader_id': uploader_id,
 131             'uploader': uploader,
 132             'like_count': like_count,
 133             'comment_count': comment_count,
 134         }
 135
 136
 137 class InstagramUserIE(InfoExtractor):
 138     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
 139     IE_DESC = 'Instagram user profile'
 140     IE_NAME = 'instagram:user'
 141     _TEST = {
 142         'url': 'https://instagram.com/porsche',
 143         'info_dict': {
 144             'id': 'porsche',
 145             'title': 'porsche',
 146         },
 147         'playlist_mincount': 2,
 148         'playlist': [{
 149             'info_dict': {
 150                 'id': '614605558512799803_462752227',
 151                 'ext': 'mp4',
 152                 'title': '#Porsche Intelligent Performance.',
 153                 'thumbnail': 're:^https?://.*\.jpg',
 154                 'uploader': 'Porsche',
 155                 'uploader_id': 'porsche',
 156                 'timestamp': 1387486713,
 157                 'upload_date': '20131219',
 158             },
 159         }],
 160         'params': {
 161             'extract_flat': True,
 162             'skip_download': True,
 163         }
 164     }
 165
 166     def _real_extract(self, url):
 167         mobj = re.match(self._VALID_URL, url)
 168         uploader_id = mobj.group('username')
 169
 170         entries = []
 171         page_count = 0
 172         media_url = 'http://instagram.com/%s/media' % uploader_id
 173         while True:
 174             page = self._download_json(
 175                 media_url, uploader_id,
 176                 note='Downloading page %d ' % (page_count + 1),
 177             )
 178             page_count += 1
 179
 180             for it in page['items']:
 181                 if it.get('type') != 'video':
 182                     continue
 183                 like_count = int_or_none(it.get('likes', {}).get('count'))
 184                 user = it.get('user', {})
 185
 186                 formats = [{
 187                     'format_id': k,
 188                     'height': v.get('height'),
 189                     'width': v.get('width'),
 190                     'url': v['url'],
 191                 } for k, v in it['videos'].items()]
 192                 self._sort_formats(formats)
 193
 194                 thumbnails_el = it.get('images', {})
 195                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 196
 197                 # In some cases caption is null, which corresponds to None
 198                 # in python. As a result, it.get('caption', {}) gives None
 199                 title = (it.get('caption') or {}).get('text', it['id'])
 200
 201                 entries.append({
 202                     'id': it['id'],
 203                     'title': limit_length(title, 80),
 204                     'formats': formats,
 205                     'thumbnail': thumbnail,
 206                     'webpage_url': it.get('link'),
 207                     'uploader': user.get('full_name'),
 208                     'uploader_id': user.get('username'),
 209                     'like_count': like_count,
 210                     'timestamp': int_or_none(it.get('created_time')),
 211                 })
 212
 213             if not page['items']:
 214                 break
 215             max_id = page['items'][-1]['id'].split('_')[0]
 216             media_url = (
 217                 'http://instagram.com/%s/media?max_id=%s' % (
 218                     uploader_id, max_id))
 219
 220         return {
 221             '_type': 'playlist',
 222             'entries': entries,
 223             'id': uploader_id,
 224             'title': uploader_id,
 225         }