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             'comments': list,
  33         },
  34     }, {
  35         # missing description
  36         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
  37         'info_dict': {
  38             'id': 'BA-pQFBG8HZ',
  39             'ext': 'mp4',
  40             'title': 'Video by britneyspears',
  41             'thumbnail': 're:^https?://.*\.jpg',
  42             'timestamp': 1453760977,
  43             'upload_date': '20160125',
  44             'uploader_id': 'britneyspears',
  45             'uploader': 'Britney Spears',
  46             'like_count': int,
  47             'comment_count': int,
  48             'comments': list,
  49         },
  50         'params': {
  51             'skip_download': True,
  52         },
  53     }, {
  54         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  55         'only_matching': True,
  56     }, {
  57         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
  58         'only_matching': True,
  59     }]
  60
  61     @staticmethod
  62     def _extract_embed_url(webpage):
  63         mobj = re.search(
  64             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
  65             webpage)
  66         if mobj:
  67             return mobj.group('url')
  68
  69         blockquote_el = get_element_by_attribute(
  70             'class', 'instagram-media', webpage)
  71         if blockquote_el is None:
  72             return
  73
  74         mobj = re.search(
  75             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  76         if mobj:
  77             return mobj.group('link')
  78
  79     def _real_extract(self, url):
  80         mobj = re.match(self._VALID_URL, url)
  81         video_id = mobj.group('id')
  82         url = mobj.group('url')
  83
  84         webpage = self._download_webpage(url, video_id)
  85
  86         (video_url, description, thumbnail, timestamp, uploader,
  87          uploader_id, like_count, comment_count, height, width) = [None] * 10
  88
  89         shared_data = self._parse_json(
  90             self._search_regex(
  91                 r'window\._sharedData\s*=\s*({.+?});',
  92                 webpage, 'shared data', default='{}'),
  93             video_id, fatal=False)
  94         if shared_data:
  95             media = try_get(
  96                 shared_data, lambda x: x['entry_data']['PostPage'][0]['media'], dict)
  97             if media:
  98                 video_url = media.get('video_url')
  99                 height = int_or_none(media.get('dimensions', {}).get('height'))
 100                 width = int_or_none(media.get('dimensions', {}).get('width'))
 101                 description = media.get('caption')
 102                 thumbnail = media.get('display_src')
 103                 timestamp = int_or_none(media.get('date'))
 104                 uploader = media.get('owner', {}).get('full_name')
 105                 uploader_id = media.get('owner', {}).get('username')
 106                 like_count = int_or_none(media.get('likes', {}).get('count'))
 107                 comment_count = int_or_none(media.get('comments', {}).get('count'))
 108                 comments = [{
 109                     'author': comment.get('user', {}).get('username'),
 110                     'author_id': comment.get('user', {}).get('id'),
 111                     'id': comment.get('id'),
 112                     'text': comment.get('text'),
 113                     'timestamp': int_or_none(comment.get('created_at')),
 114                 } for comment in media.get(
 115                     'comments', {}).get('nodes', []) if comment.get('text')]
 116
 117         if not video_url:
 118             video_url = self._og_search_video_url(webpage, secure=False)
 119
 120         formats = [{
 121             'url': video_url,
 122             'width': width,
 123             'height': height,
 124         }]
 125
 126         if not uploader_id:
 127             uploader_id = self._search_regex(
 128                 r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
 129                 webpage, 'uploader id', fatal=False)
 130
 131         if not description:
 132             description = self._search_regex(
 133                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 134             if description is not None:
 135                 description = lowercase_escape(description)
 136
 137         if not thumbnail:
 138             thumbnail = self._og_search_thumbnail(webpage)
 139
 140         return {
 141             'id': video_id,
 142             'formats': formats,
 143             'ext': 'mp4',
 144             'title': 'Video by %s' % uploader_id,
 145             'description': description,
 146             'thumbnail': thumbnail,
 147             'timestamp': timestamp,
 148             'uploader_id': uploader_id,
 149             'uploader': uploader,
 150             'like_count': like_count,
 151             'comment_count': comment_count,
 152             'comments': comments,
 153         }
 154
 155
 156 class InstagramUserIE(InfoExtractor):
 157     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
 158     IE_DESC = 'Instagram user profile'
 159     IE_NAME = 'instagram:user'
 160     _TEST = {
 161         'url': 'https://instagram.com/porsche',
 162         'info_dict': {
 163             'id': 'porsche',
 164             'title': 'porsche',
 165         },
 166         'playlist_mincount': 2,
 167         'playlist': [{
 168             'info_dict': {
 169                 'id': '614605558512799803_462752227',
 170                 'ext': 'mp4',
 171                 'title': '#Porsche Intelligent Performance.',
 172                 'thumbnail': 're:^https?://.*\.jpg',
 173                 'uploader': 'Porsche',
 174                 'uploader_id': 'porsche',
 175                 'timestamp': 1387486713,
 176                 'upload_date': '20131219',
 177             },
 178         }],
 179         'params': {
 180             'extract_flat': True,
 181             'skip_download': True,
 182         }
 183     }
 184
 185     def _real_extract(self, url):
 186         mobj = re.match(self._VALID_URL, url)
 187         uploader_id = mobj.group('username')
 188
 189         entries = []
 190         page_count = 0
 191         media_url = 'http://instagram.com/%s/media' % uploader_id
 192         while True:
 193             page = self._download_json(
 194                 media_url, uploader_id,
 195                 note='Downloading page %d ' % (page_count + 1),
 196             )
 197             page_count += 1
 198
 199             for it in page['items']:
 200                 if it.get('type') != 'video':
 201                     continue
 202                 like_count = int_or_none(it.get('likes', {}).get('count'))
 203                 user = it.get('user', {})
 204
 205                 formats = [{
 206                     'format_id': k,
 207                     'height': v.get('height'),
 208                     'width': v.get('width'),
 209                     'url': v['url'],
 210                 } for k, v in it['videos'].items()]
 211                 self._sort_formats(formats)
 212
 213                 thumbnails_el = it.get('images', {})
 214                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 215
 216                 # In some cases caption is null, which corresponds to None
 217                 # in python. As a result, it.get('caption', {}) gives None
 218                 title = (it.get('caption') or {}).get('text', it['id'])
 219
 220                 entries.append({
 221                     'id': it['id'],
 222                     'title': limit_length(title, 80),
 223                     'formats': formats,
 224                     'thumbnail': thumbnail,
 225                     'webpage_url': it.get('link'),
 226                     'uploader': user.get('full_name'),
 227                     'uploader_id': user.get('username'),
 228                     'like_count': like_count,
 229                     'timestamp': int_or_none(it.get('created_time')),
 230                 })
 231
 232             if not page['items']:
 233                 break
 234             max_id = page['items'][-1]['id'].split('_')[0]
 235             media_url = (
 236                 'http://instagram.com/%s/media?max_id=%s' % (
 237                     uploader_id, max_id))
 238
 239         return {
 240             '_type': 'playlist',
 241             'entries': entries,
 242             'id': uploader_id,
 243             'title': uploader_id,
 244         }