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             'uploader_id': 'britneyspears',
  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         },
  49         'params': {
  50             'skip_download': True,
  51         },
  52     }, {
  53         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  54         'only_matching': True,
  55     }, {
  56         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
  57         'only_matching': True,
  58     }]
  59
  60     @staticmethod
  61     def _extract_embed_url(webpage):
  62         mobj = re.search(
  63             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
  64             webpage)
  65         if mobj:
  66             return mobj.group('url')
  67
  68         blockquote_el = get_element_by_attribute(
  69             'class', 'instagram-media', webpage)
  70         if blockquote_el is None:
  71             return
  72
  73         mobj = re.search(
  74             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  75         if mobj:
  76             return mobj.group('link')
  77
  78     def _real_extract(self, url):
  79         mobj = re.match(self._VALID_URL, url)
  80         video_id = mobj.group('id')
  81         url = mobj.group('url')
  82
  83         webpage = self._download_webpage(url, video_id)
  84
  85         (video_url, description, thumbnail, timestamp, uploader,
  86          uploader_id, like_count, comment_count) = [None] * 8
  87
  88         shared_data = self._parse_json(
  89             self._search_regex(
  90                 r'window\._sharedData\s*=\s*({.+?});',
  91                 webpage, 'shared data', default='{}'),
  92             video_id, fatal=False)
  93         if shared_data:
  94             media = try_get(
  95                 shared_data, lambda x: x['entry_data']['PostPage'][0]['media'], dict)
  96             if media:
  97                 video_url = media.get('video_url')
  98                 description = media.get('caption')
  99                 thumbnail = media.get('display_src')
 100                 timestamp = int_or_none(media.get('date'))
 101                 uploader = media.get('owner', {}).get('full_name')
 102                 uploader_id = media.get('owner', {}).get('username')
 103                 like_count = int_or_none(media.get('likes', {}).get('count'))
 104                 comment_count = int_or_none(media.get('comments', {}).get('count'))
 105
 106         if not video_url:
 107             video_url = self._og_search_video_url(webpage, secure=False)
 108
 109         if not uploader_id:
 110             uploader_id = self._search_regex(
 111                 r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
 112                 webpage, 'uploader id', fatal=False)
 113
 114         if not description:
 115             description = self._search_regex(
 116                 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
 117             if description is not None:
 118                 description = lowercase_escape(description)
 119
 120         if not thumbnail:
 121             thumbnail = self._og_search_thumbnail(webpage)
 122
 123         return {
 124             'id': video_id,
 125             'url': video_url,
 126             'ext': 'mp4',
 127             'title': 'Video by %s' % uploader_id,
 128             'description': description,
 129             'thumbnail': thumbnail,
 130             'timestamp': timestamp,
 131             'uploader_id': uploader_id,
 132             'uploader': uploader,
 133             'like_count': like_count,
 134             'comment_count': comment_count,
 135         }
 136
 137
 138 class InstagramUserIE(InfoExtractor):
 139     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
 140     IE_DESC = 'Instagram user profile'
 141     IE_NAME = 'instagram:user'
 142     _TEST = {
 143         'url': 'https://instagram.com/porsche',
 144         'info_dict': {
 145             'id': 'porsche',
 146             'title': 'porsche',
 147         },
 148         'playlist_mincount': 2,
 149         'playlist': [{
 150             'info_dict': {
 151                 'id': '614605558512799803_462752227',
 152                 'ext': 'mp4',
 153                 'title': '#Porsche Intelligent Performance.',
 154                 'thumbnail': 're:^https?://.*\.jpg',
 155                 'uploader': 'Porsche',
 156                 'uploader_id': 'porsche',
 157                 'timestamp': 1387486713,
 158                 'upload_date': '20131219',
 159             },
 160         }],
 161         'params': {
 162             'extract_flat': True,
 163             'skip_download': True,
 164         }
 165     }
 166
 167     def _real_extract(self, url):
 168         mobj = re.match(self._VALID_URL, url)
 169         uploader_id = mobj.group('username')
 170
 171         entries = []
 172         page_count = 0
 173         media_url = 'http://instagram.com/%s/media' % uploader_id
 174         while True:
 175             page = self._download_json(
 176                 media_url, uploader_id,
 177                 note='Downloading page %d ' % (page_count + 1),
 178             )
 179             page_count += 1
 180
 181             for it in page['items']:
 182                 if it.get('type') != 'video':
 183                     continue
 184                 like_count = int_or_none(it.get('likes', {}).get('count'))
 185                 user = it.get('user', {})
 186
 187                 formats = [{
 188                     'format_id': k,
 189                     'height': v.get('height'),
 190                     'width': v.get('width'),
 191                     'url': v['url'],
 192                 } for k, v in it['videos'].items()]
 193                 self._sort_formats(formats)
 194
 195                 thumbnails_el = it.get('images', {})
 196                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 197
 198                 # In some cases caption is null, which corresponds to None
 199                 # in python. As a result, it.get('caption', {}) gives None
 200                 title = (it.get('caption') or {}).get('text', it['id'])
 201
 202                 entries.append({
 203                     'id': it['id'],
 204                     'title': limit_length(title, 80),
 205                     'formats': formats,
 206                     'thumbnail': thumbnail,
 207                     'webpage_url': it.get('link'),
 208                     'uploader': user.get('full_name'),
 209                     'uploader_id': user.get('username'),
 210                     'like_count': like_count,
 211                     'timestamp': int_or_none(it.get('created_time')),
 212                 })
 213
 214             if not page['items']:
 215                 break
 216             max_id = page['items'][-1]['id'].split('_')[0]
 217             media_url = (
 218                 'http://instagram.com/%s/media?max_id=%s' % (
 219                     uploader_id, max_id))
 220
 221         return {
 222             '_type': 'playlist',
 223             'entries': entries,
 224             'id': uploader_id,
 225             'title': uploader_id,
 226         }