_ Git - youtube-dl/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 )
  12
  13
  14 class InstagramIE(InfoExtractor):
  15     _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com/p/(?P<id>[^/?#&]+))'
  16     _TESTS = [{
  17         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
  18         'md5': '0d2da106a9d2631273e192b372806516',
  19         'info_dict': {
  20             'id': 'aye83DjauH',
  21             'ext': 'mp4',
  22             'uploader_id': 'naomipq',
  23             'title': 'Video by naomipq',
  24             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
  25         }
  26     }, {
  27         # missing description
  28         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
  29         'info_dict': {
  30             'id': 'BA-pQFBG8HZ',
  31             'ext': 'mp4',
  32             'uploader_id': 'britneyspears',
  33             'title': 'Video by britneyspears',
  34         },
  35         'params': {
  36             'skip_download': True,
  37         },
  38     }, {
  39         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  40         'only_matching': True,
  41     }, {
  42         'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
  43         'only_matching': True,
  44     }]
  45
  46     @staticmethod
  47     def _extract_embed_url(webpage):
  48         mobj = re.search(
  49             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
  50             webpage)
  51         if mobj:
  52             return mobj.group('url')
  53
  54         blockquote_el = get_element_by_attribute(
  55             'class', 'instagram-media', webpage)
  56         if blockquote_el is None:
  57             return
  58
  59         mobj = re.search(
  60             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  61         if mobj:
  62             return mobj.group('link')
  63
  64     def _real_extract(self, url):
  65         mobj = re.match(self._VALID_URL, url)
  66         video_id = mobj.group('id')
  67         url = mobj.group('url')
  68
  69         webpage = self._download_webpage(url, video_id)
  70         uploader_id = self._search_regex(r'"owner":{"username":"(.+?)"',
  71                                          webpage, 'uploader id', fatal=False)
  72         desc = self._search_regex(
  73             r'"caption":"(.+?)"', webpage, 'description', default=None)
  74         if desc is not None:
  75             desc = lowercase_escape(desc)
  76
  77         return {
  78             'id': video_id,
  79             'url': self._og_search_video_url(webpage, secure=False),
  80             'ext': 'mp4',
  81             'title': 'Video by %s' % uploader_id,
  82             'thumbnail': self._og_search_thumbnail(webpage),
  83             'uploader_id': uploader_id,
  84             'description': desc,
  85         }
  86
  87
  88 class InstagramUserIE(InfoExtractor):
  89     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
  90     IE_DESC = 'Instagram user profile'
  91     IE_NAME = 'instagram:user'
  92     _TEST = {
  93         'url': 'https://instagram.com/porsche',
  94         'info_dict': {
  95             'id': 'porsche',
  96             'title': 'porsche',
  97         },
  98         'playlist_mincount': 2,
  99         'playlist': [{
 100             'info_dict': {
 101                 'id': '614605558512799803_462752227',
 102                 'ext': 'mp4',
 103                 'title': '#Porsche Intelligent Performance.',
 104                 'thumbnail': 're:^https?://.*\.jpg',
 105                 'uploader': 'Porsche',
 106                 'uploader_id': 'porsche',
 107                 'timestamp': 1387486713,
 108                 'upload_date': '20131219',
 109             },
 110         }],
 111         'params': {
 112             'extract_flat': True,
 113             'skip_download': True,
 114         }
 115     }
 116
 117     def _real_extract(self, url):
 118         mobj = re.match(self._VALID_URL, url)
 119         uploader_id = mobj.group('username')
 120
 121         entries = []
 122         page_count = 0
 123         media_url = 'http://instagram.com/%s/media' % uploader_id
 124         while True:
 125             page = self._download_json(
 126                 media_url, uploader_id,
 127                 note='Downloading page %d ' % (page_count + 1),
 128             )
 129             page_count += 1
 130
 131             for it in page['items']:
 132                 if it.get('type') != 'video':
 133                     continue
 134                 like_count = int_or_none(it.get('likes', {}).get('count'))
 135                 user = it.get('user', {})
 136
 137                 formats = [{
 138                     'format_id': k,
 139                     'height': v.get('height'),
 140                     'width': v.get('width'),
 141                     'url': v['url'],
 142                 } for k, v in it['videos'].items()]
 143                 self._sort_formats(formats)
 144
 145                 thumbnails_el = it.get('images', {})
 146                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 147
 148                 # In some cases caption is null, which corresponds to None
 149                 # in python. As a result, it.get('caption', {}) gives None
 150                 title = (it.get('caption') or {}).get('text', it['id'])
 151
 152                 entries.append({
 153                     'id': it['id'],
 154                     'title': limit_length(title, 80),
 155                     'formats': formats,
 156                     'thumbnail': thumbnail,
 157                     'webpage_url': it.get('link'),
 158                     'uploader': user.get('full_name'),
 159                     'uploader_id': user.get('username'),
 160                     'like_count': like_count,
 161                     'timestamp': int_or_none(it.get('created_time')),
 162                 })
 163
 164             if not page['items']:
 165                 break
 166             max_id = page['items'][-1]['id'].split('_')[0]
 167             media_url = (
 168                 'http://instagram.com/%s/media?max_id=%s' % (
 169                     uploader_id, max_id))
 170
 171         return {
 172             '_type': 'playlist',
 173             'entries': entries,
 174             'id': uploader_id,
 175             'title': uploader_id,
 176         }