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 )
  11
  12
  13 class InstagramIE(InfoExtractor):
  14     _VALID_URL = r'https?://(?:www\.)?instagram\.com/p/(?P<id>[^/?#&]+)'
  15     _TESTS = [{
  16         'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
  17         'md5': '0d2da106a9d2631273e192b372806516',
  18         'info_dict': {
  19             'id': 'aye83DjauH',
  20             'ext': 'mp4',
  21             'uploader_id': 'naomipq',
  22             'title': 'Video by naomipq',
  23             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
  24         }
  25     }, {
  26         # missing description
  27         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
  28         'info_dict': {
  29             'id': 'BA-pQFBG8HZ',
  30             'ext': 'mp4',
  31             'uploader_id': 'britneyspears',
  32             'title': 'Video by britneyspears',
  33         },
  34         'params': {
  35             'skip_download': True,
  36         },
  37     }, {
  38         'url': 'https://instagram.com/p/-Cmh1cukG2/',
  39         'only_matching': True,
  40     }]
  41
  42     @staticmethod
  43     def _extract_embed_url(webpage):
  44         blockquote_el = get_element_by_attribute(
  45             'class', 'instagram-media', webpage)
  46         if blockquote_el is None:
  47             return
  48
  49         mobj = re.search(
  50             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  51         if mobj:
  52             return mobj.group('link')
  53
  54     def _real_extract(self, url):
  55         video_id = self._match_id(url)
  56
  57         webpage = self._download_webpage(url, video_id)
  58         uploader_id = self._search_regex(r'"owner":{"username":"(.+?)"',
  59                                          webpage, 'uploader id', fatal=False)
  60         desc = self._search_regex(
  61             r'"caption":"(.+?)"', webpage, 'description', default=None)
  62
  63         return {
  64             'id': video_id,
  65             'url': self._og_search_video_url(webpage, secure=False),
  66             'ext': 'mp4',
  67             'title': 'Video by %s' % uploader_id,
  68             'thumbnail': self._og_search_thumbnail(webpage),
  69             'uploader_id': uploader_id,
  70             'description': desc,
  71         }
  72
  73
  74 class InstagramUserIE(InfoExtractor):
  75     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
  76     IE_DESC = 'Instagram user profile'
  77     IE_NAME = 'instagram:user'
  78     _TEST = {
  79         'url': 'https://instagram.com/porsche',
  80         'info_dict': {
  81             'id': 'porsche',
  82             'title': 'porsche',
  83         },
  84         'playlist_mincount': 2,
  85         'playlist': [{
  86             'info_dict': {
  87                 'id': '614605558512799803_462752227',
  88                 'ext': 'mp4',
  89                 'title': '#Porsche Intelligent Performance.',
  90                 'thumbnail': 're:^https?://.*\.jpg',
  91                 'uploader': 'Porsche',
  92                 'uploader_id': 'porsche',
  93                 'timestamp': 1387486713,
  94                 'upload_date': '20131219',
  95             },
  96         }],
  97         'params': {
  98             'extract_flat': True,
  99             'skip_download': True,
 100         }
 101     }
 102
 103     def _real_extract(self, url):
 104         mobj = re.match(self._VALID_URL, url)
 105         uploader_id = mobj.group('username')
 106
 107         entries = []
 108         page_count = 0
 109         media_url = 'http://instagram.com/%s/media' % uploader_id
 110         while True:
 111             page = self._download_json(
 112                 media_url, uploader_id,
 113                 note='Downloading page %d ' % (page_count + 1),
 114             )
 115             page_count += 1
 116
 117             for it in page['items']:
 118                 if it.get('type') != 'video':
 119                     continue
 120                 like_count = int_or_none(it.get('likes', {}).get('count'))
 121                 user = it.get('user', {})
 122
 123                 formats = [{
 124                     'format_id': k,
 125                     'height': v.get('height'),
 126                     'width': v.get('width'),
 127                     'url': v['url'],
 128                 } for k, v in it['videos'].items()]
 129                 self._sort_formats(formats)
 130
 131                 thumbnails_el = it.get('images', {})
 132                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 133
 134                 # In some cases caption is null, which corresponds to None
 135                 # in python. As a result, it.get('caption', {}) gives None
 136                 title = (it.get('caption') or {}).get('text', it['id'])
 137
 138                 entries.append({
 139                     'id': it['id'],
 140                     'title': limit_length(title, 80),
 141                     'formats': formats,
 142                     'thumbnail': thumbnail,
 143                     'webpage_url': it.get('link'),
 144                     'uploader': user.get('full_name'),
 145                     'uploader_id': user.get('username'),
 146                     'like_count': like_count,
 147                     'timestamp': int_or_none(it.get('created_time')),
 148                 })
 149
 150             if not page['items']:
 151                 break
 152             max_id = page['items'][-1]['id']
 153             media_url = (
 154                 'http://instagram.com/%s/media?max_id=%s' % (
 155                     uploader_id, max_id))
 156
 157         return {
 158             '_type': 'playlist',
 159             'entries': entries,
 160             'id': uploader_id,
 161             'title': uploader_id,
 162         }