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