_ 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         blockquote_el = get_element_by_attribute(
  49             'class', 'instagram-media', webpage)
  50         if blockquote_el is None:
  51             return
  52
  53         mobj = re.search(
  54             r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
  55         if mobj:
  56             return mobj.group('link')
  57
  58     def _real_extract(self, url):
  59         mobj = re.match(self._VALID_URL, url)
  60         video_id = mobj.group('id')
  61         url = mobj.group('url')
  62
  63         webpage = self._download_webpage(url, video_id)
  64         uploader_id = self._search_regex(r'"owner":{"username":"(.+?)"',
  65                                          webpage, 'uploader id', fatal=False)
  66         desc = self._search_regex(
  67             r'"caption":"(.+?)"', webpage, 'description', default=None)
  68         if desc is not None:
  69             desc = lowercase_escape(desc)
  70
  71         return {
  72             'id': video_id,
  73             'url': self._og_search_video_url(webpage, secure=False),
  74             'ext': 'mp4',
  75             'title': 'Video by %s' % uploader_id,
  76             'thumbnail': self._og_search_thumbnail(webpage),
  77             'uploader_id': uploader_id,
  78             'description': desc,
  79         }
  80
  81
  82 class InstagramUserIE(InfoExtractor):
  83     _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<username>[^/]{2,})/?(?:$|[?#])'
  84     IE_DESC = 'Instagram user profile'
  85     IE_NAME = 'instagram:user'
  86     _TEST = {
  87         'url': 'https://instagram.com/porsche',
  88         'info_dict': {
  89             'id': 'porsche',
  90             'title': 'porsche',
  91         },
  92         'playlist_mincount': 2,
  93         'playlist': [{
  94             'info_dict': {
  95                 'id': '614605558512799803_462752227',
  96                 'ext': 'mp4',
  97                 'title': '#Porsche Intelligent Performance.',
  98                 'thumbnail': 're:^https?://.*\.jpg',
  99                 'uploader': 'Porsche',
 100                 'uploader_id': 'porsche',
 101                 'timestamp': 1387486713,
 102                 'upload_date': '20131219',
 103             },
 104         }],
 105         'params': {
 106             'extract_flat': True,
 107             'skip_download': True,
 108         }
 109     }
 110
 111     def _real_extract(self, url):
 112         mobj = re.match(self._VALID_URL, url)
 113         uploader_id = mobj.group('username')
 114
 115         entries = []
 116         page_count = 0
 117         media_url = 'http://instagram.com/%s/media' % uploader_id
 118         while True:
 119             page = self._download_json(
 120                 media_url, uploader_id,
 121                 note='Downloading page %d ' % (page_count + 1),
 122             )
 123             page_count += 1
 124
 125             for it in page['items']:
 126                 if it.get('type') != 'video':
 127                     continue
 128                 like_count = int_or_none(it.get('likes', {}).get('count'))
 129                 user = it.get('user', {})
 130
 131                 formats = [{
 132                     'format_id': k,
 133                     'height': v.get('height'),
 134                     'width': v.get('width'),
 135                     'url': v['url'],
 136                 } for k, v in it['videos'].items()]
 137                 self._sort_formats(formats)
 138
 139                 thumbnails_el = it.get('images', {})
 140                 thumbnail = thumbnails_el.get('thumbnail', {}).get('url')
 141
 142                 # In some cases caption is null, which corresponds to None
 143                 # in python. As a result, it.get('caption', {}) gives None
 144                 title = (it.get('caption') or {}).get('text', it['id'])
 145
 146                 entries.append({
 147                     'id': it['id'],
 148                     'title': limit_length(title, 80),
 149                     'formats': formats,
 150                     'thumbnail': thumbnail,
 151                     'webpage_url': it.get('link'),
 152                     'uploader': user.get('full_name'),
 153                     'uploader_id': user.get('username'),
 154                     'like_count': like_count,
 155                     'timestamp': int_or_none(it.get('created_time')),
 156                 })
 157
 158             if not page['items']:
 159                 break
 160             max_id = page['items'][-1]['id'].split('_')[0]
 161             media_url = (
 162                 'http://instagram.com/%s/media?max_id=%s' % (
 163                     uploader_id, max_id))
 164
 165         return {
 166             '_type': 'playlist',
 167             'entries': entries,
 168             'id': uploader_id,
 169             'title': uploader_id,
 170         }