]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/extractor/instagram.py
Update upstream source from tag 'upstream/2018.03.14'
[youtubedl] / youtube_dl / extractor / instagram.py
1 from __future__ import unicode_literals
2
3 import itertools
4 import re
5
6 from .common import InfoExtractor
7 from ..compat import compat_str
8 from ..utils import (
9 get_element_by_attribute,
10 int_or_none,
11 lowercase_escape,
12 try_get,
13 )
14
15
16 class InstagramIE(InfoExtractor):
17 _VALID_URL = r'(?P<url>https?://(?:www\.)?instagram\.com/p/(?P<id>[^/?#&]+))'
18 _TESTS = [{
19 'url': 'https://instagram.com/p/aye83DjauH/?foo=bar#abc',
20 'md5': '0d2da106a9d2631273e192b372806516',
21 'info_dict': {
22 'id': 'aye83DjauH',
23 'ext': 'mp4',
24 'title': 'Video by naomipq',
25 'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
26 'thumbnail': r're:^https?://.*\.jpg',
27 'timestamp': 1371748545,
28 'upload_date': '20130620',
29 'uploader_id': 'naomipq',
30 'uploader': 'Naomi Leonor Phan-Quang',
31 'like_count': int,
32 'comment_count': int,
33 'comments': list,
34 },
35 }, {
36 # missing description
37 'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
38 'info_dict': {
39 'id': 'BA-pQFBG8HZ',
40 'ext': 'mp4',
41 'title': 'Video by britneyspears',
42 'thumbnail': r're:^https?://.*\.jpg',
43 'timestamp': 1453760977,
44 'upload_date': '20160125',
45 'uploader_id': 'britneyspears',
46 'uploader': 'Britney Spears',
47 'like_count': int,
48 'comment_count': int,
49 'comments': list,
50 },
51 'params': {
52 'skip_download': True,
53 },
54 }, {
55 # multi video post
56 'url': 'https://www.instagram.com/p/BQ0eAlwhDrw/',
57 'playlist': [{
58 'info_dict': {
59 'id': 'BQ0dSaohpPW',
60 'ext': 'mp4',
61 'title': 'Video 1',
62 },
63 }, {
64 'info_dict': {
65 'id': 'BQ0dTpOhuHT',
66 'ext': 'mp4',
67 'title': 'Video 2',
68 },
69 }, {
70 'info_dict': {
71 'id': 'BQ0dT7RBFeF',
72 'ext': 'mp4',
73 'title': 'Video 3',
74 },
75 }],
76 'info_dict': {
77 'id': 'BQ0eAlwhDrw',
78 'title': 'Post by instagram',
79 'description': 'md5:0f9203fc6a2ce4d228da5754bcf54957',
80 },
81 }, {
82 'url': 'https://instagram.com/p/-Cmh1cukG2/',
83 'only_matching': True,
84 }, {
85 'url': 'http://instagram.com/p/9o6LshA7zy/embed/',
86 'only_matching': True,
87 }]
88
89 @staticmethod
90 def _extract_embed_url(webpage):
91 mobj = re.search(
92 r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?instagram\.com/p/[^/]+/embed.*?)\1',
93 webpage)
94 if mobj:
95 return mobj.group('url')
96
97 blockquote_el = get_element_by_attribute(
98 'class', 'instagram-media', webpage)
99 if blockquote_el is None:
100 return
101
102 mobj = re.search(
103 r'<a[^>]+href=([\'"])(?P<link>[^\'"]+)\1', blockquote_el)
104 if mobj:
105 return mobj.group('link')
106
107 def _real_extract(self, url):
108 mobj = re.match(self._VALID_URL, url)
109 video_id = mobj.group('id')
110 url = mobj.group('url')
111
112 webpage = self._download_webpage(url, video_id)
113
114 (video_url, description, thumbnail, timestamp, uploader,
115 uploader_id, like_count, comment_count, comments, height,
116 width) = [None] * 11
117
118 shared_data = self._parse_json(
119 self._search_regex(
120 r'window\._sharedData\s*=\s*({.+?});',
121 webpage, 'shared data', default='{}'),
122 video_id, fatal=False)
123 if shared_data:
124 media = try_get(
125 shared_data,
126 (lambda x: x['entry_data']['PostPage'][0]['graphql']['shortcode_media'],
127 lambda x: x['entry_data']['PostPage'][0]['media']),
128 dict)
129 if media:
130 video_url = media.get('video_url')
131 height = int_or_none(media.get('dimensions', {}).get('height'))
132 width = int_or_none(media.get('dimensions', {}).get('width'))
133 description = try_get(
134 media, lambda x: x['edge_media_to_caption']['edges'][0]['node']['text'],
135 compat_str) or media.get('caption')
136 thumbnail = media.get('display_src')
137 timestamp = int_or_none(media.get('taken_at_timestamp') or media.get('date'))
138 uploader = media.get('owner', {}).get('full_name')
139 uploader_id = media.get('owner', {}).get('username')
140
141 def get_count(key, kind):
142 return int_or_none(try_get(
143 media, (lambda x: x['edge_media_%s' % key]['count'],
144 lambda x: x['%ss' % kind]['count'])))
145 like_count = get_count('preview_like', 'like')
146 comment_count = get_count('to_comment', 'comment')
147
148 comments = [{
149 'author': comment.get('user', {}).get('username'),
150 'author_id': comment.get('user', {}).get('id'),
151 'id': comment.get('id'),
152 'text': comment.get('text'),
153 'timestamp': int_or_none(comment.get('created_at')),
154 } for comment in media.get(
155 'comments', {}).get('nodes', []) if comment.get('text')]
156 if not video_url:
157 edges = try_get(
158 media, lambda x: x['edge_sidecar_to_children']['edges'],
159 list) or []
160 if edges:
161 entries = []
162 for edge_num, edge in enumerate(edges, start=1):
163 node = try_get(edge, lambda x: x['node'], dict)
164 if not node:
165 continue
166 node_video_url = try_get(node, lambda x: x['video_url'], compat_str)
167 if not node_video_url:
168 continue
169 entries.append({
170 'id': node.get('shortcode') or node['id'],
171 'title': 'Video %d' % edge_num,
172 'url': node_video_url,
173 'thumbnail': node.get('display_url'),
174 'width': int_or_none(try_get(node, lambda x: x['dimensions']['width'])),
175 'height': int_or_none(try_get(node, lambda x: x['dimensions']['height'])),
176 'view_count': int_or_none(node.get('video_view_count')),
177 })
178 return self.playlist_result(
179 entries, video_id,
180 'Post by %s' % uploader_id if uploader_id else None,
181 description)
182
183 if not video_url:
184 video_url = self._og_search_video_url(webpage, secure=False)
185
186 formats = [{
187 'url': video_url,
188 'width': width,
189 'height': height,
190 }]
191
192 if not uploader_id:
193 uploader_id = self._search_regex(
194 r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
195 webpage, 'uploader id', fatal=False)
196
197 if not description:
198 description = self._search_regex(
199 r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
200 if description is not None:
201 description = lowercase_escape(description)
202
203 if not thumbnail:
204 thumbnail = self._og_search_thumbnail(webpage)
205
206 return {
207 'id': video_id,
208 'formats': formats,
209 'ext': 'mp4',
210 'title': 'Video by %s' % uploader_id,
211 'description': description,
212 'thumbnail': thumbnail,
213 'timestamp': timestamp,
214 'uploader_id': uploader_id,
215 'uploader': uploader,
216 'like_count': like_count,
217 'comment_count': comment_count,
218 'comments': comments,
219 }
220
221
222 class InstagramUserIE(InfoExtractor):
223 _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
224 IE_DESC = 'Instagram user profile'
225 IE_NAME = 'instagram:user'
226 _TEST = {
227 'url': 'https://instagram.com/porsche',
228 'info_dict': {
229 'id': 'porsche',
230 'title': 'porsche',
231 },
232 'playlist_count': 5,
233 'params': {
234 'extract_flat': True,
235 'skip_download': True,
236 'playlistend': 5,
237 }
238 }
239
240 def _entries(self, uploader_id):
241 query = {
242 '__a': 1,
243 }
244
245 def get_count(kind):
246 return int_or_none(try_get(
247 node, lambda x: x['%ss' % kind]['count']))
248
249 for page_num in itertools.count(1):
250 page = self._download_json(
251 'https://instagram.com/%s/' % uploader_id, uploader_id,
252 note='Downloading page %d' % page_num,
253 fatal=False, query=query)
254 if not page:
255 break
256
257 nodes = try_get(page, lambda x: x['user']['media']['nodes'], list)
258 if not nodes:
259 break
260
261 max_id = None
262
263 for node in nodes:
264 node_id = node.get('id')
265 if node_id:
266 max_id = node_id
267
268 if node.get('__typename') != 'GraphVideo' and node.get('is_video') is not True:
269 continue
270 video_id = node.get('code')
271 if not video_id:
272 continue
273
274 info = self.url_result(
275 'https://instagram.com/p/%s/' % video_id,
276 ie=InstagramIE.ie_key(), video_id=video_id)
277
278 description = try_get(
279 node, [lambda x: x['caption'], lambda x: x['text']['id']],
280 compat_str)
281 thumbnail = node.get('thumbnail_src') or node.get('display_src')
282 timestamp = int_or_none(node.get('date'))
283
284 comment_count = get_count('comment')
285 like_count = get_count('like')
286 view_count = int_or_none(node.get('video_views'))
287
288 info.update({
289 'description': description,
290 'thumbnail': thumbnail,
291 'timestamp': timestamp,
292 'comment_count': comment_count,
293 'like_count': like_count,
294 'view_count': view_count,
295 })
296
297 yield info
298
299 if not max_id:
300 break
301
302 query['max_id'] = max_id
303
304 def _real_extract(self, url):
305 uploader_id = self._match_id(url)
306 return self.playlist_result(
307 self._entries(uploader_id), uploader_id, uploader_id)