]>
Commit | Line | Data |
---|---|---|
1 | import json | |
2 | import re | |
3 | import urllib.parse | |
4 | ||
5 | from .common import InfoExtractor | |
6 | from .youtube import YoutubeBaseInfoExtractor, YoutubeIE | |
7 | from ..compat import compat_HTTPError, compat_urllib_parse_unquote | |
8 | from ..utils import ( | |
9 | KNOWN_EXTENSIONS, | |
10 | ExtractorError, | |
11 | HEADRequest, | |
12 | bug_reports_message, | |
13 | clean_html, | |
14 | dict_get, | |
15 | extract_attributes, | |
16 | get_element_by_id, | |
17 | int_or_none, | |
18 | join_nonempty, | |
19 | js_to_json, | |
20 | merge_dicts, | |
21 | mimetype2ext, | |
22 | orderedSet, | |
23 | parse_duration, | |
24 | parse_qs, | |
25 | str_or_none, | |
26 | str_to_int, | |
27 | traverse_obj, | |
28 | try_get, | |
29 | unified_strdate, | |
30 | unified_timestamp, | |
31 | url_or_none, | |
32 | urlhandle_detect_ext, | |
33 | ) | |
34 | ||
35 | ||
36 | class ArchiveOrgIE(InfoExtractor): | |
37 | IE_NAME = 'archive.org' | |
38 | IE_DESC = 'archive.org video and audio' | |
39 | _VALID_URL = r'https?://(?:www\.)?archive\.org/(?:details|embed)/(?P<id>[^?#]+)(?:[?].*)?$' | |
40 | _TESTS = [{ | |
41 | 'url': 'http://archive.org/details/XD300-23_68HighlightsAResearchCntAugHumanIntellect', | |
42 | 'md5': '8af1d4cf447933ed3c7f4871162602db', | |
43 | 'info_dict': { | |
44 | 'id': 'XD300-23_68HighlightsAResearchCntAugHumanIntellect', | |
45 | 'ext': 'ogv', | |
46 | 'title': '1968 Demo - FJCC Conference Presentation Reel #1', | |
47 | 'description': 'md5:da45c349df039f1cc8075268eb1b5c25', | |
48 | 'release_date': '19681210', | |
49 | 'timestamp': 1268695290, | |
50 | 'upload_date': '20100315', | |
51 | 'creator': 'SRI International', | |
52 | 'uploader': 'laura@archive.org', | |
53 | 'thumbnail': r're:https://archive\.org/download/.*\.jpg', | |
54 | 'release_year': 1968, | |
55 | 'display_id': 'XD300-23_68HighlightsAResearchCntAugHumanIntellect.cdr', | |
56 | 'track': 'XD300-23 68HighlightsAResearchCntAugHumanIntellect', | |
57 | ||
58 | }, | |
59 | }, { | |
60 | 'url': 'https://archive.org/details/Cops1922', | |
61 | 'md5': '0869000b4ce265e8ca62738b336b268a', | |
62 | 'info_dict': { | |
63 | 'id': 'Cops1922', | |
64 | 'ext': 'mp4', | |
65 | 'title': 'Buster Keaton\'s "Cops" (1922)', | |
66 | 'description': 'md5:cd6f9910c35aedd5fc237dbc3957e2ca', | |
67 | 'uploader': 'yorkmba99@hotmail.com', | |
68 | 'timestamp': 1387699629, | |
69 | 'upload_date': '20131222', | |
70 | 'display_id': 'Cops-v2.mp4', | |
71 | 'thumbnail': r're:https://archive\.org/download/.*\.jpg', | |
72 | 'duration': 1091.96, | |
73 | }, | |
74 | }, { | |
75 | 'url': 'http://archive.org/embed/XD300-23_68HighlightsAResearchCntAugHumanIntellect', | |
76 | 'only_matching': True, | |
77 | }, { | |
78 | 'url': 'https://archive.org/details/Election_Ads', | |
79 | 'md5': 'eec5cddebd4793c6a653b69c3b11f2e6', | |
80 | 'info_dict': { | |
81 | 'id': 'Election_Ads/Commercial-JFK1960ElectionAdCampaignJingle.mpg', | |
82 | 'title': 'Commercial-JFK1960ElectionAdCampaignJingle.mpg', | |
83 | 'ext': 'mpg', | |
84 | 'thumbnail': r're:https://archive\.org/download/.*\.jpg', | |
85 | 'duration': 59.77, | |
86 | 'display_id': 'Commercial-JFK1960ElectionAdCampaignJingle.mpg', | |
87 | }, | |
88 | }, { | |
89 | 'url': 'https://archive.org/details/Election_Ads/Commercial-Nixon1960ElectionAdToughonDefense.mpg', | |
90 | 'md5': 'ea1eed8234e7d4165f38c8c769edef38', | |
91 | 'info_dict': { | |
92 | 'id': 'Election_Ads/Commercial-Nixon1960ElectionAdToughonDefense.mpg', | |
93 | 'title': 'Commercial-Nixon1960ElectionAdToughonDefense.mpg', | |
94 | 'ext': 'mpg', | |
95 | 'timestamp': 1205588045, | |
96 | 'uploader': 'mikedavisstripmaster@yahoo.com', | |
97 | 'description': '1960 Presidential Campaign Election Commercials John F Kennedy, Richard M Nixon', | |
98 | 'upload_date': '20080315', | |
99 | 'display_id': 'Commercial-Nixon1960ElectionAdToughonDefense.mpg', | |
100 | 'duration': 59.51, | |
101 | 'license': 'http://creativecommons.org/licenses/publicdomain/', | |
102 | 'thumbnail': r're:https://archive\.org/download/.*\.jpg', | |
103 | }, | |
104 | }, { | |
105 | 'url': 'https://archive.org/details/gd1977-05-08.shure57.stevenson.29303.flac16', | |
106 | 'md5': '7d07ffb42aba6537c28e053efa4b54c9', | |
107 | 'info_dict': { | |
108 | 'id': 'gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t01.flac', | |
109 | 'title': 'Turning', | |
110 | 'ext': 'flac', | |
111 | 'track': 'Turning', | |
112 | 'creator': 'Grateful Dead', | |
113 | 'display_id': 'gd1977-05-08d01t01.flac', | |
114 | 'track_number': 1, | |
115 | 'album': '1977-05-08 - Barton Hall - Cornell University', | |
116 | 'duration': 39.8, | |
117 | }, | |
118 | }, { | |
119 | 'url': 'https://archive.org/details/gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t07.flac', | |
120 | 'md5': 'a07cd8c6ab4ee1560f8a0021717130f3', | |
121 | 'info_dict': { | |
122 | 'id': 'gd1977-05-08.shure57.stevenson.29303.flac16/gd1977-05-08d01t07.flac', | |
123 | 'title': 'Deal', | |
124 | 'ext': 'flac', | |
125 | 'timestamp': 1205895624, | |
126 | 'uploader': 'mvernon54@yahoo.com', | |
127 | 'description': 'md5:6c921464414814720c6593810a5c7e3d', | |
128 | 'upload_date': '20080319', | |
129 | 'location': 'Barton Hall - Cornell University', | |
130 | 'duration': 438.68, | |
131 | 'track': 'Deal', | |
132 | 'creator': 'Grateful Dead', | |
133 | 'album': '1977-05-08 - Barton Hall - Cornell University', | |
134 | 'release_date': '19770508', | |
135 | 'display_id': 'gd1977-05-08d01t07.flac', | |
136 | 'release_year': 1977, | |
137 | 'track_number': 7, | |
138 | }, | |
139 | }, { | |
140 | # FIXME: give a better error message than just IndexError when all available formats are restricted | |
141 | 'url': 'https://archive.org/details/lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik', | |
142 | 'md5': '7cb019baa9b332e82ea7c10403acd180', | |
143 | 'info_dict': { | |
144 | 'id': 'lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/01.01. Bells Of Rostov.mp3', | |
145 | 'title': 'Bells Of Rostov', | |
146 | 'ext': 'mp3', | |
147 | }, | |
148 | 'skip': 'restricted' | |
149 | }, { | |
150 | 'url': 'https://archive.org/details/lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/02.02.+Song+And+Chorus+In+The+Polovetsian+Camp+From+%22Prince+Igor%22+(Act+2%2C+Scene+1).mp3', | |
151 | 'md5': '1d0aabe03edca83ca58d9ed3b493a3c3', | |
152 | 'info_dict': { | |
153 | 'id': 'lp_the-music-of-russia_various-artists-a-askaryan-alexander-melik/disc1/02.02. Song And Chorus In The Polovetsian Camp From "Prince Igor" (Act 2, Scene 1).mp3', | |
154 | 'title': 'Song And Chorus In The Polovetsian Camp From "Prince Igor" (Act 2, Scene 1)', | |
155 | 'ext': 'mp3', | |
156 | 'timestamp': 1569662587, | |
157 | 'uploader': 'associate-joygen-odiongan@archive.org', | |
158 | 'description': 'md5:012b2d668ae753be36896f343d12a236', | |
159 | 'upload_date': '20190928', | |
160 | }, | |
161 | 'skip': 'restricted' | |
162 | }, { | |
163 | # Original formats are private | |
164 | 'url': 'https://archive.org/details/irelandthemakingofarepublic', | |
165 | 'info_dict': { | |
166 | 'id': 'irelandthemakingofarepublic', | |
167 | 'title': 'Ireland: The Making of a Republic', | |
168 | 'upload_date': '20160610', | |
169 | 'description': 'md5:f70956a156645a658a0dc9513d9e78b7', | |
170 | 'uploader': 'dimitrios@archive.org', | |
171 | 'creator': ['British Broadcasting Corporation', 'Time-Life Films'], | |
172 | 'timestamp': 1465594947, | |
173 | }, | |
174 | 'playlist': [ | |
175 | { | |
176 | 'md5': '0b211261b26590d49df968f71b90690d', | |
177 | 'info_dict': { | |
178 | 'id': 'irelandthemakingofarepublic/irelandthemakingofarepublicreel1_01.mov', | |
179 | 'ext': 'mp4', | |
180 | 'title': 'irelandthemakingofarepublicreel1_01.mov', | |
181 | 'duration': 130.46, | |
182 | 'thumbnail': 'https://archive.org/download/irelandthemakingofarepublic/irelandthemakingofarepublic.thumbs/irelandthemakingofarepublicreel1_01_000117.jpg', | |
183 | 'display_id': 'irelandthemakingofarepublicreel1_01.mov', | |
184 | }, | |
185 | }, { | |
186 | 'md5': '67335ee3b23a0da930841981c1e79b02', | |
187 | 'info_dict': { | |
188 | 'id': 'irelandthemakingofarepublic/irelandthemakingofarepublicreel1_02.mov', | |
189 | 'ext': 'mp4', | |
190 | 'duration': 1395.13, | |
191 | 'title': 'irelandthemakingofarepublicreel1_02.mov', | |
192 | 'display_id': 'irelandthemakingofarepublicreel1_02.mov', | |
193 | 'thumbnail': 'https://archive.org/download/irelandthemakingofarepublic/irelandthemakingofarepublic.thumbs/irelandthemakingofarepublicreel1_02_001374.jpg', | |
194 | }, | |
195 | }, { | |
196 | 'md5': 'e470e86787893603f4a341a16c281eb5', | |
197 | 'info_dict': { | |
198 | 'id': 'irelandthemakingofarepublic/irelandthemakingofarepublicreel2.mov', | |
199 | 'ext': 'mp4', | |
200 | 'duration': 1602.67, | |
201 | 'title': 'irelandthemakingofarepublicreel2.mov', | |
202 | 'thumbnail': 'https://archive.org/download/irelandthemakingofarepublic/irelandthemakingofarepublic.thumbs/irelandthemakingofarepublicreel2_001554.jpg', | |
203 | 'display_id': 'irelandthemakingofarepublicreel2.mov', | |
204 | }, | |
205 | } | |
206 | ] | |
207 | }] | |
208 | ||
209 | @staticmethod | |
210 | def _playlist_data(webpage): | |
211 | element = re.findall(r'''(?xs) | |
212 | <input | |
213 | (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*? | |
214 | \s+class=['"]?js-play8-playlist['"]? | |
215 | (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*? | |
216 | \s*/> | |
217 | ''', webpage)[0] | |
218 | ||
219 | return json.loads(extract_attributes(element)['value']) | |
220 | ||
221 | def _real_extract(self, url): | |
222 | video_id = urllib.parse.unquote_plus(self._match_id(url)) | |
223 | identifier, entry_id = (video_id.split('/', 1) + [None])[:2] | |
224 | ||
225 | # Archive.org metadata API doesn't clearly demarcate playlist entries | |
226 | # or subtitle tracks, so we get them from the embeddable player. | |
227 | embed_page = self._download_webpage(f'https://archive.org/embed/{identifier}', identifier) | |
228 | playlist = self._playlist_data(embed_page) | |
229 | ||
230 | entries = {} | |
231 | for p in playlist: | |
232 | # If the user specified a playlist entry in the URL, ignore the | |
233 | # rest of the playlist. | |
234 | if entry_id and p['orig'] != entry_id: | |
235 | continue | |
236 | ||
237 | entries[p['orig']] = { | |
238 | 'formats': [], | |
239 | 'thumbnails': [], | |
240 | 'artist': p.get('artist'), | |
241 | 'track': p.get('title'), | |
242 | 'subtitles': {}, | |
243 | } | |
244 | ||
245 | for track in p.get('tracks', []): | |
246 | if track['kind'] != 'subtitles': | |
247 | continue | |
248 | entries[p['orig']][track['label']] = { | |
249 | 'url': 'https://archive.org/' + track['file'].lstrip('/') | |
250 | } | |
251 | ||
252 | metadata = self._download_json('http://archive.org/metadata/' + identifier, identifier) | |
253 | m = metadata['metadata'] | |
254 | identifier = m['identifier'] | |
255 | ||
256 | info = { | |
257 | 'id': identifier, | |
258 | 'title': m['title'], | |
259 | 'description': clean_html(m.get('description')), | |
260 | 'uploader': dict_get(m, ['uploader', 'adder']), | |
261 | 'creator': m.get('creator'), | |
262 | 'license': m.get('licenseurl'), | |
263 | 'release_date': unified_strdate(m.get('date')), | |
264 | 'timestamp': unified_timestamp(dict_get(m, ['publicdate', 'addeddate'])), | |
265 | 'webpage_url': f'https://archive.org/details/{identifier}', | |
266 | 'location': m.get('venue'), | |
267 | 'release_year': int_or_none(m.get('year'))} | |
268 | ||
269 | for f in metadata['files']: | |
270 | if f['name'] in entries: | |
271 | entries[f['name']] = merge_dicts(entries[f['name']], { | |
272 | 'id': identifier + '/' + f['name'], | |
273 | 'title': f.get('title') or f['name'], | |
274 | 'display_id': f['name'], | |
275 | 'description': clean_html(f.get('description')), | |
276 | 'creator': f.get('creator'), | |
277 | 'duration': parse_duration(f.get('length')), | |
278 | 'track_number': int_or_none(f.get('track')), | |
279 | 'album': f.get('album'), | |
280 | 'discnumber': int_or_none(f.get('disc')), | |
281 | 'release_year': int_or_none(f.get('year'))}) | |
282 | entry = entries[f['name']] | |
283 | elif traverse_obj(f, 'original', expected_type=str) in entries: | |
284 | entry = entries[f['original']] | |
285 | else: | |
286 | continue | |
287 | ||
288 | if f.get('format') == 'Thumbnail': | |
289 | entry['thumbnails'].append({ | |
290 | 'id': f['name'], | |
291 | 'url': 'https://archive.org/download/' + identifier + '/' + f['name'], | |
292 | 'width': int_or_none(f.get('width')), | |
293 | 'height': int_or_none(f.get('width')), | |
294 | 'filesize': int_or_none(f.get('size'))}) | |
295 | ||
296 | extension = (f['name'].rsplit('.', 1) + [None])[1] | |
297 | ||
298 | # We don't want to skip private formats if the user has access to them, | |
299 | # however without access to an account with such privileges we can't implement/test this. | |
300 | # For now to be safe, we will only skip them if there is no user logged in. | |
301 | is_logged_in = bool(self._get_cookies('https://archive.org').get('logged-in-sig')) | |
302 | if extension in KNOWN_EXTENSIONS and (not f.get('private') or is_logged_in): | |
303 | entry['formats'].append({ | |
304 | 'url': 'https://archive.org/download/' + identifier + '/' + f['name'], | |
305 | 'format': f.get('format'), | |
306 | 'width': int_or_none(f.get('width')), | |
307 | 'height': int_or_none(f.get('height')), | |
308 | 'filesize': int_or_none(f.get('size')), | |
309 | 'protocol': 'https', | |
310 | 'source_preference': 0 if f.get('source') == 'original' else -1, | |
311 | 'format_note': f.get('source') | |
312 | }) | |
313 | ||
314 | for entry in entries.values(): | |
315 | entry['_format_sort_fields'] = ('source', ) | |
316 | ||
317 | if len(entries) == 1: | |
318 | # If there's only one item, use it as the main info dict | |
319 | only_video = next(iter(entries.values())) | |
320 | if entry_id: | |
321 | info = merge_dicts(only_video, info) | |
322 | else: | |
323 | info = merge_dicts(info, only_video) | |
324 | else: | |
325 | # Otherwise, we have a playlist. | |
326 | info['_type'] = 'playlist' | |
327 | info['entries'] = list(entries.values()) | |
328 | ||
329 | if metadata.get('reviews'): | |
330 | info['comments'] = [] | |
331 | for review in metadata['reviews']: | |
332 | info['comments'].append({ | |
333 | 'id': review.get('review_id'), | |
334 | 'author': review.get('reviewer'), | |
335 | 'text': str_or_none(review.get('reviewtitle'), '') + '\n\n' + review.get('reviewbody'), | |
336 | 'timestamp': unified_timestamp(review.get('createdate')), | |
337 | 'parent': 'root'}) | |
338 | ||
339 | return info | |
340 | ||
341 | ||
342 | class YoutubeWebArchiveIE(InfoExtractor): | |
343 | IE_NAME = 'web.archive:youtube' | |
344 | IE_DESC = 'web.archive.org saved youtube videos, "ytarchive:" prefix' | |
345 | _VALID_URL = r'''(?x)(?:(?P<prefix>ytarchive:)| | |
346 | (?:https?://)?web\.archive\.org/ | |
347 | (?:web/)?(?:(?P<date>[0-9]{14})?[0-9A-Za-z_*]*/)? # /web and the version index is optional | |
348 | (?:https?(?::|%3[Aa])//)?(?: | |
349 | (?:\w+\.)?youtube\.com(?::(?:80|443))?/watch(?:\.php)?(?:\?|%3[fF])(?:[^\#]+(?:&|%26))?v(?:=|%3[dD]) # Youtube URL | |
350 | |(?:wayback-fakeurl\.archive\.org/yt/) # Or the internal fake url | |
351 | ) | |
352 | )(?P<id>[0-9A-Za-z_-]{11}) | |
353 | (?(prefix) | |
354 | (?::(?P<date2>[0-9]{14}))?$| | |
355 | (?:%26|[#&]|$) | |
356 | )''' | |
357 | ||
358 | _TESTS = [ | |
359 | { | |
360 | 'url': 'https://web.archive.org/web/20150415002341/https://www.youtube.com/watch?v=aYAGB11YrSs', | |
361 | 'info_dict': { | |
362 | 'id': 'aYAGB11YrSs', | |
363 | 'ext': 'webm', | |
364 | 'title': 'Team Fortress 2 - Sandviches!', | |
365 | 'description': 'md5:4984c0f9a07f349fc5d8e82ab7af4eaf', | |
366 | 'upload_date': '20110926', | |
367 | 'uploader': 'Zeurel', | |
368 | 'channel_id': 'UCukCyHaD-bK3in_pKpfH9Eg', | |
369 | 'duration': 32, | |
370 | 'uploader_id': 'Zeurel', | |
371 | 'uploader_url': 'https://www.youtube.com/user/Zeurel', | |
372 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
373 | 'channel_url': 'https://www.youtube.com/channel/UCukCyHaD-bK3in_pKpfH9Eg', | |
374 | } | |
375 | }, { | |
376 | # Internal link | |
377 | 'url': 'https://web.archive.org/web/2oe/http://wayback-fakeurl.archive.org/yt/97t7Xj_iBv0', | |
378 | 'info_dict': { | |
379 | 'id': '97t7Xj_iBv0', | |
380 | 'ext': 'mp4', | |
381 | 'title': 'Why Machines That Bend Are Better', | |
382 | 'description': 'md5:00404df2c632d16a674ff8df1ecfbb6c', | |
383 | 'upload_date': '20190312', | |
384 | 'uploader': 'Veritasium', | |
385 | 'channel_id': 'UCHnyfMqiRRG1u-2MsSQLbXA', | |
386 | 'duration': 771, | |
387 | 'uploader_id': '1veritasium', | |
388 | 'uploader_url': 'https://www.youtube.com/user/1veritasium', | |
389 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
390 | 'channel_url': 'https://www.youtube.com/channel/UCHnyfMqiRRG1u-2MsSQLbXA', | |
391 | } | |
392 | }, { | |
393 | # Video from 2012, webm format itag 45. Newest capture is deleted video, with an invalid description. | |
394 | # Should use the date in the link. Title ends with '- Youtube'. Capture has description in eow-description | |
395 | 'url': 'https://web.archive.org/web/20120712231619/http://www.youtube.com/watch?v=AkhihxRKcrs&gl=US&hl=en', | |
396 | 'info_dict': { | |
397 | 'id': 'AkhihxRKcrs', | |
398 | 'ext': 'webm', | |
399 | 'title': 'Limited Run: Mondo\'s Modern Classic 1 of 3 (SDCC 2012)', | |
400 | 'upload_date': '20120712', | |
401 | 'duration': 398, | |
402 | 'description': 'md5:ff4de6a7980cb65d951c2f6966a4f2f3', | |
403 | 'uploader_id': 'machinima', | |
404 | 'uploader_url': 'https://www.youtube.com/user/machinima', | |
405 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
406 | 'uploader': 'machinima' | |
407 | } | |
408 | }, { | |
409 | # FLV video. Video file URL does not provide itag information | |
410 | 'url': 'https://web.archive.org/web/20081211103536/http://www.youtube.com/watch?v=jNQXAC9IVRw', | |
411 | 'info_dict': { | |
412 | 'id': 'jNQXAC9IVRw', | |
413 | 'ext': 'flv', | |
414 | 'title': 'Me at the zoo', | |
415 | 'upload_date': '20050423', | |
416 | 'channel_id': 'UC4QobU6STFB0P71PMvOGN5A', | |
417 | 'duration': 19, | |
418 | 'description': 'md5:10436b12e07ac43ff8df65287a56efb4', | |
419 | 'uploader_id': 'jawed', | |
420 | 'uploader_url': 'https://www.youtube.com/user/jawed', | |
421 | 'channel_url': 'https://www.youtube.com/channel/UC4QobU6STFB0P71PMvOGN5A', | |
422 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
423 | 'uploader': 'jawed', | |
424 | } | |
425 | }, { | |
426 | 'url': 'https://web.archive.org/web/20110712231407/http://www.youtube.com/watch?v=lTx3G6h2xyA', | |
427 | 'info_dict': { | |
428 | 'id': 'lTx3G6h2xyA', | |
429 | 'ext': 'flv', | |
430 | 'title': 'Madeon - Pop Culture (live mashup)', | |
431 | 'upload_date': '20110711', | |
432 | 'uploader': 'Madeon', | |
433 | 'channel_id': 'UCqMDNf3Pn5L7pcNkuSEeO3w', | |
434 | 'duration': 204, | |
435 | 'description': 'md5:f7535343b6eda34a314eff8b85444680', | |
436 | 'uploader_id': 'itsmadeon', | |
437 | 'uploader_url': 'https://www.youtube.com/user/itsmadeon', | |
438 | 'channel_url': 'https://www.youtube.com/channel/UCqMDNf3Pn5L7pcNkuSEeO3w', | |
439 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
440 | } | |
441 | }, { | |
442 | # First capture is of dead video, second is the oldest from CDX response. | |
443 | 'url': 'https://web.archive.org/https://www.youtube.com/watch?v=1JYutPM8O6E', | |
444 | 'info_dict': { | |
445 | 'id': '1JYutPM8O6E', | |
446 | 'ext': 'mp4', | |
447 | 'title': 'Fake Teen Doctor Strikes AGAIN! - Weekly Weird News', | |
448 | 'upload_date': '20160218', | |
449 | 'channel_id': 'UCdIaNUarhzLSXGoItz7BHVA', | |
450 | 'duration': 1235, | |
451 | 'description': 'md5:21032bae736421e89c2edf36d1936947', | |
452 | 'uploader_id': 'MachinimaETC', | |
453 | 'uploader_url': 'https://www.youtube.com/user/MachinimaETC', | |
454 | 'channel_url': 'https://www.youtube.com/channel/UCdIaNUarhzLSXGoItz7BHVA', | |
455 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
456 | 'uploader': 'ETC News', | |
457 | } | |
458 | }, { | |
459 | # First capture of dead video, capture date in link links to dead capture. | |
460 | 'url': 'https://web.archive.org/web/20180803221945/https://www.youtube.com/watch?v=6FPhZJGvf4E', | |
461 | 'info_dict': { | |
462 | 'id': '6FPhZJGvf4E', | |
463 | 'ext': 'mp4', | |
464 | 'title': 'WTF: Video Games Still Launch BROKEN?! - T.U.G.S.', | |
465 | 'upload_date': '20160219', | |
466 | 'channel_id': 'UCdIaNUarhzLSXGoItz7BHVA', | |
467 | 'duration': 797, | |
468 | 'description': 'md5:a1dbf12d9a3bd7cb4c5e33b27d77ffe7', | |
469 | 'uploader_id': 'MachinimaETC', | |
470 | 'uploader_url': 'https://www.youtube.com/user/MachinimaETC', | |
471 | 'channel_url': 'https://www.youtube.com/channel/UCdIaNUarhzLSXGoItz7BHVA', | |
472 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
473 | 'uploader': 'ETC News', | |
474 | }, | |
475 | 'expected_warnings': [ | |
476 | r'unable to download capture webpage \(it may not be archived\)' | |
477 | ] | |
478 | }, { # Very old YouTube page, has - YouTube in title. | |
479 | 'url': 'http://web.archive.org/web/20070302011044/http://youtube.com/watch?v=-06-KB9XTzg', | |
480 | 'info_dict': { | |
481 | 'id': '-06-KB9XTzg', | |
482 | 'ext': 'flv', | |
483 | 'title': 'New Coin Hack!! 100% Safe!!' | |
484 | } | |
485 | }, { | |
486 | 'url': 'web.archive.org/https://www.youtube.com/watch?v=dWW7qP423y8', | |
487 | 'info_dict': { | |
488 | 'id': 'dWW7qP423y8', | |
489 | 'ext': 'mp4', | |
490 | 'title': 'It\'s Bootleg AirPods Time.', | |
491 | 'upload_date': '20211021', | |
492 | 'channel_id': 'UC7Jwj9fkrf1adN4fMmTkpug', | |
493 | 'channel_url': 'https://www.youtube.com/channel/UC7Jwj9fkrf1adN4fMmTkpug', | |
494 | 'duration': 810, | |
495 | 'description': 'md5:7b567f898d8237b256f36c1a07d6d7bc', | |
496 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
497 | 'uploader': 'DankPods', | |
498 | } | |
499 | }, { | |
500 | # player response contains '};' See: https://github.com/ytdl-org/youtube-dl/issues/27093 | |
501 | 'url': 'https://web.archive.org/web/20200827003909if_/http://www.youtube.com/watch?v=6Dh-RL__uN4', | |
502 | 'info_dict': { | |
503 | 'id': '6Dh-RL__uN4', | |
504 | 'ext': 'mp4', | |
505 | 'title': 'bitch lasagna', | |
506 | 'upload_date': '20181005', | |
507 | 'channel_id': 'UC-lHJZR3Gqxm24_Vd_AJ5Yw', | |
508 | 'channel_url': 'https://www.youtube.com/channel/UC-lHJZR3Gqxm24_Vd_AJ5Yw', | |
509 | 'duration': 135, | |
510 | 'description': 'md5:2dbe4051feeff2dab5f41f82bb6d11d0', | |
511 | 'uploader': 'PewDiePie', | |
512 | 'uploader_id': 'PewDiePie', | |
513 | 'uploader_url': 'https://www.youtube.com/user/PewDiePie', | |
514 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
515 | } | |
516 | }, { | |
517 | # ~June 2010 Capture. swfconfig | |
518 | 'url': 'https://web.archive.org/web/0/https://www.youtube.com/watch?v=8XeW5ilk-9Y', | |
519 | 'info_dict': { | |
520 | 'id': '8XeW5ilk-9Y', | |
521 | 'ext': 'flv', | |
522 | 'title': 'Story of Stuff, The Critique Part 4 of 4', | |
523 | 'duration': 541, | |
524 | 'description': 'md5:28157da06f2c5e94c97f7f3072509972', | |
525 | 'uploader': 'HowTheWorldWorks', | |
526 | 'uploader_id': 'HowTheWorldWorks', | |
527 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
528 | 'uploader_url': 'https://www.youtube.com/user/HowTheWorldWorks', | |
529 | 'upload_date': '20090520', | |
530 | } | |
531 | }, { | |
532 | # Jan 2011: watch-video-date/eow-date surrounded by whitespace | |
533 | 'url': 'https://web.archive.org/web/20110126141719/http://www.youtube.com/watch?v=Q_yjX80U7Yc', | |
534 | 'info_dict': { | |
535 | 'id': 'Q_yjX80U7Yc', | |
536 | 'ext': 'flv', | |
537 | 'title': 'Spray Paint Art by Clay Butler: Purple Fantasy Forest', | |
538 | 'uploader_id': 'claybutlermusic', | |
539 | 'description': 'md5:4595264559e3d0a0ceb3f011f6334543', | |
540 | 'upload_date': '20090803', | |
541 | 'uploader': 'claybutlermusic', | |
542 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
543 | 'duration': 132, | |
544 | 'uploader_url': 'https://www.youtube.com/user/claybutlermusic', | |
545 | } | |
546 | }, { | |
547 | # ~May 2009 swfArgs. ytcfg is spread out over various vars | |
548 | 'url': 'https://web.archive.org/web/0/https://www.youtube.com/watch?v=c5uJgG05xUY', | |
549 | 'info_dict': { | |
550 | 'id': 'c5uJgG05xUY', | |
551 | 'ext': 'webm', | |
552 | 'title': 'Story of Stuff, The Critique Part 1 of 4', | |
553 | 'uploader_id': 'HowTheWorldWorks', | |
554 | 'uploader': 'HowTheWorldWorks', | |
555 | 'uploader_url': 'https://www.youtube.com/user/HowTheWorldWorks', | |
556 | 'upload_date': '20090513', | |
557 | 'description': 'md5:4ca77d79538064e41e4cc464e93f44f0', | |
558 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
559 | 'duration': 754, | |
560 | } | |
561 | }, { | |
562 | # ~June 2012. Upload date is in another lang so cannot extract. | |
563 | 'url': 'https://web.archive.org/web/20120607174520/http://www.youtube.com/watch?v=xWTLLl-dQaA', | |
564 | 'info_dict': { | |
565 | 'id': 'xWTLLl-dQaA', | |
566 | 'ext': 'mp4', | |
567 | 'title': 'Black Nerd eHarmony Video Bio Parody (SPOOF)', | |
568 | 'uploader_url': 'https://www.youtube.com/user/BlackNerdComedy', | |
569 | 'description': 'md5:e25f0133aaf9e6793fb81c18021d193e', | |
570 | 'uploader_id': 'BlackNerdComedy', | |
571 | 'uploader': 'BlackNerdComedy', | |
572 | 'duration': 182, | |
573 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
574 | } | |
575 | }, { | |
576 | # ~July 2013 | |
577 | 'url': 'https://web.archive.org/web/*/https://www.youtube.com/watch?v=9eO1aasHyTM', | |
578 | 'info_dict': { | |
579 | 'id': '9eO1aasHyTM', | |
580 | 'ext': 'mp4', | |
581 | 'title': 'Polar-oid', | |
582 | 'description': 'Cameras and bears are dangerous!', | |
583 | 'uploader_url': 'https://www.youtube.com/user/punkybird', | |
584 | 'uploader_id': 'punkybird', | |
585 | 'duration': 202, | |
586 | 'channel_id': 'UC62R2cBezNBOqxSerfb1nMQ', | |
587 | 'channel_url': 'https://www.youtube.com/channel/UC62R2cBezNBOqxSerfb1nMQ', | |
588 | 'upload_date': '20060428', | |
589 | 'uploader': 'punkybird', | |
590 | } | |
591 | }, { | |
592 | # April 2020: Player response in player config | |
593 | 'url': 'https://web.archive.org/web/20200416034815/https://www.youtube.com/watch?v=Cf7vS8jc7dY&gl=US&hl=en', | |
594 | 'info_dict': { | |
595 | 'id': 'Cf7vS8jc7dY', | |
596 | 'ext': 'mp4', | |
597 | 'title': 'A Dramatic Pool Story (by Jamie Spicer-Lewis) - Game Grumps Animated', | |
598 | 'duration': 64, | |
599 | 'upload_date': '20200408', | |
600 | 'uploader_id': 'GameGrumps', | |
601 | 'uploader': 'GameGrumps', | |
602 | 'channel_url': 'https://www.youtube.com/channel/UC9CuvdOVfMPvKCiwdGKL3cQ', | |
603 | 'channel_id': 'UC9CuvdOVfMPvKCiwdGKL3cQ', | |
604 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
605 | 'description': 'md5:c625bb3c02c4f5fb4205971e468fa341', | |
606 | 'uploader_url': 'https://www.youtube.com/user/GameGrumps', | |
607 | } | |
608 | }, { | |
609 | # watch7-user-header with yt-user-info | |
610 | 'url': 'ytarchive:kbh4T_b4Ixw:20160307085057', | |
611 | 'info_dict': { | |
612 | 'id': 'kbh4T_b4Ixw', | |
613 | 'ext': 'mp4', | |
614 | 'title': 'Shovel Knight OST - Strike the Earth! Plains of Passage 16 bit SNES style remake / remix', | |
615 | 'channel_url': 'https://www.youtube.com/channel/UCnTaGvsHmMy792DWeT6HbGA', | |
616 | 'uploader': 'Nelward music', | |
617 | 'duration': 213, | |
618 | 'description': 'md5:804b4a9ce37b050a5fefdbb23aeba54d', | |
619 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
620 | 'upload_date': '20150503', | |
621 | 'channel_id': 'UCnTaGvsHmMy792DWeT6HbGA', | |
622 | } | |
623 | }, { | |
624 | # April 2012 | |
625 | 'url': 'https://web.archive.org/web/0/https://www.youtube.com/watch?v=SOm7mPoPskU', | |
626 | 'info_dict': { | |
627 | 'id': 'SOm7mPoPskU', | |
628 | 'ext': 'mp4', | |
629 | 'title': 'Boyfriend - Justin Bieber Parody', | |
630 | 'uploader_url': 'https://www.youtube.com/user/thecomputernerd01', | |
631 | 'uploader': 'thecomputernerd01', | |
632 | 'thumbnail': r're:https?://.*\.(jpg|webp)', | |
633 | 'description': 'md5:dd7fa635519c2a5b4d566beaecad7491', | |
634 | 'duration': 200, | |
635 | 'upload_date': '20120407', | |
636 | 'uploader_id': 'thecomputernerd01', | |
637 | } | |
638 | }, { | |
639 | 'url': 'https://web.archive.org/web/http://www.youtube.com/watch?v=kH-G_aIBlFw', | |
640 | 'only_matching': True | |
641 | }, { | |
642 | 'url': 'https://web.archive.org/web/20050214000000_if/http://www.youtube.com/watch?v=0altSZ96U4M', | |
643 | 'only_matching': True | |
644 | }, { | |
645 | # Video not archived, only capture is unavailable video page | |
646 | 'url': 'https://web.archive.org/web/20210530071008/https://www.youtube.com/watch?v=lHJTf93HL1s&spfreload=10', | |
647 | 'only_matching': True | |
648 | }, { # Encoded url | |
649 | 'url': 'https://web.archive.org/web/20120712231619/http%3A//www.youtube.com/watch%3Fgl%3DUS%26v%3DAkhihxRKcrs%26hl%3Den', | |
650 | 'only_matching': True | |
651 | }, { | |
652 | 'url': 'https://web.archive.org/web/20120712231619/http%3A//www.youtube.com/watch%3Fv%3DAkhihxRKcrs%26gl%3DUS%26hl%3Den', | |
653 | 'only_matching': True | |
654 | }, { | |
655 | 'url': 'https://web.archive.org/web/20060527081937/http://www.youtube.com:80/watch.php?v=ELTFsLT73fA&search=soccer', | |
656 | 'only_matching': True | |
657 | }, { | |
658 | 'url': 'https://web.archive.org/http://www.youtube.com:80/watch?v=-05VVye-ffg', | |
659 | 'only_matching': True | |
660 | }, { | |
661 | 'url': 'ytarchive:BaW_jenozKc:20050214000000', | |
662 | 'only_matching': True | |
663 | }, { | |
664 | 'url': 'ytarchive:BaW_jenozKc', | |
665 | 'only_matching': True | |
666 | }, | |
667 | ] | |
668 | _YT_INITIAL_DATA_RE = YoutubeBaseInfoExtractor._YT_INITIAL_DATA_RE | |
669 | _YT_INITIAL_PLAYER_RESPONSE_RE = fr'''(?x: | |
670 | (?:window\s*\[\s*["\']ytInitialPlayerResponse["\']\s*\]|ytInitialPlayerResponse)\s*=[(\s]*| | |
671 | {YoutubeBaseInfoExtractor._YT_INITIAL_PLAYER_RESPONSE_RE} | |
672 | )''' | |
673 | ||
674 | _YT_DEFAULT_THUMB_SERVERS = ['i.ytimg.com'] # thumbnails most likely archived on these servers | |
675 | _YT_ALL_THUMB_SERVERS = orderedSet( | |
676 | _YT_DEFAULT_THUMB_SERVERS + ['img.youtube.com', *[f'{c}{n or ""}.ytimg.com' for c in ('i', 's') for n in (*range(0, 5), 9)]]) | |
677 | ||
678 | _WAYBACK_BASE_URL = 'https://web.archive.org/web/%sif_/' | |
679 | _OLDEST_CAPTURE_DATE = 20050214000000 | |
680 | _NEWEST_CAPTURE_DATE = 20500101000000 | |
681 | ||
682 | def _call_cdx_api(self, item_id, url, filters: list = None, collapse: list = None, query: dict = None, note=None, fatal=False): | |
683 | # CDX docs: https://github.com/internetarchive/wayback/blob/master/wayback-cdx-server/README.md | |
684 | query = { | |
685 | 'url': url, | |
686 | 'output': 'json', | |
687 | 'fl': 'original,mimetype,length,timestamp', | |
688 | 'limit': 500, | |
689 | 'filter': ['statuscode:200'] + (filters or []), | |
690 | 'collapse': collapse or [], | |
691 | **(query or {}) | |
692 | } | |
693 | res = self._download_json( | |
694 | 'https://web.archive.org/cdx/search/cdx', item_id, | |
695 | note or 'Downloading CDX API JSON', query=query, fatal=fatal) | |
696 | if isinstance(res, list) and len(res) >= 2: | |
697 | # format response to make it easier to use | |
698 | return list(dict(zip(res[0], v)) for v in res[1:]) | |
699 | elif not isinstance(res, list) or len(res) != 0: | |
700 | self.report_warning('Error while parsing CDX API response' + bug_reports_message()) | |
701 | ||
702 | def _extract_webpage_title(self, webpage): | |
703 | page_title = self._html_extract_title(webpage, default='') | |
704 | # YouTube video pages appear to always have either 'YouTube -' as prefix or '- YouTube' as suffix. | |
705 | return self._html_search_regex( | |
706 | r'(?:YouTube\s*-\s*(.*)$)|(?:(.*)\s*-\s*YouTube$)', | |
707 | page_title, 'title', default='') | |
708 | ||
709 | def _extract_metadata(self, video_id, webpage): | |
710 | search_meta = ((lambda x: self._html_search_meta(x, webpage, default=None)) if webpage else (lambda x: None)) | |
711 | player_response = self._search_json( | |
712 | self._YT_INITIAL_PLAYER_RESPONSE_RE, webpage, 'initial player response', | |
713 | video_id, default={}) | |
714 | initial_data = self._search_json( | |
715 | self._YT_INITIAL_DATA_RE, webpage, 'initial data', video_id, default={}) | |
716 | ||
717 | ytcfg = {} | |
718 | for j in re.findall(r'yt\.setConfig\(\s*(?P<json>{\s*(?s:.+?)\s*})\s*\);', webpage): # ~June 2010 | |
719 | ytcfg.update(self._parse_json(j, video_id, fatal=False, ignore_extra=True, transform_source=js_to_json, errnote='') or {}) | |
720 | ||
721 | # XXX: this also may contain a 'ptchn' key | |
722 | player_config = ( | |
723 | self._search_json( | |
724 | r'(?:yt\.playerConfig|ytplayer\.config|swfConfig)\s*=', | |
725 | webpage, 'player config', video_id, default=None) | |
726 | or ytcfg.get('PLAYER_CONFIG') or {}) | |
727 | ||
728 | # XXX: this may also contain a 'creator' key. | |
729 | swf_args = self._search_json(r'swfArgs\s*=', webpage, 'swf config', video_id, default={}) | |
730 | if swf_args and not traverse_obj(player_config, ('args',)): | |
731 | player_config['args'] = swf_args | |
732 | ||
733 | if not player_response: | |
734 | # April 2020 | |
735 | player_response = self._parse_json( | |
736 | traverse_obj(player_config, ('args', 'player_response')) or '{}', video_id, fatal=False) | |
737 | ||
738 | initial_data_video = traverse_obj( | |
739 | initial_data, ('contents', 'twoColumnWatchNextResults', 'results', 'results', 'contents', ..., 'videoPrimaryInfoRenderer'), | |
740 | expected_type=dict, get_all=False, default={}) | |
741 | ||
742 | video_details = traverse_obj( | |
743 | player_response, 'videoDetails', expected_type=dict, get_all=False, default={}) | |
744 | ||
745 | microformats = traverse_obj( | |
746 | player_response, ('microformat', 'playerMicroformatRenderer'), expected_type=dict, get_all=False, default={}) | |
747 | ||
748 | video_title = ( | |
749 | video_details.get('title') | |
750 | or YoutubeBaseInfoExtractor._get_text(microformats, 'title') | |
751 | or YoutubeBaseInfoExtractor._get_text(initial_data_video, 'title') | |
752 | or traverse_obj(player_config, ('args', 'title')) | |
753 | or self._extract_webpage_title(webpage) | |
754 | or search_meta(['og:title', 'twitter:title', 'title'])) | |
755 | ||
756 | def id_from_url(url, type_): | |
757 | return self._search_regex( | |
758 | rf'(?:{type_})/([^/#&?]+)', url or '', f'{type_} id', default=None) | |
759 | ||
760 | # XXX: would the get_elements_by_... functions be better suited here? | |
761 | _CHANNEL_URL_HREF_RE = r'href="[^"]*(?P<url>https?://www\.youtube\.com/(?:user|channel)/[^"]+)"' | |
762 | uploader_or_channel_url = self._search_regex( | |
763 | [fr'<(?:link\s*itemprop=\"url\"|a\s*id=\"watch-username\").*?\b{_CHANNEL_URL_HREF_RE}>', # @fd05024 | |
764 | fr'<div\s*id=\"(?:watch-channel-stats|watch-headline-user-info)\"[^>]*>\s*<a[^>]*\b{_CHANNEL_URL_HREF_RE}'], # ~ May 2009, ~June 2012 | |
765 | webpage, 'uploader or channel url', default=None) | |
766 | ||
767 | owner_profile_url = url_or_none(microformats.get('ownerProfileUrl')) # @a6211d2 | |
768 | ||
769 | # Uploader refers to the /user/ id ONLY | |
770 | uploader_id = ( | |
771 | id_from_url(owner_profile_url, 'user') | |
772 | or id_from_url(uploader_or_channel_url, 'user') | |
773 | or ytcfg.get('VIDEO_USERNAME')) | |
774 | uploader_url = f'https://www.youtube.com/user/{uploader_id}' if uploader_id else None | |
775 | ||
776 | # XXX: do we want to differentiate uploader and channel? | |
777 | uploader = ( | |
778 | self._search_regex( | |
779 | [r'<a\s*id="watch-username"[^>]*>\s*<strong>([^<]+)</strong>', # June 2010 | |
780 | r'var\s*watchUsername\s*=\s*\'(.+?)\';', # ~May 2009 | |
781 | r'<div\s*\bid=\"watch-channel-stats"[^>]*>\s*<a[^>]*>\s*(.+?)\s*</a', # ~May 2009 | |
782 | r'<a\s*id="watch-userbanner"[^>]*title="\s*(.+?)\s*"'], # ~June 2012 | |
783 | webpage, 'uploader', default=None) | |
784 | or self._html_search_regex( | |
785 | [r'(?s)<div\s*class="yt-user-info".*?<a[^>]*[^>]*>\s*(.*?)\s*</a', # March 2016 | |
786 | r'(?s)<a[^>]*yt-user-name[^>]*>\s*(.*?)\s*</a'], # july 2013 | |
787 | get_element_by_id('watch7-user-header', webpage), 'uploader', default=None) | |
788 | or self._html_search_regex( | |
789 | r'<button\s*href="/user/[^>]*>\s*<span[^>]*>\s*(.+?)\s*<', # April 2012 | |
790 | get_element_by_id('watch-headline-user-info', webpage), 'uploader', default=None) | |
791 | or traverse_obj(player_config, ('args', 'creator')) | |
792 | or video_details.get('author')) | |
793 | ||
794 | channel_id = str_or_none( | |
795 | video_details.get('channelId') | |
796 | or microformats.get('externalChannelId') | |
797 | or search_meta('channelId') | |
798 | or self._search_regex( | |
799 | r'data-channel-external-id=(["\'])(?P<id>(?:(?!\1).)+)\1', # @b45a9e6 | |
800 | webpage, 'channel id', default=None, group='id') | |
801 | or id_from_url(owner_profile_url, 'channel') | |
802 | or id_from_url(uploader_or_channel_url, 'channel') | |
803 | or traverse_obj(player_config, ('args', 'ucid'))) | |
804 | ||
805 | channel_url = f'https://www.youtube.com/channel/{channel_id}' if channel_id else None | |
806 | duration = int_or_none( | |
807 | video_details.get('lengthSeconds') | |
808 | or microformats.get('lengthSeconds') | |
809 | or traverse_obj(player_config, ('args', ('length_seconds', 'l')), get_all=False) | |
810 | or parse_duration(search_meta('duration'))) | |
811 | description = ( | |
812 | video_details.get('shortDescription') | |
813 | or YoutubeBaseInfoExtractor._get_text(microformats, 'description') | |
814 | or clean_html(get_element_by_id('eow-description', webpage)) # @9e6dd23 | |
815 | or search_meta(['description', 'og:description', 'twitter:description'])) | |
816 | ||
817 | upload_date = unified_strdate( | |
818 | dict_get(microformats, ('uploadDate', 'publishDate')) | |
819 | or search_meta(['uploadDate', 'datePublished']) | |
820 | or self._search_regex( | |
821 | [r'(?s)id="eow-date.*?>\s*(.*?)\s*</span>', | |
822 | r'(?:id="watch-uploader-info".*?>.*?|["\']simpleText["\']\s*:\s*["\'])(?:Published|Uploaded|Streamed live|Started) on (.+?)[<"\']', # @7998520 | |
823 | r'class\s*=\s*"(?:watch-video-date|watch-video-added post-date)"[^>]*>\s*([^<]+?)\s*<'], # ~June 2010, ~Jan 2009 (respectively) | |
824 | webpage, 'upload date', default=None)) | |
825 | ||
826 | return { | |
827 | 'title': video_title, | |
828 | 'description': description, | |
829 | 'upload_date': upload_date, | |
830 | 'uploader': uploader, | |
831 | 'channel_id': channel_id, | |
832 | 'channel_url': channel_url, | |
833 | 'duration': duration, | |
834 | 'uploader_url': uploader_url, | |
835 | 'uploader_id': uploader_id, | |
836 | } | |
837 | ||
838 | def _extract_thumbnails(self, video_id): | |
839 | try_all = 'thumbnails' in self._configuration_arg('check_all') | |
840 | thumbnail_base_urls = ['http://{server}/vi{webp}/{video_id}'.format( | |
841 | webp='_webp' if ext == 'webp' else '', video_id=video_id, server=server) | |
842 | for server in (self._YT_ALL_THUMB_SERVERS if try_all else self._YT_DEFAULT_THUMB_SERVERS) for ext in (('jpg', 'webp') if try_all else ('jpg',))] | |
843 | ||
844 | thumbnails = [] | |
845 | for url in thumbnail_base_urls: | |
846 | response = self._call_cdx_api( | |
847 | video_id, url, filters=['mimetype:image/(?:webp|jpeg)'], | |
848 | collapse=['urlkey'], query={'matchType': 'prefix'}) | |
849 | if not response: | |
850 | continue | |
851 | thumbnails.extend( | |
852 | { | |
853 | 'url': (self._WAYBACK_BASE_URL % (int_or_none(thumbnail_dict.get('timestamp')) or self._OLDEST_CAPTURE_DATE)) + thumbnail_dict.get('original'), | |
854 | 'filesize': int_or_none(thumbnail_dict.get('length')), | |
855 | 'preference': int_or_none(thumbnail_dict.get('length')) | |
856 | } for thumbnail_dict in response) | |
857 | if not try_all: | |
858 | break | |
859 | ||
860 | self._remove_duplicate_formats(thumbnails) | |
861 | return thumbnails | |
862 | ||
863 | def _get_capture_dates(self, video_id, url_date): | |
864 | capture_dates = [] | |
865 | # Note: CDX API will not find watch pages with extra params in the url. | |
866 | response = self._call_cdx_api( | |
867 | video_id, f'https://www.youtube.com/watch?v={video_id}', | |
868 | filters=['mimetype:text/html'], collapse=['timestamp:6', 'digest'], query={'matchType': 'prefix'}) or [] | |
869 | all_captures = sorted(int_or_none(r['timestamp']) for r in response if int_or_none(r['timestamp']) is not None) | |
870 | ||
871 | # Prefer the new polymer UI captures as we support extracting more metadata from them | |
872 | # WBM captures seem to all switch to this layout ~July 2020 | |
873 | modern_captures = [x for x in all_captures if x >= 20200701000000] | |
874 | if modern_captures: | |
875 | capture_dates.append(modern_captures[0]) | |
876 | capture_dates.append(url_date) | |
877 | if all_captures: | |
878 | capture_dates.append(all_captures[0]) | |
879 | ||
880 | if 'captures' in self._configuration_arg('check_all'): | |
881 | capture_dates.extend(modern_captures + all_captures) | |
882 | ||
883 | # Fallbacks if any of the above fail | |
884 | capture_dates.extend([self._OLDEST_CAPTURE_DATE, self._NEWEST_CAPTURE_DATE]) | |
885 | return orderedSet(filter(None, capture_dates)) | |
886 | ||
887 | def _real_extract(self, url): | |
888 | video_id, url_date, url_date_2 = self._match_valid_url(url).group('id', 'date', 'date2') | |
889 | url_date = url_date or url_date_2 | |
890 | ||
891 | urlh = None | |
892 | retry_manager = self.RetryManager(fatal=False) | |
893 | for retry in retry_manager: | |
894 | try: | |
895 | urlh = self._request_webpage( | |
896 | HEADRequest('https://web.archive.org/web/2oe_/http://wayback-fakeurl.archive.org/yt/%s' % video_id), | |
897 | video_id, note='Fetching archived video file url', expected_status=True) | |
898 | except ExtractorError as e: | |
899 | # HTTP Error 404 is expected if the video is not saved. | |
900 | if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404: | |
901 | self.raise_no_formats( | |
902 | 'The requested video is not archived, indexed, or there is an issue with web.archive.org (try again later)', expected=True) | |
903 | else: | |
904 | retry.error = e | |
905 | ||
906 | if retry_manager.error: | |
907 | self.raise_no_formats(retry_manager.error, expected=True, video_id=video_id) | |
908 | ||
909 | capture_dates = self._get_capture_dates(video_id, int_or_none(url_date)) | |
910 | self.write_debug('Captures to try: ' + join_nonempty(*capture_dates, delim=', ')) | |
911 | info = {'id': video_id} | |
912 | for capture in capture_dates: | |
913 | webpage = self._download_webpage( | |
914 | (self._WAYBACK_BASE_URL + 'http://www.youtube.com/watch?v=%s') % (capture, video_id), | |
915 | video_id=video_id, fatal=False, errnote='unable to download capture webpage (it may not be archived)', | |
916 | note='Downloading capture webpage') | |
917 | current_info = self._extract_metadata(video_id, webpage or '') | |
918 | # Try avoid getting deleted video metadata | |
919 | if current_info.get('title'): | |
920 | info = merge_dicts(info, current_info) | |
921 | if 'captures' not in self._configuration_arg('check_all'): | |
922 | break | |
923 | ||
924 | info['thumbnails'] = self._extract_thumbnails(video_id) | |
925 | ||
926 | if urlh: | |
927 | url = compat_urllib_parse_unquote(urlh.geturl()) | |
928 | video_file_url_qs = parse_qs(url) | |
929 | # Attempt to recover any ext & format info from playback url & response headers | |
930 | format = {'url': url, 'filesize': int_or_none(urlh.headers.get('x-archive-orig-content-length'))} | |
931 | itag = try_get(video_file_url_qs, lambda x: x['itag'][0]) | |
932 | if itag and itag in YoutubeIE._formats: | |
933 | format.update(YoutubeIE._formats[itag]) | |
934 | format.update({'format_id': itag}) | |
935 | else: | |
936 | mime = try_get(video_file_url_qs, lambda x: x['mime'][0]) | |
937 | ext = (mimetype2ext(mime) | |
938 | or urlhandle_detect_ext(urlh) | |
939 | or mimetype2ext(urlh.headers.get('x-archive-guessed-content-type'))) | |
940 | format.update({'ext': ext}) | |
941 | info['formats'] = [format] | |
942 | if not info.get('duration'): | |
943 | info['duration'] = str_to_int(try_get(video_file_url_qs, lambda x: x['dur'][0])) | |
944 | ||
945 | if not info.get('title'): | |
946 | info['title'] = video_id | |
947 | return info |