]> jfr.im git - yt-dlp.git/blob - youtube_dl/YoutubeDL.py
Merge remote-tracking branch 'origin/master'
[yt-dlp.git] / youtube_dl / YoutubeDL.py
1 #!/usr/bin/env python
2 # -*- coding: utf-8 -*-
3
4 from __future__ import absolute_import
5
6 import io
7 import os
8 import re
9 import shutil
10 import socket
11 import sys
12 import time
13 import traceback
14
15 from .utils import *
16 from .extractor import get_info_extractor, gen_extractors
17 from .FileDownloader import FileDownloader
18
19
20 class YoutubeDL(object):
21 """YoutubeDL class.
22
23 YoutubeDL objects are the ones responsible of downloading the
24 actual video file and writing it to disk if the user has requested
25 it, among some other tasks. In most cases there should be one per
26 program. As, given a video URL, the downloader doesn't know how to
27 extract all the needed information, task that InfoExtractors do, it
28 has to pass the URL to one of them.
29
30 For this, YoutubeDL objects have a method that allows
31 InfoExtractors to be registered in a given order. When it is passed
32 a URL, the YoutubeDL object handles it to the first InfoExtractor it
33 finds that reports being able to handle it. The InfoExtractor extracts
34 all the information about the video or videos the URL refers to, and
35 YoutubeDL process the extracted information, possibly using a File
36 Downloader to download the video.
37
38 YoutubeDL objects accept a lot of parameters. In order not to saturate
39 the object constructor with arguments, it receives a dictionary of
40 options instead. These options are available through the params
41 attribute for the InfoExtractors to use. The YoutubeDL also
42 registers itself as the downloader in charge for the InfoExtractors
43 that are added to it, so this is a "mutual registration".
44
45 Available options:
46
47 username: Username for authentication purposes.
48 password: Password for authentication purposes.
49 videopassword: Password for acces a video.
50 usenetrc: Use netrc for authentication instead.
51 verbose: Print additional info to stdout.
52 quiet: Do not print messages to stdout.
53 forceurl: Force printing final URL.
54 forcetitle: Force printing title.
55 forceid: Force printing ID.
56 forcethumbnail: Force printing thumbnail URL.
57 forcedescription: Force printing description.
58 forcefilename: Force printing final filename.
59 simulate: Do not download the video files.
60 format: Video format code.
61 format_limit: Highest quality format to try.
62 outtmpl: Template for output names.
63 restrictfilenames: Do not allow "&" and spaces in file names
64 ignoreerrors: Do not stop on download errors.
65 nooverwrites: Prevent overwriting files.
66 playliststart: Playlist item to start at.
67 playlistend: Playlist item to end at.
68 matchtitle: Download only matching titles.
69 rejecttitle: Reject downloads for matching titles.
70 logtostderr: Log messages to stderr instead of stdout.
71 writedescription: Write the video description to a .description file
72 writeinfojson: Write the video description to a .info.json file
73 writethumbnail: Write the thumbnail image to a file
74 writesubtitles: Write the video subtitles to a file
75 writeautomaticsub: Write the automatic subtitles to a file
76 allsubtitles: Downloads all the subtitles of the video
77 (requires writesubtitles or writeautomaticsub)
78 listsubtitles: Lists all available subtitles for the video
79 subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt)
80 subtitleslangs: List of languages of the subtitles to download
81 keepvideo: Keep the video file after post-processing
82 daterange: A DateRange object, download only if the upload_date is in the range.
83 skip_download: Skip the actual download of the video file
84
85 The following parameters are not used by YoutubeDL itself, they are used by
86 the FileDownloader:
87 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
88 noresizebuffer, retries, continuedl, noprogress, consoletitle
89 """
90
91 params = None
92 _ies = []
93 _pps = []
94 _download_retcode = None
95 _num_downloads = None
96 _screen_file = None
97
98 def __init__(self, params):
99 """Create a FileDownloader object with the given options."""
100 self._ies = []
101 self._ies_instances = {}
102 self._pps = []
103 self._progress_hooks = []
104 self._download_retcode = 0
105 self._num_downloads = 0
106 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
107 self.params = params
108 self.fd = FileDownloader(self, self.params)
109
110 if '%(stitle)s' in self.params['outtmpl']:
111 self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
112
113 def add_info_extractor(self, ie):
114 """Add an InfoExtractor object to the end of the list."""
115 self._ies.append(ie)
116 self._ies_instances[ie.ie_key()] = ie
117 ie.set_downloader(self)
118
119 def get_info_extractor(self, ie_key):
120 """
121 Get an instance of an IE with name ie_key, it will try to get one from
122 the _ies list, if there's no instance it will create a new one and add
123 it to the extractor list.
124 """
125 ie = self._ies_instances.get(ie_key)
126 if ie is None:
127 ie = get_info_extractor(ie_key)()
128 self.add_info_extractor(ie)
129 return ie
130
131 def add_default_info_extractors(self):
132 """
133 Add the InfoExtractors returned by gen_extractors to the end of the list
134 """
135 for ie in gen_extractors():
136 self.add_info_extractor(ie)
137
138 def add_post_processor(self, pp):
139 """Add a PostProcessor object to the end of the chain."""
140 self._pps.append(pp)
141 pp.set_downloader(self)
142
143 def to_screen(self, message, skip_eol=False):
144 """Print message to stdout if not in quiet mode."""
145 assert type(message) == type(u'')
146 if not self.params.get('quiet', False):
147 terminator = [u'\n', u''][skip_eol]
148 output = message + terminator
149 if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr
150 output = output.encode(preferredencoding(), 'ignore')
151 self._screen_file.write(output)
152 self._screen_file.flush()
153
154 def to_stderr(self, message):
155 """Print message to stderr."""
156 assert type(message) == type(u'')
157 output = message + u'\n'
158 if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr
159 output = output.encode(preferredencoding())
160 sys.stderr.write(output)
161
162 def fixed_template(self):
163 """Checks if the output template is fixed."""
164 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
165
166 def trouble(self, message=None, tb=None):
167 """Determine action to take when a download problem appears.
168
169 Depending on if the downloader has been configured to ignore
170 download errors or not, this method may throw an exception or
171 not when errors are found, after printing the message.
172
173 tb, if given, is additional traceback information.
174 """
175 if message is not None:
176 self.to_stderr(message)
177 if self.params.get('verbose'):
178 if tb is None:
179 if sys.exc_info()[0]: # if .trouble has been called from an except block
180 tb = u''
181 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
182 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
183 tb += compat_str(traceback.format_exc())
184 else:
185 tb_data = traceback.format_list(traceback.extract_stack())
186 tb = u''.join(tb_data)
187 self.to_stderr(tb)
188 if not self.params.get('ignoreerrors', False):
189 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
190 exc_info = sys.exc_info()[1].exc_info
191 else:
192 exc_info = sys.exc_info()
193 raise DownloadError(message, exc_info)
194 self._download_retcode = 1
195
196 def report_warning(self, message):
197 '''
198 Print the message to stderr, it will be prefixed with 'WARNING:'
199 If stderr is a tty file the 'WARNING:' will be colored
200 '''
201 if sys.stderr.isatty() and os.name != 'nt':
202 _msg_header=u'\033[0;33mWARNING:\033[0m'
203 else:
204 _msg_header=u'WARNING:'
205 warning_message=u'%s %s' % (_msg_header,message)
206 self.to_stderr(warning_message)
207
208 def report_error(self, message, tb=None):
209 '''
210 Do the same as trouble, but prefixes the message with 'ERROR:', colored
211 in red if stderr is a tty file.
212 '''
213 if sys.stderr.isatty() and os.name != 'nt':
214 _msg_header = u'\033[0;31mERROR:\033[0m'
215 else:
216 _msg_header = u'ERROR:'
217 error_message = u'%s %s' % (_msg_header, message)
218 self.trouble(error_message, tb)
219
220 def slow_down(self, start_time, byte_counter):
221 """Sleep if the download speed is over the rate limit."""
222 rate_limit = self.params.get('ratelimit', None)
223 if rate_limit is None or byte_counter == 0:
224 return
225 now = time.time()
226 elapsed = now - start_time
227 if elapsed <= 0.0:
228 return
229 speed = float(byte_counter) / elapsed
230 if speed > rate_limit:
231 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
232
233 def report_writedescription(self, descfn):
234 """ Report that the description file is being written """
235 self.to_screen(u'[info] Writing video description to: ' + descfn)
236
237 def report_writesubtitles(self, sub_filename):
238 """ Report that the subtitles file is being written """
239 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
240
241 def report_writeinfojson(self, infofn):
242 """ Report that the metadata file has been written """
243 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
244
245 def report_file_already_downloaded(self, file_name):
246 """Report file has already been fully downloaded."""
247 try:
248 self.to_screen(u'[download] %s has already been downloaded' % file_name)
249 except (UnicodeEncodeError) as err:
250 self.to_screen(u'[download] The file has already been downloaded')
251
252 def increment_downloads(self):
253 """Increment the ordinal that assigns a number to each file."""
254 self._num_downloads += 1
255
256 def prepare_filename(self, info_dict):
257 """Generate the output filename."""
258 try:
259 template_dict = dict(info_dict)
260
261 template_dict['epoch'] = int(time.time())
262 autonumber_size = self.params.get('autonumber_size')
263 if autonumber_size is None:
264 autonumber_size = 5
265 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
266 template_dict['autonumber'] = autonumber_templ % self._num_downloads
267 if template_dict['playlist_index'] is not None:
268 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
269
270 sanitize = lambda k,v: sanitize_filename(
271 u'NA' if v is None else compat_str(v),
272 restricted=self.params.get('restrictfilenames'),
273 is_id=(k==u'id'))
274 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
275
276 filename = self.params['outtmpl'] % template_dict
277 return filename
278 except KeyError as err:
279 self.report_error(u'Erroneous output template')
280 return None
281 except ValueError as err:
282 self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
283 return None
284
285 def _match_entry(self, info_dict):
286 """ Returns None iff the file should be downloaded """
287
288 title = info_dict['title']
289 matchtitle = self.params.get('matchtitle', False)
290 if matchtitle:
291 if not re.search(matchtitle, title, re.IGNORECASE):
292 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
293 rejecttitle = self.params.get('rejecttitle', False)
294 if rejecttitle:
295 if re.search(rejecttitle, title, re.IGNORECASE):
296 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
297 date = info_dict.get('upload_date', None)
298 if date is not None:
299 dateRange = self.params.get('daterange', DateRange())
300 if date not in dateRange:
301 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
302 return None
303
304 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
305 '''
306 Returns a list with a dictionary for each video we find.
307 If 'download', also downloads the videos.
308 extra_info is a dict containing the extra values to add to each result
309 '''
310
311 if ie_key:
312 ies = [self.get_info_extractor(ie_key)]
313 else:
314 ies = self._ies
315
316 for ie in ies:
317 if not ie.suitable(url):
318 continue
319
320 if not ie.working():
321 self.report_warning(u'The program functionality for this site has been marked as broken, '
322 u'and will probably not work.')
323
324 try:
325 ie_result = ie.extract(url)
326 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
327 break
328 if isinstance(ie_result, list):
329 # Backwards compatibility: old IE result format
330 for result in ie_result:
331 result.update(extra_info)
332 ie_result = {
333 '_type': 'compat_list',
334 'entries': ie_result,
335 }
336 else:
337 ie_result.update(extra_info)
338 if 'extractor' not in ie_result:
339 ie_result['extractor'] = ie.IE_NAME
340 return self.process_ie_result(ie_result, download=download)
341 except ExtractorError as de: # An error we somewhat expected
342 self.report_error(compat_str(de), de.format_traceback())
343 break
344 except Exception as e:
345 if self.params.get('ignoreerrors', False):
346 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
347 break
348 else:
349 raise
350 else:
351 self.report_error(u'no suitable InfoExtractor: %s' % url)
352
353 def process_ie_result(self, ie_result, download=True, extra_info={}):
354 """
355 Take the result of the ie(may be modified) and resolve all unresolved
356 references (URLs, playlist items).
357
358 It will also download the videos if 'download'.
359 Returns the resolved ie_result.
360 """
361
362 result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
363 if result_type == 'video':
364 ie_result.update(extra_info)
365 if 'playlist' not in ie_result:
366 # It isn't part of a playlist
367 ie_result['playlist'] = None
368 ie_result['playlist_index'] = None
369 if download:
370 self.process_info(ie_result)
371 return ie_result
372 elif result_type == 'url':
373 # We have to add extra_info to the results because it may be
374 # contained in a playlist
375 return self.extract_info(ie_result['url'],
376 download,
377 ie_key=ie_result.get('ie_key'),
378 extra_info=extra_info)
379 elif result_type == 'playlist':
380 # We process each entry in the playlist
381 playlist = ie_result.get('title', None) or ie_result.get('id', None)
382 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
383
384 playlist_results = []
385
386 n_all_entries = len(ie_result['entries'])
387 playliststart = self.params.get('playliststart', 1) - 1
388 playlistend = self.params.get('playlistend', -1)
389
390 if playlistend == -1:
391 entries = ie_result['entries'][playliststart:]
392 else:
393 entries = ie_result['entries'][playliststart:playlistend]
394
395 n_entries = len(entries)
396
397 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
398 (ie_result['extractor'], playlist, n_all_entries, n_entries))
399
400 for i,entry in enumerate(entries,1):
401 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
402 extra = {
403 'playlist': playlist,
404 'playlist_index': i + playliststart,
405 }
406 if not 'extractor' in entry:
407 # We set the extractor, if it's an url it will be set then to
408 # the new extractor, but if it's already a video we must make
409 # sure it's present: see issue #877
410 entry['extractor'] = ie_result['extractor']
411 entry_result = self.process_ie_result(entry,
412 download=download,
413 extra_info=extra)
414 playlist_results.append(entry_result)
415 ie_result['entries'] = playlist_results
416 return ie_result
417 elif result_type == 'compat_list':
418 def _fixup(r):
419 r.setdefault('extractor', ie_result['extractor'])
420 return r
421 ie_result['entries'] = [
422 self.process_ie_result(_fixup(r), download=download)
423 for r in ie_result['entries']
424 ]
425 return ie_result
426 else:
427 raise Exception('Invalid result type: %s' % result_type)
428
429 def process_info(self, info_dict):
430 """Process a single resolved IE result."""
431
432 assert info_dict.get('_type', 'video') == 'video'
433 #We increment the download the download count here to match the previous behaviour.
434 self.increment_downloads()
435
436 info_dict['fulltitle'] = info_dict['title']
437 if len(info_dict['title']) > 200:
438 info_dict['title'] = info_dict['title'][:197] + u'...'
439
440 # Keep for backwards compatibility
441 info_dict['stitle'] = info_dict['title']
442
443 if not 'format' in info_dict:
444 info_dict['format'] = info_dict['ext']
445
446 reason = self._match_entry(info_dict)
447 if reason is not None:
448 self.to_screen(u'[download] ' + reason)
449 return
450
451 max_downloads = self.params.get('max_downloads')
452 if max_downloads is not None:
453 if self._num_downloads > int(max_downloads):
454 raise MaxDownloadsReached()
455
456 filename = self.prepare_filename(info_dict)
457
458 # Forced printings
459 if self.params.get('forcetitle', False):
460 compat_print(info_dict['title'])
461 if self.params.get('forceid', False):
462 compat_print(info_dict['id'])
463 if self.params.get('forceurl', False):
464 # For RTMP URLs, also include the playpath
465 compat_print(info_dict['url'] + info_dict.get('play_path', u''))
466 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
467 compat_print(info_dict['thumbnail'])
468 if self.params.get('forcedescription', False) and 'description' in info_dict:
469 compat_print(info_dict['description'])
470 if self.params.get('forcefilename', False) and filename is not None:
471 compat_print(filename)
472 if self.params.get('forceformat', False):
473 compat_print(info_dict['format'])
474
475 # Do nothing else if in simulate mode
476 if self.params.get('simulate', False):
477 return
478
479 if filename is None:
480 return
481
482 try:
483 dn = os.path.dirname(encodeFilename(filename))
484 if dn != '' and not os.path.exists(dn):
485 os.makedirs(dn)
486 except (OSError, IOError) as err:
487 self.report_error(u'unable to create directory ' + compat_str(err))
488 return
489
490 if self.params.get('writedescription', False):
491 try:
492 descfn = filename + u'.description'
493 self.report_writedescription(descfn)
494 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
495 descfile.write(info_dict['description'])
496 except (KeyError, TypeError):
497 self.report_warning(u'There\'s no description to write.')
498 except (OSError, IOError):
499 self.report_error(u'Cannot write description file ' + descfn)
500 return
501
502 subtitles_are_requested = any([self.params.get('writesubtitles', False),
503 self.params.get('writeautomaticsub')])
504
505 if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
506 # subtitles download errors are already managed as troubles in relevant IE
507 # that way it will silently go on when used with unsupporting IE
508 subtitles = info_dict['subtitles']
509 sub_format = self.params.get('subtitlesformat')
510 for sub_lang in subtitles.keys():
511 sub = subtitles[sub_lang]
512 if sub is None:
513 continue
514 try:
515 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
516 self.report_writesubtitles(sub_filename)
517 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
518 subfile.write(sub)
519 except (OSError, IOError):
520 self.report_error(u'Cannot write subtitles file ' + descfn)
521 return
522
523 if self.params.get('writeinfojson', False):
524 infofn = filename + u'.info.json'
525 self.report_writeinfojson(infofn)
526 try:
527 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
528 write_json_file(json_info_dict, encodeFilename(infofn))
529 except (OSError, IOError):
530 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
531 return
532
533 if self.params.get('writethumbnail', False):
534 if info_dict.get('thumbnail') is not None:
535 thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
536 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
537 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
538 (info_dict['extractor'], info_dict['id']))
539 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
540 with open(thumb_filename, 'wb') as thumbf:
541 shutil.copyfileobj(uf, thumbf)
542 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
543 (info_dict['extractor'], info_dict['id'], thumb_filename))
544
545 if not self.params.get('skip_download', False):
546 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
547 success = True
548 else:
549 try:
550 success = self.fd._do_download(filename, info_dict)
551 except (OSError, IOError) as err:
552 raise UnavailableVideoError(err)
553 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
554 self.report_error(u'unable to download video data: %s' % str(err))
555 return
556 except (ContentTooShortError, ) as err:
557 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
558 return
559
560 if success:
561 try:
562 self.post_process(filename, info_dict)
563 except (PostProcessingError) as err:
564 self.report_error(u'postprocessing: %s' % str(err))
565 return
566
567 def download(self, url_list):
568 """Download a given list of URLs."""
569 if len(url_list) > 1 and self.fixed_template():
570 raise SameFileError(self.params['outtmpl'])
571
572 for url in url_list:
573 try:
574 #It also downloads the videos
575 videos = self.extract_info(url)
576 except UnavailableVideoError:
577 self.report_error(u'unable to download video')
578 except MaxDownloadsReached:
579 self.to_screen(u'[info] Maximum number of downloaded files reached.')
580 raise
581
582 return self._download_retcode
583
584 def post_process(self, filename, ie_info):
585 """Run all the postprocessors on the given file."""
586 info = dict(ie_info)
587 info['filepath'] = filename
588 keep_video = None
589 for pp in self._pps:
590 try:
591 keep_video_wish,new_info = pp.run(info)
592 if keep_video_wish is not None:
593 if keep_video_wish:
594 keep_video = keep_video_wish
595 elif keep_video is None:
596 # No clear decision yet, let IE decide
597 keep_video = keep_video_wish
598 except PostProcessingError as e:
599 self.report_error(e.msg)
600 if keep_video is False and not self.params.get('keepvideo', False):
601 try:
602 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
603 os.remove(encodeFilename(filename))
604 except (IOError, OSError):
605 self.report_warning(u'Unable to remove downloaded video file')