]> jfr.im git - yt-dlp.git/blob - youtube_dl/YoutubeDL.py
Clarify a couple of calls
[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
108 if (sys.version_info >= (3,) and sys.platform != 'win32' and
109 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
110 and not params['restrictfilenames']):
111 # On Python 3, the Unicode filesystem API will throw errors (#1474)
112 self.report_warning(
113 u'Assuming --restrict-filenames isnce file system encoding '
114 u'cannot encode all charactes. '
115 u'Set the LC_ALL environment variable to fix this.')
116 params['restrictfilenames'] = True
117
118 self.params = params
119 self.fd = FileDownloader(self, self.params)
120
121 if '%(stitle)s' in self.params['outtmpl']:
122 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.')
123
124 def add_info_extractor(self, ie):
125 """Add an InfoExtractor object to the end of the list."""
126 self._ies.append(ie)
127 self._ies_instances[ie.ie_key()] = ie
128 ie.set_downloader(self)
129
130 def get_info_extractor(self, ie_key):
131 """
132 Get an instance of an IE with name ie_key, it will try to get one from
133 the _ies list, if there's no instance it will create a new one and add
134 it to the extractor list.
135 """
136 ie = self._ies_instances.get(ie_key)
137 if ie is None:
138 ie = get_info_extractor(ie_key)()
139 self.add_info_extractor(ie)
140 return ie
141
142 def add_default_info_extractors(self):
143 """
144 Add the InfoExtractors returned by gen_extractors to the end of the list
145 """
146 for ie in gen_extractors():
147 self.add_info_extractor(ie)
148
149 def add_post_processor(self, pp):
150 """Add a PostProcessor object to the end of the chain."""
151 self._pps.append(pp)
152 pp.set_downloader(self)
153
154 def to_screen(self, message, skip_eol=False):
155 """Print message to stdout if not in quiet mode."""
156 if not self.params.get('quiet', False):
157 terminator = [u'\n', u''][skip_eol]
158 output = message + terminator
159 write_string(output, self._screen_file)
160
161 def to_stderr(self, message):
162 """Print message to stderr."""
163 assert type(message) == type(u'')
164 output = message + u'\n'
165 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
166 output = output.encode(preferredencoding())
167 sys.stderr.write(output)
168
169 def fixed_template(self):
170 """Checks if the output template is fixed."""
171 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
172
173 def trouble(self, message=None, tb=None):
174 """Determine action to take when a download problem appears.
175
176 Depending on if the downloader has been configured to ignore
177 download errors or not, this method may throw an exception or
178 not when errors are found, after printing the message.
179
180 tb, if given, is additional traceback information.
181 """
182 if message is not None:
183 self.to_stderr(message)
184 if self.params.get('verbose'):
185 if tb is None:
186 if sys.exc_info()[0]: # if .trouble has been called from an except block
187 tb = u''
188 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
189 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
190 tb += compat_str(traceback.format_exc())
191 else:
192 tb_data = traceback.format_list(traceback.extract_stack())
193 tb = u''.join(tb_data)
194 self.to_stderr(tb)
195 if not self.params.get('ignoreerrors', False):
196 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
197 exc_info = sys.exc_info()[1].exc_info
198 else:
199 exc_info = sys.exc_info()
200 raise DownloadError(message, exc_info)
201 self._download_retcode = 1
202
203 def report_warning(self, message):
204 '''
205 Print the message to stderr, it will be prefixed with 'WARNING:'
206 If stderr is a tty file the 'WARNING:' will be colored
207 '''
208 if sys.stderr.isatty() and os.name != 'nt':
209 _msg_header=u'\033[0;33mWARNING:\033[0m'
210 else:
211 _msg_header=u'WARNING:'
212 warning_message=u'%s %s' % (_msg_header,message)
213 self.to_stderr(warning_message)
214
215 def report_error(self, message, tb=None):
216 '''
217 Do the same as trouble, but prefixes the message with 'ERROR:', colored
218 in red if stderr is a tty file.
219 '''
220 if sys.stderr.isatty() and os.name != 'nt':
221 _msg_header = u'\033[0;31mERROR:\033[0m'
222 else:
223 _msg_header = u'ERROR:'
224 error_message = u'%s %s' % (_msg_header, message)
225 self.trouble(error_message, tb)
226
227 def slow_down(self, start_time, byte_counter):
228 """Sleep if the download speed is over the rate limit."""
229 rate_limit = self.params.get('ratelimit', None)
230 if rate_limit is None or byte_counter == 0:
231 return
232 now = time.time()
233 elapsed = now - start_time
234 if elapsed <= 0.0:
235 return
236 speed = float(byte_counter) / elapsed
237 if speed > rate_limit:
238 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
239
240 def report_writedescription(self, descfn):
241 """ Report that the description file is being written """
242 self.to_screen(u'[info] Writing video description to: ' + descfn)
243
244 def report_writesubtitles(self, sub_filename):
245 """ Report that the subtitles file is being written """
246 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
247
248 def report_writeinfojson(self, infofn):
249 """ Report that the metadata file has been written """
250 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
251
252 def report_file_already_downloaded(self, file_name):
253 """Report file has already been fully downloaded."""
254 try:
255 self.to_screen(u'[download] %s has already been downloaded' % file_name)
256 except (UnicodeEncodeError) as err:
257 self.to_screen(u'[download] The file has already been downloaded')
258
259 def increment_downloads(self):
260 """Increment the ordinal that assigns a number to each file."""
261 self._num_downloads += 1
262
263 def prepare_filename(self, info_dict):
264 """Generate the output filename."""
265 try:
266 template_dict = dict(info_dict)
267
268 template_dict['epoch'] = int(time.time())
269 autonumber_size = self.params.get('autonumber_size')
270 if autonumber_size is None:
271 autonumber_size = 5
272 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
273 template_dict['autonumber'] = autonumber_templ % self._num_downloads
274 if template_dict['playlist_index'] is not None:
275 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
276
277 sanitize = lambda k,v: sanitize_filename(
278 u'NA' if v is None else compat_str(v),
279 restricted=self.params.get('restrictfilenames'),
280 is_id=(k==u'id'))
281 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
282
283 filename = self.params['outtmpl'] % template_dict
284 return filename
285 except KeyError as err:
286 self.report_error(u'Erroneous output template')
287 return None
288 except ValueError as err:
289 self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
290 return None
291
292 def _match_entry(self, info_dict):
293 """ Returns None iff the file should be downloaded """
294
295 title = info_dict['title']
296 matchtitle = self.params.get('matchtitle', False)
297 if matchtitle:
298 if not re.search(matchtitle, title, re.IGNORECASE):
299 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
300 rejecttitle = self.params.get('rejecttitle', False)
301 if rejecttitle:
302 if re.search(rejecttitle, title, re.IGNORECASE):
303 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
304 date = info_dict.get('upload_date', None)
305 if date is not None:
306 dateRange = self.params.get('daterange', DateRange())
307 if date not in dateRange:
308 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
309 return None
310
311 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
312 '''
313 Returns a list with a dictionary for each video we find.
314 If 'download', also downloads the videos.
315 extra_info is a dict containing the extra values to add to each result
316 '''
317
318 if ie_key:
319 ies = [self.get_info_extractor(ie_key)]
320 else:
321 ies = self._ies
322
323 for ie in ies:
324 if not ie.suitable(url):
325 continue
326
327 if not ie.working():
328 self.report_warning(u'The program functionality for this site has been marked as broken, '
329 u'and will probably not work.')
330
331 try:
332 ie_result = ie.extract(url)
333 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
334 break
335 if isinstance(ie_result, list):
336 # Backwards compatibility: old IE result format
337 for result in ie_result:
338 result.update(extra_info)
339 ie_result = {
340 '_type': 'compat_list',
341 'entries': ie_result,
342 }
343 else:
344 ie_result.update(extra_info)
345 if 'extractor' not in ie_result:
346 ie_result['extractor'] = ie.IE_NAME
347 return self.process_ie_result(ie_result, download=download)
348 except ExtractorError as de: # An error we somewhat expected
349 self.report_error(compat_str(de), de.format_traceback())
350 break
351 except Exception as e:
352 if self.params.get('ignoreerrors', False):
353 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
354 break
355 else:
356 raise
357 else:
358 self.report_error(u'no suitable InfoExtractor: %s' % url)
359
360 def process_ie_result(self, ie_result, download=True, extra_info={}):
361 """
362 Take the result of the ie(may be modified) and resolve all unresolved
363 references (URLs, playlist items).
364
365 It will also download the videos if 'download'.
366 Returns the resolved ie_result.
367 """
368
369 result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
370 if result_type == 'video':
371 ie_result.update(extra_info)
372 if 'playlist' not in ie_result:
373 # It isn't part of a playlist
374 ie_result['playlist'] = None
375 ie_result['playlist_index'] = None
376 if download:
377 self.process_info(ie_result)
378 return ie_result
379 elif result_type == 'url':
380 # We have to add extra_info to the results because it may be
381 # contained in a playlist
382 return self.extract_info(ie_result['url'],
383 download,
384 ie_key=ie_result.get('ie_key'),
385 extra_info=extra_info)
386 elif result_type == 'playlist':
387 # We process each entry in the playlist
388 playlist = ie_result.get('title', None) or ie_result.get('id', None)
389 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
390
391 playlist_results = []
392
393 n_all_entries = len(ie_result['entries'])
394 playliststart = self.params.get('playliststart', 1) - 1
395 playlistend = self.params.get('playlistend', -1)
396
397 if playlistend == -1:
398 entries = ie_result['entries'][playliststart:]
399 else:
400 entries = ie_result['entries'][playliststart:playlistend]
401
402 n_entries = len(entries)
403
404 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
405 (ie_result['extractor'], playlist, n_all_entries, n_entries))
406
407 for i,entry in enumerate(entries,1):
408 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
409 extra = {
410 'playlist': playlist,
411 'playlist_index': i + playliststart,
412 }
413 if not 'extractor' in entry:
414 # We set the extractor, if it's an url it will be set then to
415 # the new extractor, but if it's already a video we must make
416 # sure it's present: see issue #877
417 entry['extractor'] = ie_result['extractor']
418 entry_result = self.process_ie_result(entry,
419 download=download,
420 extra_info=extra)
421 playlist_results.append(entry_result)
422 ie_result['entries'] = playlist_results
423 return ie_result
424 elif result_type == 'compat_list':
425 def _fixup(r):
426 r.setdefault('extractor', ie_result['extractor'])
427 return r
428 ie_result['entries'] = [
429 self.process_ie_result(_fixup(r), download=download)
430 for r in ie_result['entries']
431 ]
432 return ie_result
433 else:
434 raise Exception('Invalid result type: %s' % result_type)
435
436 def process_info(self, info_dict):
437 """Process a single resolved IE result."""
438
439 assert info_dict.get('_type', 'video') == 'video'
440 #We increment the download the download count here to match the previous behaviour.
441 self.increment_downloads()
442
443 info_dict['fulltitle'] = info_dict['title']
444 if len(info_dict['title']) > 200:
445 info_dict['title'] = info_dict['title'][:197] + u'...'
446
447 # Keep for backwards compatibility
448 info_dict['stitle'] = info_dict['title']
449
450 if not 'format' in info_dict:
451 info_dict['format'] = info_dict['ext']
452
453 reason = self._match_entry(info_dict)
454 if reason is not None:
455 self.to_screen(u'[download] ' + reason)
456 return
457
458 max_downloads = self.params.get('max_downloads')
459 if max_downloads is not None:
460 if self._num_downloads > int(max_downloads):
461 raise MaxDownloadsReached()
462
463 filename = self.prepare_filename(info_dict)
464
465 # Forced printings
466 if self.params.get('forcetitle', False):
467 compat_print(info_dict['title'])
468 if self.params.get('forceid', False):
469 compat_print(info_dict['id'])
470 if self.params.get('forceurl', False):
471 # For RTMP URLs, also include the playpath
472 compat_print(info_dict['url'] + info_dict.get('play_path', u''))
473 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
474 compat_print(info_dict['thumbnail'])
475 if self.params.get('forcedescription', False) and 'description' in info_dict:
476 compat_print(info_dict['description'])
477 if self.params.get('forcefilename', False) and filename is not None:
478 compat_print(filename)
479 if self.params.get('forceformat', False):
480 compat_print(info_dict['format'])
481
482 # Do nothing else if in simulate mode
483 if self.params.get('simulate', False):
484 return
485
486 if filename is None:
487 return
488
489 try:
490 dn = os.path.dirname(encodeFilename(filename))
491 if dn != '' and not os.path.exists(dn):
492 os.makedirs(dn)
493 except (OSError, IOError) as err:
494 self.report_error(u'unable to create directory ' + compat_str(err))
495 return
496
497 if self.params.get('writedescription', False):
498 try:
499 descfn = filename + u'.description'
500 self.report_writedescription(descfn)
501 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
502 descfile.write(info_dict['description'])
503 except (KeyError, TypeError):
504 self.report_warning(u'There\'s no description to write.')
505 except (OSError, IOError):
506 self.report_error(u'Cannot write description file ' + descfn)
507 return
508
509 subtitles_are_requested = any([self.params.get('writesubtitles', False),
510 self.params.get('writeautomaticsub')])
511
512 if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
513 # subtitles download errors are already managed as troubles in relevant IE
514 # that way it will silently go on when used with unsupporting IE
515 subtitles = info_dict['subtitles']
516 sub_format = self.params.get('subtitlesformat')
517 for sub_lang in subtitles.keys():
518 sub = subtitles[sub_lang]
519 if sub is None:
520 continue
521 try:
522 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
523 self.report_writesubtitles(sub_filename)
524 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
525 subfile.write(sub)
526 except (OSError, IOError):
527 self.report_error(u'Cannot write subtitles file ' + descfn)
528 return
529
530 if self.params.get('writeinfojson', False):
531 infofn = filename + u'.info.json'
532 self.report_writeinfojson(infofn)
533 try:
534 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
535 write_json_file(json_info_dict, encodeFilename(infofn))
536 except (OSError, IOError):
537 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
538 return
539
540 if self.params.get('writethumbnail', False):
541 if info_dict.get('thumbnail') is not None:
542 thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
543 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
544 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
545 (info_dict['extractor'], info_dict['id']))
546 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
547 with open(thumb_filename, 'wb') as thumbf:
548 shutil.copyfileobj(uf, thumbf)
549 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
550 (info_dict['extractor'], info_dict['id'], thumb_filename))
551
552 if not self.params.get('skip_download', False):
553 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
554 success = True
555 else:
556 try:
557 success = self.fd._do_download(filename, info_dict)
558 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
559 self.report_error(u'unable to download video data: %s' % str(err))
560 return
561 except (OSError, IOError) as err:
562 raise UnavailableVideoError(err)
563 except (ContentTooShortError, ) as err:
564 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
565 return
566
567 if success:
568 try:
569 self.post_process(filename, info_dict)
570 except (PostProcessingError) as err:
571 self.report_error(u'postprocessing: %s' % str(err))
572 return
573
574 def download(self, url_list):
575 """Download a given list of URLs."""
576 if len(url_list) > 1 and self.fixed_template():
577 raise SameFileError(self.params['outtmpl'])
578
579 for url in url_list:
580 try:
581 #It also downloads the videos
582 videos = self.extract_info(url)
583 except UnavailableVideoError:
584 self.report_error(u'unable to download video')
585 except MaxDownloadsReached:
586 self.to_screen(u'[info] Maximum number of downloaded files reached.')
587 raise
588
589 return self._download_retcode
590
591 def post_process(self, filename, ie_info):
592 """Run all the postprocessors on the given file."""
593 info = dict(ie_info)
594 info['filepath'] = filename
595 keep_video = None
596 for pp in self._pps:
597 try:
598 keep_video_wish,new_info = pp.run(info)
599 if keep_video_wish is not None:
600 if keep_video_wish:
601 keep_video = keep_video_wish
602 elif keep_video is None:
603 # No clear decision yet, let IE decide
604 keep_video = keep_video_wish
605 except PostProcessingError as e:
606 self.report_error(e.msg)
607 if keep_video is False and not self.params.get('keepvideo', False):
608 try:
609 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
610 os.remove(encodeFilename(filename))
611 except (IOError, OSError):
612 self.report_warning(u'Unable to remove downloaded video file')