2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import
16 from .extractor import get_info_extractor, gen_extractors
17 from .FileDownloader import FileDownloader
20 class YoutubeDL(object):
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.
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.
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".
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 listsubtitles: Lists all available subtitles for the video
78 subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt)
79 subtitleslang: Language of the subtitles to download
80 keepvideo: Keep the video file after post-processing
81 daterange: A DateRange object, download only if the upload_date is in the range.
82 skip_download: Skip the actual download of the video file
84 The following parameters are not used by YoutubeDL itself, they are used by
86 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
87 noresizebuffer, retries, continuedl, noprogress, consoletitle
93 _download_retcode = None
97 def __init__(self, params):
98 """Create a FileDownloader object with the given options."""
100 self._ies_instances = {}
102 self._progress_hooks = []
103 self._download_retcode = 0
104 self._num_downloads = 0
105 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
107 self.fd = FileDownloader(self, self.params)
109 if '%(stitle)s' in self.params['outtmpl']:
110 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 def add_info_extractor(self, ie):
113 """Add an InfoExtractor object to the end of the list."""
115 self._ies_instances[ie.ie_key()] = ie
116 ie.set_downloader(self)
118 def get_info_extractor(self, ie_key):
120 Get an instance of an IE with name ie_key, it will try to get one from
121 the _ies list, if there's no instance it will create a new one and add
122 it to the extractor list.
124 ie = self._ies_instances.get(ie_key)
126 ie = get_info_extractor(ie_key)()
127 self.add_info_extractor(ie)
130 def add_default_info_extractors(self):
132 Add the InfoExtractors returned by gen_extractors to the end of the list
134 for ie in gen_extractors():
135 self.add_info_extractor(ie)
137 def add_post_processor(self, pp):
138 """Add a PostProcessor object to the end of the chain."""
140 pp.set_downloader(self)
142 def to_screen(self, message, skip_eol=False):
143 """Print message to stdout if not in quiet mode."""
144 assert type(message) == type(u'')
145 if not self.params.get('quiet', False):
146 terminator = [u'\n', u''][skip_eol]
147 output = message + terminator
148 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
149 output = output.encode(preferredencoding(), 'ignore')
150 self._screen_file.write(output)
151 self._screen_file.flush()
153 def to_stderr(self, message):
154 """Print message to stderr."""
155 assert type(message) == type(u'')
156 output = message + u'\n'
157 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
158 output = output.encode(preferredencoding())
159 sys.stderr.write(output)
161 def fixed_template(self):
162 """Checks if the output template is fixed."""
163 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
165 def trouble(self, message=None, tb=None):
166 """Determine action to take when a download problem appears.
168 Depending on if the downloader has been configured to ignore
169 download errors or not, this method may throw an exception or
170 not when errors are found, after printing the message.
172 tb, if given, is additional traceback information.
174 if message is not None:
175 self.to_stderr(message)
176 if self.params.get('verbose'):
178 if sys.exc_info()[0]: # if .trouble has been called from an except block
180 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
181 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
182 tb += compat_str(traceback.format_exc())
184 tb_data = traceback.format_list(traceback.extract_stack())
185 tb = u''.join(tb_data)
187 if not self.params.get('ignoreerrors', False):
188 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
189 exc_info = sys.exc_info()[1].exc_info
191 exc_info = sys.exc_info()
192 raise DownloadError(message, exc_info)
193 self._download_retcode = 1
195 def report_warning(self, message):
197 Print the message to stderr, it will be prefixed with 'WARNING:'
198 If stderr is a tty file the 'WARNING:' will be colored
200 if sys.stderr.isatty() and os.name != 'nt':
201 _msg_header=u'\033[0;33mWARNING:\033[0m'
203 _msg_header=u'WARNING:'
204 warning_message=u'%s %s' % (_msg_header,message)
205 self.to_stderr(warning_message)
207 def report_error(self, message, tb=None):
209 Do the same as trouble, but prefixes the message with 'ERROR:', colored
210 in red if stderr is a tty file.
212 if sys.stderr.isatty() and os.name != 'nt':
213 _msg_header = u'\033[0;31mERROR:\033[0m'
215 _msg_header = u'ERROR:'
216 error_message = u'%s %s' % (_msg_header, message)
217 self.trouble(error_message, tb)
219 def slow_down(self, start_time, byte_counter):
220 """Sleep if the download speed is over the rate limit."""
221 rate_limit = self.params.get('ratelimit', None)
222 if rate_limit is None or byte_counter == 0:
225 elapsed = now - start_time
228 speed = float(byte_counter) / elapsed
229 if speed > rate_limit:
230 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
232 def report_writedescription(self, descfn):
233 """ Report that the description file is being written """
234 self.to_screen(u'[info] Writing video description to: ' + descfn)
236 def report_writesubtitles(self, sub_filename):
237 """ Report that the subtitles file is being written """
238 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
240 def report_writeinfojson(self, infofn):
241 """ Report that the metadata file has been written """
242 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
244 def report_file_already_downloaded(self, file_name):
245 """Report file has already been fully downloaded."""
247 self.to_screen(u'[download] %s has already been downloaded' % file_name)
248 except (UnicodeEncodeError) as err:
249 self.to_screen(u'[download] The file has already been downloaded')
251 def increment_downloads(self):
252 """Increment the ordinal that assigns a number to each file."""
253 self._num_downloads += 1
255 def prepare_filename(self, info_dict):
256 """Generate the output filename."""
258 template_dict = dict(info_dict)
260 template_dict['epoch'] = int(time.time())
261 autonumber_size = self.params.get('autonumber_size')
262 if autonumber_size is None:
264 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
265 template_dict['autonumber'] = autonumber_templ % self._num_downloads
266 if template_dict['playlist_index'] is not None:
267 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
269 sanitize = lambda k,v: sanitize_filename(
270 u'NA' if v is None else compat_str(v),
271 restricted=self.params.get('restrictfilenames'),
273 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
275 filename = self.params['outtmpl'] % template_dict
277 except KeyError as err:
278 self.report_error(u'Erroneous output template')
280 except ValueError as err:
281 self.report_error(u'Insufficient system charset ' + repr(preferredencoding()))
284 def _match_entry(self, info_dict):
285 """ Returns None iff the file should be downloaded """
287 title = info_dict['title']
288 matchtitle = self.params.get('matchtitle', False)
290 if not re.search(matchtitle, title, re.IGNORECASE):
291 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
292 rejecttitle = self.params.get('rejecttitle', False)
294 if re.search(rejecttitle, title, re.IGNORECASE):
295 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
296 date = info_dict.get('upload_date', None)
298 dateRange = self.params.get('daterange', DateRange())
299 if date not in dateRange:
300 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
303 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
305 Returns a list with a dictionary for each video we find.
306 If 'download', also downloads the videos.
307 extra_info is a dict containing the extra values to add to each result
311 ies = [self.get_info_extractor(ie_key)]
316 if not ie.suitable(url):
320 self.report_warning(u'The program functionality for this site has been marked as broken, '
321 u'and will probably not work.')
324 ie_result = ie.extract(url)
325 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
327 if isinstance(ie_result, list):
328 # Backwards compatibility: old IE result format
329 for result in ie_result:
330 result.update(extra_info)
332 '_type': 'compat_list',
333 'entries': ie_result,
336 ie_result.update(extra_info)
337 if 'extractor' not in ie_result:
338 ie_result['extractor'] = ie.IE_NAME
339 return self.process_ie_result(ie_result, download=download)
340 except ExtractorError as de: # An error we somewhat expected
341 self.report_error(compat_str(de), de.format_traceback())
343 except Exception as e:
344 if self.params.get('ignoreerrors', False):
345 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
350 self.report_error(u'no suitable InfoExtractor: %s' % url)
352 def process_ie_result(self, ie_result, download=True, extra_info={}):
354 Take the result of the ie(may be modified) and resolve all unresolved
355 references (URLs, playlist items).
357 It will also download the videos if 'download'.
358 Returns the resolved ie_result.
361 result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
362 if result_type == 'video':
363 if 'playlist' not in ie_result:
364 # It isn't part of a playlist
365 ie_result['playlist'] = None
366 ie_result['playlist_index'] = None
368 self.process_info(ie_result)
370 elif result_type == 'url':
371 # We have to add extra_info to the results because it may be
372 # contained in a playlist
373 return self.extract_info(ie_result['url'],
375 ie_key=ie_result.get('ie_key'),
376 extra_info=extra_info)
377 elif result_type == 'playlist':
378 # We process each entry in the playlist
379 playlist = ie_result.get('title', None) or ie_result.get('id', None)
380 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
382 playlist_results = []
384 n_all_entries = len(ie_result['entries'])
385 playliststart = self.params.get('playliststart', 1) - 1
386 playlistend = self.params.get('playlistend', -1)
388 if playlistend == -1:
389 entries = ie_result['entries'][playliststart:]
391 entries = ie_result['entries'][playliststart:playlistend]
393 n_entries = len(entries)
395 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
396 (ie_result['extractor'], playlist, n_all_entries, n_entries))
398 for i,entry in enumerate(entries,1):
399 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
401 'playlist': playlist,
402 'playlist_index': i + playliststart,
404 if not 'extractor' in entry:
405 # We set the extractor, if it's an url it will be set then to
406 # the new extractor, but if it's already a video we must make
407 # sure it's present: see issue #877
408 entry['extractor'] = ie_result['extractor']
409 entry_result = self.process_ie_result(entry,
412 playlist_results.append(entry_result)
413 ie_result['entries'] = playlist_results
415 elif result_type == 'compat_list':
417 r.setdefault('extractor', ie_result['extractor'])
419 ie_result['entries'] = [
420 self.process_ie_result(_fixup(r), download=download)
421 for r in ie_result['entries']
425 raise Exception('Invalid result type: %s' % result_type)
427 def process_info(self, info_dict):
428 """Process a single resolved IE result."""
430 assert info_dict.get('_type', 'video') == 'video'
431 #We increment the download the download count here to match the previous behaviour.
432 self.increment_downloads()
434 info_dict['fulltitle'] = info_dict['title']
435 if len(info_dict['title']) > 200:
436 info_dict['title'] = info_dict['title'][:197] + u'...'
438 # Keep for backwards compatibility
439 info_dict['stitle'] = info_dict['title']
441 if not 'format' in info_dict:
442 info_dict['format'] = info_dict['ext']
444 reason = self._match_entry(info_dict)
445 if reason is not None:
446 self.to_screen(u'[download] ' + reason)
449 max_downloads = self.params.get('max_downloads')
450 if max_downloads is not None:
451 if self._num_downloads > int(max_downloads):
452 raise MaxDownloadsReached()
454 filename = self.prepare_filename(info_dict)
457 if self.params.get('forcetitle', False):
458 compat_print(info_dict['title'])
459 if self.params.get('forceid', False):
460 compat_print(info_dict['id'])
461 if self.params.get('forceurl', False):
462 compat_print(info_dict['url'])
463 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
464 compat_print(info_dict['thumbnail'])
465 if self.params.get('forcedescription', False) and 'description' in info_dict:
466 compat_print(info_dict['description'])
467 if self.params.get('forcefilename', False) and filename is not None:
468 compat_print(filename)
469 if self.params.get('forceformat', False):
470 compat_print(info_dict['format'])
472 # Do nothing else if in simulate mode
473 if self.params.get('simulate', False):
480 dn = os.path.dirname(encodeFilename(filename))
481 if dn != '' and not os.path.exists(dn):
483 except (OSError, IOError) as err:
484 self.report_error(u'unable to create directory ' + compat_str(err))
487 if self.params.get('writedescription', False):
489 descfn = filename + u'.description'
490 self.report_writedescription(descfn)
491 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
492 descfile.write(info_dict['description'])
493 except (OSError, IOError):
494 self.report_error(u'Cannot write description file ' + descfn)
497 if (self.params.get('writesubtitles', False) or self.params.get('writeautomaticsub')) and 'subtitles' in info_dict and info_dict['subtitles']:
498 # subtitles download errors are already managed as troubles in relevant IE
499 # that way it will silently go on when used with unsupporting IE
500 subtitle = info_dict['subtitles'][0]
501 (sub_error, sub_lang, sub) = subtitle
502 sub_format = self.params.get('subtitlesformat')
504 self.report_warning("Some error while getting the subtitles")
507 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
508 self.report_writesubtitles(sub_filename)
509 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
511 except (OSError, IOError):
512 self.report_error(u'Cannot write subtitles file ' + descfn)
515 if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
516 subtitles = info_dict['subtitles']
517 sub_format = self.params.get('subtitlesformat')
518 for subtitle in subtitles:
519 (sub_error, sub_lang, sub) = subtitle
521 self.report_warning("Some error while getting the subtitles")
524 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
525 self.report_writesubtitles(sub_filename)
526 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
528 except (OSError, IOError):
529 self.report_error(u'Cannot write subtitles file ' + descfn)
532 if self.params.get('writeinfojson', False):
533 infofn = filename + u'.info.json'
534 self.report_writeinfojson(infofn)
536 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
537 write_json_file(json_info_dict, encodeFilename(infofn))
538 except (OSError, IOError):
539 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
542 if self.params.get('writethumbnail', False):
543 if 'thumbnail' in info_dict:
544 thumb_format = info_dict['thumbnail'].rpartition(u'/')[2].rpartition(u'.')[2]
547 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
548 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
549 (info_dict['extractor'], info_dict['id']))
550 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
551 with open(thumb_filename, 'wb') as thumbf:
552 shutil.copyfileobj(uf, thumbf)
553 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
554 (info_dict['extractor'], info_dict['id'], thumb_filename))
556 if not self.params.get('skip_download', False):
557 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
561 success = self.fd._do_download(filename, info_dict)
562 except (OSError, IOError) as err:
563 raise UnavailableVideoError()
564 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
565 self.report_error(u'unable to download video data: %s' % str(err))
567 except (ContentTooShortError, ) as err:
568 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
573 self.post_process(filename, info_dict)
574 except (PostProcessingError) as err:
575 self.report_error(u'postprocessing: %s' % str(err))
578 def download(self, url_list):
579 """Download a given list of URLs."""
580 if len(url_list) > 1 and self.fixed_template():
581 raise SameFileError(self.params['outtmpl'])
585 #It also downloads the videos
586 videos = self.extract_info(url)
587 except UnavailableVideoError:
588 self.report_error(u'unable to download video')
589 except MaxDownloadsReached:
590 self.to_screen(u'[info] Maximum number of downloaded files reached.')
593 return self._download_retcode
595 def post_process(self, filename, ie_info):
596 """Run all the postprocessors on the given file."""
598 info['filepath'] = filename
602 keep_video_wish,new_info = pp.run(info)
603 if keep_video_wish is not None:
605 keep_video = keep_video_wish
606 elif keep_video is None:
607 # No clear decision yet, let IE decide
608 keep_video = keep_video_wish
609 except PostProcessingError as e:
610 self.to_stderr(u'ERROR: ' + e.msg)
611 if keep_video is False and not self.params.get('keepvideo', False):
613 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
614 os.remove(encodeFilename(filename))
615 except (IOError, OSError):
616 self.report_warning(u'Unable to remove downloaded video file')