]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
Imported Upstream version 2013.10.01
[youtubedl] / 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 cachedir: Location of the cache files in the filesystem.
85 None to disable filesystem cache.
86
87 The following parameters are not used by YoutubeDL itself, they are used by
88 the FileDownloader:
89 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
90 noresizebuffer, retries, continuedl, noprogress, consoletitle
91 """
92
93 params = None
94 _ies = []
95 _pps = []
96 _download_retcode = None
97 _num_downloads = None
98 _screen_file = None
99
100 def __init__(self, params):
101 """Create a FileDownloader object with the given options."""
102 self._ies = []
103 self._ies_instances = {}
104 self._pps = []
105 self._progress_hooks = []
106 self._download_retcode = 0
107 self._num_downloads = 0
108 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
109
110 if (sys.version_info >= (3,) and sys.platform != 'win32' and
111 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968']
112 and not params['restrictfilenames']):
113 # On Python 3, the Unicode filesystem API will throw errors (#1474)
114 self.report_warning(
115 u'Assuming --restrict-filenames isnce file system encoding '
116 u'cannot encode all charactes. '
117 u'Set the LC_ALL environment variable to fix this.')
118 params['restrictfilenames'] = True
119
120 self.params = params
121 self.fd = FileDownloader(self, self.params)
122
123 if '%(stitle)s' in self.params['outtmpl']:
124 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.')
125
126 def add_info_extractor(self, ie):
127 """Add an InfoExtractor object to the end of the list."""
128 self._ies.append(ie)
129 self._ies_instances[ie.ie_key()] = ie
130 ie.set_downloader(self)
131
132 def get_info_extractor(self, ie_key):
133 """
134 Get an instance of an IE with name ie_key, it will try to get one from
135 the _ies list, if there's no instance it will create a new one and add
136 it to the extractor list.
137 """
138 ie = self._ies_instances.get(ie_key)
139 if ie is None:
140 ie = get_info_extractor(ie_key)()
141 self.add_info_extractor(ie)
142 return ie
143
144 def add_default_info_extractors(self):
145 """
146 Add the InfoExtractors returned by gen_extractors to the end of the list
147 """
148 for ie in gen_extractors():
149 self.add_info_extractor(ie)
150
151 def add_post_processor(self, pp):
152 """Add a PostProcessor object to the end of the chain."""
153 self._pps.append(pp)
154 pp.set_downloader(self)
155
156 def to_screen(self, message, skip_eol=False):
157 """Print message to stdout if not in quiet mode."""
158 if not self.params.get('quiet', False):
159 terminator = [u'\n', u''][skip_eol]
160 output = message + terminator
161 write_string(output, self._screen_file)
162
163 def to_stderr(self, message):
164 """Print message to stderr."""
165 assert type(message) == type(u'')
166 output = message + u'\n'
167 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
168 output = output.encode(preferredencoding())
169 sys.stderr.write(output)
170
171 def fixed_template(self):
172 """Checks if the output template is fixed."""
173 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
174
175 def trouble(self, message=None, tb=None):
176 """Determine action to take when a download problem appears.
177
178 Depending on if the downloader has been configured to ignore
179 download errors or not, this method may throw an exception or
180 not when errors are found, after printing the message.
181
182 tb, if given, is additional traceback information.
183 """
184 if message is not None:
185 self.to_stderr(message)
186 if self.params.get('verbose'):
187 if tb is None:
188 if sys.exc_info()[0]: # if .trouble has been called from an except block
189 tb = u''
190 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
191 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
192 tb += compat_str(traceback.format_exc())
193 else:
194 tb_data = traceback.format_list(traceback.extract_stack())
195 tb = u''.join(tb_data)
196 self.to_stderr(tb)
197 if not self.params.get('ignoreerrors', False):
198 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
199 exc_info = sys.exc_info()[1].exc_info
200 else:
201 exc_info = sys.exc_info()
202 raise DownloadError(message, exc_info)
203 self._download_retcode = 1
204
205 def report_warning(self, message):
206 '''
207 Print the message to stderr, it will be prefixed with 'WARNING:'
208 If stderr is a tty file the 'WARNING:' will be colored
209 '''
210 if sys.stderr.isatty() and os.name != 'nt':
211 _msg_header=u'\033[0;33mWARNING:\033[0m'
212 else:
213 _msg_header=u'WARNING:'
214 warning_message=u'%s %s' % (_msg_header,message)
215 self.to_stderr(warning_message)
216
217 def report_error(self, message, tb=None):
218 '''
219 Do the same as trouble, but prefixes the message with 'ERROR:', colored
220 in red if stderr is a tty file.
221 '''
222 if sys.stderr.isatty() and os.name != 'nt':
223 _msg_header = u'\033[0;31mERROR:\033[0m'
224 else:
225 _msg_header = u'ERROR:'
226 error_message = u'%s %s' % (_msg_header, message)
227 self.trouble(error_message, tb)
228
229 def slow_down(self, start_time, byte_counter):
230 """Sleep if the download speed is over the rate limit."""
231 rate_limit = self.params.get('ratelimit', None)
232 if rate_limit is None or byte_counter == 0:
233 return
234 now = time.time()
235 elapsed = now - start_time
236 if elapsed <= 0.0:
237 return
238 speed = float(byte_counter) / elapsed
239 if speed > rate_limit:
240 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
241
242 def report_writedescription(self, descfn):
243 """ Report that the description file is being written """
244 self.to_screen(u'[info] Writing video description to: ' + descfn)
245
246 def report_writesubtitles(self, sub_filename):
247 """ Report that the subtitles file is being written """
248 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
249
250 def report_writeinfojson(self, infofn):
251 """ Report that the metadata file has been written """
252 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
253
254 def report_file_already_downloaded(self, file_name):
255 """Report file has already been fully downloaded."""
256 try:
257 self.to_screen(u'[download] %s has already been downloaded' % file_name)
258 except (UnicodeEncodeError) as err:
259 self.to_screen(u'[download] The file has already been downloaded')
260
261 def increment_downloads(self):
262 """Increment the ordinal that assigns a number to each file."""
263 self._num_downloads += 1
264
265 def prepare_filename(self, info_dict):
266 """Generate the output filename."""
267 try:
268 template_dict = dict(info_dict)
269
270 template_dict['epoch'] = int(time.time())
271 autonumber_size = self.params.get('autonumber_size')
272 if autonumber_size is None:
273 autonumber_size = 5
274 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
275 template_dict['autonumber'] = autonumber_templ % self._num_downloads
276 if template_dict['playlist_index'] is not None:
277 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
278
279 sanitize = lambda k,v: sanitize_filename(
280 u'NA' if v is None else compat_str(v),
281 restricted=self.params.get('restrictfilenames'),
282 is_id=(k==u'id'))
283 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
284
285 filename = self.params['outtmpl'] % template_dict
286 return filename
287 except KeyError as err:
288 self.report_error(u'Erroneous output template')
289 return None
290 except ValueError as err:
291 self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
292 return None
293
294 def _match_entry(self, info_dict):
295 """ Returns None iff the file should be downloaded """
296
297 title = info_dict['title']
298 matchtitle = self.params.get('matchtitle', False)
299 if matchtitle:
300 if not re.search(matchtitle, title, re.IGNORECASE):
301 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
302 rejecttitle = self.params.get('rejecttitle', False)
303 if rejecttitle:
304 if re.search(rejecttitle, title, re.IGNORECASE):
305 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
306 date = info_dict.get('upload_date', None)
307 if date is not None:
308 dateRange = self.params.get('daterange', DateRange())
309 if date not in dateRange:
310 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
311 return None
312
313 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
314 '''
315 Returns a list with a dictionary for each video we find.
316 If 'download', also downloads the videos.
317 extra_info is a dict containing the extra values to add to each result
318 '''
319
320 if ie_key:
321 ies = [self.get_info_extractor(ie_key)]
322 else:
323 ies = self._ies
324
325 for ie in ies:
326 if not ie.suitable(url):
327 continue
328
329 if not ie.working():
330 self.report_warning(u'The program functionality for this site has been marked as broken, '
331 u'and will probably not work.')
332
333 try:
334 ie_result = ie.extract(url)
335 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
336 break
337 if isinstance(ie_result, list):
338 # Backwards compatibility: old IE result format
339 for result in ie_result:
340 result.update(extra_info)
341 ie_result = {
342 '_type': 'compat_list',
343 'entries': ie_result,
344 }
345 else:
346 ie_result.update(extra_info)
347 if 'extractor' not in ie_result:
348 ie_result['extractor'] = ie.IE_NAME
349 return self.process_ie_result(ie_result, download=download)
350 except ExtractorError as de: # An error we somewhat expected
351 self.report_error(compat_str(de), de.format_traceback())
352 break
353 except Exception as e:
354 if self.params.get('ignoreerrors', False):
355 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
356 break
357 else:
358 raise
359 else:
360 self.report_error(u'no suitable InfoExtractor: %s' % url)
361
362 def process_ie_result(self, ie_result, download=True, extra_info={}):
363 """
364 Take the result of the ie(may be modified) and resolve all unresolved
365 references (URLs, playlist items).
366
367 It will also download the videos if 'download'.
368 Returns the resolved ie_result.
369 """
370
371 result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
372 if result_type == 'video':
373 ie_result.update(extra_info)
374 if 'playlist' not in ie_result:
375 # It isn't part of a playlist
376 ie_result['playlist'] = None
377 ie_result['playlist_index'] = None
378 if download:
379 self.process_info(ie_result)
380 return ie_result
381 elif result_type == 'url':
382 # We have to add extra_info to the results because it may be
383 # contained in a playlist
384 return self.extract_info(ie_result['url'],
385 download,
386 ie_key=ie_result.get('ie_key'),
387 extra_info=extra_info)
388 elif result_type == 'playlist':
389 # We process each entry in the playlist
390 playlist = ie_result.get('title', None) or ie_result.get('id', None)
391 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
392
393 playlist_results = []
394
395 n_all_entries = len(ie_result['entries'])
396 playliststart = self.params.get('playliststart', 1) - 1
397 playlistend = self.params.get('playlistend', -1)
398
399 if playlistend == -1:
400 entries = ie_result['entries'][playliststart:]
401 else:
402 entries = ie_result['entries'][playliststart:playlistend]
403
404 n_entries = len(entries)
405
406 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
407 (ie_result['extractor'], playlist, n_all_entries, n_entries))
408
409 for i,entry in enumerate(entries,1):
410 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
411 extra = {
412 'playlist': playlist,
413 'playlist_index': i + playliststart,
414 }
415 if not 'extractor' in entry:
416 # We set the extractor, if it's an url it will be set then to
417 # the new extractor, but if it's already a video we must make
418 # sure it's present: see issue #877
419 entry['extractor'] = ie_result['extractor']
420 entry_result = self.process_ie_result(entry,
421 download=download,
422 extra_info=extra)
423 playlist_results.append(entry_result)
424 ie_result['entries'] = playlist_results
425 return ie_result
426 elif result_type == 'compat_list':
427 def _fixup(r):
428 r.setdefault('extractor', ie_result['extractor'])
429 return r
430 ie_result['entries'] = [
431 self.process_ie_result(_fixup(r), download=download)
432 for r in ie_result['entries']
433 ]
434 return ie_result
435 else:
436 raise Exception('Invalid result type: %s' % result_type)
437
438 def process_info(self, info_dict):
439 """Process a single resolved IE result."""
440
441 assert info_dict.get('_type', 'video') == 'video'
442 #We increment the download the download count here to match the previous behaviour.
443 self.increment_downloads()
444
445 info_dict['fulltitle'] = info_dict['title']
446 if len(info_dict['title']) > 200:
447 info_dict['title'] = info_dict['title'][:197] + u'...'
448
449 # Keep for backwards compatibility
450 info_dict['stitle'] = info_dict['title']
451
452 if not 'format' in info_dict:
453 info_dict['format'] = info_dict['ext']
454
455 reason = self._match_entry(info_dict)
456 if reason is not None:
457 self.to_screen(u'[download] ' + reason)
458 return
459
460 max_downloads = self.params.get('max_downloads')
461 if max_downloads is not None:
462 if self._num_downloads > int(max_downloads):
463 raise MaxDownloadsReached()
464
465 filename = self.prepare_filename(info_dict)
466
467 # Forced printings
468 if self.params.get('forcetitle', False):
469 compat_print(info_dict['title'])
470 if self.params.get('forceid', False):
471 compat_print(info_dict['id'])
472 if self.params.get('forceurl', False):
473 # For RTMP URLs, also include the playpath
474 compat_print(info_dict['url'] + info_dict.get('play_path', u''))
475 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
476 compat_print(info_dict['thumbnail'])
477 if self.params.get('forcedescription', False) and 'description' in info_dict:
478 compat_print(info_dict['description'])
479 if self.params.get('forcefilename', False) and filename is not None:
480 compat_print(filename)
481 if self.params.get('forceformat', False):
482 compat_print(info_dict['format'])
483
484 # Do nothing else if in simulate mode
485 if self.params.get('simulate', False):
486 return
487
488 if filename is None:
489 return
490
491 try:
492 dn = os.path.dirname(encodeFilename(filename))
493 if dn != '' and not os.path.exists(dn):
494 os.makedirs(dn)
495 except (OSError, IOError) as err:
496 self.report_error(u'unable to create directory ' + compat_str(err))
497 return
498
499 if self.params.get('writedescription', False):
500 try:
501 descfn = filename + u'.description'
502 self.report_writedescription(descfn)
503 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
504 descfile.write(info_dict['description'])
505 except (KeyError, TypeError):
506 self.report_warning(u'There\'s no description to write.')
507 except (OSError, IOError):
508 self.report_error(u'Cannot write description file ' + descfn)
509 return
510
511 subtitles_are_requested = any([self.params.get('writesubtitles', False),
512 self.params.get('writeautomaticsub')])
513
514 if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
515 # subtitles download errors are already managed as troubles in relevant IE
516 # that way it will silently go on when used with unsupporting IE
517 subtitles = info_dict['subtitles']
518 sub_format = self.params.get('subtitlesformat')
519 for sub_lang in subtitles.keys():
520 sub = subtitles[sub_lang]
521 if sub is None:
522 continue
523 try:
524 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
525 self.report_writesubtitles(sub_filename)
526 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
527 subfile.write(sub)
528 except (OSError, IOError):
529 self.report_error(u'Cannot write subtitles file ' + descfn)
530 return
531
532 if self.params.get('writeinfojson', False):
533 infofn = filename + u'.info.json'
534 self.report_writeinfojson(infofn)
535 try:
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)
540 return
541
542 if self.params.get('writethumbnail', False):
543 if info_dict.get('thumbnail') is not None:
544 thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
545 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
546 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
547 (info_dict['extractor'], info_dict['id']))
548 try:
549 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
550 with open(thumb_filename, 'wb') as thumbf:
551 shutil.copyfileobj(uf, thumbf)
552 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
553 (info_dict['extractor'], info_dict['id'], thumb_filename))
554 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
555 self.report_warning(u'Unable to download thumbnail "%s": %s' %
556 (info_dict['thumbnail'], compat_str(err)))
557
558 if not self.params.get('skip_download', False):
559 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
560 success = True
561 else:
562 try:
563 success = self.fd._do_download(filename, info_dict)
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))
566 return
567 except (OSError, IOError) as err:
568 raise UnavailableVideoError(err)
569 except (ContentTooShortError, ) as err:
570 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
571 return
572
573 if success:
574 try:
575 self.post_process(filename, info_dict)
576 except (PostProcessingError) as err:
577 self.report_error(u'postprocessing: %s' % str(err))
578 return
579
580 def download(self, url_list):
581 """Download a given list of URLs."""
582 if len(url_list) > 1 and self.fixed_template():
583 raise SameFileError(self.params['outtmpl'])
584
585 for url in url_list:
586 try:
587 #It also downloads the videos
588 videos = self.extract_info(url)
589 except UnavailableVideoError:
590 self.report_error(u'unable to download video')
591 except MaxDownloadsReached:
592 self.to_screen(u'[info] Maximum number of downloaded files reached.')
593 raise
594
595 return self._download_retcode
596
597 def post_process(self, filename, ie_info):
598 """Run all the postprocessors on the given file."""
599 info = dict(ie_info)
600 info['filepath'] = filename
601 keep_video = None
602 for pp in self._pps:
603 try:
604 keep_video_wish,new_info = pp.run(info)
605 if keep_video_wish is not None:
606 if keep_video_wish:
607 keep_video = keep_video_wish
608 elif keep_video is None:
609 # No clear decision yet, let IE decide
610 keep_video = keep_video_wish
611 except PostProcessingError as e:
612 self.report_error(e.msg)
613 if keep_video is False and not self.params.get('keepvideo', False):
614 try:
615 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
616 os.remove(encodeFilename(filename))
617 except (IOError, OSError):
618 self.report_warning(u'Unable to remove downloaded video file')