]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/YoutubeDL.py
Imported Upstream version 2013.08.29
[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 listsubtitles: Lists all available subtitles for the video
78 subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt)
79 subtitleslangs: List of languages 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
83
84 The following parameters are not used by YoutubeDL itself, they are used by
85 the FileDownloader:
86 nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test,
87 noresizebuffer, retries, continuedl, noprogress, consoletitle
88 """
89
90 params = None
91 _ies = []
92 _pps = []
93 _download_retcode = None
94 _num_downloads = None
95 _screen_file = None
96
97 def __init__(self, params):
98 """Create a FileDownloader object with the given options."""
99 self._ies = []
100 self._ies_instances = {}
101 self._pps = []
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)]
106 self.params = params
107 self.fd = FileDownloader(self, self.params)
108
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.')
111
112 def add_info_extractor(self, ie):
113 """Add an InfoExtractor object to the end of the list."""
114 self._ies.append(ie)
115 self._ies_instances[ie.ie_key()] = ie
116 ie.set_downloader(self)
117
118 def get_info_extractor(self, ie_key):
119 """
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.
123 """
124 ie = self._ies_instances.get(ie_key)
125 if ie is None:
126 ie = get_info_extractor(ie_key)()
127 self.add_info_extractor(ie)
128 return ie
129
130 def add_default_info_extractors(self):
131 """
132 Add the InfoExtractors returned by gen_extractors to the end of the list
133 """
134 for ie in gen_extractors():
135 self.add_info_extractor(ie)
136
137 def add_post_processor(self, pp):
138 """Add a PostProcessor object to the end of the chain."""
139 self._pps.append(pp)
140 pp.set_downloader(self)
141
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()
152
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)
160
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)
164
165 def trouble(self, message=None, tb=None):
166 """Determine action to take when a download problem appears.
167
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.
171
172 tb, if given, is additional traceback information.
173 """
174 if message is not None:
175 self.to_stderr(message)
176 if self.params.get('verbose'):
177 if tb is None:
178 if sys.exc_info()[0]: # if .trouble has been called from an except block
179 tb = u''
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())
183 else:
184 tb_data = traceback.format_list(traceback.extract_stack())
185 tb = u''.join(tb_data)
186 self.to_stderr(tb)
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
190 else:
191 exc_info = sys.exc_info()
192 raise DownloadError(message, exc_info)
193 self._download_retcode = 1
194
195 def report_warning(self, message):
196 '''
197 Print the message to stderr, it will be prefixed with 'WARNING:'
198 If stderr is a tty file the 'WARNING:' will be colored
199 '''
200 if sys.stderr.isatty() and os.name != 'nt':
201 _msg_header=u'\033[0;33mWARNING:\033[0m'
202 else:
203 _msg_header=u'WARNING:'
204 warning_message=u'%s %s' % (_msg_header,message)
205 self.to_stderr(warning_message)
206
207 def report_error(self, message, tb=None):
208 '''
209 Do the same as trouble, but prefixes the message with 'ERROR:', colored
210 in red if stderr is a tty file.
211 '''
212 if sys.stderr.isatty() and os.name != 'nt':
213 _msg_header = u'\033[0;31mERROR:\033[0m'
214 else:
215 _msg_header = u'ERROR:'
216 error_message = u'%s %s' % (_msg_header, message)
217 self.trouble(error_message, tb)
218
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:
223 return
224 now = time.time()
225 elapsed = now - start_time
226 if elapsed <= 0.0:
227 return
228 speed = float(byte_counter) / elapsed
229 if speed > rate_limit:
230 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
231
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)
235
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)
239
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)
243
244 def report_file_already_downloaded(self, file_name):
245 """Report file has already been fully downloaded."""
246 try:
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')
250
251 def increment_downloads(self):
252 """Increment the ordinal that assigns a number to each file."""
253 self._num_downloads += 1
254
255 def prepare_filename(self, info_dict):
256 """Generate the output filename."""
257 try:
258 template_dict = dict(info_dict)
259
260 template_dict['epoch'] = int(time.time())
261 autonumber_size = self.params.get('autonumber_size')
262 if autonumber_size is None:
263 autonumber_size = 5
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']
268
269 sanitize = lambda k,v: sanitize_filename(
270 u'NA' if v is None else compat_str(v),
271 restricted=self.params.get('restrictfilenames'),
272 is_id=(k==u'id'))
273 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
274
275 filename = self.params['outtmpl'] % template_dict
276 return filename
277 except KeyError as err:
278 self.report_error(u'Erroneous output template')
279 return None
280 except ValueError as err:
281 self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')')
282 return None
283
284 def _match_entry(self, info_dict):
285 """ Returns None iff the file should be downloaded """
286
287 title = info_dict['title']
288 matchtitle = self.params.get('matchtitle', False)
289 if matchtitle:
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)
293 if rejecttitle:
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)
297 if date is not 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)
301 return None
302
303 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
304 '''
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
308 '''
309
310 if ie_key:
311 ies = [self.get_info_extractor(ie_key)]
312 else:
313 ies = self._ies
314
315 for ie in ies:
316 if not ie.suitable(url):
317 continue
318
319 if not ie.working():
320 self.report_warning(u'The program functionality for this site has been marked as broken, '
321 u'and will probably not work.')
322
323 try:
324 ie_result = ie.extract(url)
325 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
326 break
327 if isinstance(ie_result, list):
328 # Backwards compatibility: old IE result format
329 for result in ie_result:
330 result.update(extra_info)
331 ie_result = {
332 '_type': 'compat_list',
333 'entries': ie_result,
334 }
335 else:
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())
342 break
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()))
346 break
347 else:
348 raise
349 else:
350 self.report_error(u'no suitable InfoExtractor: %s' % url)
351
352 def process_ie_result(self, ie_result, download=True, extra_info={}):
353 """
354 Take the result of the ie(may be modified) and resolve all unresolved
355 references (URLs, playlist items).
356
357 It will also download the videos if 'download'.
358 Returns the resolved ie_result.
359 """
360
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 ie_result.update(extra_info)
364 if 'playlist' not in ie_result:
365 # It isn't part of a playlist
366 ie_result['playlist'] = None
367 ie_result['playlist_index'] = None
368 if download:
369 self.process_info(ie_result)
370 return ie_result
371 elif result_type == 'url':
372 # We have to add extra_info to the results because it may be
373 # contained in a playlist
374 return self.extract_info(ie_result['url'],
375 download,
376 ie_key=ie_result.get('ie_key'),
377 extra_info=extra_info)
378 elif result_type == 'playlist':
379 # We process each entry in the playlist
380 playlist = ie_result.get('title', None) or ie_result.get('id', None)
381 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
382
383 playlist_results = []
384
385 n_all_entries = len(ie_result['entries'])
386 playliststart = self.params.get('playliststart', 1) - 1
387 playlistend = self.params.get('playlistend', -1)
388
389 if playlistend == -1:
390 entries = ie_result['entries'][playliststart:]
391 else:
392 entries = ie_result['entries'][playliststart:playlistend]
393
394 n_entries = len(entries)
395
396 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
397 (ie_result['extractor'], playlist, n_all_entries, n_entries))
398
399 for i,entry in enumerate(entries,1):
400 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
401 extra = {
402 'playlist': playlist,
403 'playlist_index': i + playliststart,
404 }
405 if not 'extractor' in entry:
406 # We set the extractor, if it's an url it will be set then to
407 # the new extractor, but if it's already a video we must make
408 # sure it's present: see issue #877
409 entry['extractor'] = ie_result['extractor']
410 entry_result = self.process_ie_result(entry,
411 download=download,
412 extra_info=extra)
413 playlist_results.append(entry_result)
414 ie_result['entries'] = playlist_results
415 return ie_result
416 elif result_type == 'compat_list':
417 def _fixup(r):
418 r.setdefault('extractor', ie_result['extractor'])
419 return r
420 ie_result['entries'] = [
421 self.process_ie_result(_fixup(r), download=download)
422 for r in ie_result['entries']
423 ]
424 return ie_result
425 else:
426 raise Exception('Invalid result type: %s' % result_type)
427
428 def process_info(self, info_dict):
429 """Process a single resolved IE result."""
430
431 assert info_dict.get('_type', 'video') == 'video'
432 #We increment the download the download count here to match the previous behaviour.
433 self.increment_downloads()
434
435 info_dict['fulltitle'] = info_dict['title']
436 if len(info_dict['title']) > 200:
437 info_dict['title'] = info_dict['title'][:197] + u'...'
438
439 # Keep for backwards compatibility
440 info_dict['stitle'] = info_dict['title']
441
442 if not 'format' in info_dict:
443 info_dict['format'] = info_dict['ext']
444
445 reason = self._match_entry(info_dict)
446 if reason is not None:
447 self.to_screen(u'[download] ' + reason)
448 return
449
450 max_downloads = self.params.get('max_downloads')
451 if max_downloads is not None:
452 if self._num_downloads > int(max_downloads):
453 raise MaxDownloadsReached()
454
455 filename = self.prepare_filename(info_dict)
456
457 # Forced printings
458 if self.params.get('forcetitle', False):
459 compat_print(info_dict['title'])
460 if self.params.get('forceid', False):
461 compat_print(info_dict['id'])
462 if self.params.get('forceurl', False):
463 # For RTMP URLs, also include the playpath
464 compat_print(info_dict['url'] + info_dict.get('play_path', u''))
465 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
466 compat_print(info_dict['thumbnail'])
467 if self.params.get('forcedescription', False) and 'description' in info_dict:
468 compat_print(info_dict['description'])
469 if self.params.get('forcefilename', False) and filename is not None:
470 compat_print(filename)
471 if self.params.get('forceformat', False):
472 compat_print(info_dict['format'])
473
474 # Do nothing else if in simulate mode
475 if self.params.get('simulate', False):
476 return
477
478 if filename is None:
479 return
480
481 try:
482 dn = os.path.dirname(encodeFilename(filename))
483 if dn != '' and not os.path.exists(dn):
484 os.makedirs(dn)
485 except (OSError, IOError) as err:
486 self.report_error(u'unable to create directory ' + compat_str(err))
487 return
488
489 if self.params.get('writedescription', False):
490 try:
491 descfn = filename + u'.description'
492 self.report_writedescription(descfn)
493 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
494 descfile.write(info_dict['description'])
495 except (OSError, IOError):
496 self.report_error(u'Cannot write description file ' + descfn)
497 return
498
499 subtitles_are_requested = any([self.params.get('writesubtitles', False),
500 self.params.get('writeautomaticsub'),
501 self.params.get('allsubtitles', False)])
502
503 if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:
504 # subtitles download errors are already managed as troubles in relevant IE
505 # that way it will silently go on when used with unsupporting IE
506 subtitles = info_dict['subtitles']
507 sub_format = self.params.get('subtitlesformat')
508 for sub_lang in subtitles.keys():
509 sub = subtitles[sub_lang]
510 if sub is None:
511 continue
512 try:
513 sub_filename = subtitles_filename(filename, sub_lang, sub_format)
514 self.report_writesubtitles(sub_filename)
515 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
516 subfile.write(sub)
517 except (OSError, IOError):
518 self.report_error(u'Cannot write subtitles file ' + descfn)
519 return
520
521 if self.params.get('writeinfojson', False):
522 infofn = filename + u'.info.json'
523 self.report_writeinfojson(infofn)
524 try:
525 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
526 write_json_file(json_info_dict, encodeFilename(infofn))
527 except (OSError, IOError):
528 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
529 return
530
531 if self.params.get('writethumbnail', False):
532 if info_dict.get('thumbnail') is not None:
533 thumb_format = determine_ext(info_dict['thumbnail'], u'jpg')
534 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
535 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
536 (info_dict['extractor'], info_dict['id']))
537 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
538 with open(thumb_filename, 'wb') as thumbf:
539 shutil.copyfileobj(uf, thumbf)
540 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
541 (info_dict['extractor'], info_dict['id'], thumb_filename))
542
543 if not self.params.get('skip_download', False):
544 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
545 success = True
546 else:
547 try:
548 success = self.fd._do_download(filename, info_dict)
549 except (OSError, IOError) as err:
550 raise UnavailableVideoError(err)
551 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
552 self.report_error(u'unable to download video data: %s' % str(err))
553 return
554 except (ContentTooShortError, ) as err:
555 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
556 return
557
558 if success:
559 try:
560 self.post_process(filename, info_dict)
561 except (PostProcessingError) as err:
562 self.report_error(u'postprocessing: %s' % str(err))
563 return
564
565 def download(self, url_list):
566 """Download a given list of URLs."""
567 if len(url_list) > 1 and self.fixed_template():
568 raise SameFileError(self.params['outtmpl'])
569
570 for url in url_list:
571 try:
572 #It also downloads the videos
573 videos = self.extract_info(url)
574 except UnavailableVideoError:
575 self.report_error(u'unable to download video')
576 except MaxDownloadsReached:
577 self.to_screen(u'[info] Maximum number of downloaded files reached.')
578 raise
579
580 return self._download_retcode
581
582 def post_process(self, filename, ie_info):
583 """Run all the postprocessors on the given file."""
584 info = dict(ie_info)
585 info['filepath'] = filename
586 keep_video = None
587 for pp in self._pps:
588 try:
589 keep_video_wish,new_info = pp.run(info)
590 if keep_video_wish is not None:
591 if keep_video_wish:
592 keep_video = keep_video_wish
593 elif keep_video is None:
594 # No clear decision yet, let IE decide
595 keep_video = keep_video_wish
596 except PostProcessingError as e:
597 self.report_error(e.msg)
598 if keep_video is False and not self.params.get('keepvideo', False):
599 try:
600 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
601 os.remove(encodeFilename(filename))
602 except (IOError, OSError):
603 self.report_warning(u'Unable to remove downloaded video file')