2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import
20 from .InfoExtractors import get_info_extractor
23 class FileDownloader(object):
24 """File Downloader class.
26 File downloader objects are the ones responsible of downloading the
27 actual video file and writing it to disk if the user has requested
28 it, among some other tasks. In most cases there should be one per
29 program. As, given a video URL, the downloader doesn't know how to
30 extract all the needed information, task that InfoExtractors do, it
31 has to pass the URL to one of them.
33 For this, file downloader objects have a method that allows
34 InfoExtractors to be registered in a given order. When it is passed
35 a URL, the file downloader handles it to the first InfoExtractor it
36 finds that reports being able to handle it. The InfoExtractor extracts
37 all the information about the video or videos the URL refers to, and
38 asks the FileDownloader to process the video information, possibly
39 downloading the video.
41 File downloaders accept a lot of parameters. In order not to saturate
42 the object constructor with arguments, it receives a dictionary of
43 options instead. These options are available through the params
44 attribute for the InfoExtractors to use. The FileDownloader also
45 registers itself as the downloader in charge for the InfoExtractors
46 that are added to it, so this is a "mutual registration".
50 username: Username for authentication purposes.
51 password: Password for authentication purposes.
52 usenetrc: Use netrc for authentication instead.
53 quiet: Do not print messages to stdout.
54 forceurl: Force printing final URL.
55 forcetitle: Force printing title.
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 ratelimit: Download speed limit, in bytes/sec.
66 nooverwrites: Prevent overwriting files.
67 retries: Number of times to retry for HTTP error 5xx
68 buffersize: Size of download buffer in bytes.
69 noresizebuffer: Do not automatically resize the download buffer.
70 continuedl: Try to continue downloads if possible.
71 noprogress: Do not print the progress bar.
72 playliststart: Playlist item to start at.
73 playlistend: Playlist item to end at.
74 matchtitle: Download only matching titles.
75 rejecttitle: Reject downloads for matching titles.
76 logtostderr: Log messages to stderr instead of stdout.
77 consoletitle: Display progress in console window's titlebar.
78 nopart: Do not use temporary .part files.
79 updatetime: Use the Last-modified header to set output file timestamps.
80 writedescription: Write the video description to a .description file
81 writeinfojson: Write the video description to a .info.json file
82 writesubtitles: Write the video subtitles to a file
83 onlysubtitles: Downloads only the subtitles of the video
84 allsubtitles: Downloads all the subtitles of the video
85 listsubtitles: Lists all available subtitles for the video
86 subtitlesformat: Subtitle format [sbv/srt] (default=srt)
87 subtitleslang: Language of the subtitles to download
88 test: Download only first bytes to test the downloader.
89 keepvideo: Keep the video file after post-processing
90 min_filesize: Skip files smaller than this size
91 max_filesize: Skip files larger than this size
92 daterange: A DateRange object, download only if the upload_date is in the range.
98 _download_retcode = None
102 def __init__(self, params):
103 """Create a FileDownloader object with the given options."""
106 self._progress_hooks = []
107 self._download_retcode = 0
108 self._num_downloads = 0
109 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
112 if '%(stitle)s' in self.params['outtmpl']:
113 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.')
116 def format_bytes(bytes):
119 if type(bytes) is str:
124 exponent = int(math.log(bytes, 1024.0))
125 suffix = ['B','KiB','MiB','GiB','TiB','PiB','EiB','ZiB','YiB'][exponent]
126 converted = float(bytes) / float(1024 ** exponent)
127 return '%.2f%s' % (converted, suffix)
130 def calc_percent(byte_counter, data_len):
133 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
136 def calc_eta(start, now, total, current):
140 if current == 0 or dif < 0.001: # One millisecond
142 rate = float(current) / dif
143 eta = int((float(total) - float(current)) / rate)
144 (eta_mins, eta_secs) = divmod(eta, 60)
147 return '%02d:%02d' % (eta_mins, eta_secs)
150 def calc_speed(start, now, bytes):
152 if bytes == 0 or dif < 0.001: # One millisecond
153 return '%10s' % '---b/s'
154 return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif))
157 def best_block_size(elapsed_time, bytes):
158 new_min = max(bytes / 2.0, 1.0)
159 new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
160 if elapsed_time < 0.001:
162 rate = bytes / elapsed_time
170 def parse_bytes(bytestr):
171 """Parse a string indicating a byte quantity into an integer."""
172 matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
175 number = float(matchobj.group(1))
176 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
177 return int(round(number * multiplier))
179 def add_info_extractor(self, ie):
180 """Add an InfoExtractor object to the end of the list."""
182 ie.set_downloader(self)
184 def add_post_processor(self, pp):
185 """Add a PostProcessor object to the end of the chain."""
187 pp.set_downloader(self)
189 def to_screen(self, message, skip_eol=False):
190 """Print message to stdout if not in quiet mode."""
191 assert type(message) == type(u'')
192 if not self.params.get('quiet', False):
193 terminator = [u'\n', u''][skip_eol]
194 output = message + terminator
195 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
196 output = output.encode(preferredencoding(), 'ignore')
197 self._screen_file.write(output)
198 self._screen_file.flush()
200 def to_stderr(self, message):
201 """Print message to stderr."""
202 assert type(message) == type(u'')
203 output = message + u'\n'
204 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
205 output = output.encode(preferredencoding())
206 sys.stderr.write(output)
208 def to_cons_title(self, message):
209 """Set console/terminal window title to message."""
210 if not self.params.get('consoletitle', False):
212 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
213 # c_wchar_p() might not be necessary if `message` is
214 # already of type unicode()
215 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
216 elif 'TERM' in os.environ:
217 self.to_screen('\033]0;%s\007' % message, skip_eol=True)
219 def fixed_template(self):
220 """Checks if the output template is fixed."""
221 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
223 def trouble(self, message=None, tb=None):
224 """Determine action to take when a download problem appears.
226 Depending on if the downloader has been configured to ignore
227 download errors or not, this method may throw an exception or
228 not when errors are found, after printing the message.
230 tb, if given, is additional traceback information.
232 if message is not None:
233 self.to_stderr(message)
234 if self.params.get('verbose'):
236 if sys.exc_info()[0]: # if .trouble has been called from an except block
238 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
239 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
240 tb += compat_str(traceback.format_exc())
242 tb_data = traceback.format_list(traceback.extract_stack())
243 tb = u''.join(tb_data)
245 if not self.params.get('ignoreerrors', False):
246 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
247 exc_info = sys.exc_info()[1].exc_info
249 exc_info = sys.exc_info()
250 raise DownloadError(message, exc_info)
251 self._download_retcode = 1
253 def report_warning(self, message):
255 Print the message to stderr, it will be prefixed with 'WARNING:'
256 If stderr is a tty file the 'WARNING:' will be colored
258 if sys.stderr.isatty() and os.name != 'nt':
259 _msg_header=u'\033[0;33mWARNING:\033[0m'
261 _msg_header=u'WARNING:'
262 warning_message=u'%s %s' % (_msg_header,message)
263 self.to_stderr(warning_message)
265 def report_error(self, message, tb=None):
267 Do the same as trouble, but prefixes the message with 'ERROR:', colored
268 in red if stderr is a tty file.
270 if sys.stderr.isatty() and os.name != 'nt':
271 _msg_header = u'\033[0;31mERROR:\033[0m'
273 _msg_header = u'ERROR:'
274 error_message = u'%s %s' % (_msg_header, message)
275 self.trouble(error_message, tb)
277 def slow_down(self, start_time, byte_counter):
278 """Sleep if the download speed is over the rate limit."""
279 rate_limit = self.params.get('ratelimit', None)
280 if rate_limit is None or byte_counter == 0:
283 elapsed = now - start_time
286 speed = float(byte_counter) / elapsed
287 if speed > rate_limit:
288 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
290 def temp_name(self, filename):
291 """Returns a temporary filename for the given filename."""
292 if self.params.get('nopart', False) or filename == u'-' or \
293 (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
295 return filename + u'.part'
297 def undo_temp_name(self, filename):
298 if filename.endswith(u'.part'):
299 return filename[:-len(u'.part')]
302 def try_rename(self, old_filename, new_filename):
304 if old_filename == new_filename:
306 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
307 except (IOError, OSError) as err:
308 self.report_error(u'unable to rename file')
310 def try_utime(self, filename, last_modified_hdr):
311 """Try to set the last-modified time of the given file."""
312 if last_modified_hdr is None:
314 if not os.path.isfile(encodeFilename(filename)):
316 timestr = last_modified_hdr
319 filetime = timeconvert(timestr)
323 os.utime(filename, (time.time(), filetime))
328 def report_writedescription(self, descfn):
329 """ Report that the description file is being written """
330 self.to_screen(u'[info] Writing video description to: ' + descfn)
332 def report_writesubtitles(self, sub_filename):
333 """ Report that the subtitles file is being written """
334 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
336 def report_writeinfojson(self, infofn):
337 """ Report that the metadata file has been written """
338 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
340 def report_destination(self, filename):
341 """Report destination filename."""
342 self.to_screen(u'[download] Destination: ' + filename)
344 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
345 """Report download progress."""
346 if self.params.get('noprogress', False):
348 if self.params.get('progress_with_newline', False):
349 self.to_screen(u'[download] %s of %s at %s ETA %s' %
350 (percent_str, data_len_str, speed_str, eta_str))
352 self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
353 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
354 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
355 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
357 def report_resuming_byte(self, resume_len):
358 """Report attempt to resume at given byte."""
359 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
361 def report_retry(self, count, retries):
362 """Report retry in case of HTTP error 5xx"""
363 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
365 def report_file_already_downloaded(self, file_name):
366 """Report file has already been fully downloaded."""
368 self.to_screen(u'[download] %s has already been downloaded' % file_name)
369 except (UnicodeEncodeError) as err:
370 self.to_screen(u'[download] The file has already been downloaded')
372 def report_unable_to_resume(self):
373 """Report it was impossible to resume download."""
374 self.to_screen(u'[download] Unable to resume')
376 def report_finish(self):
377 """Report download finished."""
378 if self.params.get('noprogress', False):
379 self.to_screen(u'[download] Download completed')
383 def increment_downloads(self):
384 """Increment the ordinal that assigns a number to each file."""
385 self._num_downloads += 1
387 def prepare_filename(self, info_dict):
388 """Generate the output filename."""
390 template_dict = dict(info_dict)
392 template_dict['epoch'] = int(time.time())
393 autonumber_size = self.params.get('autonumber_size')
394 if autonumber_size is None:
396 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
397 template_dict['autonumber'] = autonumber_templ % self._num_downloads
398 if template_dict['playlist_index'] is not None:
399 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
401 sanitize = lambda k,v: sanitize_filename(
402 u'NA' if v is None else compat_str(v),
403 restricted=self.params.get('restrictfilenames'),
405 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
407 filename = self.params['outtmpl'] % template_dict
409 except KeyError as err:
410 self.report_error(u'Erroneous output template')
412 except ValueError as err:
413 self.report_error(u'Insufficient system charset ' + repr(preferredencoding()))
416 def _match_entry(self, info_dict):
417 """ Returns None iff the file should be downloaded """
419 title = info_dict['title']
420 matchtitle = self.params.get('matchtitle', False)
422 if not re.search(matchtitle, title, re.IGNORECASE):
423 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
424 rejecttitle = self.params.get('rejecttitle', False)
426 if re.search(rejecttitle, title, re.IGNORECASE):
427 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
428 date = info_dict.get('upload_date', None)
430 dateRange = self.params.get('daterange', DateRange())
431 if date not in dateRange:
432 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
435 def extract_info(self, url, download = True, ie_name = None):
437 Returns a list with a dictionary for each video we find.
438 If 'download', also downloads the videos.
440 suitable_found = False
442 #We copy the original list
443 ies = list(self._ies)
445 if ie_name is not None:
446 #We put in the first place the given info extractor
447 first_ie = get_info_extractor(ie_name)()
448 first_ie.set_downloader(self)
449 ies.insert(0, first_ie)
452 # Go to next InfoExtractor if not suitable
453 if not ie.suitable(url):
456 # Warn if the _WORKING attribute is False
458 self.report_warning(u'the program functionality for this site has been marked as broken, '
459 u'and will probably not work. If you want to go on, use the -i option.')
461 # Suitable InfoExtractor found
462 suitable_found = True
464 # Extract information from URL and process it
466 ie_results = ie.extract(url)
467 if ie_results is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
470 for ie_result in ie_results:
471 if not 'extractor' in ie_result:
472 #The extractor has already been set somewhere else
473 ie_result['extractor'] = ie.IE_NAME
474 results.append(self.process_ie_result(ie_result, download))
476 except ExtractorError as de: # An error we somewhat expected
477 self.report_error(compat_str(de), de.format_traceback())
479 except Exception as e:
480 if self.params.get('ignoreerrors', False):
481 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
485 if not suitable_found:
486 self.report_error(u'no suitable InfoExtractor: %s' % url)
488 def process_ie_result(self, ie_result, download = True):
490 Take the result of the ie and return a list of videos.
491 For url elements it will search the suitable ie and get the videos
492 For playlist elements it will process each of the elements of the 'entries' key
494 It will also download the videos if 'download'.
496 result_type = ie_result.get('_type', 'video') #If not given we suppose it's a video, support the dafault old system
497 if result_type == 'video':
498 if 'playlist' not in ie_result:
499 #It isn't part of a playlist
500 ie_result['playlist'] = None
501 ie_result['playlist_index'] = None
504 self.process_info(ie_result)
506 elif result_type == 'url':
507 #We get the video pointed by the url
508 result = self.extract_info(ie_result['url'], download, ie_name = ie_result['ie_key'])[0]
510 elif result_type == 'playlist':
511 #We process each entry in the playlist
512 playlist = ie_result.get('title', None) or ie_result.get('id', None)
513 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
515 playlist_results = []
517 n_all_entries = len(ie_result['entries'])
518 playliststart = self.params.get('playliststart', 1) - 1
519 playlistend = self.params.get('playlistend', -1)
521 if playlistend == -1:
522 entries = ie_result['entries'][playliststart:]
524 entries = ie_result['entries'][playliststart:playlistend]
526 n_entries = len(entries)
528 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
529 (ie_result['extractor'], playlist, n_all_entries, n_entries))
531 for i,entry in enumerate(entries,1):
532 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
533 entry_result = self.process_ie_result(entry, False)
534 entry_result['playlist'] = playlist
535 entry_result['playlist_index'] = i + playliststart
536 #We must do the download here to correctly set the 'playlist' key
538 self.process_info(entry_result)
539 playlist_results.append(entry_result)
540 result = ie_result.copy()
541 result['entries'] = playlist_results
544 def process_info(self, info_dict):
545 """Process a single dictionary returned by an InfoExtractor."""
547 #We increment the download the download count here to match the previous behaviour.
548 self.increment_downloads()
550 info_dict['fulltitle'] = info_dict['title']
551 if len(info_dict['title']) > 200:
552 info_dict['title'] = info_dict['title'][:197] + u'...'
554 # Keep for backwards compatibility
555 info_dict['stitle'] = info_dict['title']
557 if not 'format' in info_dict:
558 info_dict['format'] = info_dict['ext']
560 reason = self._match_entry(info_dict)
561 if reason is not None:
562 self.to_screen(u'[download] ' + reason)
565 max_downloads = self.params.get('max_downloads')
566 if max_downloads is not None:
567 if self._num_downloads > int(max_downloads):
568 raise MaxDownloadsReached()
570 filename = self.prepare_filename(info_dict)
573 if self.params.get('forcetitle', False):
574 compat_print(info_dict['title'])
575 if self.params.get('forceurl', False):
576 compat_print(info_dict['url'])
577 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
578 compat_print(info_dict['thumbnail'])
579 if self.params.get('forcedescription', False) and 'description' in info_dict:
580 compat_print(info_dict['description'])
581 if self.params.get('forcefilename', False) and filename is not None:
582 compat_print(filename)
583 if self.params.get('forceformat', False):
584 compat_print(info_dict['format'])
586 # Do nothing else if in simulate mode
587 if self.params.get('simulate', False):
594 dn = os.path.dirname(encodeFilename(filename))
595 if dn != '' and not os.path.exists(dn): # dn is already encoded
597 except (OSError, IOError) as err:
598 self.report_error(u'unable to create directory ' + compat_str(err))
601 if self.params.get('writedescription', False):
603 descfn = filename + u'.description'
604 self.report_writedescription(descfn)
605 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
606 descfile.write(info_dict['description'])
607 except (OSError, IOError):
608 self.report_error(u'Cannot write description file ' + descfn)
611 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
612 # subtitles download errors are already managed as troubles in relevant IE
613 # that way it will silently go on when used with unsupporting IE
614 subtitle = info_dict['subtitles'][0]
615 (sub_error, sub_lang, sub) = subtitle
616 sub_format = self.params.get('subtitlesformat')
618 self.report_warning("Some error while getting the subtitles")
621 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
622 self.report_writesubtitles(sub_filename)
623 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
625 except (OSError, IOError):
626 self.report_error(u'Cannot write subtitles file ' + descfn)
628 if self.params.get('onlysubtitles', False):
631 if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
632 subtitles = info_dict['subtitles']
633 sub_format = self.params.get('subtitlesformat')
634 for subtitle in subtitles:
635 (sub_error, sub_lang, sub) = subtitle
637 self.report_warning("Some error while getting the subtitles")
640 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
641 self.report_writesubtitles(sub_filename)
642 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
644 except (OSError, IOError):
645 self.report_error(u'Cannot write subtitles file ' + descfn)
647 if self.params.get('onlysubtitles', False):
650 if self.params.get('writeinfojson', False):
651 infofn = filename + u'.info.json'
652 self.report_writeinfojson(infofn)
654 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
655 write_json_file(json_info_dict, encodeFilename(infofn))
656 except (OSError, IOError):
657 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
660 if not self.params.get('skip_download', False):
661 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
665 success = self._do_download(filename, info_dict)
666 except (OSError, IOError) as err:
667 raise UnavailableVideoError()
668 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
669 self.report_error(u'unable to download video data: %s' % str(err))
671 except (ContentTooShortError, ) as err:
672 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
677 self.post_process(filename, info_dict)
678 except (PostProcessingError) as err:
679 self.report_error(u'postprocessing: %s' % str(err))
682 def download(self, url_list):
683 """Download a given list of URLs."""
684 if len(url_list) > 1 and self.fixed_template():
685 raise SameFileError(self.params['outtmpl'])
689 #It also downloads the videos
690 videos = self.extract_info(url)
691 except UnavailableVideoError:
692 self.report_error(u'unable to download video')
693 except MaxDownloadsReached:
694 self.to_screen(u'[info] Maximum number of downloaded files reached.')
697 return self._download_retcode
699 def post_process(self, filename, ie_info):
700 """Run all the postprocessors on the given file."""
702 info['filepath'] = filename
706 keep_video_wish,new_info = pp.run(info)
707 if keep_video_wish is not None:
709 keep_video = keep_video_wish
710 elif keep_video is None:
711 # No clear decision yet, let IE decide
712 keep_video = keep_video_wish
713 except PostProcessingError as e:
714 self.to_stderr(u'ERROR: ' + e.msg)
715 if keep_video is False and not self.params.get('keepvideo', False):
717 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
718 os.remove(encodeFilename(filename))
719 except (IOError, OSError):
720 self.report_warning(u'Unable to remove downloaded video file')
722 def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path):
723 self.report_destination(filename)
724 tmpfilename = self.temp_name(filename)
726 # Check for rtmpdump first
728 subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
729 except (OSError, IOError):
730 self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
733 # Download using rtmpdump. rtmpdump returns exit code 2 when
734 # the connection was interrumpted and resuming appears to be
735 # possible. This is part of rtmpdump's normal usage, AFAIK.
736 basic_args = ['rtmpdump', '-q', '-r', url, '-o', tmpfilename]
737 if player_url is not None:
738 basic_args += ['-W', player_url]
739 if page_url is not None:
740 basic_args += ['--pageUrl', page_url]
741 if play_path is not None:
742 basic_args += ['-y', play_path]
743 args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
744 if self.params.get('verbose', False):
747 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
750 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
751 retval = subprocess.call(args)
752 while retval == 2 or retval == 1:
753 prevsize = os.path.getsize(encodeFilename(tmpfilename))
754 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
755 time.sleep(5.0) # This seems to be needed
756 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
757 cursize = os.path.getsize(encodeFilename(tmpfilename))
758 if prevsize == cursize and retval == 1:
760 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
761 if prevsize == cursize and retval == 2 and cursize > 1024:
762 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
766 fsize = os.path.getsize(encodeFilename(tmpfilename))
767 self.to_screen(u'\r[rtmpdump] %s bytes' % fsize)
768 self.try_rename(tmpfilename, filename)
769 self._hook_progress({
770 'downloaded_bytes': fsize,
771 'total_bytes': fsize,
772 'filename': filename,
773 'status': 'finished',
777 self.to_stderr(u"\n")
778 self.report_error(u'rtmpdump exited with code %d' % retval)
781 def _do_download(self, filename, info_dict):
782 url = info_dict['url']
784 # Check file already present
785 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
786 self.report_file_already_downloaded(filename)
787 self._hook_progress({
788 'filename': filename,
789 'status': 'finished',
793 # Attempt to download using rtmpdump
794 if url.startswith('rtmp'):
795 return self._download_with_rtmpdump(filename, url,
796 info_dict.get('player_url', None),
797 info_dict.get('page_url', None),
798 info_dict.get('play_path', None))
800 tmpfilename = self.temp_name(filename)
803 # Do not include the Accept-Encoding header
804 headers = {'Youtubedl-no-compression': 'True'}
805 if 'user_agent' in info_dict:
806 headers['Youtubedl-user-agent'] = info_dict['user_agent']
807 basic_request = compat_urllib_request.Request(url, None, headers)
808 request = compat_urllib_request.Request(url, None, headers)
810 if self.params.get('test', False):
811 request.add_header('Range','bytes=0-10240')
813 # Establish possible resume length
814 if os.path.isfile(encodeFilename(tmpfilename)):
815 resume_len = os.path.getsize(encodeFilename(tmpfilename))
821 if self.params.get('continuedl', False):
822 self.report_resuming_byte(resume_len)
823 request.add_header('Range','bytes=%d-' % resume_len)
829 retries = self.params.get('retries', 0)
830 while count <= retries:
831 # Establish connection
833 if count == 0 and 'urlhandle' in info_dict:
834 data = info_dict['urlhandle']
835 data = compat_urllib_request.urlopen(request)
837 except (compat_urllib_error.HTTPError, ) as err:
838 if (err.code < 500 or err.code >= 600) and err.code != 416:
839 # Unexpected HTTP error
841 elif err.code == 416:
842 # Unable to resume (requested range not satisfiable)
844 # Open the connection again without the range header
845 data = compat_urllib_request.urlopen(basic_request)
846 content_length = data.info()['Content-Length']
847 except (compat_urllib_error.HTTPError, ) as err:
848 if err.code < 500 or err.code >= 600:
851 # Examine the reported length
852 if (content_length is not None and
853 (resume_len - 100 < int(content_length) < resume_len + 100)):
854 # The file had already been fully downloaded.
855 # Explanation to the above condition: in issue #175 it was revealed that
856 # YouTube sometimes adds or removes a few bytes from the end of the file,
857 # changing the file size slightly and causing problems for some users. So
858 # I decided to implement a suggested change and consider the file
859 # completely downloaded if the file size differs less than 100 bytes from
860 # the one in the hard drive.
861 self.report_file_already_downloaded(filename)
862 self.try_rename(tmpfilename, filename)
863 self._hook_progress({
864 'filename': filename,
865 'status': 'finished',
869 # The length does not match, we start the download over
870 self.report_unable_to_resume()
876 self.report_retry(count, retries)
879 self.report_error(u'giving up after %s retries' % retries)
882 data_len = data.info().get('Content-length', None)
883 if data_len is not None:
884 data_len = int(data_len) + resume_len
885 min_data_len = self.params.get("min_filesize", None)
886 max_data_len = self.params.get("max_filesize", None)
887 if min_data_len is not None and data_len < min_data_len:
888 self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
890 if max_data_len is not None and data_len > max_data_len:
891 self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
894 data_len_str = self.format_bytes(data_len)
895 byte_counter = 0 + resume_len
896 block_size = self.params.get('buffersize', 1024)
901 data_block = data.read(block_size)
903 if len(data_block) == 0:
905 byte_counter += len(data_block)
907 # Open file just in time
910 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
911 assert stream is not None
912 filename = self.undo_temp_name(tmpfilename)
913 self.report_destination(filename)
914 except (OSError, IOError) as err:
915 self.report_error(u'unable to open for writing: %s' % str(err))
918 stream.write(data_block)
919 except (IOError, OSError) as err:
920 self.to_stderr(u"\n")
921 self.report_error(u'unable to write data: %s' % str(err))
923 if not self.params.get('noresizebuffer', False):
924 block_size = self.best_block_size(after - before, len(data_block))
927 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
929 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
931 percent_str = self.calc_percent(byte_counter, data_len)
932 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
933 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
935 self._hook_progress({
936 'downloaded_bytes': byte_counter,
937 'total_bytes': data_len,
938 'tmpfilename': tmpfilename,
939 'filename': filename,
940 'status': 'downloading',
944 self.slow_down(start, byte_counter - resume_len)
947 self.to_stderr(u"\n")
948 self.report_error(u'Did not get any data blocks')
952 if data_len is not None and byte_counter != data_len:
953 raise ContentTooShortError(byte_counter, int(data_len))
954 self.try_rename(tmpfilename, filename)
956 # Update file modification time
957 if self.params.get('updatetime', True):
958 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
960 self._hook_progress({
961 'downloaded_bytes': byte_counter,
962 'total_bytes': byte_counter,
963 'filename': filename,
964 'status': 'finished',
969 def _hook_progress(self, status):
970 for ph in self._progress_hooks:
973 def add_progress_hook(self, ph):
974 """ ph gets called on download progress, with a dictionary with the entries
975 * filename: The final filename
976 * status: One of "downloading" and "finished"
978 It can also have some of the following entries:
980 * downloaded_bytes: Bytes on disks
981 * total_bytes: Total bytes, None if unknown
982 * tmpfilename: The filename we're currently writing to
984 Hooks are guaranteed to be called at least once (with status "finished")
985 if the download is successful.
987 self._progress_hooks.append(ph)