2 # -*- coding: utf-8 -*-
4 from __future__ import absolute_import
22 class FileDownloader(object):
23 """File Downloader class.
25 File downloader objects are the ones responsible of downloading the
26 actual video file and writing it to disk if the user has requested
27 it, among some other tasks. In most cases there should be one per
28 program. As, given a video URL, the downloader doesn't know how to
29 extract all the needed information, task that InfoExtractors do, it
30 has to pass the URL to one of them.
32 For this, file downloader objects have a method that allows
33 InfoExtractors to be registered in a given order. When it is passed
34 a URL, the file downloader handles it to the first InfoExtractor it
35 finds that reports being able to handle it. The InfoExtractor extracts
36 all the information about the video or videos the URL refers to, and
37 asks the FileDownloader to process the video information, possibly
38 downloading the video.
40 File downloaders accept a lot of parameters. In order not to saturate
41 the object constructor with arguments, it receives a dictionary of
42 options instead. These options are available through the params
43 attribute for the InfoExtractors to use. The FileDownloader also
44 registers itself as the downloader in charge for the InfoExtractors
45 that are added to it, so this is a "mutual registration".
49 username: Username for authentication purposes.
50 password: Password for authentication purposes.
51 usenetrc: Use netrc for authentication instead.
52 quiet: Do not print messages to stdout.
53 forceurl: Force printing final URL.
54 forcetitle: Force printing title.
55 forcethumbnail: Force printing thumbnail URL.
56 forcedescription: Force printing description.
57 forcefilename: Force printing final filename.
58 simulate: Do not download the video files.
59 format: Video format code.
60 format_limit: Highest quality format to try.
61 outtmpl: Template for output names.
62 restrictfilenames: Do not allow "&" and spaces in file names
63 ignoreerrors: Do not stop on download errors.
64 ratelimit: Download speed limit, in bytes/sec.
65 nooverwrites: Prevent overwriting files.
66 retries: Number of times to retry for HTTP error 5xx
67 buffersize: Size of download buffer in bytes.
68 noresizebuffer: Do not automatically resize the download buffer.
69 continuedl: Try to continue downloads if possible.
70 noprogress: Do not print the progress bar.
71 playliststart: Playlist item to start at.
72 playlistend: Playlist item to end at.
73 matchtitle: Download only matching titles.
74 rejecttitle: Reject downloads for matching titles.
75 logtostderr: Log messages to stderr instead of stdout.
76 consoletitle: Display progress in console window's titlebar.
77 nopart: Do not use temporary .part files.
78 updatetime: Use the Last-modified header to set output file timestamps.
79 writedescription: Write the video description to a .description file
80 writeinfojson: Write the video description to a .info.json file
81 writesubtitles: Write the video subtitles to a file
82 onlysubtitles: Downloads only the subtitles of the video
83 allsubtitles: Downloads all the subtitles of the video
84 listsubtitles: Lists all available subtitles for the video
85 subtitlesformat: Subtitle format [sbv/srt] (default=srt)
86 subtitleslang: Language of the subtitles to download
87 test: Download only first bytes to test the downloader.
88 keepvideo: Keep the video file after post-processing
89 min_filesize: Skip files smaller than this size
90 max_filesize: Skip files larger than this size
96 _download_retcode = None
100 def __init__(self, params):
101 """Create a FileDownloader object with the given options."""
104 self._progress_hooks = []
105 self._download_retcode = 0
106 self._num_downloads = 0
107 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
110 if '%(stitle)s' in self.params['outtmpl']:
111 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.')
114 def format_bytes(bytes):
117 if type(bytes) is str:
122 exponent = int(math.log(bytes, 1024.0))
123 suffix = 'bkMGTPEZY'[exponent]
124 converted = float(bytes) / float(1024 ** exponent)
125 return '%.2f%s' % (converted, suffix)
128 def calc_percent(byte_counter, data_len):
131 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
134 def calc_eta(start, now, total, current):
138 if current == 0 or dif < 0.001: # One millisecond
140 rate = float(current) / dif
141 eta = int((float(total) - float(current)) / rate)
142 (eta_mins, eta_secs) = divmod(eta, 60)
145 return '%02d:%02d' % (eta_mins, eta_secs)
148 def calc_speed(start, now, bytes):
150 if bytes == 0 or dif < 0.001: # One millisecond
151 return '%10s' % '---b/s'
152 return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif))
155 def best_block_size(elapsed_time, bytes):
156 new_min = max(bytes / 2.0, 1.0)
157 new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
158 if elapsed_time < 0.001:
160 rate = bytes / elapsed_time
168 def parse_bytes(bytestr):
169 """Parse a string indicating a byte quantity into an integer."""
170 matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
173 number = float(matchobj.group(1))
174 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
175 return int(round(number * multiplier))
177 def add_info_extractor(self, ie):
178 """Add an InfoExtractor object to the end of the list."""
180 ie.set_downloader(self)
182 def add_post_processor(self, pp):
183 """Add a PostProcessor object to the end of the chain."""
185 pp.set_downloader(self)
187 def to_screen(self, message, skip_eol=False):
188 """Print message to stdout if not in quiet mode."""
189 assert type(message) == type(u'')
190 if not self.params.get('quiet', False):
191 terminator = [u'\n', u''][skip_eol]
192 output = message + terminator
193 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
194 output = output.encode(preferredencoding(), 'ignore')
195 self._screen_file.write(output)
196 self._screen_file.flush()
198 def to_stderr(self, message):
199 """Print message to stderr."""
200 assert type(message) == type(u'')
201 output = message + u'\n'
202 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
203 output = output.encode(preferredencoding())
204 sys.stderr.write(output)
206 def to_cons_title(self, message):
207 """Set console/terminal window title to message."""
208 if not self.params.get('consoletitle', False):
210 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
211 # c_wchar_p() might not be necessary if `message` is
212 # already of type unicode()
213 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
214 elif 'TERM' in os.environ:
215 self.to_screen('\033]0;%s\007' % message, skip_eol=True)
217 def fixed_template(self):
218 """Checks if the output template is fixed."""
219 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
221 def trouble(self, message=None, tb=None):
222 """Determine action to take when a download problem appears.
224 Depending on if the downloader has been configured to ignore
225 download errors or not, this method may throw an exception or
226 not when errors are found, after printing the message.
228 tb, if given, is additional traceback information.
230 if message is not None:
231 self.to_stderr(message)
232 if self.params.get('verbose'):
234 if sys.exc_info()[0]: # if .trouble has been called from an except block
236 if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
237 tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info))
238 tb += compat_str(traceback.format_exc())
240 tb_data = traceback.format_list(traceback.extract_stack())
241 tb = u''.join(tb_data)
243 if not self.params.get('ignoreerrors', False):
244 if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
245 exc_info = sys.exc_info()[1].exc_info
247 exc_info = sys.exc_info()
248 raise DownloadError(message, exc_info)
249 self._download_retcode = 1
251 def report_warning(self, message):
253 Print the message to stderr, it will be prefixed with 'WARNING:'
254 If stderr is a tty file the 'WARNING:' will be colored
256 if sys.stderr.isatty():
257 _msg_header=u'\033[0;33mWARNING:\033[0m'
259 _msg_header=u'WARNING:'
260 warning_message=u'%s %s' % (_msg_header,message)
261 self.to_stderr(warning_message)
263 def report_error(self, message, tb=None):
265 Do the same as trouble, but prefixes the message with 'ERROR:', colored
266 in red if stderr is a tty file.
268 if sys.stderr.isatty():
269 _msg_header = u'\033[0;31mERROR:\033[0m'
271 _msg_header = u'ERROR:'
272 error_message = u'%s %s' % (_msg_header, message)
273 self.trouble(error_message, tb)
275 def slow_down(self, start_time, byte_counter):
276 """Sleep if the download speed is over the rate limit."""
277 rate_limit = self.params.get('ratelimit', None)
278 if rate_limit is None or byte_counter == 0:
281 elapsed = now - start_time
284 speed = float(byte_counter) / elapsed
285 if speed > rate_limit:
286 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
288 def temp_name(self, filename):
289 """Returns a temporary filename for the given filename."""
290 if self.params.get('nopart', False) or filename == u'-' or \
291 (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
293 return filename + u'.part'
295 def undo_temp_name(self, filename):
296 if filename.endswith(u'.part'):
297 return filename[:-len(u'.part')]
300 def try_rename(self, old_filename, new_filename):
302 if old_filename == new_filename:
304 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
305 except (IOError, OSError) as err:
306 self.report_error(u'unable to rename file')
308 def try_utime(self, filename, last_modified_hdr):
309 """Try to set the last-modified time of the given file."""
310 if last_modified_hdr is None:
312 if not os.path.isfile(encodeFilename(filename)):
314 timestr = last_modified_hdr
317 filetime = timeconvert(timestr)
321 os.utime(filename, (time.time(), filetime))
326 def report_writedescription(self, descfn):
327 """ Report that the description file is being written """
328 self.to_screen(u'[info] Writing video description to: ' + descfn)
330 def report_writesubtitles(self, sub_filename):
331 """ Report that the subtitles file is being written """
332 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
334 def report_writeinfojson(self, infofn):
335 """ Report that the metadata file has been written """
336 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
338 def report_destination(self, filename):
339 """Report destination filename."""
340 self.to_screen(u'[download] Destination: ' + filename)
342 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
343 """Report download progress."""
344 if self.params.get('noprogress', False):
346 if self.params.get('progress_with_newline', False):
347 self.to_screen(u'[download] %s of %s at %s ETA %s' %
348 (percent_str, data_len_str, speed_str, eta_str))
350 self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
351 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
352 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
353 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
355 def report_resuming_byte(self, resume_len):
356 """Report attempt to resume at given byte."""
357 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
359 def report_retry(self, count, retries):
360 """Report retry in case of HTTP error 5xx"""
361 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
363 def report_file_already_downloaded(self, file_name):
364 """Report file has already been fully downloaded."""
366 self.to_screen(u'[download] %s has already been downloaded' % file_name)
367 except (UnicodeEncodeError) as err:
368 self.to_screen(u'[download] The file has already been downloaded')
370 def report_unable_to_resume(self):
371 """Report it was impossible to resume download."""
372 self.to_screen(u'[download] Unable to resume')
374 def report_finish(self):
375 """Report download finished."""
376 if self.params.get('noprogress', False):
377 self.to_screen(u'[download] Download completed')
381 def increment_downloads(self):
382 """Increment the ordinal that assigns a number to each file."""
383 self._num_downloads += 1
385 def prepare_filename(self, info_dict):
386 """Generate the output filename."""
388 template_dict = dict(info_dict)
390 template_dict['epoch'] = int(time.time())
391 template_dict['autonumber'] = u'%05d' % self._num_downloads
392 if template_dict['playlist_index'] is not None:
393 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
395 sanitize = lambda k,v: sanitize_filename(
396 u'NA' if v is None else compat_str(v),
397 restricted=self.params.get('restrictfilenames'),
399 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
401 filename = self.params['outtmpl'] % template_dict
403 except KeyError as err:
404 self.trouble(u'ERROR: Erroneous output template')
406 except ValueError as err:
407 self.trouble(u'ERROR: Insufficient system charset ' + repr(preferredencoding()))
410 def _match_entry(self, info_dict):
411 """ Returns None iff the file should be downloaded """
413 title = info_dict['title']
414 matchtitle = self.params.get('matchtitle', False)
416 if not re.search(matchtitle, title, re.IGNORECASE):
417 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
418 rejecttitle = self.params.get('rejecttitle', False)
420 if re.search(rejecttitle, title, re.IGNORECASE):
421 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
424 def extract_info(self, url, download = True):
426 Returns a list with a dictionary for each video we find.
427 If 'download', also downloads the videos.
429 suitable_found = False
431 # Go to next InfoExtractor if not suitable
432 if not ie.suitable(url):
435 # Warn if the _WORKING attribute is False
437 self.to_stderr(u'WARNING: the program functionality for this site has been marked as broken, '
438 u'and will probably not work. If you want to go on, use the -i option.')
440 # Suitable InfoExtractor found
441 suitable_found = True
443 # Extract information from URL and process it
445 ie_results = ie.extract(url)
447 for ie_result in ie_results:
448 if not 'extractor' in ie_result:
449 #The extractor has already been set somewhere else
450 ie_result['extractor'] = ie.IE_NAME
451 results.append(self.process_ie_result(ie_result, download))
453 except ExtractorError as de: # An error we somewhat expected
454 self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
456 except Exception as e:
457 if self.params.get('ignoreerrors', False):
458 self.trouble(u'ERROR: ' + compat_str(e), tb=compat_str(traceback.format_exc()))
462 if not suitable_found:
463 self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
465 def process_ie_result(self, ie_result, download = True):
467 Take the result of the ie and return a list of videos.
468 For url elements it will search the suitable ie and get the videos
469 For playlist elements it will process each of the elements of the 'entries' key
471 It will also download the videos if 'download'.
473 result_type = ie_result.get('_type', 'video') #If not given we suppose it's a video, support the dafault old system
474 if result_type == 'video':
475 if 'playlist' not in ie_result:
476 #It isn't part of a playlist
477 ie_result['playlist'] = None
478 ie_result['playlist_index'] = None
481 self.process_info(ie_result)
483 elif result_type == 'url':
484 #We get the video pointed by the url
485 result = self.extract_info(ie_result['url'], download)[0]
487 elif result_type == 'playlist':
488 #We process each entry in the playlist
489 playlist = ie_result.get('title', None) or ie_result.get('id', None)
490 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
492 playlist_results = []
494 n_all_entries = len(ie_result['entries'])
495 playliststart = self.params.get('playliststart', 1) - 1
496 playlistend = self.params.get('playlistend', -1)
498 if playlistend == -1:
499 entries = ie_result['entries'][playliststart:]
501 entries = ie_result['entries'][playliststart:playlistend]
503 n_entries = len(entries)
505 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
506 (ie_result['extractor'], playlist, n_all_entries, n_entries))
508 for i,entry in enumerate(entries,1):
509 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
510 entry_result = self.process_ie_result(entry, False)
511 entry_result['playlist'] = playlist
512 entry_result['playlist_index'] = i + playliststart
513 #We must do the download here to correctly set the 'playlist' key
515 self.process_info(entry_result)
516 playlist_results.append(entry_result)
517 result = ie_result.copy()
518 result['entries'] = playlist_results
521 def process_info(self, info_dict):
522 """Process a single dictionary returned by an InfoExtractor."""
524 #We increment the download the download count here to match the previous behaviour.
525 self.increment_downloads()
527 # Keep for backwards compatibility
528 info_dict['stitle'] = info_dict['title']
530 if not 'format' in info_dict:
531 info_dict['format'] = info_dict['ext']
533 reason = self._match_entry(info_dict)
534 if reason is not None:
535 self.to_screen(u'[download] ' + reason)
538 max_downloads = self.params.get('max_downloads')
539 if max_downloads is not None:
540 if self._num_downloads > int(max_downloads):
541 raise MaxDownloadsReached()
543 filename = self.prepare_filename(info_dict)
546 if self.params.get('forcetitle', False):
547 compat_print(info_dict['title'])
548 if self.params.get('forceurl', False):
549 compat_print(info_dict['url'])
550 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
551 compat_print(info_dict['thumbnail'])
552 if self.params.get('forcedescription', False) and 'description' in info_dict:
553 compat_print(info_dict['description'])
554 if self.params.get('forcefilename', False) and filename is not None:
555 compat_print(filename)
556 if self.params.get('forceformat', False):
557 compat_print(info_dict['format'])
559 # Do nothing else if in simulate mode
560 if self.params.get('simulate', False):
567 dn = os.path.dirname(encodeFilename(filename))
568 if dn != '' and not os.path.exists(dn): # dn is already encoded
570 except (OSError, IOError) as err:
571 self.report_error(u'unable to create directory ' + compat_str(err))
574 if self.params.get('writedescription', False):
576 descfn = filename + u'.description'
577 self.report_writedescription(descfn)
578 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
579 descfile.write(info_dict['description'])
580 except (OSError, IOError):
581 self.report_error(u'Cannot write description file ' + descfn)
584 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
585 # subtitles download errors are already managed as troubles in relevant IE
586 # that way it will silently go on when used with unsupporting IE
587 subtitle = info_dict['subtitles'][0]
588 (sub_error, sub_lang, sub) = subtitle
589 sub_format = self.params.get('subtitlesformat')
591 self.report_warning("Some error while getting the subtitles")
594 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
595 self.report_writesubtitles(sub_filename)
596 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
598 except (OSError, IOError):
599 self.report_error(u'Cannot write subtitles file ' + descfn)
601 if self.params.get('onlysubtitles', False):
604 if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
605 subtitles = info_dict['subtitles']
606 sub_format = self.params.get('subtitlesformat')
607 for subtitle in subtitles:
608 (sub_error, sub_lang, sub) = subtitle
610 self.report_warning("Some error while getting the subtitles")
613 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
614 self.report_writesubtitles(sub_filename)
615 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
617 except (OSError, IOError):
618 self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
620 if self.params.get('onlysubtitles', False):
623 if self.params.get('writeinfojson', False):
624 infofn = filename + u'.info.json'
625 self.report_writeinfojson(infofn)
627 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
628 write_json_file(json_info_dict, encodeFilename(infofn))
629 except (OSError, IOError):
630 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
633 if not self.params.get('skip_download', False):
634 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
638 success = self._do_download(filename, info_dict)
639 except (OSError, IOError) as err:
640 raise UnavailableVideoError()
641 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
642 self.report_error(u'unable to download video data: %s' % str(err))
644 except (ContentTooShortError, ) as err:
645 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
650 self.post_process(filename, info_dict)
651 except (PostProcessingError) as err:
652 self.report_error(u'postprocessing: %s' % str(err))
655 def download(self, url_list):
656 """Download a given list of URLs."""
657 if len(url_list) > 1 and self.fixed_template():
658 raise SameFileError(self.params['outtmpl'])
662 #It also downloads the videos
663 videos = self.extract_info(url)
664 except UnavailableVideoError:
665 self.trouble(u'\nERROR: unable to download video')
666 except MaxDownloadsReached:
667 self.to_screen(u'[info] Maximum number of downloaded files reached.')
670 return self._download_retcode
672 def post_process(self, filename, ie_info):
673 """Run all the postprocessors on the given file."""
675 info['filepath'] = filename
679 keep_video_wish,new_info = pp.run(info)
680 if keep_video_wish is not None:
682 keep_video = keep_video_wish
683 elif keep_video is None:
684 # No clear decision yet, let IE decide
685 keep_video = keep_video_wish
686 except PostProcessingError as e:
687 self.to_stderr(u'ERROR: ' + e.msg)
688 if keep_video is False and not self.params.get('keepvideo', False):
690 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
691 os.remove(encodeFilename(filename))
692 except (IOError, OSError):
693 self.report_warning(u'Unable to remove downloaded video file')
695 def _download_with_rtmpdump(self, filename, url, player_url, page_url):
696 self.report_destination(filename)
697 tmpfilename = self.temp_name(filename)
699 # Check for rtmpdump first
701 subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
702 except (OSError, IOError):
703 self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
706 # Download using rtmpdump. rtmpdump returns exit code 2 when
707 # the connection was interrumpted and resuming appears to be
708 # possible. This is part of rtmpdump's normal usage, AFAIK.
709 basic_args = ['rtmpdump', '-q', '-r', url, '-o', tmpfilename]
710 if player_url is not None:
711 basic_args += ['-W', player_url]
712 if page_url is not None:
713 basic_args += ['--pageUrl', page_url]
714 args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
715 if self.params.get('verbose', False):
718 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
721 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
722 retval = subprocess.call(args)
723 while retval == 2 or retval == 1:
724 prevsize = os.path.getsize(encodeFilename(tmpfilename))
725 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
726 time.sleep(5.0) # This seems to be needed
727 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
728 cursize = os.path.getsize(encodeFilename(tmpfilename))
729 if prevsize == cursize and retval == 1:
731 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
732 if prevsize == cursize and retval == 2 and cursize > 1024:
733 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
737 fsize = os.path.getsize(encodeFilename(tmpfilename))
738 self.to_screen(u'\r[rtmpdump] %s bytes' % fsize)
739 self.try_rename(tmpfilename, filename)
740 self._hook_progress({
741 'downloaded_bytes': fsize,
742 'total_bytes': fsize,
743 'filename': filename,
744 'status': 'finished',
748 self.to_stderr(u"\n")
749 self.report_error(u'rtmpdump exited with code %d' % retval)
752 def _do_download(self, filename, info_dict):
753 url = info_dict['url']
755 # Check file already present
756 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
757 self.report_file_already_downloaded(filename)
758 self._hook_progress({
759 'filename': filename,
760 'status': 'finished',
764 # Attempt to download using rtmpdump
765 if url.startswith('rtmp'):
766 return self._download_with_rtmpdump(filename, url,
767 info_dict.get('player_url', None),
768 info_dict.get('page_url', None))
770 tmpfilename = self.temp_name(filename)
773 # Do not include the Accept-Encoding header
774 headers = {'Youtubedl-no-compression': 'True'}
775 if 'user_agent' in info_dict:
776 headers['Youtubedl-user-agent'] = info_dict['user_agent']
777 basic_request = compat_urllib_request.Request(url, None, headers)
778 request = compat_urllib_request.Request(url, None, headers)
780 if self.params.get('test', False):
781 request.add_header('Range','bytes=0-10240')
783 # Establish possible resume length
784 if os.path.isfile(encodeFilename(tmpfilename)):
785 resume_len = os.path.getsize(encodeFilename(tmpfilename))
791 if self.params.get('continuedl', False):
792 self.report_resuming_byte(resume_len)
793 request.add_header('Range','bytes=%d-' % resume_len)
799 retries = self.params.get('retries', 0)
800 while count <= retries:
801 # Establish connection
803 if count == 0 and 'urlhandle' in info_dict:
804 data = info_dict['urlhandle']
805 data = compat_urllib_request.urlopen(request)
807 except (compat_urllib_error.HTTPError, ) as err:
808 if (err.code < 500 or err.code >= 600) and err.code != 416:
809 # Unexpected HTTP error
811 elif err.code == 416:
812 # Unable to resume (requested range not satisfiable)
814 # Open the connection again without the range header
815 data = compat_urllib_request.urlopen(basic_request)
816 content_length = data.info()['Content-Length']
817 except (compat_urllib_error.HTTPError, ) as err:
818 if err.code < 500 or err.code >= 600:
821 # Examine the reported length
822 if (content_length is not None and
823 (resume_len - 100 < int(content_length) < resume_len + 100)):
824 # The file had already been fully downloaded.
825 # Explanation to the above condition: in issue #175 it was revealed that
826 # YouTube sometimes adds or removes a few bytes from the end of the file,
827 # changing the file size slightly and causing problems for some users. So
828 # I decided to implement a suggested change and consider the file
829 # completely downloaded if the file size differs less than 100 bytes from
830 # the one in the hard drive.
831 self.report_file_already_downloaded(filename)
832 self.try_rename(tmpfilename, filename)
833 self._hook_progress({
834 'filename': filename,
835 'status': 'finished',
839 # The length does not match, we start the download over
840 self.report_unable_to_resume()
846 self.report_retry(count, retries)
849 self.report_error(u'giving up after %s retries' % retries)
852 data_len = data.info().get('Content-length', None)
853 if data_len is not None:
854 data_len = int(data_len) + resume_len
855 min_data_len = self.params.get("min_filesize", None)
856 max_data_len = self.params.get("max_filesize", None)
857 if min_data_len is not None and data_len < min_data_len:
858 self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
860 if max_data_len is not None and data_len > max_data_len:
861 self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
864 data_len_str = self.format_bytes(data_len)
865 byte_counter = 0 + resume_len
866 block_size = self.params.get('buffersize', 1024)
871 data_block = data.read(block_size)
873 if len(data_block) == 0:
875 byte_counter += len(data_block)
877 # Open file just in time
880 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
881 assert stream is not None
882 filename = self.undo_temp_name(tmpfilename)
883 self.report_destination(filename)
884 except (OSError, IOError) as err:
885 self.report_error(u'unable to open for writing: %s' % str(err))
888 stream.write(data_block)
889 except (IOError, OSError) as err:
890 self.to_stderr(u"\n")
891 self.report_error(u'unable to write data: %s' % str(err))
893 if not self.params.get('noresizebuffer', False):
894 block_size = self.best_block_size(after - before, len(data_block))
897 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
899 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
901 percent_str = self.calc_percent(byte_counter, data_len)
902 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
903 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
905 self._hook_progress({
906 'downloaded_bytes': byte_counter,
907 'total_bytes': data_len,
908 'tmpfilename': tmpfilename,
909 'filename': filename,
910 'status': 'downloading',
914 self.slow_down(start, byte_counter - resume_len)
917 self.to_stderr(u"\n")
918 self.report_error(u'Did not get any data blocks')
922 if data_len is not None and byte_counter != data_len:
923 raise ContentTooShortError(byte_counter, int(data_len))
924 self.try_rename(tmpfilename, filename)
926 # Update file modification time
927 if self.params.get('updatetime', True):
928 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
930 self._hook_progress({
931 'downloaded_bytes': byte_counter,
932 'total_bytes': byte_counter,
933 'filename': filename,
934 'status': 'finished',
939 def _hook_progress(self, status):
940 for ph in self._progress_hooks:
943 def add_progress_hook(self, ph):
944 """ ph gets called on download progress, with a dictionary with the entries
945 * filename: The final filename
946 * status: One of "downloading" and "finished"
948 It can also have some of the following entries:
950 * downloaded_bytes: Bytes on disks
951 * total_bytes: Total bytes, None if unknown
952 * tmpfilename: The filename we're currently writing to
954 Hooks are guaranteed to be called at least once (with status "finished")
955 if the download is successful.
957 self._progress_hooks.append(ph)