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 tb_data = traceback.format_list(traceback.extract_stack())
235 tb = u''.join(tb_data)
237 if not self.params.get('ignoreerrors', False):
238 raise DownloadError(message)
239 self._download_retcode = 1
241 def report_warning(self, message):
243 Print the message to stderr, it will be prefixed with 'WARNING:'
244 If stderr is a tty file the 'WARNING:' will be colored
246 if sys.stderr.isatty():
247 _msg_header=u'\033[0;33mWARNING:\033[0m'
249 _msg_header=u'WARNING:'
250 warning_message=u'%s %s' % (_msg_header,message)
251 self.to_stderr(warning_message)
253 def report_error(self, message, tb=None):
255 Do the same as trouble, but prefixes the message with 'ERROR:', colored
256 in red if stderr is a tty file.
258 if sys.stderr.isatty():
259 _msg_header = u'\033[0;31mERROR:\033[0m'
261 _msg_header = u'ERROR:'
262 error_message = u'%s %s' % (_msg_header, message)
263 self.trouble(error_message, tb)
265 def slow_down(self, start_time, byte_counter):
266 """Sleep if the download speed is over the rate limit."""
267 rate_limit = self.params.get('ratelimit', None)
268 if rate_limit is None or byte_counter == 0:
271 elapsed = now - start_time
274 speed = float(byte_counter) / elapsed
275 if speed > rate_limit:
276 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
278 def temp_name(self, filename):
279 """Returns a temporary filename for the given filename."""
280 if self.params.get('nopart', False) or filename == u'-' or \
281 (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
283 return filename + u'.part'
285 def undo_temp_name(self, filename):
286 if filename.endswith(u'.part'):
287 return filename[:-len(u'.part')]
290 def try_rename(self, old_filename, new_filename):
292 if old_filename == new_filename:
294 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
295 except (IOError, OSError) as err:
296 self.report_error(u'unable to rename file')
298 def try_utime(self, filename, last_modified_hdr):
299 """Try to set the last-modified time of the given file."""
300 if last_modified_hdr is None:
302 if not os.path.isfile(encodeFilename(filename)):
304 timestr = last_modified_hdr
307 filetime = timeconvert(timestr)
311 os.utime(filename, (time.time(), filetime))
316 def report_writedescription(self, descfn):
317 """ Report that the description file is being written """
318 self.to_screen(u'[info] Writing video description to: ' + descfn)
320 def report_writesubtitles(self, sub_filename):
321 """ Report that the subtitles file is being written """
322 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
324 def report_writeinfojson(self, infofn):
325 """ Report that the metadata file has been written """
326 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
328 def report_destination(self, filename):
329 """Report destination filename."""
330 self.to_screen(u'[download] Destination: ' + filename)
332 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
333 """Report download progress."""
334 if self.params.get('noprogress', False):
336 if self.params.get('progress_with_newline', False):
337 self.to_screen(u'[download] %s of %s at %s ETA %s' %
338 (percent_str, data_len_str, speed_str, eta_str))
340 self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
341 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
342 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
343 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
345 def report_resuming_byte(self, resume_len):
346 """Report attempt to resume at given byte."""
347 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
349 def report_retry(self, count, retries):
350 """Report retry in case of HTTP error 5xx"""
351 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
353 def report_file_already_downloaded(self, file_name):
354 """Report file has already been fully downloaded."""
356 self.to_screen(u'[download] %s has already been downloaded' % file_name)
357 except (UnicodeEncodeError) as err:
358 self.to_screen(u'[download] The file has already been downloaded')
360 def report_unable_to_resume(self):
361 """Report it was impossible to resume download."""
362 self.to_screen(u'[download] Unable to resume')
364 def report_finish(self):
365 """Report download finished."""
366 if self.params.get('noprogress', False):
367 self.to_screen(u'[download] Download completed')
371 def increment_downloads(self):
372 """Increment the ordinal that assigns a number to each file."""
373 self._num_downloads += 1
375 def prepare_filename(self, info_dict):
376 """Generate the output filename."""
378 template_dict = dict(info_dict)
380 template_dict['epoch'] = int(time.time())
381 template_dict['autonumber'] = u'%05d' % self._num_downloads
383 sanitize = lambda k,v: sanitize_filename(
384 u'NA' if v is None else compat_str(v),
385 restricted=self.params.get('restrictfilenames'),
387 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
389 filename = self.params['outtmpl'] % template_dict
391 except KeyError as err:
392 self.trouble(u'ERROR: Erroneous output template')
394 except ValueError as err:
395 self.trouble(u'ERROR: Insufficient system charset ' + repr(preferredencoding()))
398 def _match_entry(self, info_dict):
399 """ Returns None iff the file should be downloaded """
401 title = info_dict['title']
402 matchtitle = self.params.get('matchtitle', False)
404 if not re.search(matchtitle, title, re.IGNORECASE):
405 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
406 rejecttitle = self.params.get('rejecttitle', False)
408 if re.search(rejecttitle, title, re.IGNORECASE):
409 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
412 def process_info(self, info_dict):
413 """Process a single dictionary returned by an InfoExtractor."""
415 # Keep for backwards compatibility
416 info_dict['stitle'] = info_dict['title']
418 if not 'format' in info_dict:
419 info_dict['format'] = info_dict['ext']
421 reason = self._match_entry(info_dict)
422 if reason is not None:
423 self.to_screen(u'[download] ' + reason)
426 max_downloads = self.params.get('max_downloads')
427 if max_downloads is not None:
428 if self._num_downloads > int(max_downloads):
429 raise MaxDownloadsReached()
431 filename = self.prepare_filename(info_dict)
434 if self.params.get('forcetitle', False):
435 compat_print(info_dict['title'])
436 if self.params.get('forceurl', False):
437 compat_print(info_dict['url'])
438 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
439 compat_print(info_dict['thumbnail'])
440 if self.params.get('forcedescription', False) and 'description' in info_dict:
441 compat_print(info_dict['description'])
442 if self.params.get('forcefilename', False) and filename is not None:
443 compat_print(filename)
444 if self.params.get('forceformat', False):
445 compat_print(info_dict['format'])
447 # Do nothing else if in simulate mode
448 if self.params.get('simulate', False):
455 dn = os.path.dirname(encodeFilename(filename))
456 if dn != '' and not os.path.exists(dn): # dn is already encoded
458 except (OSError, IOError) as err:
459 self.report_error(u'unable to create directory ' + compat_str(err))
462 if self.params.get('writedescription', False):
464 descfn = filename + u'.description'
465 self.report_writedescription(descfn)
466 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
467 descfile.write(info_dict['description'])
468 except (OSError, IOError):
469 self.report_error(u'Cannot write description file ' + descfn)
472 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
473 # subtitles download errors are already managed as troubles in relevant IE
474 # that way it will silently go on when used with unsupporting IE
475 subtitle = info_dict['subtitles'][0]
476 (sub_error, sub_lang, sub) = subtitle
477 sub_format = self.params.get('subtitlesformat')
479 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
480 self.report_writesubtitles(sub_filename)
481 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
483 except (OSError, IOError):
484 self.report_error(u'Cannot write subtitles file ' + descfn)
486 if self.params.get('onlysubtitles', False):
489 if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
490 subtitles = info_dict['subtitles']
491 sub_format = self.params.get('subtitlesformat')
492 for subtitle in subtitles:
493 (sub_error, sub_lang, sub) = subtitle
495 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
496 self.report_writesubtitles(sub_filename)
497 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
499 except (OSError, IOError):
500 self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
502 if self.params.get('onlysubtitles', False):
505 if self.params.get('writeinfojson', False):
506 infofn = filename + u'.info.json'
507 self.report_writeinfojson(infofn)
509 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
510 write_json_file(json_info_dict, encodeFilename(infofn))
511 except (OSError, IOError):
512 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
515 if not self.params.get('skip_download', False):
516 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
520 success = self._do_download(filename, info_dict)
521 except (OSError, IOError) as err:
522 raise UnavailableVideoError()
523 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
524 self.report_error(u'unable to download video data: %s' % str(err))
526 except (ContentTooShortError, ) as err:
527 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
532 self.post_process(filename, info_dict)
533 except (PostProcessingError) as err:
534 self.report_error(u'postprocessing: %s' % str(err))
537 def download(self, url_list):
538 """Download a given list of URLs."""
539 if len(url_list) > 1 and self.fixed_template():
540 raise SameFileError(self.params['outtmpl'])
543 suitable_found = False
545 # Go to next InfoExtractor if not suitable
546 if not ie.suitable(url):
549 # Warn if the _WORKING attribute is False
551 self.report_warning(u'the program functionality for this site has been marked as broken, '
552 u'and will probably not work. If you want to go on, use the -i option.')
554 # Suitable InfoExtractor found
555 suitable_found = True
557 # Extract information from URL and process it
559 videos = ie.extract(url)
560 except ExtractorError as de: # An error we somewhat expected
561 self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
563 except MaxDownloadsReached:
564 self.to_screen(u'[info] Maximum number of downloaded files reached.')
566 except Exception as e:
567 if self.params.get('ignoreerrors', False):
568 self.report_error(u'' + compat_str(e), tb=compat_str(traceback.format_exc()))
573 if len(videos or []) > 1 and self.fixed_template():
574 raise SameFileError(self.params['outtmpl'])
576 for video in videos or []:
577 video['extractor'] = ie.IE_NAME
579 self.increment_downloads()
580 self.process_info(video)
581 except UnavailableVideoError:
582 self.to_stderr(u"\n")
583 self.report_error(u'unable to download video')
585 # Suitable InfoExtractor had been found; go to next URL
588 if not suitable_found:
589 self.report_error(u'no suitable InfoExtractor: %s' % url)
591 return self._download_retcode
593 def post_process(self, filename, ie_info):
594 """Run all the postprocessors on the given file."""
596 info['filepath'] = filename
600 keep_video_wish,new_info = pp.run(info)
601 if keep_video_wish is not None:
603 keep_video = keep_video_wish
604 elif keep_video is None:
605 # No clear decision yet, let IE decide
606 keep_video = keep_video_wish
607 except PostProcessingError as e:
608 self.to_stderr(u'ERROR: ' + e.msg)
609 if keep_video is False and not self.params.get('keepvideo', False):
611 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
612 os.remove(encodeFilename(filename))
613 except (IOError, OSError):
614 self.report_warning(u'Unable to remove downloaded video file')
616 def _download_with_rtmpdump(self, filename, url, player_url, page_url):
617 self.report_destination(filename)
618 tmpfilename = self.temp_name(filename)
620 # Check for rtmpdump first
622 subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
623 except (OSError, IOError):
624 self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
627 # Download using rtmpdump. rtmpdump returns exit code 2 when
628 # the connection was interrumpted and resuming appears to be
629 # possible. This is part of rtmpdump's normal usage, AFAIK.
630 basic_args = ['rtmpdump', '-q', '-r', url, '-o', tmpfilename]
631 if player_url is not None:
632 basic_args += ['-W', player_url]
633 if page_url is not None:
634 basic_args += ['--pageUrl', page_url]
635 args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
636 if self.params.get('verbose', False):
639 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
642 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
643 retval = subprocess.call(args)
644 while retval == 2 or retval == 1:
645 prevsize = os.path.getsize(encodeFilename(tmpfilename))
646 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
647 time.sleep(5.0) # This seems to be needed
648 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
649 cursize = os.path.getsize(encodeFilename(tmpfilename))
650 if prevsize == cursize and retval == 1:
652 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
653 if prevsize == cursize and retval == 2 and cursize > 1024:
654 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
658 fsize = os.path.getsize(encodeFilename(tmpfilename))
659 self.to_screen(u'\r[rtmpdump] %s bytes' % fsize)
660 self.try_rename(tmpfilename, filename)
661 self._hook_progress({
662 'downloaded_bytes': fsize,
663 'total_bytes': fsize,
664 'filename': filename,
665 'status': 'finished',
669 self.to_stderr(u"\n")
670 self.report_error(u'rtmpdump exited with code %d' % retval)
673 def _do_download(self, filename, info_dict):
674 url = info_dict['url']
676 # Check file already present
677 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
678 self.report_file_already_downloaded(filename)
679 self._hook_progress({
680 'filename': filename,
681 'status': 'finished',
685 # Attempt to download using rtmpdump
686 if url.startswith('rtmp'):
687 return self._download_with_rtmpdump(filename, url,
688 info_dict.get('player_url', None),
689 info_dict.get('page_url', None))
691 tmpfilename = self.temp_name(filename)
694 # Do not include the Accept-Encoding header
695 headers = {'Youtubedl-no-compression': 'True'}
696 if 'user_agent' in info_dict:
697 headers['Youtubedl-user-agent'] = info_dict['user_agent']
698 basic_request = compat_urllib_request.Request(url, None, headers)
699 request = compat_urllib_request.Request(url, None, headers)
701 if self.params.get('test', False):
702 request.add_header('Range','bytes=0-10240')
704 # Establish possible resume length
705 if os.path.isfile(encodeFilename(tmpfilename)):
706 resume_len = os.path.getsize(encodeFilename(tmpfilename))
712 if self.params.get('continuedl', False):
713 self.report_resuming_byte(resume_len)
714 request.add_header('Range','bytes=%d-' % resume_len)
720 retries = self.params.get('retries', 0)
721 while count <= retries:
722 # Establish connection
724 if count == 0 and 'urlhandle' in info_dict:
725 data = info_dict['urlhandle']
726 data = compat_urllib_request.urlopen(request)
728 except (compat_urllib_error.HTTPError, ) as err:
729 if (err.code < 500 or err.code >= 600) and err.code != 416:
730 # Unexpected HTTP error
732 elif err.code == 416:
733 # Unable to resume (requested range not satisfiable)
735 # Open the connection again without the range header
736 data = compat_urllib_request.urlopen(basic_request)
737 content_length = data.info()['Content-Length']
738 except (compat_urllib_error.HTTPError, ) as err:
739 if err.code < 500 or err.code >= 600:
742 # Examine the reported length
743 if (content_length is not None and
744 (resume_len - 100 < int(content_length) < resume_len + 100)):
745 # The file had already been fully downloaded.
746 # Explanation to the above condition: in issue #175 it was revealed that
747 # YouTube sometimes adds or removes a few bytes from the end of the file,
748 # changing the file size slightly and causing problems for some users. So
749 # I decided to implement a suggested change and consider the file
750 # completely downloaded if the file size differs less than 100 bytes from
751 # the one in the hard drive.
752 self.report_file_already_downloaded(filename)
753 self.try_rename(tmpfilename, filename)
754 self._hook_progress({
755 'filename': filename,
756 'status': 'finished',
760 # The length does not match, we start the download over
761 self.report_unable_to_resume()
767 self.report_retry(count, retries)
770 self.report_error(u'giving up after %s retries' % retries)
773 data_len = data.info().get('Content-length', None)
774 if data_len is not None:
775 data_len = int(data_len) + resume_len
776 min_data_len = self.params.get("min_filesize", None)
777 max_data_len = self.params.get("max_filesize", None)
778 if min_data_len is not None and data_len < min_data_len:
779 self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
781 if max_data_len is not None and data_len > max_data_len:
782 self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
785 data_len_str = self.format_bytes(data_len)
786 byte_counter = 0 + resume_len
787 block_size = self.params.get('buffersize', 1024)
792 data_block = data.read(block_size)
794 if len(data_block) == 0:
796 byte_counter += len(data_block)
798 # Open file just in time
801 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
802 assert stream is not None
803 filename = self.undo_temp_name(tmpfilename)
804 self.report_destination(filename)
805 except (OSError, IOError) as err:
806 self.report_error(u'unable to open for writing: %s' % str(err))
809 stream.write(data_block)
810 except (IOError, OSError) as err:
811 self.to_stderr(u"\n")
812 self.report_error(u'unable to write data: %s' % str(err))
814 if not self.params.get('noresizebuffer', False):
815 block_size = self.best_block_size(after - before, len(data_block))
818 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
820 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
822 percent_str = self.calc_percent(byte_counter, data_len)
823 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
824 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
826 self._hook_progress({
827 'downloaded_bytes': byte_counter,
828 'total_bytes': data_len,
829 'tmpfilename': tmpfilename,
830 'filename': filename,
831 'status': 'downloading',
835 self.slow_down(start, byte_counter - resume_len)
838 self.to_stderr(u"\n")
839 self.report_error(u'Did not get any data blocks')
843 if data_len is not None and byte_counter != data_len:
844 raise ContentTooShortError(byte_counter, int(data_len))
845 self.try_rename(tmpfilename, filename)
847 # Update file modification time
848 if self.params.get('updatetime', True):
849 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
851 self._hook_progress({
852 'downloaded_bytes': byte_counter,
853 'total_bytes': byte_counter,
854 'filename': filename,
855 'status': 'finished',
860 def _hook_progress(self, status):
861 for ph in self._progress_hooks:
864 def add_progress_hook(self, ph):
865 """ ph gets called on download progress, with a dictionary with the entries
866 * filename: The final filename
867 * status: One of "downloading" and "finished"
869 It can also have some of the following entries:
871 * downloaded_bytes: Bytes on disks
872 * total_bytes: Total bytes, None if unknown
873 * tmpfilename: The filename we're currently writing to
875 Hooks are guaranteed to be called at least once (with status "finished")
876 if the download is successful.
878 self._progress_hooks.append(ph)