]> jfr.im git - yt-dlp.git/blob - youtube_dl/FileDownloader.py
Merge remote-tracking branch 'jefftimesten/master'
[yt-dlp.git] / youtube_dl / FileDownloader.py
1 #!/usr/bin/env python
2 # -*- coding: utf-8 -*-
3
4 from __future__ import absolute_import
5
6 import math
7 import io
8 import os
9 import re
10 import socket
11 import subprocess
12 import sys
13 import time
14 import traceback
15
16 if os.name == 'nt':
17 import ctypes
18
19 from .utils import *
20
21
22 class FileDownloader(object):
23 """File Downloader class.
24
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.
31
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.
39
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".
46
47 Available options:
48
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 .srt file
82 subtitleslang: Language of the subtitles to download
83 test: Download only first bytes to test the downloader.
84 keepvideo: Keep the video file after post-processing
85 """
86
87 params = None
88 _ies = []
89 _pps = []
90 _download_retcode = None
91 _num_downloads = None
92 _screen_file = None
93
94 def __init__(self, params):
95 """Create a FileDownloader object with the given options."""
96 self._ies = []
97 self._pps = []
98 self._download_retcode = 0
99 self._num_downloads = 0
100 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
101 self.params = params
102
103 if '%(stitle)s' in self.params['outtmpl']:
104 self.to_stderr(u'WARNING: %(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.')
105
106 @staticmethod
107 def format_bytes(bytes):
108 if bytes is None:
109 return 'N/A'
110 if type(bytes) is str:
111 bytes = float(bytes)
112 if bytes == 0.0:
113 exponent = 0
114 else:
115 exponent = int(math.log(bytes, 1024.0))
116 suffix = 'bkMGTPEZY'[exponent]
117 converted = float(bytes) / float(1024 ** exponent)
118 return '%.2f%s' % (converted, suffix)
119
120 @staticmethod
121 def calc_percent(byte_counter, data_len):
122 if data_len is None:
123 return '---.-%'
124 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
125
126 @staticmethod
127 def calc_eta(start, now, total, current):
128 if total is None:
129 return '--:--'
130 dif = now - start
131 if current == 0 or dif < 0.001: # One millisecond
132 return '--:--'
133 rate = float(current) / dif
134 eta = int((float(total) - float(current)) / rate)
135 (eta_mins, eta_secs) = divmod(eta, 60)
136 if eta_mins > 99:
137 return '--:--'
138 return '%02d:%02d' % (eta_mins, eta_secs)
139
140 @staticmethod
141 def calc_speed(start, now, bytes):
142 dif = now - start
143 if bytes == 0 or dif < 0.001: # One millisecond
144 return '%10s' % '---b/s'
145 return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif))
146
147 @staticmethod
148 def best_block_size(elapsed_time, bytes):
149 new_min = max(bytes / 2.0, 1.0)
150 new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
151 if elapsed_time < 0.001:
152 return int(new_max)
153 rate = bytes / elapsed_time
154 if rate > new_max:
155 return int(new_max)
156 if rate < new_min:
157 return int(new_min)
158 return int(rate)
159
160 @staticmethod
161 def parse_bytes(bytestr):
162 """Parse a string indicating a byte quantity into an integer."""
163 matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
164 if matchobj is None:
165 return None
166 number = float(matchobj.group(1))
167 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
168 return int(round(number * multiplier))
169
170 def add_info_extractor(self, ie):
171 """Add an InfoExtractor object to the end of the list."""
172 self._ies.append(ie)
173 ie.set_downloader(self)
174
175 def add_post_processor(self, pp):
176 """Add a PostProcessor object to the end of the chain."""
177 self._pps.append(pp)
178 pp.set_downloader(self)
179
180 def to_screen(self, message, skip_eol=False):
181 """Print message to stdout if not in quiet mode."""
182 assert type(message) == type(u'')
183 if not self.params.get('quiet', False):
184 terminator = [u'\n', u''][skip_eol]
185 output = message + terminator
186 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
187 output = output.encode(preferredencoding(), 'ignore')
188 self._screen_file.write(output)
189 self._screen_file.flush()
190
191 def to_stderr(self, message):
192 """Print message to stderr."""
193 assert type(message) == type(u'')
194 output = message + u'\n'
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())
197 sys.stderr.write(output)
198
199 def to_cons_title(self, message):
200 """Set console/terminal window title to message."""
201 if not self.params.get('consoletitle', False):
202 return
203 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
204 # c_wchar_p() might not be necessary if `message` is
205 # already of type unicode()
206 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
207 elif 'TERM' in os.environ:
208 sys.stderr.write('\033]0;%s\007' % message.encode(preferredencoding()))
209
210 def fixed_template(self):
211 """Checks if the output template is fixed."""
212 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
213
214 def trouble(self, message=None, tb=None):
215 """Determine action to take when a download problem appears.
216
217 Depending on if the downloader has been configured to ignore
218 download errors or not, this method may throw an exception or
219 not when errors are found, after printing the message.
220
221 tb, if given, is additional traceback information.
222 """
223 if message is not None:
224 self.to_stderr(message)
225 if self.params.get('verbose'):
226 if tb is None:
227 tb_data = traceback.format_list(traceback.extract_stack())
228 tb = u''.join(tb_data)
229 self.to_stderr(tb)
230 if not self.params.get('ignoreerrors', False):
231 raise DownloadError(message)
232 self._download_retcode = 1
233
234 def slow_down(self, start_time, byte_counter):
235 """Sleep if the download speed is over the rate limit."""
236 rate_limit = self.params.get('ratelimit', None)
237 if rate_limit is None or byte_counter == 0:
238 return
239 now = time.time()
240 elapsed = now - start_time
241 if elapsed <= 0.0:
242 return
243 speed = float(byte_counter) / elapsed
244 if speed > rate_limit:
245 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
246
247 def temp_name(self, filename):
248 """Returns a temporary filename for the given filename."""
249 if self.params.get('nopart', False) or filename == u'-' or \
250 (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
251 return filename
252 return filename + u'.part'
253
254 def undo_temp_name(self, filename):
255 if filename.endswith(u'.part'):
256 return filename[:-len(u'.part')]
257 return filename
258
259 def try_rename(self, old_filename, new_filename):
260 try:
261 if old_filename == new_filename:
262 return
263 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
264 except (IOError, OSError) as err:
265 self.trouble(u'ERROR: unable to rename file')
266
267 def try_utime(self, filename, last_modified_hdr):
268 """Try to set the last-modified time of the given file."""
269 if last_modified_hdr is None:
270 return
271 if not os.path.isfile(encodeFilename(filename)):
272 return
273 timestr = last_modified_hdr
274 if timestr is None:
275 return
276 filetime = timeconvert(timestr)
277 if filetime is None:
278 return filetime
279 try:
280 os.utime(filename, (time.time(), filetime))
281 except:
282 pass
283 return filetime
284
285 def report_writedescription(self, descfn):
286 """ Report that the description file is being written """
287 self.to_screen(u'[info] Writing video description to: ' + descfn)
288
289 def report_writesubtitles(self, srtfn):
290 """ Report that the subtitles file is being written """
291 self.to_screen(u'[info] Writing video subtitles to: ' + srtfn)
292
293 def report_writeinfojson(self, infofn):
294 """ Report that the metadata file has been written """
295 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
296
297 def report_destination(self, filename):
298 """Report destination filename."""
299 self.to_screen(u'[download] Destination: ' + filename)
300
301 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
302 """Report download progress."""
303 if self.params.get('noprogress', False):
304 return
305 self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
306 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
307 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
308 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
309
310 def report_resuming_byte(self, resume_len):
311 """Report attempt to resume at given byte."""
312 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
313
314 def report_retry(self, count, retries):
315 """Report retry in case of HTTP error 5xx"""
316 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
317
318 def report_file_already_downloaded(self, file_name):
319 """Report file has already been fully downloaded."""
320 try:
321 self.to_screen(u'[download] %s has already been downloaded' % file_name)
322 except (UnicodeEncodeError) as err:
323 self.to_screen(u'[download] The file has already been downloaded')
324
325 def report_unable_to_resume(self):
326 """Report it was impossible to resume download."""
327 self.to_screen(u'[download] Unable to resume')
328
329 def report_finish(self):
330 """Report download finished."""
331 if self.params.get('noprogress', False):
332 self.to_screen(u'[download] Download completed')
333 else:
334 self.to_screen(u'')
335
336 def increment_downloads(self):
337 """Increment the ordinal that assigns a number to each file."""
338 self._num_downloads += 1
339
340 def prepare_filename(self, info_dict):
341 """Generate the output filename."""
342 try:
343 template_dict = dict(info_dict)
344
345 template_dict['epoch'] = int(time.time())
346 template_dict['autonumber'] = u'%05d' % self._num_downloads
347
348 sanitize = lambda k,v: sanitize_filename(
349 u'NA' if v is None else compat_str(v),
350 restricted=self.params.get('restrictfilenames'),
351 is_id=(k==u'id'))
352 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
353
354 filename = self.params['outtmpl'] % template_dict
355 return filename
356 except (ValueError, KeyError) as err:
357 self.trouble(u'ERROR: invalid system charset or erroneous output template')
358 return None
359
360 def _match_entry(self, info_dict):
361 """ Returns None iff the file should be downloaded """
362
363 title = info_dict['title']
364 matchtitle = self.params.get('matchtitle', False)
365 if matchtitle:
366 matchtitle = matchtitle.decode('utf8')
367 if not re.search(matchtitle, title, re.IGNORECASE):
368 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
369 rejecttitle = self.params.get('rejecttitle', False)
370 if rejecttitle:
371 rejecttitle = rejecttitle.decode('utf8')
372 if re.search(rejecttitle, title, re.IGNORECASE):
373 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
374 return None
375
376 def process_info(self, info_dict):
377 """Process a single dictionary returned by an InfoExtractor."""
378
379 # Keep for backwards compatibility
380 info_dict['stitle'] = info_dict['title']
381
382 if not 'format' in info_dict:
383 info_dict['format'] = info_dict['ext']
384
385 reason = self._match_entry(info_dict)
386 if reason is not None:
387 self.to_screen(u'[download] ' + reason)
388 return
389
390 max_downloads = self.params.get('max_downloads')
391 if max_downloads is not None:
392 if self._num_downloads > int(max_downloads):
393 raise MaxDownloadsReached()
394
395 filename = self.prepare_filename(info_dict)
396
397 # Forced printings
398 if self.params.get('forcetitle', False):
399 compat_print(info_dict['title'])
400 if self.params.get('forceurl', False):
401 compat_print(info_dict['url'])
402 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
403 compat_print(info_dict['thumbnail'])
404 if self.params.get('forcedescription', False) and 'description' in info_dict:
405 compat_print(info_dict['description'])
406 if self.params.get('forcefilename', False) and filename is not None:
407 compat_print(filename)
408 if self.params.get('forceformat', False):
409 compat_print(info_dict['format'])
410
411 # Do nothing else if in simulate mode
412 if self.params.get('simulate', False):
413 return
414
415 if filename is None:
416 return
417
418 try:
419 dn = os.path.dirname(encodeFilename(filename))
420 if dn != '' and not os.path.exists(dn): # dn is already encoded
421 os.makedirs(dn)
422 except (OSError, IOError) as err:
423 self.trouble(u'ERROR: unable to create directory ' + compat_str(err))
424 return
425
426 if self.params.get('writedescription', False):
427 try:
428 descfn = filename + u'.description'
429 self.report_writedescription(descfn)
430 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
431 descfile.write(info_dict['description'])
432 except (OSError, IOError):
433 self.trouble(u'ERROR: Cannot write description file ' + descfn)
434 return
435
436 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
437 # subtitles download errors are already managed as troubles in relevant IE
438 # that way it will silently go on when used with unsupporting IE
439 try:
440 srtfn = filename.rsplit('.', 1)[0] + u'.srt'
441 self.report_writesubtitles(srtfn)
442 with io.open(encodeFilename(srtfn), 'w', encoding='utf-8') as srtfile:
443 srtfile.write(info_dict['subtitles'])
444 except (OSError, IOError):
445 self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
446 return
447
448 if self.params.get('writeinfojson', False):
449 infofn = filename + u'.info.json'
450 self.report_writeinfojson(infofn)
451 try:
452 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
453 write_json_file(json_info_dict, encodeFilename(infofn))
454 except (OSError, IOError):
455 self.trouble(u'ERROR: Cannot write metadata to JSON file ' + infofn)
456 return
457
458 if not self.params.get('skip_download', False):
459 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
460 success = True
461 else:
462 try:
463 success = self._do_download(filename, info_dict)
464 except (OSError, IOError) as err:
465 raise UnavailableVideoError()
466 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
467 self.trouble(u'ERROR: unable to download video data: %s' % str(err))
468 return
469 except (ContentTooShortError, ) as err:
470 self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
471 return
472
473 if success:
474 try:
475 self.post_process(filename, info_dict)
476 except (PostProcessingError) as err:
477 self.trouble(u'ERROR: postprocessing: %s' % str(err))
478 return
479
480 def download(self, url_list):
481 """Download a given list of URLs."""
482 if len(url_list) > 1 and self.fixed_template():
483 raise SameFileError(self.params['outtmpl'])
484
485 for url in url_list:
486 suitable_found = False
487 for ie in self._ies:
488 # Go to next InfoExtractor if not suitable
489 if not ie.suitable(url):
490 continue
491
492 # Warn if the _WORKING attribute is False
493 if not ie.working():
494 self.to_stderr(u'WARNING: the program functionality for this site has been marked as broken, '
495 u'and will probably not work. If you want to go on, use the -i option.')
496
497 # Suitable InfoExtractor found
498 suitable_found = True
499
500 # Extract information from URL and process it
501 try:
502 videos = ie.extract(url)
503 except ExtractorError as de: # An error we somewhat expected
504 self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
505 break
506 except Exception as e:
507 if self.params.get('ignoreerrors', False):
508 self.trouble(u'ERROR: ' + compat_str(e), tb=compat_str(traceback.format_exc()))
509 break
510 else:
511 raise
512
513 if len(videos or []) > 1 and self.fixed_template():
514 raise SameFileError(self.params['outtmpl'])
515
516 for video in videos or []:
517 video['extractor'] = ie.IE_NAME
518 try:
519 self.increment_downloads()
520 self.process_info(video)
521 except UnavailableVideoError:
522 self.trouble(u'\nERROR: unable to download video')
523
524 # Suitable InfoExtractor had been found; go to next URL
525 break
526
527 if not suitable_found:
528 self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
529
530 return self._download_retcode
531
532 def post_process(self, filename, ie_info):
533 """Run all the postprocessors on the given file."""
534 info = dict(ie_info)
535 info['filepath'] = filename
536 keep_video = None
537 for pp in self._pps:
538 try:
539 keep_video_wish,new_info = pp.run(info)
540 if keep_video_wish is not None:
541 if keep_video_wish:
542 keep_video = keep_video_wish
543 elif keep_video is None:
544 # No clear decision yet, let IE decide
545 keep_video = keep_video_wish
546 except PostProcessingError as e:
547 self.to_stderr(u'ERROR: ' + e.msg)
548 if keep_video is False and not self.params.get('keepvideo', False):
549 try:
550 self.to_stderr(u'Deleting original file %s (pass -k to keep)' % filename)
551 os.remove(encodeFilename(filename))
552 except (IOError, OSError):
553 self.to_stderr(u'WARNING: Unable to remove downloaded video file')
554
555 def _download_with_rtmpdump(self, filename, url, player_url, page_url):
556 self.report_destination(filename)
557 tmpfilename = self.temp_name(filename)
558
559 # Check for rtmpdump first
560 try:
561 subprocess.call(['rtmpdump', '-h'], stdout=(file(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
562 except (OSError, IOError):
563 self.trouble(u'ERROR: RTMP download detected but "rtmpdump" could not be run')
564 return False
565
566 # Download using rtmpdump. rtmpdump returns exit code 2 when
567 # the connection was interrumpted and resuming appears to be
568 # possible. This is part of rtmpdump's normal usage, AFAIK.
569 basic_args = ['rtmpdump', '-q', '-r', url, '-o', tmpfilename]
570 if player_url is not None:
571 basic_args += ['-W', player_url]
572 if page_url is not None:
573 basic_args += ['--pageUrl', page_url]
574 args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
575 if self.params.get('verbose', False):
576 try:
577 import pipes
578 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
579 except ImportError:
580 shell_quote = repr
581 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
582 retval = subprocess.call(args)
583 while retval == 2 or retval == 1:
584 prevsize = os.path.getsize(encodeFilename(tmpfilename))
585 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
586 time.sleep(5.0) # This seems to be needed
587 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
588 cursize = os.path.getsize(encodeFilename(tmpfilename))
589 if prevsize == cursize and retval == 1:
590 break
591 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
592 if prevsize == cursize and retval == 2 and cursize > 1024:
593 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
594 retval = 0
595 break
596 if retval == 0:
597 self.to_screen(u'\r[rtmpdump] %s bytes' % os.path.getsize(encodeFilename(tmpfilename)))
598 self.try_rename(tmpfilename, filename)
599 return True
600 else:
601 self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval)
602 return False
603
604 def _do_download(self, filename, info_dict):
605 url = info_dict['url']
606
607 # Check file already present
608 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
609 self.report_file_already_downloaded(filename)
610 return True
611
612 # Attempt to download using rtmpdump
613 if url.startswith('rtmp'):
614 return self._download_with_rtmpdump(filename, url,
615 info_dict.get('player_url', None),
616 info_dict.get('page_url', None))
617
618 tmpfilename = self.temp_name(filename)
619 stream = None
620
621 # Do not include the Accept-Encoding header
622 headers = {'Youtubedl-no-compression': 'True'}
623 basic_request = compat_urllib_request.Request(url, None, headers)
624 request = compat_urllib_request.Request(url, None, headers)
625
626 if self.params.get('test', False):
627 request.add_header('Range','bytes=0-10240')
628
629 # Establish possible resume length
630 if os.path.isfile(encodeFilename(tmpfilename)):
631 resume_len = os.path.getsize(encodeFilename(tmpfilename))
632 else:
633 resume_len = 0
634
635 open_mode = 'wb'
636 if resume_len != 0:
637 if self.params.get('continuedl', False):
638 self.report_resuming_byte(resume_len)
639 request.add_header('Range','bytes=%d-' % resume_len)
640 open_mode = 'ab'
641 else:
642 resume_len = 0
643
644 count = 0
645 retries = self.params.get('retries', 0)
646 while count <= retries:
647 # Establish connection
648 try:
649 if count == 0 and 'urlhandle' in info_dict:
650 data = info_dict['urlhandle']
651 data = compat_urllib_request.urlopen(request)
652 break
653 except (compat_urllib_error.HTTPError, ) as err:
654 if (err.code < 500 or err.code >= 600) and err.code != 416:
655 # Unexpected HTTP error
656 raise
657 elif err.code == 416:
658 # Unable to resume (requested range not satisfiable)
659 try:
660 # Open the connection again without the range header
661 data = compat_urllib_request.urlopen(basic_request)
662 content_length = data.info()['Content-Length']
663 except (compat_urllib_error.HTTPError, ) as err:
664 if err.code < 500 or err.code >= 600:
665 raise
666 else:
667 # Examine the reported length
668 if (content_length is not None and
669 (resume_len - 100 < int(content_length) < resume_len + 100)):
670 # The file had already been fully downloaded.
671 # Explanation to the above condition: in issue #175 it was revealed that
672 # YouTube sometimes adds or removes a few bytes from the end of the file,
673 # changing the file size slightly and causing problems for some users. So
674 # I decided to implement a suggested change and consider the file
675 # completely downloaded if the file size differs less than 100 bytes from
676 # the one in the hard drive.
677 self.report_file_already_downloaded(filename)
678 self.try_rename(tmpfilename, filename)
679 return True
680 else:
681 # The length does not match, we start the download over
682 self.report_unable_to_resume()
683 open_mode = 'wb'
684 break
685 # Retry
686 count += 1
687 if count <= retries:
688 self.report_retry(count, retries)
689
690 if count > retries:
691 self.trouble(u'ERROR: giving up after %s retries' % retries)
692 return False
693
694 data_len = data.info().get('Content-length', None)
695 if data_len is not None:
696 data_len = int(data_len) + resume_len
697 data_len_str = self.format_bytes(data_len)
698 byte_counter = 0 + resume_len
699 block_size = self.params.get('buffersize', 1024)
700 start = time.time()
701 while True:
702 # Download and write
703 before = time.time()
704 data_block = data.read(block_size)
705 after = time.time()
706 if len(data_block) == 0:
707 break
708 byte_counter += len(data_block)
709
710 # Open file just in time
711 if stream is None:
712 try:
713 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
714 assert stream is not None
715 filename = self.undo_temp_name(tmpfilename)
716 self.report_destination(filename)
717 except (OSError, IOError) as err:
718 self.trouble(u'ERROR: unable to open for writing: %s' % str(err))
719 return False
720 try:
721 stream.write(data_block)
722 except (IOError, OSError) as err:
723 self.trouble(u'\nERROR: unable to write data: %s' % str(err))
724 return False
725 if not self.params.get('noresizebuffer', False):
726 block_size = self.best_block_size(after - before, len(data_block))
727
728 # Progress message
729 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
730 if data_len is None:
731 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
732 else:
733 percent_str = self.calc_percent(byte_counter, data_len)
734 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
735 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
736
737 # Apply rate limit
738 self.slow_down(start, byte_counter - resume_len)
739
740 if stream is None:
741 self.trouble(u'\nERROR: Did not get any data blocks')
742 return False
743 stream.close()
744 self.report_finish()
745 if data_len is not None and byte_counter != data_len:
746 raise ContentTooShortError(byte_counter, int(data_len))
747 self.try_rename(tmpfilename, filename)
748
749 # Update file modification time
750 if self.params.get('updatetime', True):
751 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
752
753 return True