]> jfr.im git - yt-dlp.git/blame - youtube_dl/FileDownloader.py
add info_dict testing to test_download
[yt-dlp.git] / youtube_dl / FileDownloader.py
CommitLineData
d77c3dfd
FV
1#!/usr/bin/env python
2# -*- coding: utf-8 -*-
3
9e8056d5
PH
4from __future__ import absolute_import
5
d77c3dfd 6import math
ce4be3a9 7import io
d77c3dfd
FV
8import os
9import re
10import socket
11import subprocess
12import sys
13import time
59ce2019 14import traceback
d77c3dfd
FV
15
16if os.name == 'nt':
59ae15a5 17 import ctypes
3eec021a 18
9e8056d5 19from .utils import *
d77c3dfd
FV
20
21
22class FileDownloader(object):
59ae15a5
PH
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
37c8fd48 83 test: Download only first bytes to test the downloader.
59ae15a5
PH
84 """
85
86 params = None
87 _ies = []
88 _pps = []
89 _download_retcode = None
90 _num_downloads = None
91 _screen_file = None
92
93 def __init__(self, params):
94 """Create a FileDownloader object with the given options."""
95 self._ies = []
96 self._pps = []
97 self._download_retcode = 0
98 self._num_downloads = 0
99 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)]
100 self.params = params
101
102 if '%(stitle)s' in self.params['outtmpl']:
103 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.')
104
105 @staticmethod
106 def format_bytes(bytes):
107 if bytes is None:
108 return 'N/A'
109 if type(bytes) is str:
110 bytes = float(bytes)
111 if bytes == 0.0:
112 exponent = 0
113 else:
114 exponent = int(math.log(bytes, 1024.0))
115 suffix = 'bkMGTPEZY'[exponent]
116 converted = float(bytes) / float(1024 ** exponent)
117 return '%.2f%s' % (converted, suffix)
118
119 @staticmethod
120 def calc_percent(byte_counter, data_len):
121 if data_len is None:
122 return '---.-%'
123 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
124
125 @staticmethod
126 def calc_eta(start, now, total, current):
127 if total is None:
128 return '--:--'
129 dif = now - start
130 if current == 0 or dif < 0.001: # One millisecond
131 return '--:--'
132 rate = float(current) / dif
133 eta = int((float(total) - float(current)) / rate)
134 (eta_mins, eta_secs) = divmod(eta, 60)
135 if eta_mins > 99:
136 return '--:--'
137 return '%02d:%02d' % (eta_mins, eta_secs)
138
139 @staticmethod
140 def calc_speed(start, now, bytes):
141 dif = now - start
142 if bytes == 0 or dif < 0.001: # One millisecond
143 return '%10s' % '---b/s'
144 return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif))
145
146 @staticmethod
147 def best_block_size(elapsed_time, bytes):
148 new_min = max(bytes / 2.0, 1.0)
149 new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
150 if elapsed_time < 0.001:
151 return int(new_max)
152 rate = bytes / elapsed_time
153 if rate > new_max:
154 return int(new_max)
155 if rate < new_min:
156 return int(new_min)
157 return int(rate)
158
159 @staticmethod
160 def parse_bytes(bytestr):
161 """Parse a string indicating a byte quantity into an integer."""
162 matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr)
163 if matchobj is None:
164 return None
165 number = float(matchobj.group(1))
166 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
167 return int(round(number * multiplier))
168
169 def add_info_extractor(self, ie):
170 """Add an InfoExtractor object to the end of the list."""
171 self._ies.append(ie)
172 ie.set_downloader(self)
173
174 def add_post_processor(self, pp):
175 """Add a PostProcessor object to the end of the chain."""
176 self._pps.append(pp)
177 pp.set_downloader(self)
178
179 def to_screen(self, message, skip_eol=False):
180 """Print message to stdout if not in quiet mode."""
181 assert type(message) == type(u'')
182 if not self.params.get('quiet', False):
183 terminator = [u'\n', u''][skip_eol]
184 output = message + terminator
185 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
186 output = output.encode(preferredencoding(), 'ignore')
187 self._screen_file.write(output)
188 self._screen_file.flush()
189
190 def to_stderr(self, message):
191 """Print message to stderr."""
192 assert type(message) == type(u'')
193 output = message + u'\n'
194 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
195 output = output.encode(preferredencoding())
196 sys.stderr.write(output)
197
198 def to_cons_title(self, message):
199 """Set console/terminal window title to message."""
200 if not self.params.get('consoletitle', False):
201 return
202 if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
203 # c_wchar_p() might not be necessary if `message` is
204 # already of type unicode()
205 ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
206 elif 'TERM' in os.environ:
207 sys.stderr.write('\033]0;%s\007' % message.encode(preferredencoding()))
208
209 def fixed_template(self):
210 """Checks if the output template is fixed."""
211 return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None)
212
213 def trouble(self, message=None):
214 """Determine action to take when a download problem appears.
215
216 Depending on if the downloader has been configured to ignore
217 download errors or not, this method may throw an exception or
218 not when errors are found, after printing the message.
219 """
220 if message is not None:
221 self.to_stderr(message)
59ce2019
FV
222 if self.params.get('verbose'):
223 self.to_stderr(u''.join(traceback.format_list(traceback.extract_stack())))
59ae15a5
PH
224 if not self.params.get('ignoreerrors', False):
225 raise DownloadError(message)
226 self._download_retcode = 1
227
228 def slow_down(self, start_time, byte_counter):
229 """Sleep if the download speed is over the rate limit."""
230 rate_limit = self.params.get('ratelimit', None)
231 if rate_limit is None or byte_counter == 0:
232 return
233 now = time.time()
234 elapsed = now - start_time
235 if elapsed <= 0.0:
236 return
237 speed = float(byte_counter) / elapsed
238 if speed > rate_limit:
239 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
240
241 def temp_name(self, filename):
242 """Returns a temporary filename for the given filename."""
243 if self.params.get('nopart', False) or filename == u'-' or \
244 (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
245 return filename
246 return filename + u'.part'
247
248 def undo_temp_name(self, filename):
249 if filename.endswith(u'.part'):
250 return filename[:-len(u'.part')]
251 return filename
252
253 def try_rename(self, old_filename, new_filename):
254 try:
255 if old_filename == new_filename:
256 return
257 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
258 except (IOError, OSError) as err:
259 self.trouble(u'ERROR: unable to rename file')
260
261 def try_utime(self, filename, last_modified_hdr):
262 """Try to set the last-modified time of the given file."""
263 if last_modified_hdr is None:
264 return
265 if not os.path.isfile(encodeFilename(filename)):
266 return
267 timestr = last_modified_hdr
268 if timestr is None:
269 return
270 filetime = timeconvert(timestr)
271 if filetime is None:
272 return filetime
273 try:
274 os.utime(filename, (time.time(), filetime))
275 except:
276 pass
277 return filetime
278
279 def report_writedescription(self, descfn):
280 """ Report that the description file is being written """
281 self.to_screen(u'[info] Writing video description to: ' + descfn)
282
283 def report_writesubtitles(self, srtfn):
284 """ Report that the subtitles file is being written """
285 self.to_screen(u'[info] Writing video subtitles to: ' + srtfn)
286
287 def report_writeinfojson(self, infofn):
288 """ Report that the metadata file has been written """
289 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
290
291 def report_destination(self, filename):
292 """Report destination filename."""
293 self.to_screen(u'[download] Destination: ' + filename)
294
295 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
296 """Report download progress."""
297 if self.params.get('noprogress', False):
298 return
299 self.to_screen(u'\r[download] %s of %s at %s ETA %s' %
300 (percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
301 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
302 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
303
304 def report_resuming_byte(self, resume_len):
305 """Report attempt to resume at given byte."""
306 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
307
308 def report_retry(self, count, retries):
309 """Report retry in case of HTTP error 5xx"""
310 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
311
312 def report_file_already_downloaded(self, file_name):
313 """Report file has already been fully downloaded."""
314 try:
315 self.to_screen(u'[download] %s has already been downloaded' % file_name)
316 except (UnicodeEncodeError) as err:
317 self.to_screen(u'[download] The file has already been downloaded')
318
319 def report_unable_to_resume(self):
320 """Report it was impossible to resume download."""
321 self.to_screen(u'[download] Unable to resume')
322
323 def report_finish(self):
324 """Report download finished."""
325 if self.params.get('noprogress', False):
326 self.to_screen(u'[download] Download completed')
327 else:
328 self.to_screen(u'')
329
330 def increment_downloads(self):
331 """Increment the ordinal that assigns a number to each file."""
332 self._num_downloads += 1
333
334 def prepare_filename(self, info_dict):
335 """Generate the output filename."""
336 try:
337 template_dict = dict(info_dict)
338
339 template_dict['epoch'] = int(time.time())
340 template_dict['autonumber'] = u'%05d' % self._num_downloads
341
796173d0
PH
342 sanitize = lambda k,v: sanitize_filename(
343 u'NA' if v is None else compat_str(v),
344 restricted=self.params.get('restrictfilenames'),
345 is_id=(k==u'id'))
346 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
59ae15a5
PH
347
348 filename = self.params['outtmpl'] % template_dict
349 return filename
350 except (ValueError, KeyError) as err:
351 self.trouble(u'ERROR: invalid system charset or erroneous output template')
352 return None
353
354 def _match_entry(self, info_dict):
355 """ Returns None iff the file should be downloaded """
356
357 title = info_dict['title']
358 matchtitle = self.params.get('matchtitle', False)
359 if matchtitle:
360 matchtitle = matchtitle.decode('utf8')
361 if not re.search(matchtitle, title, re.IGNORECASE):
362 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
363 rejecttitle = self.params.get('rejecttitle', False)
364 if rejecttitle:
365 rejecttitle = rejecttitle.decode('utf8')
366 if re.search(rejecttitle, title, re.IGNORECASE):
367 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
368 return None
369
370 def process_info(self, info_dict):
371 """Process a single dictionary returned by an InfoExtractor."""
372
373 # Keep for backwards compatibility
374 info_dict['stitle'] = info_dict['title']
375
376 if not 'format' in info_dict:
377 info_dict['format'] = info_dict['ext']
378
379 reason = self._match_entry(info_dict)
380 if reason is not None:
381 self.to_screen(u'[download] ' + reason)
382 return
383
384 max_downloads = self.params.get('max_downloads')
385 if max_downloads is not None:
386 if self._num_downloads > int(max_downloads):
387 raise MaxDownloadsReached()
388
389 filename = self.prepare_filename(info_dict)
390
391 # Forced printings
392 if self.params.get('forcetitle', False):
393 compat_print(info_dict['title'])
394 if self.params.get('forceurl', False):
395 compat_print(info_dict['url'])
396 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
397 compat_print(info_dict['thumbnail'])
398 if self.params.get('forcedescription', False) and 'description' in info_dict:
399 compat_print(info_dict['description'])
400 if self.params.get('forcefilename', False) and filename is not None:
401 compat_print(filename)
402 if self.params.get('forceformat', False):
403 compat_print(info_dict['format'])
404
405 # Do nothing else if in simulate mode
406 if self.params.get('simulate', False):
407 return
408
409 if filename is None:
410 return
411
412 try:
413 dn = os.path.dirname(encodeFilename(filename))
414 if dn != '' and not os.path.exists(dn): # dn is already encoded
415 os.makedirs(dn)
416 except (OSError, IOError) as err:
417 self.trouble(u'ERROR: unable to create directory ' + compat_str(err))
418 return
419
420 if self.params.get('writedescription', False):
421 try:
422 descfn = filename + u'.description'
423 self.report_writedescription(descfn)
bfa6389b
PH
424 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
425 descfile.write(info_dict['description'])
59ae15a5
PH
426 except (OSError, IOError):
427 self.trouble(u'ERROR: Cannot write description file ' + descfn)
428 return
429
430 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
431 # subtitles download errors are already managed as troubles in relevant IE
432 # that way it will silently go on when used with unsupporting IE
433 try:
434 srtfn = filename.rsplit('.', 1)[0] + u'.srt'
435 self.report_writesubtitles(srtfn)
436 srtfile = open(encodeFilename(srtfn), 'wb')
437 try:
438 srtfile.write(info_dict['subtitles'].encode('utf-8'))
439 finally:
440 srtfile.close()
441 except (OSError, IOError):
442 self.trouble(u'ERROR: Cannot write subtitles file ' + descfn)
443 return
444
445 if self.params.get('writeinfojson', False):
446 infofn = filename + u'.info.json'
447 self.report_writeinfojson(infofn)
59ae15a5 448 try:
f4bfd65f
PH
449 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
450 write_json_file(json_info_dict, encodeFilename(infofn))
59ae15a5
PH
451 except (OSError, IOError):
452 self.trouble(u'ERROR: Cannot write metadata to JSON file ' + infofn)
453 return
454
455 if not self.params.get('skip_download', False):
456 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
457 success = True
458 else:
459 try:
460 success = self._do_download(filename, info_dict)
461 except (OSError, IOError) as err:
6ad98fb3 462 raise UnavailableVideoError()
59ae15a5
PH
463 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
464 self.trouble(u'ERROR: unable to download video data: %s' % str(err))
465 return
466 except (ContentTooShortError, ) as err:
467 self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
468 return
469
470 if success:
471 try:
472 self.post_process(filename, info_dict)
473 except (PostProcessingError) as err:
474 self.trouble(u'ERROR: postprocessing: %s' % str(err))
475 return
476
477 def download(self, url_list):
478 """Download a given list of URLs."""
479 if len(url_list) > 1 and self.fixed_template():
480 raise SameFileError(self.params['outtmpl'])
481
482 for url in url_list:
483 suitable_found = False
484 for ie in self._ies:
485 # Go to next InfoExtractor if not suitable
486 if not ie.suitable(url):
487 continue
488
489 # Warn if the _WORKING attribute is False
490 if not ie.working():
491 self.trouble(u'WARNING: the program functionality for this site has been marked as broken, '
492 u'and will probably not work. If you want to go on, use the -i option.')
493
494 # Suitable InfoExtractor found
495 suitable_found = True
496
497 # Extract information from URL and process it
498 videos = ie.extract(url)
499 for video in videos or []:
500 video['extractor'] = ie.IE_NAME
501 try:
502 self.increment_downloads()
503 self.process_info(video)
504 except UnavailableVideoError:
505 self.trouble(u'\nERROR: unable to download video')
506
507 # Suitable InfoExtractor had been found; go to next URL
508 break
509
510 if not suitable_found:
511 self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url)
512
513 return self._download_retcode
514
515 def post_process(self, filename, ie_info):
516 """Run the postprocessing chain on the given file."""
517 info = dict(ie_info)
518 info['filepath'] = filename
519 for pp in self._pps:
520 info = pp.run(info)
521 if info is None:
522 break
523
524 def _download_with_rtmpdump(self, filename, url, player_url):
525 self.report_destination(filename)
526 tmpfilename = self.temp_name(filename)
527
528 # Check for rtmpdump first
529 try:
530 subprocess.call(['rtmpdump', '-h'], stdout=(file(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
531 except (OSError, IOError):
532 self.trouble(u'ERROR: RTMP download detected but "rtmpdump" could not be run')
533 return False
534
535 # Download using rtmpdump. rtmpdump returns exit code 2 when
536 # the connection was interrumpted and resuming appears to be
537 # possible. This is part of rtmpdump's normal usage, AFAIK.
538 basic_args = ['rtmpdump', '-q'] + [[], ['-W', player_url]][player_url is not None] + ['-r', url, '-o', tmpfilename]
539 args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)]
540 if self.params.get('verbose', False):
541 try:
542 import pipes
543 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
544 except ImportError:
545 shell_quote = repr
546 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
547 retval = subprocess.call(args)
548 while retval == 2 or retval == 1:
549 prevsize = os.path.getsize(encodeFilename(tmpfilename))
550 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
551 time.sleep(5.0) # This seems to be needed
552 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
553 cursize = os.path.getsize(encodeFilename(tmpfilename))
554 if prevsize == cursize and retval == 1:
555 break
556 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
557 if prevsize == cursize and retval == 2 and cursize > 1024:
558 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
559 retval = 0
560 break
561 if retval == 0:
562 self.to_screen(u'\r[rtmpdump] %s bytes' % os.path.getsize(encodeFilename(tmpfilename)))
563 self.try_rename(tmpfilename, filename)
564 return True
565 else:
566 self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval)
567 return False
568
569 def _do_download(self, filename, info_dict):
570 url = info_dict['url']
571 player_url = info_dict.get('player_url', None)
572
573 # Check file already present
574 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
575 self.report_file_already_downloaded(filename)
576 return True
577
578 # Attempt to download using rtmpdump
579 if url.startswith('rtmp'):
580 return self._download_with_rtmpdump(filename, url, player_url)
581
582 tmpfilename = self.temp_name(filename)
583 stream = None
584
585 # Do not include the Accept-Encoding header
586 headers = {'Youtubedl-no-compression': 'True'}
587 basic_request = compat_urllib_request.Request(url, None, headers)
588 request = compat_urllib_request.Request(url, None, headers)
589
37c8fd48
FV
590 if self.params.get('test', False):
591 request.add_header('Range','bytes=0-10240')
592
59ae15a5
PH
593 # Establish possible resume length
594 if os.path.isfile(encodeFilename(tmpfilename)):
595 resume_len = os.path.getsize(encodeFilename(tmpfilename))
596 else:
597 resume_len = 0
598
599 open_mode = 'wb'
600 if resume_len != 0:
601 if self.params.get('continuedl', False):
602 self.report_resuming_byte(resume_len)
603 request.add_header('Range','bytes=%d-' % resume_len)
604 open_mode = 'ab'
605 else:
606 resume_len = 0
607
608 count = 0
609 retries = self.params.get('retries', 0)
610 while count <= retries:
611 # Establish connection
612 try:
613 if count == 0 and 'urlhandle' in info_dict:
614 data = info_dict['urlhandle']
615 data = compat_urllib_request.urlopen(request)
616 break
617 except (compat_urllib_error.HTTPError, ) as err:
618 if (err.code < 500 or err.code >= 600) and err.code != 416:
619 # Unexpected HTTP error
620 raise
621 elif err.code == 416:
622 # Unable to resume (requested range not satisfiable)
623 try:
624 # Open the connection again without the range header
625 data = compat_urllib_request.urlopen(basic_request)
626 content_length = data.info()['Content-Length']
627 except (compat_urllib_error.HTTPError, ) as err:
628 if err.code < 500 or err.code >= 600:
629 raise
630 else:
631 # Examine the reported length
632 if (content_length is not None and
633 (resume_len - 100 < int(content_length) < resume_len + 100)):
634 # The file had already been fully downloaded.
635 # Explanation to the above condition: in issue #175 it was revealed that
636 # YouTube sometimes adds or removes a few bytes from the end of the file,
637 # changing the file size slightly and causing problems for some users. So
638 # I decided to implement a suggested change and consider the file
639 # completely downloaded if the file size differs less than 100 bytes from
640 # the one in the hard drive.
641 self.report_file_already_downloaded(filename)
642 self.try_rename(tmpfilename, filename)
643 return True
644 else:
645 # The length does not match, we start the download over
646 self.report_unable_to_resume()
647 open_mode = 'wb'
648 break
649 # Retry
650 count += 1
651 if count <= retries:
652 self.report_retry(count, retries)
653
654 if count > retries:
655 self.trouble(u'ERROR: giving up after %s retries' % retries)
656 return False
657
658 data_len = data.info().get('Content-length', None)
659 if data_len is not None:
660 data_len = int(data_len) + resume_len
661 data_len_str = self.format_bytes(data_len)
662 byte_counter = 0 + resume_len
663 block_size = self.params.get('buffersize', 1024)
664 start = time.time()
665 while True:
666 # Download and write
667 before = time.time()
668 data_block = data.read(block_size)
669 after = time.time()
670 if len(data_block) == 0:
671 break
672 byte_counter += len(data_block)
673
674 # Open file just in time
675 if stream is None:
676 try:
677 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
678 assert stream is not None
679 filename = self.undo_temp_name(tmpfilename)
680 self.report_destination(filename)
681 except (OSError, IOError) as err:
682 self.trouble(u'ERROR: unable to open for writing: %s' % str(err))
683 return False
684 try:
685 stream.write(data_block)
686 except (IOError, OSError) as err:
687 self.trouble(u'\nERROR: unable to write data: %s' % str(err))
688 return False
689 if not self.params.get('noresizebuffer', False):
690 block_size = self.best_block_size(after - before, len(data_block))
691
692 # Progress message
693 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
694 if data_len is None:
695 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
696 else:
697 percent_str = self.calc_percent(byte_counter, data_len)
698 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
699 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
700
701 # Apply rate limit
702 self.slow_down(start, byte_counter - resume_len)
703
704 if stream is None:
705 self.trouble(u'\nERROR: Did not get any data blocks')
706 return False
707 stream.close()
708 self.report_finish()
709 if data_len is not None and byte_counter != data_len:
710 raise ContentTooShortError(byte_counter, int(data_len))
711 self.try_rename(tmpfilename, filename)
712
713 # Update file modification time
714 if self.params.get('updatetime', True):
715 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
716
717 return True