]> Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/FileDownloader.py
f21c66c29d74a332a7eba535365fc0192ae7ad23
[youtubedl] / youtube_dl / FileDownloader.py
1 import math
2 import io
3 import os
4 import re
5 import shutil
6 import socket
7 import subprocess
8 import sys
9 import time
10 import traceback
11
12 if os.name == 'nt':
13 import ctypes
14
15 from .utils import *
16 from .extractor import get_info_extractor
17
18
19 class FileDownloader(object):
20 """File Downloader class.
21
22 File downloader objects are the ones responsible of downloading the
23 actual video file and writing it to disk if the user has requested
24 it, among some other tasks. In most cases there should be one per
25 program. As, given a video URL, the downloader doesn't know how to
26 extract all the needed information, task that InfoExtractors do, it
27 has to pass the URL to one of them.
28
29 For this, file downloader objects have a method that allows
30 InfoExtractors to be registered in a given order. When it is passed
31 a URL, the file downloader handles it to the first InfoExtractor it
32 finds that reports being able to handle it. The InfoExtractor extracts
33 all the information about the video or videos the URL refers to, and
34 asks the FileDownloader to process the video information, possibly
35 downloading the video.
36
37 File downloaders accept a lot of parameters. In order not to saturate
38 the object constructor with arguments, it receives a dictionary of
39 options instead. These options are available through the params
40 attribute for the InfoExtractors to use. The FileDownloader also
41 registers itself as the downloader in charge for the InfoExtractors
42 that are added to it, so this is a "mutual registration".
43
44 Available options:
45
46 username: Username for authentication purposes.
47 password: Password for authentication purposes.
48 usenetrc: Use netrc for authentication instead.
49 quiet: Do not print messages to stdout.
50 forceurl: Force printing final URL.
51 forcetitle: Force printing title.
52 forceid: Force printing ID.
53 forcethumbnail: Force printing thumbnail URL.
54 forcedescription: Force printing description.
55 forcefilename: Force printing final filename.
56 simulate: Do not download the video files.
57 format: Video format code.
58 format_limit: Highest quality format to try.
59 outtmpl: Template for output names.
60 restrictfilenames: Do not allow "&" and spaces in file names
61 ignoreerrors: Do not stop on download errors.
62 ratelimit: Download speed limit, in bytes/sec.
63 nooverwrites: Prevent overwriting files.
64 retries: Number of times to retry for HTTP error 5xx
65 buffersize: Size of download buffer in bytes.
66 noresizebuffer: Do not automatically resize the download buffer.
67 continuedl: Try to continue downloads if possible.
68 noprogress: Do not print the progress bar.
69 playliststart: Playlist item to start at.
70 playlistend: Playlist item to end at.
71 matchtitle: Download only matching titles.
72 rejecttitle: Reject downloads for matching titles.
73 logtostderr: Log messages to stderr instead of stdout.
74 consoletitle: Display progress in console window's titlebar.
75 nopart: Do not use temporary .part files.
76 updatetime: Use the Last-modified header to set output file timestamps.
77 writedescription: Write the video description to a .description file
78 writeinfojson: Write the video description to a .info.json file
79 writethumbnail: Write the thumbnail image to a file
80 writesubtitles: Write the video subtitles to a file
81 allsubtitles: Downloads all the subtitles of the video
82 listsubtitles: Lists all available subtitles for the video
83 subtitlesformat: Subtitle format [sbv/srt] (default=srt)
84 subtitleslang: Language of the subtitles to download
85 test: Download only first bytes to test the downloader.
86 keepvideo: Keep the video file after post-processing
87 min_filesize: Skip files smaller than this size
88 max_filesize: Skip files larger than this size
89 daterange: A DateRange object, download only if the upload_date is in the range.
90 skip_download: Skip the actual download of the video file
91 """
92
93 params = None
94 _ies = []
95 _pps = []
96 _download_retcode = None
97 _num_downloads = None
98 _screen_file = None
99
100 def __init__(self, params):
101 """Create a FileDownloader object with the given options."""
102 self._ies = []
103 self._pps = []
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)]
108 self.params = params
109
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.')
112
113 @staticmethod
114 def format_bytes(bytes):
115 if bytes is None:
116 return 'N/A'
117 if type(bytes) is str:
118 bytes = float(bytes)
119 if bytes == 0.0:
120 exponent = 0
121 else:
122 exponent = int(math.log(bytes, 1024.0))
123 suffix = ['B','KiB','MiB','GiB','TiB','PiB','EiB','ZiB','YiB'][exponent]
124 converted = float(bytes) / float(1024 ** exponent)
125 return '%.2f%s' % (converted, suffix)
126
127 @staticmethod
128 def calc_percent(byte_counter, data_len):
129 if data_len is None:
130 return '---.-%'
131 return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0))
132
133 @staticmethod
134 def calc_eta(start, now, total, current):
135 if total is None:
136 return '--:--'
137 dif = now - start
138 if current == 0 or dif < 0.001: # One millisecond
139 return '--:--'
140 rate = float(current) / dif
141 eta = int((float(total) - float(current)) / rate)
142 (eta_mins, eta_secs) = divmod(eta, 60)
143 if eta_mins > 99:
144 return '--:--'
145 return '%02d:%02d' % (eta_mins, eta_secs)
146
147 @staticmethod
148 def calc_speed(start, now, bytes):
149 dif = now - start
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))
153
154 @staticmethod
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:
159 return int(new_max)
160 rate = bytes / elapsed_time
161 if rate > new_max:
162 return int(new_max)
163 if rate < new_min:
164 return int(new_min)
165 return int(rate)
166
167 @staticmethod
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)
171 if matchobj is None:
172 return None
173 number = float(matchobj.group(1))
174 multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower())
175 return int(round(number * multiplier))
176
177 def add_info_extractor(self, ie):
178 """Add an InfoExtractor object to the end of the list."""
179 self._ies.append(ie)
180 ie.set_downloader(self)
181
182 def add_post_processor(self, pp):
183 """Add a PostProcessor object to the end of the chain."""
184 self._pps.append(pp)
185 pp.set_downloader(self)
186
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()
197
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)
205
206 def to_cons_title(self, message):
207 """Set console/terminal window title to message."""
208 if not self.params.get('consoletitle', False):
209 return
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)
216
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)
220
221 def trouble(self, message=None, tb=None):
222 """Determine action to take when a download problem appears.
223
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.
227
228 tb, if given, is additional traceback information.
229 """
230 if message is not None:
231 self.to_stderr(message)
232 if self.params.get('verbose'):
233 if tb is None:
234 if sys.exc_info()[0]: # if .trouble has been called from an except block
235 tb = u''
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())
239 else:
240 tb_data = traceback.format_list(traceback.extract_stack())
241 tb = u''.join(tb_data)
242 self.to_stderr(tb)
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
246 else:
247 exc_info = sys.exc_info()
248 raise DownloadError(message, exc_info)
249 self._download_retcode = 1
250
251 def report_warning(self, message):
252 '''
253 Print the message to stderr, it will be prefixed with 'WARNING:'
254 If stderr is a tty file the 'WARNING:' will be colored
255 '''
256 if sys.stderr.isatty() and os.name != 'nt':
257 _msg_header=u'\033[0;33mWARNING:\033[0m'
258 else:
259 _msg_header=u'WARNING:'
260 warning_message=u'%s %s' % (_msg_header,message)
261 self.to_stderr(warning_message)
262
263 def report_error(self, message, tb=None):
264 '''
265 Do the same as trouble, but prefixes the message with 'ERROR:', colored
266 in red if stderr is a tty file.
267 '''
268 if sys.stderr.isatty() and os.name != 'nt':
269 _msg_header = u'\033[0;31mERROR:\033[0m'
270 else:
271 _msg_header = u'ERROR:'
272 error_message = u'%s %s' % (_msg_header, message)
273 self.trouble(error_message, tb)
274
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:
279 return
280 now = time.time()
281 elapsed = now - start_time
282 if elapsed <= 0.0:
283 return
284 speed = float(byte_counter) / elapsed
285 if speed > rate_limit:
286 time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
287
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))):
292 return filename
293 return filename + u'.part'
294
295 def undo_temp_name(self, filename):
296 if filename.endswith(u'.part'):
297 return filename[:-len(u'.part')]
298 return filename
299
300 def try_rename(self, old_filename, new_filename):
301 try:
302 if old_filename == new_filename:
303 return
304 os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
305 except (IOError, OSError) as err:
306 self.report_error(u'unable to rename file')
307
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:
311 return
312 if not os.path.isfile(encodeFilename(filename)):
313 return
314 timestr = last_modified_hdr
315 if timestr is None:
316 return
317 filetime = timeconvert(timestr)
318 if filetime is None:
319 return filetime
320 # Ignore obviously invalid dates
321 if filetime == 0:
322 return
323 try:
324 os.utime(filename, (time.time(), filetime))
325 except:
326 pass
327 return filetime
328
329 def report_writedescription(self, descfn):
330 """ Report that the description file is being written """
331 self.to_screen(u'[info] Writing video description to: ' + descfn)
332
333 def report_writesubtitles(self, sub_filename):
334 """ Report that the subtitles file is being written """
335 self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename)
336
337 def report_writeinfojson(self, infofn):
338 """ Report that the metadata file has been written """
339 self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn)
340
341 def report_destination(self, filename):
342 """Report destination filename."""
343 self.to_screen(u'[download] Destination: ' + filename)
344
345 def report_progress(self, percent_str, data_len_str, speed_str, eta_str):
346 """Report download progress."""
347 if self.params.get('noprogress', False):
348 return
349 clear_line = (u'\x1b[K' if sys.stderr.isatty() and os.name != 'nt' else u'')
350 if self.params.get('progress_with_newline', False):
351 self.to_screen(u'[download] %s of %s at %s ETA %s' %
352 (percent_str, data_len_str, speed_str, eta_str))
353 else:
354 self.to_screen(u'\r%s[download] %s of %s at %s ETA %s' %
355 (clear_line, percent_str, data_len_str, speed_str, eta_str), skip_eol=True)
356 self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' %
357 (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip()))
358
359 def report_resuming_byte(self, resume_len):
360 """Report attempt to resume at given byte."""
361 self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
362
363 def report_retry(self, count, retries):
364 """Report retry in case of HTTP error 5xx"""
365 self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
366
367 def report_file_already_downloaded(self, file_name):
368 """Report file has already been fully downloaded."""
369 try:
370 self.to_screen(u'[download] %s has already been downloaded' % file_name)
371 except (UnicodeEncodeError) as err:
372 self.to_screen(u'[download] The file has already been downloaded')
373
374 def report_unable_to_resume(self):
375 """Report it was impossible to resume download."""
376 self.to_screen(u'[download] Unable to resume')
377
378 def report_finish(self):
379 """Report download finished."""
380 if self.params.get('noprogress', False):
381 self.to_screen(u'[download] Download completed')
382 else:
383 self.to_screen(u'')
384
385 def increment_downloads(self):
386 """Increment the ordinal that assigns a number to each file."""
387 self._num_downloads += 1
388
389 def prepare_filename(self, info_dict):
390 """Generate the output filename."""
391 try:
392 template_dict = dict(info_dict)
393
394 template_dict['epoch'] = int(time.time())
395 autonumber_size = self.params.get('autonumber_size')
396 if autonumber_size is None:
397 autonumber_size = 5
398 autonumber_templ = u'%0' + str(autonumber_size) + u'd'
399 template_dict['autonumber'] = autonumber_templ % self._num_downloads
400 if template_dict['playlist_index'] is not None:
401 template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index']
402
403 sanitize = lambda k,v: sanitize_filename(
404 u'NA' if v is None else compat_str(v),
405 restricted=self.params.get('restrictfilenames'),
406 is_id=(k==u'id'))
407 template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items())
408
409 filename = self.params['outtmpl'] % template_dict
410 return filename
411 except KeyError as err:
412 self.report_error(u'Erroneous output template')
413 return None
414 except ValueError as err:
415 self.report_error(u'Insufficient system charset ' + repr(preferredencoding()))
416 return None
417
418 def _match_entry(self, info_dict):
419 """ Returns None iff the file should be downloaded """
420
421 title = info_dict['title']
422 matchtitle = self.params.get('matchtitle', False)
423 if matchtitle:
424 if not re.search(matchtitle, title, re.IGNORECASE):
425 return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"'
426 rejecttitle = self.params.get('rejecttitle', False)
427 if rejecttitle:
428 if re.search(rejecttitle, title, re.IGNORECASE):
429 return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'
430 date = info_dict.get('upload_date', None)
431 if date is not None:
432 dateRange = self.params.get('daterange', DateRange())
433 if date not in dateRange:
434 return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange)
435 return None
436
437 def extract_info(self, url, download=True, ie_key=None, extra_info={}):
438 '''
439 Returns a list with a dictionary for each video we find.
440 If 'download', also downloads the videos.
441 extra_info is a dict containing the extra values to add to each result
442 '''
443
444 if ie_key:
445 ie = get_info_extractor(ie_key)()
446 ie.set_downloader(self)
447 ies = [ie]
448 else:
449 ies = self._ies
450
451 for ie in ies:
452 if not ie.suitable(url):
453 continue
454
455 if not ie.working():
456 self.report_warning(u'The program functionality for this site has been marked as broken, '
457 u'and will probably not work.')
458
459 try:
460 ie_result = ie.extract(url)
461 if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here)
462 break
463 if isinstance(ie_result, list):
464 # Backwards compatibility: old IE result format
465 for result in ie_result:
466 result.update(extra_info)
467 ie_result = {
468 '_type': 'compat_list',
469 'entries': ie_result,
470 }
471 else:
472 ie_result.update(extra_info)
473 if 'extractor' not in ie_result:
474 ie_result['extractor'] = ie.IE_NAME
475 return self.process_ie_result(ie_result, download=download)
476 except ExtractorError as de: # An error we somewhat expected
477 self.report_error(compat_str(de), de.format_traceback())
478 break
479 except Exception as e:
480 if self.params.get('ignoreerrors', False):
481 self.report_error(compat_str(e), tb=compat_str(traceback.format_exc()))
482 break
483 else:
484 raise
485 else:
486 self.report_error(u'no suitable InfoExtractor: %s' % url)
487
488 def process_ie_result(self, ie_result, download=True, extra_info={}):
489 """
490 Take the result of the ie(may be modified) and resolve all unresolved
491 references (URLs, playlist items).
492
493 It will also download the videos if 'download'.
494 Returns the resolved ie_result.
495 """
496
497 result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system
498 if result_type == 'video':
499 if 'playlist' not in ie_result:
500 # It isn't part of a playlist
501 ie_result['playlist'] = None
502 ie_result['playlist_index'] = None
503 if download:
504 self.process_info(ie_result)
505 return ie_result
506 elif result_type == 'url':
507 # We have to add extra_info to the results because it may be
508 # contained in a playlist
509 return self.extract_info(ie_result['url'],
510 download,
511 ie_key=ie_result.get('ie_key'),
512 extra_info=extra_info)
513 elif result_type == 'playlist':
514 # We process each entry in the playlist
515 playlist = ie_result.get('title', None) or ie_result.get('id', None)
516 self.to_screen(u'[download] Downloading playlist: %s' % playlist)
517
518 playlist_results = []
519
520 n_all_entries = len(ie_result['entries'])
521 playliststart = self.params.get('playliststart', 1) - 1
522 playlistend = self.params.get('playlistend', -1)
523
524 if playlistend == -1:
525 entries = ie_result['entries'][playliststart:]
526 else:
527 entries = ie_result['entries'][playliststart:playlistend]
528
529 n_entries = len(entries)
530
531 self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
532 (ie_result['extractor'], playlist, n_all_entries, n_entries))
533
534 for i,entry in enumerate(entries,1):
535 self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries))
536 extra = {
537 'playlist': playlist,
538 'playlist_index': i + playliststart,
539 }
540 if not 'extractor' in entry:
541 # We set the extractor, if it's an url it will be set then to
542 # the new extractor, but if it's already a video we must make
543 # sure it's present: see issue #877
544 entry['extractor'] = ie_result['extractor']
545 entry_result = self.process_ie_result(entry,
546 download=download,
547 extra_info=extra)
548 playlist_results.append(entry_result)
549 ie_result['entries'] = playlist_results
550 return ie_result
551 elif result_type == 'compat_list':
552 def _fixup(r):
553 r.setdefault('extractor', ie_result['extractor'])
554 return r
555 ie_result['entries'] = [
556 self.process_ie_result(_fixup(r), download=download)
557 for r in ie_result['entries']
558 ]
559 return ie_result
560 else:
561 raise Exception('Invalid result type: %s' % result_type)
562
563 def process_info(self, info_dict):
564 """Process a single resolved IE result."""
565
566 assert info_dict.get('_type', 'video') == 'video'
567 #We increment the download the download count here to match the previous behaviour.
568 self.increment_downloads()
569
570 info_dict['fulltitle'] = info_dict['title']
571 if len(info_dict['title']) > 200:
572 info_dict['title'] = info_dict['title'][:197] + u'...'
573
574 # Keep for backwards compatibility
575 info_dict['stitle'] = info_dict['title']
576
577 if not 'format' in info_dict:
578 info_dict['format'] = info_dict['ext']
579
580 reason = self._match_entry(info_dict)
581 if reason is not None:
582 self.to_screen(u'[download] ' + reason)
583 return
584
585 max_downloads = self.params.get('max_downloads')
586 if max_downloads is not None:
587 if self._num_downloads > int(max_downloads):
588 raise MaxDownloadsReached()
589
590 filename = self.prepare_filename(info_dict)
591
592 # Forced printings
593 if self.params.get('forcetitle', False):
594 compat_print(info_dict['title'])
595 if self.params.get('forceid', False):
596 compat_print(info_dict['id'])
597 if self.params.get('forceurl', False):
598 compat_print(info_dict['url'])
599 if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict:
600 compat_print(info_dict['thumbnail'])
601 if self.params.get('forcedescription', False) and 'description' in info_dict:
602 compat_print(info_dict['description'])
603 if self.params.get('forcefilename', False) and filename is not None:
604 compat_print(filename)
605 if self.params.get('forceformat', False):
606 compat_print(info_dict['format'])
607
608 # Do nothing else if in simulate mode
609 if self.params.get('simulate', False):
610 return
611
612 if filename is None:
613 return
614
615 try:
616 dn = os.path.dirname(encodeFilename(filename))
617 if dn != '' and not os.path.exists(dn):
618 os.makedirs(dn)
619 except (OSError, IOError) as err:
620 self.report_error(u'unable to create directory ' + compat_str(err))
621 return
622
623 if self.params.get('writedescription', False):
624 try:
625 descfn = filename + u'.description'
626 self.report_writedescription(descfn)
627 with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
628 descfile.write(info_dict['description'])
629 except (OSError, IOError):
630 self.report_error(u'Cannot write description file ' + descfn)
631 return
632
633 if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
634 # subtitles download errors are already managed as troubles in relevant IE
635 # that way it will silently go on when used with unsupporting IE
636 subtitle = info_dict['subtitles'][0]
637 (sub_error, sub_lang, sub) = subtitle
638 sub_format = self.params.get('subtitlesformat')
639 if sub_error:
640 self.report_warning("Some error while getting the subtitles")
641 else:
642 try:
643 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
644 self.report_writesubtitles(sub_filename)
645 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
646 subfile.write(sub)
647 except (OSError, IOError):
648 self.report_error(u'Cannot write subtitles file ' + descfn)
649 return
650
651 if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']:
652 subtitles = info_dict['subtitles']
653 sub_format = self.params.get('subtitlesformat')
654 for subtitle in subtitles:
655 (sub_error, sub_lang, sub) = subtitle
656 if sub_error:
657 self.report_warning("Some error while getting the subtitles")
658 else:
659 try:
660 sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format
661 self.report_writesubtitles(sub_filename)
662 with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile:
663 subfile.write(sub)
664 except (OSError, IOError):
665 self.report_error(u'Cannot write subtitles file ' + descfn)
666 return
667
668 if self.params.get('writeinfojson', False):
669 infofn = filename + u'.info.json'
670 self.report_writeinfojson(infofn)
671 try:
672 json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle'])
673 write_json_file(json_info_dict, encodeFilename(infofn))
674 except (OSError, IOError):
675 self.report_error(u'Cannot write metadata to JSON file ' + infofn)
676 return
677
678 if self.params.get('writethumbnail', False):
679 if 'thumbnail' in info_dict:
680 thumb_format = info_dict['thumbnail'].rpartition(u'/')[2].rpartition(u'.')[2]
681 if not thumb_format:
682 thumb_format = 'jpg'
683 thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format
684 self.to_screen(u'[%s] %s: Downloading thumbnail ...' %
685 (info_dict['extractor'], info_dict['id']))
686 uf = compat_urllib_request.urlopen(info_dict['thumbnail'])
687 with open(thumb_filename, 'wb') as thumbf:
688 shutil.copyfileobj(uf, thumbf)
689 self.to_screen(u'[%s] %s: Writing thumbnail to: %s' %
690 (info_dict['extractor'], info_dict['id'], thumb_filename))
691
692 if not self.params.get('skip_download', False):
693 if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)):
694 success = True
695 else:
696 try:
697 success = self._do_download(filename, info_dict)
698 except (OSError, IOError) as err:
699 raise UnavailableVideoError()
700 except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
701 self.report_error(u'unable to download video data: %s' % str(err))
702 return
703 except (ContentTooShortError, ) as err:
704 self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded))
705 return
706
707 if success:
708 try:
709 self.post_process(filename, info_dict)
710 except (PostProcessingError) as err:
711 self.report_error(u'postprocessing: %s' % str(err))
712 return
713
714 def download(self, url_list):
715 """Download a given list of URLs."""
716 if len(url_list) > 1 and self.fixed_template():
717 raise SameFileError(self.params['outtmpl'])
718
719 for url in url_list:
720 try:
721 #It also downloads the videos
722 videos = self.extract_info(url)
723 except UnavailableVideoError:
724 self.report_error(u'unable to download video')
725 except MaxDownloadsReached:
726 self.to_screen(u'[info] Maximum number of downloaded files reached.')
727 raise
728
729 return self._download_retcode
730
731 def post_process(self, filename, ie_info):
732 """Run all the postprocessors on the given file."""
733 info = dict(ie_info)
734 info['filepath'] = filename
735 keep_video = None
736 for pp in self._pps:
737 try:
738 keep_video_wish,new_info = pp.run(info)
739 if keep_video_wish is not None:
740 if keep_video_wish:
741 keep_video = keep_video_wish
742 elif keep_video is None:
743 # No clear decision yet, let IE decide
744 keep_video = keep_video_wish
745 except PostProcessingError as e:
746 self.to_stderr(u'ERROR: ' + e.msg)
747 if keep_video is False and not self.params.get('keepvideo', False):
748 try:
749 self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename)
750 os.remove(encodeFilename(filename))
751 except (IOError, OSError):
752 self.report_warning(u'Unable to remove downloaded video file')
753
754 def _download_with_rtmpdump(self, filename, url, player_url, page_url, play_path, tc_url):
755 self.report_destination(filename)
756 tmpfilename = self.temp_name(filename)
757
758 # Check for rtmpdump first
759 try:
760 subprocess.call(['rtmpdump', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
761 except (OSError, IOError):
762 self.report_error(u'RTMP download detected but "rtmpdump" could not be run')
763 return False
764 verbosity_option = '--verbose' if self.params.get('verbose', False) else '--quiet'
765
766 # Download using rtmpdump. rtmpdump returns exit code 2 when
767 # the connection was interrumpted and resuming appears to be
768 # possible. This is part of rtmpdump's normal usage, AFAIK.
769 basic_args = ['rtmpdump', verbosity_option, '-r', url, '-o', tmpfilename]
770 if player_url is not None:
771 basic_args += ['--swfVfy', player_url]
772 if page_url is not None:
773 basic_args += ['--pageUrl', page_url]
774 if play_path is not None:
775 basic_args += ['--playpath', play_path]
776 if tc_url is not None:
777 basic_args += ['--tcUrl', url]
778 args = basic_args + [[], ['--resume', '--skip', '1']][self.params.get('continuedl', False)]
779 if self.params.get('verbose', False):
780 try:
781 import pipes
782 shell_quote = lambda args: ' '.join(map(pipes.quote, args))
783 except ImportError:
784 shell_quote = repr
785 self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args))
786 retval = subprocess.call(args)
787 while retval == 2 or retval == 1:
788 prevsize = os.path.getsize(encodeFilename(tmpfilename))
789 self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True)
790 time.sleep(5.0) # This seems to be needed
791 retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1])
792 cursize = os.path.getsize(encodeFilename(tmpfilename))
793 if prevsize == cursize and retval == 1:
794 break
795 # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those
796 if prevsize == cursize and retval == 2 and cursize > 1024:
797 self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.')
798 retval = 0
799 break
800 if retval == 0:
801 fsize = os.path.getsize(encodeFilename(tmpfilename))
802 self.to_screen(u'\r[rtmpdump] %s bytes' % fsize)
803 self.try_rename(tmpfilename, filename)
804 self._hook_progress({
805 'downloaded_bytes': fsize,
806 'total_bytes': fsize,
807 'filename': filename,
808 'status': 'finished',
809 })
810 return True
811 else:
812 self.to_stderr(u"\n")
813 self.report_error(u'rtmpdump exited with code %d' % retval)
814 return False
815
816 def _download_with_mplayer(self, filename, url):
817 self.report_destination(filename)
818 tmpfilename = self.temp_name(filename)
819
820 args = ['mplayer', '-really-quiet', '-vo', 'null', '-vc', 'dummy', '-dumpstream', '-dumpfile', tmpfilename, url]
821 # Check for mplayer first
822 try:
823 subprocess.call(['mplayer', '-h'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
824 except (OSError, IOError):
825 self.report_error(u'MMS or RTSP download detected but "%s" could not be run' % args[0] )
826 return False
827
828 # Download using mplayer.
829 retval = subprocess.call(args)
830 if retval == 0:
831 fsize = os.path.getsize(encodeFilename(tmpfilename))
832 self.to_screen(u'\r[%s] %s bytes' % (args[0], fsize))
833 self.try_rename(tmpfilename, filename)
834 self._hook_progress({
835 'downloaded_bytes': fsize,
836 'total_bytes': fsize,
837 'filename': filename,
838 'status': 'finished',
839 })
840 return True
841 else:
842 self.to_stderr(u"\n")
843 self.report_error(u'mplayer exited with code %d' % retval)
844 return False
845
846
847 def _do_download(self, filename, info_dict):
848 url = info_dict['url']
849
850 # Check file already present
851 if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
852 self.report_file_already_downloaded(filename)
853 self._hook_progress({
854 'filename': filename,
855 'status': 'finished',
856 })
857 return True
858
859 # Attempt to download using rtmpdump
860 if url.startswith('rtmp'):
861 return self._download_with_rtmpdump(filename, url,
862 info_dict.get('player_url', None),
863 info_dict.get('page_url', None),
864 info_dict.get('play_path', None),
865 info_dict.get('tc_url', None))
866
867 # Attempt to download using mplayer
868 if url.startswith('mms') or url.startswith('rtsp'):
869 return self._download_with_mplayer(filename, url)
870
871 tmpfilename = self.temp_name(filename)
872 stream = None
873
874 # Do not include the Accept-Encoding header
875 headers = {'Youtubedl-no-compression': 'True'}
876 if 'user_agent' in info_dict:
877 headers['Youtubedl-user-agent'] = info_dict['user_agent']
878 basic_request = compat_urllib_request.Request(url, None, headers)
879 request = compat_urllib_request.Request(url, None, headers)
880
881 if self.params.get('test', False):
882 request.add_header('Range','bytes=0-10240')
883
884 # Establish possible resume length
885 if os.path.isfile(encodeFilename(tmpfilename)):
886 resume_len = os.path.getsize(encodeFilename(tmpfilename))
887 else:
888 resume_len = 0
889
890 open_mode = 'wb'
891 if resume_len != 0:
892 if self.params.get('continuedl', False):
893 self.report_resuming_byte(resume_len)
894 request.add_header('Range','bytes=%d-' % resume_len)
895 open_mode = 'ab'
896 else:
897 resume_len = 0
898
899 count = 0
900 retries = self.params.get('retries', 0)
901 while count <= retries:
902 # Establish connection
903 try:
904 if count == 0 and 'urlhandle' in info_dict:
905 data = info_dict['urlhandle']
906 data = compat_urllib_request.urlopen(request)
907 break
908 except (compat_urllib_error.HTTPError, ) as err:
909 if (err.code < 500 or err.code >= 600) and err.code != 416:
910 # Unexpected HTTP error
911 raise
912 elif err.code == 416:
913 # Unable to resume (requested range not satisfiable)
914 try:
915 # Open the connection again without the range header
916 data = compat_urllib_request.urlopen(basic_request)
917 content_length = data.info()['Content-Length']
918 except (compat_urllib_error.HTTPError, ) as err:
919 if err.code < 500 or err.code >= 600:
920 raise
921 else:
922 # Examine the reported length
923 if (content_length is not None and
924 (resume_len - 100 < int(content_length) < resume_len + 100)):
925 # The file had already been fully downloaded.
926 # Explanation to the above condition: in issue #175 it was revealed that
927 # YouTube sometimes adds or removes a few bytes from the end of the file,
928 # changing the file size slightly and causing problems for some users. So
929 # I decided to implement a suggested change and consider the file
930 # completely downloaded if the file size differs less than 100 bytes from
931 # the one in the hard drive.
932 self.report_file_already_downloaded(filename)
933 self.try_rename(tmpfilename, filename)
934 self._hook_progress({
935 'filename': filename,
936 'status': 'finished',
937 })
938 return True
939 else:
940 # The length does not match, we start the download over
941 self.report_unable_to_resume()
942 open_mode = 'wb'
943 break
944 # Retry
945 count += 1
946 if count <= retries:
947 self.report_retry(count, retries)
948
949 if count > retries:
950 self.report_error(u'giving up after %s retries' % retries)
951 return False
952
953 data_len = data.info().get('Content-length', None)
954 if data_len is not None:
955 data_len = int(data_len) + resume_len
956 min_data_len = self.params.get("min_filesize", None)
957 max_data_len = self.params.get("max_filesize", None)
958 if min_data_len is not None and data_len < min_data_len:
959 self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
960 return False
961 if max_data_len is not None and data_len > max_data_len:
962 self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
963 return False
964
965 data_len_str = self.format_bytes(data_len)
966 byte_counter = 0 + resume_len
967 block_size = self.params.get('buffersize', 1024)
968 start = time.time()
969 while True:
970 # Download and write
971 before = time.time()
972 data_block = data.read(block_size)
973 after = time.time()
974 if len(data_block) == 0:
975 break
976 byte_counter += len(data_block)
977
978 # Open file just in time
979 if stream is None:
980 try:
981 (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
982 assert stream is not None
983 filename = self.undo_temp_name(tmpfilename)
984 self.report_destination(filename)
985 except (OSError, IOError) as err:
986 self.report_error(u'unable to open for writing: %s' % str(err))
987 return False
988 try:
989 stream.write(data_block)
990 except (IOError, OSError) as err:
991 self.to_stderr(u"\n")
992 self.report_error(u'unable to write data: %s' % str(err))
993 return False
994 if not self.params.get('noresizebuffer', False):
995 block_size = self.best_block_size(after - before, len(data_block))
996
997 # Progress message
998 speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len)
999 if data_len is None:
1000 self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA')
1001 else:
1002 percent_str = self.calc_percent(byte_counter, data_len)
1003 eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
1004 self.report_progress(percent_str, data_len_str, speed_str, eta_str)
1005
1006 self._hook_progress({
1007 'downloaded_bytes': byte_counter,
1008 'total_bytes': data_len,
1009 'tmpfilename': tmpfilename,
1010 'filename': filename,
1011 'status': 'downloading',
1012 })
1013
1014 # Apply rate limit
1015 self.slow_down(start, byte_counter - resume_len)
1016
1017 if stream is None:
1018 self.to_stderr(u"\n")
1019 self.report_error(u'Did not get any data blocks')
1020 return False
1021 stream.close()
1022 self.report_finish()
1023 if data_len is not None and byte_counter != data_len:
1024 raise ContentTooShortError(byte_counter, int(data_len))
1025 self.try_rename(tmpfilename, filename)
1026
1027 # Update file modification time
1028 if self.params.get('updatetime', True):
1029 info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
1030
1031 self._hook_progress({
1032 'downloaded_bytes': byte_counter,
1033 'total_bytes': byte_counter,
1034 'filename': filename,
1035 'status': 'finished',
1036 })
1037
1038 return True
1039
1040 def _hook_progress(self, status):
1041 for ph in self._progress_hooks:
1042 ph(status)
1043
1044 def add_progress_hook(self, ph):
1045 """ ph gets called on download progress, with a dictionary with the entries
1046 * filename: The final filename
1047 * status: One of "downloading" and "finished"
1048
1049 It can also have some of the following entries:
1050
1051 * downloaded_bytes: Bytes on disks
1052 * total_bytes: Total bytes, None if unknown
1053 * tmpfilename: The filename we're currently writing to
1054
1055 Hooks are guaranteed to be called at least once (with status "finished")
1056 if the download is successful.
1057 """
1058 self._progress_hooks.append(ph)