]>
Commit | Line | Data |
---|---|---|
d77c3dfd FV |
1 | #!/usr/bin/env python |
2 | # -*- coding: utf-8 -*- | |
3 | ||
9e8056d5 PH |
4 | from __future__ import absolute_import |
5 | ||
d77c3dfd FV |
6 | import math |
7 | import os | |
8 | import re | |
9 | import socket | |
10 | import subprocess | |
11 | import sys | |
12 | import time | |
59ce2019 | 13 | import traceback |
d77c3dfd FV |
14 | |
15 | if os.name == 'nt': | |
59ae15a5 | 16 | import ctypes |
3eec021a | 17 | |
9e8056d5 | 18 | from .utils import * |
d77c3dfd FV |
19 | |
20 | ||
21 | class FileDownloader(object): | |
59ae15a5 PH |
22 | """File Downloader class. |
23 | ||
24 | File downloader objects are the ones responsible of downloading the | |
25 | actual video file and writing it to disk if the user has requested | |
26 | it, among some other tasks. In most cases there should be one per | |
27 | program. As, given a video URL, the downloader doesn't know how to | |
28 | extract all the needed information, task that InfoExtractors do, it | |
29 | has to pass the URL to one of them. | |
30 | ||
31 | For this, file downloader objects have a method that allows | |
32 | InfoExtractors to be registered in a given order. When it is passed | |
33 | a URL, the file downloader handles it to the first InfoExtractor it | |
34 | finds that reports being able to handle it. The InfoExtractor extracts | |
35 | all the information about the video or videos the URL refers to, and | |
36 | asks the FileDownloader to process the video information, possibly | |
37 | downloading the video. | |
38 | ||
39 | File downloaders accept a lot of parameters. In order not to saturate | |
40 | the object constructor with arguments, it receives a dictionary of | |
41 | options instead. These options are available through the params | |
42 | attribute for the InfoExtractors to use. The FileDownloader also | |
43 | registers itself as the downloader in charge for the InfoExtractors | |
44 | that are added to it, so this is a "mutual registration". | |
45 | ||
46 | Available options: | |
47 | ||
48 | username: Username for authentication purposes. | |
49 | password: Password for authentication purposes. | |
50 | usenetrc: Use netrc for authentication instead. | |
51 | quiet: Do not print messages to stdout. | |
52 | forceurl: Force printing final URL. | |
53 | forcetitle: Force printing title. | |
54 | forcethumbnail: Force printing thumbnail URL. | |
55 | forcedescription: Force printing description. | |
56 | forcefilename: Force printing final filename. | |
57 | simulate: Do not download the video files. | |
58 | format: Video format code. | |
59 | format_limit: Highest quality format to try. | |
60 | outtmpl: Template for output names. | |
61 | restrictfilenames: Do not allow "&" and spaces in file names | |
62 | ignoreerrors: Do not stop on download errors. | |
63 | ratelimit: Download speed limit, in bytes/sec. | |
64 | nooverwrites: Prevent overwriting files. | |
65 | retries: Number of times to retry for HTTP error 5xx | |
66 | buffersize: Size of download buffer in bytes. | |
67 | noresizebuffer: Do not automatically resize the download buffer. | |
68 | continuedl: Try to continue downloads if possible. | |
69 | noprogress: Do not print the progress bar. | |
70 | playliststart: Playlist item to start at. | |
71 | playlistend: Playlist item to end at. | |
72 | matchtitle: Download only matching titles. | |
73 | rejecttitle: Reject downloads for matching titles. | |
74 | logtostderr: Log messages to stderr instead of stdout. | |
75 | consoletitle: Display progress in console window's titlebar. | |
76 | nopart: Do not use temporary .part files. | |
77 | updatetime: Use the Last-modified header to set output file timestamps. | |
78 | writedescription: Write the video description to a .description file | |
79 | writeinfojson: Write the video description to a .info.json file | |
80 | writesubtitles: Write the video subtitles to a .srt file | |
81 | subtitleslang: Language of the subtitles to download | |
37c8fd48 | 82 | test: Download only first bytes to test the downloader. |
59ae15a5 PH |
83 | """ |
84 | ||
85 | params = None | |
86 | _ies = [] | |
87 | _pps = [] | |
88 | _download_retcode = None | |
89 | _num_downloads = None | |
90 | _screen_file = None | |
91 | ||
92 | def __init__(self, params): | |
93 | """Create a FileDownloader object with the given options.""" | |
94 | self._ies = [] | |
95 | self._pps = [] | |
96 | self._download_retcode = 0 | |
97 | self._num_downloads = 0 | |
98 | self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)] | |
99 | self.params = params | |
100 | ||
101 | if '%(stitle)s' in self.params['outtmpl']: | |
102 | 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.') | |
103 | ||
104 | @staticmethod | |
105 | def format_bytes(bytes): | |
106 | if bytes is None: | |
107 | return 'N/A' | |
108 | if type(bytes) is str: | |
109 | bytes = float(bytes) | |
110 | if bytes == 0.0: | |
111 | exponent = 0 | |
112 | else: | |
113 | exponent = int(math.log(bytes, 1024.0)) | |
114 | suffix = 'bkMGTPEZY'[exponent] | |
115 | converted = float(bytes) / float(1024 ** exponent) | |
116 | return '%.2f%s' % (converted, suffix) | |
117 | ||
118 | @staticmethod | |
119 | def calc_percent(byte_counter, data_len): | |
120 | if data_len is None: | |
121 | return '---.-%' | |
122 | return '%6s' % ('%3.1f%%' % (float(byte_counter) / float(data_len) * 100.0)) | |
123 | ||
124 | @staticmethod | |
125 | def calc_eta(start, now, total, current): | |
126 | if total is None: | |
127 | return '--:--' | |
128 | dif = now - start | |
129 | if current == 0 or dif < 0.001: # One millisecond | |
130 | return '--:--' | |
131 | rate = float(current) / dif | |
132 | eta = int((float(total) - float(current)) / rate) | |
133 | (eta_mins, eta_secs) = divmod(eta, 60) | |
134 | if eta_mins > 99: | |
135 | return '--:--' | |
136 | return '%02d:%02d' % (eta_mins, eta_secs) | |
137 | ||
138 | @staticmethod | |
139 | def calc_speed(start, now, bytes): | |
140 | dif = now - start | |
141 | if bytes == 0 or dif < 0.001: # One millisecond | |
142 | return '%10s' % '---b/s' | |
143 | return '%10s' % ('%s/s' % FileDownloader.format_bytes(float(bytes) / dif)) | |
144 | ||
145 | @staticmethod | |
146 | def best_block_size(elapsed_time, bytes): | |
147 | new_min = max(bytes / 2.0, 1.0) | |
148 | new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB | |
149 | if elapsed_time < 0.001: | |
150 | return int(new_max) | |
151 | rate = bytes / elapsed_time | |
152 | if rate > new_max: | |
153 | return int(new_max) | |
154 | if rate < new_min: | |
155 | return int(new_min) | |
156 | return int(rate) | |
157 | ||
158 | @staticmethod | |
159 | def parse_bytes(bytestr): | |
160 | """Parse a string indicating a byte quantity into an integer.""" | |
161 | matchobj = re.match(r'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr) | |
162 | if matchobj is None: | |
163 | return None | |
164 | number = float(matchobj.group(1)) | |
165 | multiplier = 1024.0 ** 'bkmgtpezy'.index(matchobj.group(2).lower()) | |
166 | return int(round(number * multiplier)) | |
167 | ||
168 | def add_info_extractor(self, ie): | |
169 | """Add an InfoExtractor object to the end of the list.""" | |
170 | self._ies.append(ie) | |
171 | ie.set_downloader(self) | |
172 | ||
173 | def add_post_processor(self, pp): | |
174 | """Add a PostProcessor object to the end of the chain.""" | |
175 | self._pps.append(pp) | |
176 | pp.set_downloader(self) | |
177 | ||
178 | def to_screen(self, message, skip_eol=False): | |
179 | """Print message to stdout if not in quiet mode.""" | |
180 | assert type(message) == type(u'') | |
181 | if not self.params.get('quiet', False): | |
182 | terminator = [u'\n', u''][skip_eol] | |
183 | output = message + terminator | |
184 | 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 | |
185 | output = output.encode(preferredencoding(), 'ignore') | |
186 | self._screen_file.write(output) | |
187 | self._screen_file.flush() | |
188 | ||
189 | def to_stderr(self, message): | |
190 | """Print message to stderr.""" | |
191 | assert type(message) == type(u'') | |
192 | output = message + u'\n' | |
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()) | |
195 | sys.stderr.write(output) | |
196 | ||
197 | def to_cons_title(self, message): | |
198 | """Set console/terminal window title to message.""" | |
199 | if not self.params.get('consoletitle', False): | |
200 | return | |
201 | if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow(): | |
202 | # c_wchar_p() might not be necessary if `message` is | |
203 | # already of type unicode() | |
204 | ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message)) | |
205 | elif 'TERM' in os.environ: | |
206 | sys.stderr.write('\033]0;%s\007' % message.encode(preferredencoding())) | |
207 | ||
208 | def fixed_template(self): | |
209 | """Checks if the output template is fixed.""" | |
210 | return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None) | |
211 | ||
212 | def trouble(self, message=None): | |
213 | """Determine action to take when a download problem appears. | |
214 | ||
215 | Depending on if the downloader has been configured to ignore | |
216 | download errors or not, this method may throw an exception or | |
217 | not when errors are found, after printing the message. | |
218 | """ | |
219 | if message is not None: | |
220 | self.to_stderr(message) | |
59ce2019 FV |
221 | if self.params.get('verbose'): |
222 | self.to_stderr(u''.join(traceback.format_list(traceback.extract_stack()))) | |
59ae15a5 PH |
223 | if not self.params.get('ignoreerrors', False): |
224 | raise DownloadError(message) | |
225 | self._download_retcode = 1 | |
226 | ||
227 | def slow_down(self, start_time, byte_counter): | |
228 | """Sleep if the download speed is over the rate limit.""" | |
229 | rate_limit = self.params.get('ratelimit', None) | |
230 | if rate_limit is None or byte_counter == 0: | |
231 | return | |
232 | now = time.time() | |
233 | elapsed = now - start_time | |
234 | if elapsed <= 0.0: | |
235 | return | |
236 | speed = float(byte_counter) / elapsed | |
237 | if speed > rate_limit: | |
238 | time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit) | |
239 | ||
240 | def temp_name(self, filename): | |
241 | """Returns a temporary filename for the given filename.""" | |
242 | if self.params.get('nopart', False) or filename == u'-' or \ | |
243 | (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))): | |
244 | return filename | |
245 | return filename + u'.part' | |
246 | ||
247 | def undo_temp_name(self, filename): | |
248 | if filename.endswith(u'.part'): | |
249 | return filename[:-len(u'.part')] | |
250 | return filename | |
251 | ||
252 | def try_rename(self, old_filename, new_filename): | |
253 | try: | |
254 | if old_filename == new_filename: | |
255 | return | |
256 | os.rename(encodeFilename(old_filename), encodeFilename(new_filename)) | |
257 | except (IOError, OSError) as err: | |
258 | self.trouble(u'ERROR: unable to rename file') | |
259 | ||
260 | def try_utime(self, filename, last_modified_hdr): | |
261 | """Try to set the last-modified time of the given file.""" | |
262 | if last_modified_hdr is None: | |
263 | return | |
264 | if not os.path.isfile(encodeFilename(filename)): | |
265 | return | |
266 | timestr = last_modified_hdr | |
267 | if timestr is None: | |
268 | return | |
269 | filetime = timeconvert(timestr) | |
270 | if filetime is None: | |
271 | return filetime | |
272 | try: | |
273 | os.utime(filename, (time.time(), filetime)) | |
274 | except: | |
275 | pass | |
276 | return filetime | |
277 | ||
278 | def report_writedescription(self, descfn): | |
279 | """ Report that the description file is being written """ | |
280 | self.to_screen(u'[info] Writing video description to: ' + descfn) | |
281 | ||
282 | def report_writesubtitles(self, srtfn): | |
283 | """ Report that the subtitles file is being written """ | |
284 | self.to_screen(u'[info] Writing video subtitles to: ' + srtfn) | |
285 | ||
286 | def report_writeinfojson(self, infofn): | |
287 | """ Report that the metadata file has been written """ | |
288 | self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn) | |
289 | ||
290 | def report_destination(self, filename): | |
291 | """Report destination filename.""" | |
292 | self.to_screen(u'[download] Destination: ' + filename) | |
293 | ||
294 | def report_progress(self, percent_str, data_len_str, speed_str, eta_str): | |
295 | """Report download progress.""" | |
296 | if self.params.get('noprogress', False): | |
297 | return | |
298 | self.to_screen(u'\r[download] %s of %s at %s ETA %s' % | |
299 | (percent_str, data_len_str, speed_str, eta_str), skip_eol=True) | |
300 | self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' % | |
301 | (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip())) | |
302 | ||
303 | def report_resuming_byte(self, resume_len): | |
304 | """Report attempt to resume at given byte.""" | |
305 | self.to_screen(u'[download] Resuming download at byte %s' % resume_len) | |
306 | ||
307 | def report_retry(self, count, retries): | |
308 | """Report retry in case of HTTP error 5xx""" | |
309 | self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries)) | |
310 | ||
311 | def report_file_already_downloaded(self, file_name): | |
312 | """Report file has already been fully downloaded.""" | |
313 | try: | |
314 | self.to_screen(u'[download] %s has already been downloaded' % file_name) | |
315 | except (UnicodeEncodeError) as err: | |
316 | self.to_screen(u'[download] The file has already been downloaded') | |
317 | ||
318 | def report_unable_to_resume(self): | |
319 | """Report it was impossible to resume download.""" | |
320 | self.to_screen(u'[download] Unable to resume') | |
321 | ||
322 | def report_finish(self): | |
323 | """Report download finished.""" | |
324 | if self.params.get('noprogress', False): | |
325 | self.to_screen(u'[download] Download completed') | |
326 | else: | |
327 | self.to_screen(u'') | |
328 | ||
329 | def increment_downloads(self): | |
330 | """Increment the ordinal that assigns a number to each file.""" | |
331 | self._num_downloads += 1 | |
332 | ||
333 | def prepare_filename(self, info_dict): | |
334 | """Generate the output filename.""" | |
335 | try: | |
336 | template_dict = dict(info_dict) | |
337 | ||
338 | template_dict['epoch'] = int(time.time()) | |
339 | template_dict['autonumber'] = u'%05d' % self._num_downloads | |
340 | ||
796173d0 PH |
341 | sanitize = lambda k,v: sanitize_filename( |
342 | u'NA' if v is None else compat_str(v), | |
343 | restricted=self.params.get('restrictfilenames'), | |
344 | is_id=(k==u'id')) | |
345 | template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items()) | |
59ae15a5 PH |
346 | |
347 | filename = self.params['outtmpl'] % template_dict | |
348 | return filename | |
349 | except (ValueError, KeyError) as err: | |
350 | self.trouble(u'ERROR: invalid system charset or erroneous output template') | |
351 | return None | |
352 | ||
353 | def _match_entry(self, info_dict): | |
354 | """ Returns None iff the file should be downloaded """ | |
355 | ||
356 | title = info_dict['title'] | |
357 | matchtitle = self.params.get('matchtitle', False) | |
358 | if matchtitle: | |
359 | matchtitle = matchtitle.decode('utf8') | |
360 | if not re.search(matchtitle, title, re.IGNORECASE): | |
361 | return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"' | |
362 | rejecttitle = self.params.get('rejecttitle', False) | |
363 | if rejecttitle: | |
364 | rejecttitle = rejecttitle.decode('utf8') | |
365 | if re.search(rejecttitle, title, re.IGNORECASE): | |
366 | return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"' | |
367 | return None | |
368 | ||
369 | def process_info(self, info_dict): | |
370 | """Process a single dictionary returned by an InfoExtractor.""" | |
371 | ||
372 | # Keep for backwards compatibility | |
373 | info_dict['stitle'] = info_dict['title'] | |
374 | ||
375 | if not 'format' in info_dict: | |
376 | info_dict['format'] = info_dict['ext'] | |
377 | ||
378 | reason = self._match_entry(info_dict) | |
379 | if reason is not None: | |
380 | self.to_screen(u'[download] ' + reason) | |
381 | return | |
382 | ||
383 | max_downloads = self.params.get('max_downloads') | |
384 | if max_downloads is not None: | |
385 | if self._num_downloads > int(max_downloads): | |
386 | raise MaxDownloadsReached() | |
387 | ||
388 | filename = self.prepare_filename(info_dict) | |
389 | ||
390 | # Forced printings | |
391 | if self.params.get('forcetitle', False): | |
392 | compat_print(info_dict['title']) | |
393 | if self.params.get('forceurl', False): | |
394 | compat_print(info_dict['url']) | |
395 | if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict: | |
396 | compat_print(info_dict['thumbnail']) | |
397 | if self.params.get('forcedescription', False) and 'description' in info_dict: | |
398 | compat_print(info_dict['description']) | |
399 | if self.params.get('forcefilename', False) and filename is not None: | |
400 | compat_print(filename) | |
401 | if self.params.get('forceformat', False): | |
402 | compat_print(info_dict['format']) | |
403 | ||
404 | # Do nothing else if in simulate mode | |
405 | if self.params.get('simulate', False): | |
406 | return | |
407 | ||
408 | if filename is None: | |
409 | return | |
410 | ||
411 | try: | |
412 | dn = os.path.dirname(encodeFilename(filename)) | |
413 | if dn != '' and not os.path.exists(dn): # dn is already encoded | |
414 | os.makedirs(dn) | |
415 | except (OSError, IOError) as err: | |
416 | self.trouble(u'ERROR: unable to create directory ' + compat_str(err)) | |
417 | return | |
418 | ||
419 | if self.params.get('writedescription', False): | |
420 | try: | |
421 | descfn = filename + u'.description' | |
422 | self.report_writedescription(descfn) | |
423 | descfile = open(encodeFilename(descfn), 'wb') | |
424 | try: | |
425 | descfile.write(info_dict['description'].encode('utf-8')) | |
426 | finally: | |
427 | descfile.close() | |
428 | except (OSError, IOError): | |
429 | self.trouble(u'ERROR: Cannot write description file ' + descfn) | |
430 | return | |
431 | ||
432 | if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']: | |
433 | # subtitles download errors are already managed as troubles in relevant IE | |
434 | # that way it will silently go on when used with unsupporting IE | |
435 | try: | |
436 | srtfn = filename.rsplit('.', 1)[0] + u'.srt' | |
437 | self.report_writesubtitles(srtfn) | |
438 | srtfile = open(encodeFilename(srtfn), 'wb') | |
439 | try: | |
440 | srtfile.write(info_dict['subtitles'].encode('utf-8')) | |
441 | finally: | |
442 | srtfile.close() | |
443 | except (OSError, IOError): | |
444 | self.trouble(u'ERROR: Cannot write subtitles file ' + descfn) | |
445 | return | |
446 | ||
447 | if self.params.get('writeinfojson', False): | |
448 | infofn = filename + u'.info.json' | |
449 | self.report_writeinfojson(infofn) | |
450 | try: | |
451 | json.dump | |
452 | except (NameError,AttributeError): | |
453 | self.trouble(u'ERROR: No JSON encoder found. Update to Python 2.6+, setup a json module, or leave out --write-info-json.') | |
454 | return | |
455 | try: | |
456 | infof = open(encodeFilename(infofn), 'wb') | |
457 | try: | |
458 | json_info_dict = dict((k,v) for k,v in info_dict.iteritems() if not k in ('urlhandle',)) | |
459 | json.dump(json_info_dict, infof) | |
460 | finally: | |
461 | infof.close() | |
462 | except (OSError, IOError): | |
463 | self.trouble(u'ERROR: Cannot write metadata to JSON file ' + infofn) | |
464 | return | |
465 | ||
466 | if not self.params.get('skip_download', False): | |
467 | if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)): | |
468 | success = True | |
469 | else: | |
470 | try: | |
471 | success = self._do_download(filename, info_dict) | |
472 | except (OSError, IOError) as err: | |
6ad98fb3 | 473 | raise UnavailableVideoError() |
59ae15a5 PH |
474 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: |
475 | self.trouble(u'ERROR: unable to download video data: %s' % str(err)) | |
476 | return | |
477 | except (ContentTooShortError, ) as err: | |
478 | self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) | |
479 | return | |
480 | ||
481 | if success: | |
482 | try: | |
483 | self.post_process(filename, info_dict) | |
484 | except (PostProcessingError) as err: | |
485 | self.trouble(u'ERROR: postprocessing: %s' % str(err)) | |
486 | return | |
487 | ||
488 | def download(self, url_list): | |
489 | """Download a given list of URLs.""" | |
490 | if len(url_list) > 1 and self.fixed_template(): | |
491 | raise SameFileError(self.params['outtmpl']) | |
492 | ||
493 | for url in url_list: | |
494 | suitable_found = False | |
495 | for ie in self._ies: | |
496 | # Go to next InfoExtractor if not suitable | |
497 | if not ie.suitable(url): | |
498 | continue | |
499 | ||
500 | # Warn if the _WORKING attribute is False | |
501 | if not ie.working(): | |
502 | self.trouble(u'WARNING: the program functionality for this site has been marked as broken, ' | |
503 | u'and will probably not work. If you want to go on, use the -i option.') | |
504 | ||
505 | # Suitable InfoExtractor found | |
506 | suitable_found = True | |
507 | ||
508 | # Extract information from URL and process it | |
509 | videos = ie.extract(url) | |
510 | for video in videos or []: | |
511 | video['extractor'] = ie.IE_NAME | |
512 | try: | |
513 | self.increment_downloads() | |
514 | self.process_info(video) | |
515 | except UnavailableVideoError: | |
516 | self.trouble(u'\nERROR: unable to download video') | |
517 | ||
518 | # Suitable InfoExtractor had been found; go to next URL | |
519 | break | |
520 | ||
521 | if not suitable_found: | |
522 | self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url) | |
523 | ||
524 | return self._download_retcode | |
525 | ||
526 | def post_process(self, filename, ie_info): | |
527 | """Run the postprocessing chain on the given file.""" | |
528 | info = dict(ie_info) | |
529 | info['filepath'] = filename | |
530 | for pp in self._pps: | |
531 | info = pp.run(info) | |
532 | if info is None: | |
533 | break | |
534 | ||
535 | def _download_with_rtmpdump(self, filename, url, player_url): | |
536 | self.report_destination(filename) | |
537 | tmpfilename = self.temp_name(filename) | |
538 | ||
539 | # Check for rtmpdump first | |
540 | try: | |
541 | subprocess.call(['rtmpdump', '-h'], stdout=(file(os.path.devnull, 'w')), stderr=subprocess.STDOUT) | |
542 | except (OSError, IOError): | |
543 | self.trouble(u'ERROR: RTMP download detected but "rtmpdump" could not be run') | |
544 | return False | |
545 | ||
546 | # Download using rtmpdump. rtmpdump returns exit code 2 when | |
547 | # the connection was interrumpted and resuming appears to be | |
548 | # possible. This is part of rtmpdump's normal usage, AFAIK. | |
549 | basic_args = ['rtmpdump', '-q'] + [[], ['-W', player_url]][player_url is not None] + ['-r', url, '-o', tmpfilename] | |
550 | args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)] | |
551 | if self.params.get('verbose', False): | |
552 | try: | |
553 | import pipes | |
554 | shell_quote = lambda args: ' '.join(map(pipes.quote, args)) | |
555 | except ImportError: | |
556 | shell_quote = repr | |
557 | self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args)) | |
558 | retval = subprocess.call(args) | |
559 | while retval == 2 or retval == 1: | |
560 | prevsize = os.path.getsize(encodeFilename(tmpfilename)) | |
561 | self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True) | |
562 | time.sleep(5.0) # This seems to be needed | |
563 | retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1]) | |
564 | cursize = os.path.getsize(encodeFilename(tmpfilename)) | |
565 | if prevsize == cursize and retval == 1: | |
566 | break | |
567 | # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those | |
568 | if prevsize == cursize and retval == 2 and cursize > 1024: | |
569 | self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.') | |
570 | retval = 0 | |
571 | break | |
572 | if retval == 0: | |
573 | self.to_screen(u'\r[rtmpdump] %s bytes' % os.path.getsize(encodeFilename(tmpfilename))) | |
574 | self.try_rename(tmpfilename, filename) | |
575 | return True | |
576 | else: | |
577 | self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval) | |
578 | return False | |
579 | ||
580 | def _do_download(self, filename, info_dict): | |
581 | url = info_dict['url'] | |
582 | player_url = info_dict.get('player_url', None) | |
583 | ||
584 | # Check file already present | |
585 | if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False): | |
586 | self.report_file_already_downloaded(filename) | |
587 | return True | |
588 | ||
589 | # Attempt to download using rtmpdump | |
590 | if url.startswith('rtmp'): | |
591 | return self._download_with_rtmpdump(filename, url, player_url) | |
592 | ||
593 | tmpfilename = self.temp_name(filename) | |
594 | stream = None | |
595 | ||
596 | # Do not include the Accept-Encoding header | |
597 | headers = {'Youtubedl-no-compression': 'True'} | |
598 | basic_request = compat_urllib_request.Request(url, None, headers) | |
599 | request = compat_urllib_request.Request(url, None, headers) | |
600 | ||
37c8fd48 FV |
601 | if self.params.get('test', False): |
602 | request.add_header('Range','bytes=0-10240') | |
603 | ||
59ae15a5 PH |
604 | # Establish possible resume length |
605 | if os.path.isfile(encodeFilename(tmpfilename)): | |
606 | resume_len = os.path.getsize(encodeFilename(tmpfilename)) | |
607 | else: | |
608 | resume_len = 0 | |
609 | ||
610 | open_mode = 'wb' | |
611 | if resume_len != 0: | |
612 | if self.params.get('continuedl', False): | |
613 | self.report_resuming_byte(resume_len) | |
614 | request.add_header('Range','bytes=%d-' % resume_len) | |
615 | open_mode = 'ab' | |
616 | else: | |
617 | resume_len = 0 | |
618 | ||
619 | count = 0 | |
620 | retries = self.params.get('retries', 0) | |
621 | while count <= retries: | |
622 | # Establish connection | |
623 | try: | |
624 | if count == 0 and 'urlhandle' in info_dict: | |
625 | data = info_dict['urlhandle'] | |
626 | data = compat_urllib_request.urlopen(request) | |
627 | break | |
628 | except (compat_urllib_error.HTTPError, ) as err: | |
629 | if (err.code < 500 or err.code >= 600) and err.code != 416: | |
630 | # Unexpected HTTP error | |
631 | raise | |
632 | elif err.code == 416: | |
633 | # Unable to resume (requested range not satisfiable) | |
634 | try: | |
635 | # Open the connection again without the range header | |
636 | data = compat_urllib_request.urlopen(basic_request) | |
637 | content_length = data.info()['Content-Length'] | |
638 | except (compat_urllib_error.HTTPError, ) as err: | |
639 | if err.code < 500 or err.code >= 600: | |
640 | raise | |
641 | else: | |
642 | # Examine the reported length | |
643 | if (content_length is not None and | |
644 | (resume_len - 100 < int(content_length) < resume_len + 100)): | |
645 | # The file had already been fully downloaded. | |
646 | # Explanation to the above condition: in issue #175 it was revealed that | |
647 | # YouTube sometimes adds or removes a few bytes from the end of the file, | |
648 | # changing the file size slightly and causing problems for some users. So | |
649 | # I decided to implement a suggested change and consider the file | |
650 | # completely downloaded if the file size differs less than 100 bytes from | |
651 | # the one in the hard drive. | |
652 | self.report_file_already_downloaded(filename) | |
653 | self.try_rename(tmpfilename, filename) | |
654 | return True | |
655 | else: | |
656 | # The length does not match, we start the download over | |
657 | self.report_unable_to_resume() | |
658 | open_mode = 'wb' | |
659 | break | |
660 | # Retry | |
661 | count += 1 | |
662 | if count <= retries: | |
663 | self.report_retry(count, retries) | |
664 | ||
665 | if count > retries: | |
666 | self.trouble(u'ERROR: giving up after %s retries' % retries) | |
667 | return False | |
668 | ||
669 | data_len = data.info().get('Content-length', None) | |
670 | if data_len is not None: | |
671 | data_len = int(data_len) + resume_len | |
672 | data_len_str = self.format_bytes(data_len) | |
673 | byte_counter = 0 + resume_len | |
674 | block_size = self.params.get('buffersize', 1024) | |
675 | start = time.time() | |
676 | while True: | |
677 | # Download and write | |
678 | before = time.time() | |
679 | data_block = data.read(block_size) | |
680 | after = time.time() | |
681 | if len(data_block) == 0: | |
682 | break | |
683 | byte_counter += len(data_block) | |
684 | ||
685 | # Open file just in time | |
686 | if stream is None: | |
687 | try: | |
688 | (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode) | |
689 | assert stream is not None | |
690 | filename = self.undo_temp_name(tmpfilename) | |
691 | self.report_destination(filename) | |
692 | except (OSError, IOError) as err: | |
693 | self.trouble(u'ERROR: unable to open for writing: %s' % str(err)) | |
694 | return False | |
695 | try: | |
696 | stream.write(data_block) | |
697 | except (IOError, OSError) as err: | |
698 | self.trouble(u'\nERROR: unable to write data: %s' % str(err)) | |
699 | return False | |
700 | if not self.params.get('noresizebuffer', False): | |
701 | block_size = self.best_block_size(after - before, len(data_block)) | |
702 | ||
703 | # Progress message | |
704 | speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len) | |
705 | if data_len is None: | |
706 | self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA') | |
707 | else: | |
708 | percent_str = self.calc_percent(byte_counter, data_len) | |
709 | eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len) | |
710 | self.report_progress(percent_str, data_len_str, speed_str, eta_str) | |
711 | ||
712 | # Apply rate limit | |
713 | self.slow_down(start, byte_counter - resume_len) | |
714 | ||
715 | if stream is None: | |
716 | self.trouble(u'\nERROR: Did not get any data blocks') | |
717 | return False | |
718 | stream.close() | |
719 | self.report_finish() | |
720 | if data_len is not None and byte_counter != data_len: | |
721 | raise ContentTooShortError(byte_counter, int(data_len)) | |
722 | self.try_rename(tmpfilename, filename) | |
723 | ||
724 | # Update file modification time | |
725 | if self.params.get('updatetime', True): | |
726 | info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None)) | |
727 | ||
728 | return True |