]>
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 | 6 | import math |
ce4be3a9 | 7 | import io |
d77c3dfd FV |
8 | import os |
9 | import re | |
10 | import socket | |
11 | import subprocess | |
12 | import sys | |
13 | import time | |
59ce2019 | 14 | import traceback |
d77c3dfd FV |
15 | |
16 | if os.name == 'nt': | |
59ae15a5 | 17 | import ctypes |
3eec021a | 18 | |
9e8056d5 | 19 | from .utils import * |
d77c3dfd FV |
20 | |
21 | ||
22 | class 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 | |
9e62bc44 | 81 | writesubtitles: Write the video subtitles to a file |
cdb130b0 | 82 | onlysubtitles: Downloads only the subtitles of the video |
ae608b80 | 83 | allsubtitles: Downloads all the subtitles of the video |
2a4093ea | 84 | listsubtitles: Lists all available subtitles for the video |
9e62bc44 | 85 | subtitlesformat: Subtitle format [sbv/srt] (default=srt) |
59ae15a5 | 86 | subtitleslang: Language of the subtitles to download |
37c8fd48 | 87 | test: Download only first bytes to test the downloader. |
7851b379 | 88 | keepvideo: Keep the video file after post-processing |
9e982f9e JC |
89 | min_filesize: Skip files smaller than this size |
90 | max_filesize: Skip files larger than this size | |
59ae15a5 PH |
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 = [] | |
bffbd5f0 | 104 | self._progress_hooks = [] |
59ae15a5 PH |
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.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.') | |
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 = 'bkMGTPEZY'[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: | |
906417c7 | 215 | self.to_screen('\033]0;%s\007' % message, skip_eol=True) |
59ae15a5 PH |
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 | ||
1c256f70 | 221 | def trouble(self, message=None, tb=None): |
59ae15a5 PH |
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. | |
01951dda PH |
227 | |
228 | tb, if given, is additional traceback information. | |
59ae15a5 PH |
229 | """ |
230 | if message is not None: | |
231 | self.to_stderr(message) | |
59ce2019 | 232 | if self.params.get('verbose'): |
1c256f70 | 233 | if tb is None: |
01951dda PH |
234 | tb_data = traceback.format_list(traceback.extract_stack()) |
235 | tb = u''.join(tb_data) | |
1c256f70 | 236 | self.to_stderr(tb) |
59ae15a5 PH |
237 | if not self.params.get('ignoreerrors', False): |
238 | raise DownloadError(message) | |
239 | self._download_retcode = 1 | |
240 | ||
241 | def slow_down(self, start_time, byte_counter): | |
242 | """Sleep if the download speed is over the rate limit.""" | |
243 | rate_limit = self.params.get('ratelimit', None) | |
244 | if rate_limit is None or byte_counter == 0: | |
245 | return | |
246 | now = time.time() | |
247 | elapsed = now - start_time | |
248 | if elapsed <= 0.0: | |
249 | return | |
250 | speed = float(byte_counter) / elapsed | |
251 | if speed > rate_limit: | |
252 | time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit) | |
253 | ||
254 | def temp_name(self, filename): | |
255 | """Returns a temporary filename for the given filename.""" | |
256 | if self.params.get('nopart', False) or filename == u'-' or \ | |
257 | (os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))): | |
258 | return filename | |
259 | return filename + u'.part' | |
260 | ||
261 | def undo_temp_name(self, filename): | |
262 | if filename.endswith(u'.part'): | |
263 | return filename[:-len(u'.part')] | |
264 | return filename | |
265 | ||
266 | def try_rename(self, old_filename, new_filename): | |
267 | try: | |
268 | if old_filename == new_filename: | |
269 | return | |
270 | os.rename(encodeFilename(old_filename), encodeFilename(new_filename)) | |
271 | except (IOError, OSError) as err: | |
272 | self.trouble(u'ERROR: unable to rename file') | |
273 | ||
274 | def try_utime(self, filename, last_modified_hdr): | |
275 | """Try to set the last-modified time of the given file.""" | |
276 | if last_modified_hdr is None: | |
277 | return | |
278 | if not os.path.isfile(encodeFilename(filename)): | |
279 | return | |
280 | timestr = last_modified_hdr | |
281 | if timestr is None: | |
282 | return | |
283 | filetime = timeconvert(timestr) | |
284 | if filetime is None: | |
285 | return filetime | |
286 | try: | |
287 | os.utime(filename, (time.time(), filetime)) | |
288 | except: | |
289 | pass | |
290 | return filetime | |
291 | ||
292 | def report_writedescription(self, descfn): | |
293 | """ Report that the description file is being written """ | |
294 | self.to_screen(u'[info] Writing video description to: ' + descfn) | |
295 | ||
553d0974 | 296 | def report_writesubtitles(self, sub_filename): |
59ae15a5 | 297 | """ Report that the subtitles file is being written """ |
553d0974 | 298 | self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename) |
59ae15a5 PH |
299 | |
300 | def report_writeinfojson(self, infofn): | |
301 | """ Report that the metadata file has been written """ | |
302 | self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn) | |
303 | ||
304 | def report_destination(self, filename): | |
305 | """Report destination filename.""" | |
306 | self.to_screen(u'[download] Destination: ' + filename) | |
307 | ||
308 | def report_progress(self, percent_str, data_len_str, speed_str, eta_str): | |
309 | """Report download progress.""" | |
310 | if self.params.get('noprogress', False): | |
311 | return | |
5717d91a | 312 | if self.params.get('progress_with_newline', False): |
1528d664 | 313 | self.to_screen(u'[download] %s of %s at %s ETA %s' % |
7311fef8 | 314 | (percent_str, data_len_str, speed_str, eta_str)) |
5717d91a PH |
315 | else: |
316 | self.to_screen(u'\r[download] %s of %s at %s ETA %s' % | |
59ae15a5 PH |
317 | (percent_str, data_len_str, speed_str, eta_str), skip_eol=True) |
318 | self.to_cons_title(u'youtube-dl - %s of %s at %s ETA %s' % | |
319 | (percent_str.strip(), data_len_str.strip(), speed_str.strip(), eta_str.strip())) | |
320 | ||
321 | def report_resuming_byte(self, resume_len): | |
322 | """Report attempt to resume at given byte.""" | |
323 | self.to_screen(u'[download] Resuming download at byte %s' % resume_len) | |
324 | ||
325 | def report_retry(self, count, retries): | |
326 | """Report retry in case of HTTP error 5xx""" | |
327 | self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries)) | |
328 | ||
329 | def report_file_already_downloaded(self, file_name): | |
330 | """Report file has already been fully downloaded.""" | |
331 | try: | |
332 | self.to_screen(u'[download] %s has already been downloaded' % file_name) | |
333 | except (UnicodeEncodeError) as err: | |
334 | self.to_screen(u'[download] The file has already been downloaded') | |
335 | ||
336 | def report_unable_to_resume(self): | |
337 | """Report it was impossible to resume download.""" | |
338 | self.to_screen(u'[download] Unable to resume') | |
339 | ||
340 | def report_finish(self): | |
341 | """Report download finished.""" | |
342 | if self.params.get('noprogress', False): | |
343 | self.to_screen(u'[download] Download completed') | |
344 | else: | |
345 | self.to_screen(u'') | |
346 | ||
347 | def increment_downloads(self): | |
348 | """Increment the ordinal that assigns a number to each file.""" | |
349 | self._num_downloads += 1 | |
350 | ||
351 | def prepare_filename(self, info_dict): | |
352 | """Generate the output filename.""" | |
353 | try: | |
354 | template_dict = dict(info_dict) | |
355 | ||
356 | template_dict['epoch'] = int(time.time()) | |
357 | template_dict['autonumber'] = u'%05d' % self._num_downloads | |
358 | ||
796173d0 PH |
359 | sanitize = lambda k,v: sanitize_filename( |
360 | u'NA' if v is None else compat_str(v), | |
361 | restricted=self.params.get('restrictfilenames'), | |
362 | is_id=(k==u'id')) | |
363 | template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items()) | |
59ae15a5 PH |
364 | |
365 | filename = self.params['outtmpl'] % template_dict | |
366 | return filename | |
367 | except (ValueError, KeyError) as err: | |
368 | self.trouble(u'ERROR: invalid system charset or erroneous output template') | |
369 | return None | |
370 | ||
371 | def _match_entry(self, info_dict): | |
372 | """ Returns None iff the file should be downloaded """ | |
373 | ||
374 | title = info_dict['title'] | |
375 | matchtitle = self.params.get('matchtitle', False) | |
376 | if matchtitle: | |
59ae15a5 PH |
377 | if not re.search(matchtitle, title, re.IGNORECASE): |
378 | return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"' | |
379 | rejecttitle = self.params.get('rejecttitle', False) | |
380 | if rejecttitle: | |
59ae15a5 PH |
381 | if re.search(rejecttitle, title, re.IGNORECASE): |
382 | return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"' | |
383 | return None | |
384 | ||
385 | def process_info(self, info_dict): | |
386 | """Process a single dictionary returned by an InfoExtractor.""" | |
387 | ||
388 | # Keep for backwards compatibility | |
389 | info_dict['stitle'] = info_dict['title'] | |
390 | ||
391 | if not 'format' in info_dict: | |
392 | info_dict['format'] = info_dict['ext'] | |
393 | ||
394 | reason = self._match_entry(info_dict) | |
395 | if reason is not None: | |
396 | self.to_screen(u'[download] ' + reason) | |
397 | return | |
398 | ||
399 | max_downloads = self.params.get('max_downloads') | |
400 | if max_downloads is not None: | |
401 | if self._num_downloads > int(max_downloads): | |
402 | raise MaxDownloadsReached() | |
403 | ||
404 | filename = self.prepare_filename(info_dict) | |
405 | ||
406 | # Forced printings | |
407 | if self.params.get('forcetitle', False): | |
408 | compat_print(info_dict['title']) | |
409 | if self.params.get('forceurl', False): | |
410 | compat_print(info_dict['url']) | |
411 | if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict: | |
412 | compat_print(info_dict['thumbnail']) | |
413 | if self.params.get('forcedescription', False) and 'description' in info_dict: | |
414 | compat_print(info_dict['description']) | |
415 | if self.params.get('forcefilename', False) and filename is not None: | |
416 | compat_print(filename) | |
417 | if self.params.get('forceformat', False): | |
418 | compat_print(info_dict['format']) | |
419 | ||
420 | # Do nothing else if in simulate mode | |
421 | if self.params.get('simulate', False): | |
422 | return | |
423 | ||
424 | if filename is None: | |
425 | return | |
426 | ||
427 | try: | |
428 | dn = os.path.dirname(encodeFilename(filename)) | |
429 | if dn != '' and not os.path.exists(dn): # dn is already encoded | |
430 | os.makedirs(dn) | |
431 | except (OSError, IOError) as err: | |
432 | self.trouble(u'ERROR: unable to create directory ' + compat_str(err)) | |
433 | return | |
434 | ||
435 | if self.params.get('writedescription', False): | |
436 | try: | |
437 | descfn = filename + u'.description' | |
438 | self.report_writedescription(descfn) | |
bfa6389b PH |
439 | with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: |
440 | descfile.write(info_dict['description']) | |
59ae15a5 PH |
441 | except (OSError, IOError): |
442 | self.trouble(u'ERROR: Cannot write description file ' + descfn) | |
443 | return | |
444 | ||
445 | if self.params.get('writesubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']: | |
446 | # subtitles download errors are already managed as troubles in relevant IE | |
447 | # that way it will silently go on when used with unsupporting IE | |
ae608b80 | 448 | subtitle = info_dict['subtitles'][0] |
553d0974 | 449 | (sub_error, sub_lang, sub) = subtitle |
9e62bc44 | 450 | sub_format = self.params.get('subtitlesformat') |
59ae15a5 | 451 | try: |
9e62bc44 | 452 | sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format |
553d0974 IM |
453 | self.report_writesubtitles(sub_filename) |
454 | with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile: | |
455 | subfile.write(sub) | |
59ae15a5 PH |
456 | except (OSError, IOError): |
457 | self.trouble(u'ERROR: Cannot write subtitles file ' + descfn) | |
458 | return | |
ae608b80 IM |
459 | if self.params.get('onlysubtitles', False): |
460 | return | |
461 | ||
462 | if self.params.get('allsubtitles', False) and 'subtitles' in info_dict and info_dict['subtitles']: | |
463 | subtitles = info_dict['subtitles'] | |
9e62bc44 | 464 | sub_format = self.params.get('subtitlesformat') |
ae608b80 | 465 | for subtitle in subtitles: |
553d0974 | 466 | (sub_error, sub_lang, sub) = subtitle |
ae608b80 | 467 | try: |
9e62bc44 | 468 | sub_filename = filename.rsplit('.', 1)[0] + u'.' + sub_lang + u'.' + sub_format |
553d0974 IM |
469 | self.report_writesubtitles(sub_filename) |
470 | with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile: | |
471 | subfile.write(sub) | |
ae608b80 IM |
472 | except (OSError, IOError): |
473 | self.trouble(u'ERROR: Cannot write subtitles file ' + descfn) | |
474 | return | |
475 | if self.params.get('onlysubtitles', False): | |
476 | return | |
59ae15a5 PH |
477 | |
478 | if self.params.get('writeinfojson', False): | |
479 | infofn = filename + u'.info.json' | |
480 | self.report_writeinfojson(infofn) | |
59ae15a5 | 481 | try: |
f4bfd65f PH |
482 | json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle']) |
483 | write_json_file(json_info_dict, encodeFilename(infofn)) | |
59ae15a5 PH |
484 | except (OSError, IOError): |
485 | self.trouble(u'ERROR: Cannot write metadata to JSON file ' + infofn) | |
486 | return | |
487 | ||
488 | if not self.params.get('skip_download', False): | |
489 | if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)): | |
490 | success = True | |
491 | else: | |
492 | try: | |
493 | success = self._do_download(filename, info_dict) | |
494 | except (OSError, IOError) as err: | |
6ad98fb3 | 495 | raise UnavailableVideoError() |
59ae15a5 PH |
496 | except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: |
497 | self.trouble(u'ERROR: unable to download video data: %s' % str(err)) | |
498 | return | |
499 | except (ContentTooShortError, ) as err: | |
500 | self.trouble(u'ERROR: content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) | |
501 | return | |
502 | ||
503 | if success: | |
504 | try: | |
505 | self.post_process(filename, info_dict) | |
506 | except (PostProcessingError) as err: | |
507 | self.trouble(u'ERROR: postprocessing: %s' % str(err)) | |
508 | return | |
509 | ||
510 | def download(self, url_list): | |
511 | """Download a given list of URLs.""" | |
512 | if len(url_list) > 1 and self.fixed_template(): | |
513 | raise SameFileError(self.params['outtmpl']) | |
514 | ||
515 | for url in url_list: | |
516 | suitable_found = False | |
517 | for ie in self._ies: | |
518 | # Go to next InfoExtractor if not suitable | |
519 | if not ie.suitable(url): | |
520 | continue | |
521 | ||
522 | # Warn if the _WORKING attribute is False | |
523 | if not ie.working(): | |
1c256f70 PH |
524 | self.to_stderr(u'WARNING: the program functionality for this site has been marked as broken, ' |
525 | u'and will probably not work. If you want to go on, use the -i option.') | |
59ae15a5 PH |
526 | |
527 | # Suitable InfoExtractor found | |
528 | suitable_found = True | |
529 | ||
530 | # Extract information from URL and process it | |
1c256f70 PH |
531 | try: |
532 | videos = ie.extract(url) | |
533 | except ExtractorError as de: # An error we somewhat expected | |
01951dda | 534 | self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback()) |
1c256f70 PH |
535 | break |
536 | except Exception as e: | |
537 | if self.params.get('ignoreerrors', False): | |
538 | self.trouble(u'ERROR: ' + compat_str(e), tb=compat_str(traceback.format_exc())) | |
539 | break | |
540 | else: | |
541 | raise | |
95fedbf8 | 542 | |
0214ce7c | 543 | if len(videos or []) > 1 and self.fixed_template(): |
95fedbf8 FV |
544 | raise SameFileError(self.params['outtmpl']) |
545 | ||
59ae15a5 PH |
546 | for video in videos or []: |
547 | video['extractor'] = ie.IE_NAME | |
548 | try: | |
549 | self.increment_downloads() | |
550 | self.process_info(video) | |
551 | except UnavailableVideoError: | |
552 | self.trouble(u'\nERROR: unable to download video') | |
553 | ||
554 | # Suitable InfoExtractor had been found; go to next URL | |
555 | break | |
556 | ||
557 | if not suitable_found: | |
558 | self.trouble(u'ERROR: no suitable InfoExtractor: %s' % url) | |
559 | ||
560 | return self._download_retcode | |
561 | ||
562 | def post_process(self, filename, ie_info): | |
7851b379 | 563 | """Run all the postprocessors on the given file.""" |
59ae15a5 PH |
564 | info = dict(ie_info) |
565 | info['filepath'] = filename | |
7851b379 | 566 | keep_video = None |
59ae15a5 | 567 | for pp in self._pps: |
7851b379 PH |
568 | try: |
569 | keep_video_wish,new_info = pp.run(info) | |
570 | if keep_video_wish is not None: | |
571 | if keep_video_wish: | |
572 | keep_video = keep_video_wish | |
573 | elif keep_video is None: | |
574 | # No clear decision yet, let IE decide | |
575 | keep_video = keep_video_wish | |
576 | except PostProcessingError as e: | |
577 | self.to_stderr(u'ERROR: ' + e.msg) | |
1d16b0c3 | 578 | if keep_video is False and not self.params.get('keepvideo', False): |
7851b379 PH |
579 | try: |
580 | self.to_stderr(u'Deleting original file %s (pass -k to keep)' % filename) | |
581 | os.remove(encodeFilename(filename)) | |
582 | except (IOError, OSError): | |
583 | self.to_stderr(u'WARNING: Unable to remove downloaded video file') | |
59ae15a5 | 584 | |
f5ebb614 | 585 | def _download_with_rtmpdump(self, filename, url, player_url, page_url): |
59ae15a5 PH |
586 | self.report_destination(filename) |
587 | tmpfilename = self.temp_name(filename) | |
588 | ||
589 | # Check for rtmpdump first | |
590 | try: | |
591 | subprocess.call(['rtmpdump', '-h'], stdout=(file(os.path.devnull, 'w')), stderr=subprocess.STDOUT) | |
592 | except (OSError, IOError): | |
593 | self.trouble(u'ERROR: RTMP download detected but "rtmpdump" could not be run') | |
594 | return False | |
595 | ||
596 | # Download using rtmpdump. rtmpdump returns exit code 2 when | |
597 | # the connection was interrumpted and resuming appears to be | |
598 | # possible. This is part of rtmpdump's normal usage, AFAIK. | |
f5ebb614 PH |
599 | basic_args = ['rtmpdump', '-q', '-r', url, '-o', tmpfilename] |
600 | if player_url is not None: | |
601 | basic_args += ['-W', player_url] | |
602 | if page_url is not None: | |
603 | basic_args += ['--pageUrl', page_url] | |
59ae15a5 PH |
604 | args = basic_args + [[], ['-e', '-k', '1']][self.params.get('continuedl', False)] |
605 | if self.params.get('verbose', False): | |
606 | try: | |
607 | import pipes | |
608 | shell_quote = lambda args: ' '.join(map(pipes.quote, args)) | |
609 | except ImportError: | |
610 | shell_quote = repr | |
611 | self.to_screen(u'[debug] rtmpdump command line: ' + shell_quote(args)) | |
612 | retval = subprocess.call(args) | |
613 | while retval == 2 or retval == 1: | |
614 | prevsize = os.path.getsize(encodeFilename(tmpfilename)) | |
615 | self.to_screen(u'\r[rtmpdump] %s bytes' % prevsize, skip_eol=True) | |
616 | time.sleep(5.0) # This seems to be needed | |
617 | retval = subprocess.call(basic_args + ['-e'] + [[], ['-k', '1']][retval == 1]) | |
618 | cursize = os.path.getsize(encodeFilename(tmpfilename)) | |
619 | if prevsize == cursize and retval == 1: | |
620 | break | |
621 | # Some rtmp streams seem abort after ~ 99.8%. Don't complain for those | |
622 | if prevsize == cursize and retval == 2 and cursize > 1024: | |
623 | self.to_screen(u'\r[rtmpdump] Could not download the whole video. This can happen for some advertisements.') | |
624 | retval = 0 | |
625 | break | |
626 | if retval == 0: | |
bffbd5f0 PH |
627 | fsize = os.path.getsize(encodeFilename(tmpfilename)) |
628 | self.to_screen(u'\r[rtmpdump] %s bytes' % fsize) | |
59ae15a5 | 629 | self.try_rename(tmpfilename, filename) |
bffbd5f0 PH |
630 | self._hook_progress({ |
631 | 'downloaded_bytes': fsize, | |
632 | 'total_bytes': fsize, | |
633 | 'filename': filename, | |
634 | 'status': 'finished', | |
635 | }) | |
59ae15a5 PH |
636 | return True |
637 | else: | |
638 | self.trouble(u'\nERROR: rtmpdump exited with code %d' % retval) | |
639 | return False | |
640 | ||
641 | def _do_download(self, filename, info_dict): | |
642 | url = info_dict['url'] | |
59ae15a5 PH |
643 | |
644 | # Check file already present | |
645 | if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False): | |
646 | self.report_file_already_downloaded(filename) | |
bffbd5f0 PH |
647 | self._hook_progress({ |
648 | 'filename': filename, | |
649 | 'status': 'finished', | |
650 | }) | |
59ae15a5 PH |
651 | return True |
652 | ||
653 | # Attempt to download using rtmpdump | |
654 | if url.startswith('rtmp'): | |
f5ebb614 PH |
655 | return self._download_with_rtmpdump(filename, url, |
656 | info_dict.get('player_url', None), | |
657 | info_dict.get('page_url', None)) | |
59ae15a5 PH |
658 | |
659 | tmpfilename = self.temp_name(filename) | |
660 | stream = None | |
661 | ||
662 | # Do not include the Accept-Encoding header | |
663 | headers = {'Youtubedl-no-compression': 'True'} | |
3446dfb7 PH |
664 | if 'user_agent' in info_dict: |
665 | headers['Youtubedl-user-agent'] = info_dict['user_agent'] | |
59ae15a5 PH |
666 | basic_request = compat_urllib_request.Request(url, None, headers) |
667 | request = compat_urllib_request.Request(url, None, headers) | |
668 | ||
37c8fd48 FV |
669 | if self.params.get('test', False): |
670 | request.add_header('Range','bytes=0-10240') | |
671 | ||
59ae15a5 PH |
672 | # Establish possible resume length |
673 | if os.path.isfile(encodeFilename(tmpfilename)): | |
674 | resume_len = os.path.getsize(encodeFilename(tmpfilename)) | |
675 | else: | |
676 | resume_len = 0 | |
677 | ||
678 | open_mode = 'wb' | |
679 | if resume_len != 0: | |
680 | if self.params.get('continuedl', False): | |
681 | self.report_resuming_byte(resume_len) | |
682 | request.add_header('Range','bytes=%d-' % resume_len) | |
683 | open_mode = 'ab' | |
684 | else: | |
685 | resume_len = 0 | |
686 | ||
687 | count = 0 | |
688 | retries = self.params.get('retries', 0) | |
689 | while count <= retries: | |
690 | # Establish connection | |
691 | try: | |
692 | if count == 0 and 'urlhandle' in info_dict: | |
693 | data = info_dict['urlhandle'] | |
694 | data = compat_urllib_request.urlopen(request) | |
695 | break | |
696 | except (compat_urllib_error.HTTPError, ) as err: | |
697 | if (err.code < 500 or err.code >= 600) and err.code != 416: | |
698 | # Unexpected HTTP error | |
699 | raise | |
700 | elif err.code == 416: | |
701 | # Unable to resume (requested range not satisfiable) | |
702 | try: | |
703 | # Open the connection again without the range header | |
704 | data = compat_urllib_request.urlopen(basic_request) | |
705 | content_length = data.info()['Content-Length'] | |
706 | except (compat_urllib_error.HTTPError, ) as err: | |
707 | if err.code < 500 or err.code >= 600: | |
708 | raise | |
709 | else: | |
710 | # Examine the reported length | |
711 | if (content_length is not None and | |
712 | (resume_len - 100 < int(content_length) < resume_len + 100)): | |
713 | # The file had already been fully downloaded. | |
714 | # Explanation to the above condition: in issue #175 it was revealed that | |
715 | # YouTube sometimes adds or removes a few bytes from the end of the file, | |
716 | # changing the file size slightly and causing problems for some users. So | |
717 | # I decided to implement a suggested change and consider the file | |
718 | # completely downloaded if the file size differs less than 100 bytes from | |
719 | # the one in the hard drive. | |
720 | self.report_file_already_downloaded(filename) | |
721 | self.try_rename(tmpfilename, filename) | |
bffbd5f0 PH |
722 | self._hook_progress({ |
723 | 'filename': filename, | |
724 | 'status': 'finished', | |
725 | }) | |
59ae15a5 PH |
726 | return True |
727 | else: | |
728 | # The length does not match, we start the download over | |
729 | self.report_unable_to_resume() | |
730 | open_mode = 'wb' | |
731 | break | |
732 | # Retry | |
733 | count += 1 | |
734 | if count <= retries: | |
735 | self.report_retry(count, retries) | |
736 | ||
737 | if count > retries: | |
738 | self.trouble(u'ERROR: giving up after %s retries' % retries) | |
739 | return False | |
740 | ||
741 | data_len = data.info().get('Content-length', None) | |
742 | if data_len is not None: | |
743 | data_len = int(data_len) + resume_len | |
9e982f9e JC |
744 | min_data_len = self.params.get("min_filesize", None) |
745 | max_data_len = self.params.get("max_filesize", None) | |
746 | if min_data_len is not None and data_len < min_data_len: | |
747 | self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len)) | |
748 | return False | |
749 | if max_data_len is not None and data_len > max_data_len: | |
750 | self.to_screen(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len)) | |
751 | return False | |
752 | ||
59ae15a5 PH |
753 | data_len_str = self.format_bytes(data_len) |
754 | byte_counter = 0 + resume_len | |
755 | block_size = self.params.get('buffersize', 1024) | |
756 | start = time.time() | |
757 | while True: | |
758 | # Download and write | |
759 | before = time.time() | |
760 | data_block = data.read(block_size) | |
761 | after = time.time() | |
762 | if len(data_block) == 0: | |
763 | break | |
764 | byte_counter += len(data_block) | |
765 | ||
766 | # Open file just in time | |
767 | if stream is None: | |
768 | try: | |
769 | (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode) | |
770 | assert stream is not None | |
771 | filename = self.undo_temp_name(tmpfilename) | |
772 | self.report_destination(filename) | |
773 | except (OSError, IOError) as err: | |
774 | self.trouble(u'ERROR: unable to open for writing: %s' % str(err)) | |
775 | return False | |
776 | try: | |
777 | stream.write(data_block) | |
778 | except (IOError, OSError) as err: | |
779 | self.trouble(u'\nERROR: unable to write data: %s' % str(err)) | |
780 | return False | |
781 | if not self.params.get('noresizebuffer', False): | |
782 | block_size = self.best_block_size(after - before, len(data_block)) | |
783 | ||
784 | # Progress message | |
785 | speed_str = self.calc_speed(start, time.time(), byte_counter - resume_len) | |
786 | if data_len is None: | |
787 | self.report_progress('Unknown %', data_len_str, speed_str, 'Unknown ETA') | |
788 | else: | |
789 | percent_str = self.calc_percent(byte_counter, data_len) | |
790 | eta_str = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len) | |
791 | self.report_progress(percent_str, data_len_str, speed_str, eta_str) | |
792 | ||
bffbd5f0 PH |
793 | self._hook_progress({ |
794 | 'downloaded_bytes': byte_counter, | |
795 | 'total_bytes': data_len, | |
796 | 'tmpfilename': tmpfilename, | |
797 | 'filename': filename, | |
798 | 'status': 'downloading', | |
799 | }) | |
800 | ||
59ae15a5 PH |
801 | # Apply rate limit |
802 | self.slow_down(start, byte_counter - resume_len) | |
803 | ||
804 | if stream is None: | |
805 | self.trouble(u'\nERROR: Did not get any data blocks') | |
806 | return False | |
807 | stream.close() | |
808 | self.report_finish() | |
809 | if data_len is not None and byte_counter != data_len: | |
810 | raise ContentTooShortError(byte_counter, int(data_len)) | |
811 | self.try_rename(tmpfilename, filename) | |
812 | ||
813 | # Update file modification time | |
814 | if self.params.get('updatetime', True): | |
815 | info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None)) | |
816 | ||
bffbd5f0 PH |
817 | self._hook_progress({ |
818 | 'downloaded_bytes': byte_counter, | |
819 | 'total_bytes': byte_counter, | |
820 | 'filename': filename, | |
821 | 'status': 'finished', | |
822 | }) | |
823 | ||
59ae15a5 | 824 | return True |
bffbd5f0 PH |
825 | |
826 | def _hook_progress(self, status): | |
827 | for ph in self._progress_hooks: | |
828 | ph(status) | |
829 | ||
830 | def add_progress_hook(self, ph): | |
831 | """ ph gets called on download progress, with a dictionary with the entries | |
832 | * filename: The final filename | |
833 | * status: One of "downloading" and "finished" | |
834 | ||
835 | It can also have some of the following entries: | |
836 | ||
837 | * downloaded_bytes: Bytes on disks | |
838 | * total_bytes: Total bytes, None if unknown | |
839 | * tmpfilename: The filename we're currently writing to | |
840 | ||
841 | Hooks are guaranteed to be called at least once (with status "finished") | |
842 | if the download is successful. | |
843 | """ | |
844 | self._progress_hooks.append(ph) |