]>
Raphaƫl G. Git Repositories - youtubedl/blob - youtube_dl/downloader/common.py
1 from __future__
import division
, unicode_literals
9 from ..compat
import compat_os_name
20 class FileDownloader(object):
21 """File Downloader class.
23 File downloader objects are the ones responsible of downloading the
24 actual video file and writing it to disk.
26 File downloaders accept a lot of parameters. In order not to saturate
27 the object constructor with arguments, it receives a dictionary of
32 verbose: Print additional info to stdout.
33 quiet: Do not print messages to stdout.
34 ratelimit: Download speed limit, in bytes/sec.
35 retries: Number of times to retry for HTTP error 5xx
36 buffersize: Size of download buffer in bytes.
37 noresizebuffer: Do not automatically resize the download buffer.
38 continuedl: Try to continue downloads if possible.
39 noprogress: Do not print the progress bar.
40 logtostderr: Log messages to stderr instead of stdout.
41 consoletitle: Display progress in console window's titlebar.
42 nopart: Do not use temporary .part files.
43 updatetime: Use the Last-modified header to set output file timestamps.
44 test: Download only first bytes to test the downloader.
45 min_filesize: Skip files smaller than this size
46 max_filesize: Skip files larger than this size
47 xattr_set_filesize: Set ytdl.filesize user xattribute with expected size.
48 external_downloader_args: A list of additional command-line arguments for the
50 hls_use_mpegts: Use the mpegts container for HLS videos.
51 http_chunk_size: Size of a chunk for chunk-based HTTP downloading. May be
52 useful for bypassing bandwidth throttling imposed by
53 a webserver (experimental)
55 Subclasses of this one must re-define the real_download method.
58 _TEST_FILE_SIZE
= 10241
61 def __init__(self
, ydl
, params
):
62 """Create a FileDownloader object with the given options."""
64 self
._progress
_hooks
= []
66 self
.add_progress_hook(self
.report_progress
)
69 def format_seconds(seconds
):
70 (mins
, secs
) = divmod(seconds
, 60)
71 (hours
, mins
) = divmod(mins
, 60)
75 return '%02d:%02d' % (mins
, secs
)
77 return '%02d:%02d:%02d' % (hours
, mins
, secs
)
80 def calc_percent(byte_counter
, data_len
):
83 return float(byte_counter
) / float(data_len
) * 100.0
86 def format_percent(percent
):
89 return '%6s' % ('%3.1f%%' % percent
)
92 def calc_eta(start
, now
, total
, current
):
98 if current
== 0 or dif
< 0.001: # One millisecond
100 rate
= float(current
) / dif
101 return int((float(total
) - float(current
)) / rate
)
107 return FileDownloader
.format_seconds(eta
)
110 def calc_speed(start
, now
, bytes):
112 if bytes == 0 or dif
< 0.001: # One millisecond
114 return float(bytes) / dif
117 def format_speed(speed
):
119 return '%10s' % '---b/s'
120 return '%10s' % ('%s/s' % format_bytes(speed
))
123 def format_retries(retries
):
124 return 'inf' if retries
== float('inf') else '%.0f' % retries
127 def best_block_size(elapsed_time
, bytes):
128 new_min
= max(bytes / 2.0, 1.0)
129 new_max
= min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
130 if elapsed_time
< 0.001:
132 rate
= bytes / elapsed_time
140 def parse_bytes(bytestr
):
141 """Parse a string indicating a byte quantity into an integer."""
142 matchobj
= re
.match(r
'(?i)^(\d+(?:\.\d+)?)([kMGTPEZY]?)$', bytestr
)
145 number
= float(matchobj
.group(1))
146 multiplier
= 1024.0 ** 'bkmgtpezy'.index(matchobj
.group(2).lower())
147 return int(round(number
* multiplier
))
149 def to_screen(self
, *args
, **kargs
):
150 self
.ydl
.to_screen(*args
, **kargs
)
152 def to_stderr(self
, message
):
153 self
.ydl
.to_screen(message
)
155 def to_console_title(self
, message
):
156 self
.ydl
.to_console_title(message
)
158 def trouble(self
, *args
, **kargs
):
159 self
.ydl
.trouble(*args
, **kargs
)
161 def report_warning(self
, *args
, **kargs
):
162 self
.ydl
.report_warning(*args
, **kargs
)
164 def report_error(self
, *args
, **kargs
):
165 self
.ydl
.report_error(*args
, **kargs
)
167 def slow_down(self
, start_time
, now
, byte_counter
):
168 """Sleep if the download speed is over the rate limit."""
169 rate_limit
= self
.params
.get('ratelimit')
170 if rate_limit
is None or byte_counter
== 0:
174 elapsed
= now
- start_time
177 speed
= float(byte_counter
) / elapsed
178 if speed
> rate_limit
:
179 time
.sleep(max((byte_counter
// rate_limit
) - elapsed
, 0))
181 def temp_name(self
, filename
):
182 """Returns a temporary filename for the given filename."""
183 if self
.params
.get('nopart', False) or filename
== '-' or \
184 (os
.path
.exists(encodeFilename(filename
)) and not os
.path
.isfile(encodeFilename(filename
))):
186 return filename
+ '.part'
188 def undo_temp_name(self
, filename
):
189 if filename
.endswith('.part'):
190 return filename
[:-len('.part')]
193 def ytdl_filename(self
, filename
):
194 return filename
+ '.ytdl'
196 def try_rename(self
, old_filename
, new_filename
):
198 if old_filename
== new_filename
:
200 os
.rename(encodeFilename(old_filename
), encodeFilename(new_filename
))
201 except (IOError, OSError) as err
:
202 self
.report_error('unable to rename file: %s' % error_to_compat_str(err
))
204 def try_utime(self
, filename
, last_modified_hdr
):
205 """Try to set the last-modified time of the given file."""
206 if last_modified_hdr
is None:
208 if not os
.path
.isfile(encodeFilename(filename
)):
210 timestr
= last_modified_hdr
213 filetime
= timeconvert(timestr
)
216 # Ignore obviously invalid dates
220 os
.utime(filename
, (time
.time(), filetime
))
225 def report_destination(self
, filename
):
226 """Report destination filename."""
227 self
.to_screen('[download] Destination: ' + filename
)
229 def _report_progress_status(self
, msg
, is_last_line
=False):
230 fullmsg
= '[download] ' + msg
231 if self
.params
.get('progress_with_newline', False):
232 self
.to_screen(fullmsg
)
234 if compat_os_name
== 'nt':
235 prev_len
= getattr(self
, '_report_progress_prev_line_length',
237 if prev_len
> len(fullmsg
):
238 fullmsg
+= ' ' * (prev_len
- len(fullmsg
))
239 self
._report
_progress
_prev
_line
_length
= len(fullmsg
)
242 clear_line
= ('\r\x1b[K' if sys
.stderr
.isatty() else '\r')
243 self
.to_screen(clear_line
+ fullmsg
, skip_eol
=not is_last_line
)
244 self
.to_console_title('youtube-dl ' + msg
)
246 def report_progress(self
, s
):
247 if s
['status'] == 'finished':
248 if self
.params
.get('noprogress', False):
249 self
.to_screen('[download] Download completed')
251 msg_template
= '100%%'
252 if s
.get('total_bytes') is not None:
253 s
['_total_bytes_str'] = format_bytes(s
['total_bytes'])
254 msg_template
+= ' of %(_total_bytes_str)s'
255 if s
.get('elapsed') is not None:
256 s
['_elapsed_str'] = self
.format_seconds(s
['elapsed'])
257 msg_template
+= ' in %(_elapsed_str)s'
258 self
._report
_progress
_status
(
259 msg_template
% s
, is_last_line
=True)
261 if self
.params
.get('noprogress'):
264 if s
['status'] != 'downloading':
267 if s
.get('eta') is not None:
268 s
['_eta_str'] = self
.format_eta(s
['eta'])
270 s
['_eta_str'] = 'Unknown ETA'
272 if s
.get('total_bytes') and s
.get('downloaded_bytes') is not None:
273 s
['_percent_str'] = self
.format_percent(100 * s
['downloaded_bytes'] / s
['total_bytes'])
274 elif s
.get('total_bytes_estimate') and s
.get('downloaded_bytes') is not None:
275 s
['_percent_str'] = self
.format_percent(100 * s
['downloaded_bytes'] / s
['total_bytes_estimate'])
277 if s
.get('downloaded_bytes') == 0:
278 s
['_percent_str'] = self
.format_percent(0)
280 s
['_percent_str'] = 'Unknown %'
282 if s
.get('speed') is not None:
283 s
['_speed_str'] = self
.format_speed(s
['speed'])
285 s
['_speed_str'] = 'Unknown speed'
287 if s
.get('total_bytes') is not None:
288 s
['_total_bytes_str'] = format_bytes(s
['total_bytes'])
289 msg_template
= '%(_percent_str)s of %(_total_bytes_str)s at %(_speed_str)s ETA %(_eta_str)s'
290 elif s
.get('total_bytes_estimate') is not None:
291 s
['_total_bytes_estimate_str'] = format_bytes(s
['total_bytes_estimate'])
292 msg_template
= '%(_percent_str)s of ~%(_total_bytes_estimate_str)s at %(_speed_str)s ETA %(_eta_str)s'
294 if s
.get('downloaded_bytes') is not None:
295 s
['_downloaded_bytes_str'] = format_bytes(s
['downloaded_bytes'])
297 s
['_elapsed_str'] = self
.format_seconds(s
['elapsed'])
298 msg_template
= '%(_downloaded_bytes_str)s at %(_speed_str)s (%(_elapsed_str)s)'
300 msg_template
= '%(_downloaded_bytes_str)s at %(_speed_str)s'
302 msg_template
= '%(_percent_str)s % at %(_speed_str)s ETA %(_eta_str)s'
304 self
._report
_progress
_status
(msg_template
% s
)
306 def report_resuming_byte(self
, resume_len
):
307 """Report attempt to resume at given byte."""
308 self
.to_screen('[download] Resuming download at byte %s' % resume_len
)
310 def report_retry(self
, err
, count
, retries
):
311 """Report retry in case of HTTP error 5xx"""
313 '[download] Got server HTTP error: %s. Retrying (attempt %d of %s)...'
314 % (error_to_compat_str(err
), count
, self
.format_retries(retries
)))
316 def report_file_already_downloaded(self
, file_name
):
317 """Report file has already been fully downloaded."""
319 self
.to_screen('[download] %s has already been downloaded' % file_name
)
320 except UnicodeEncodeError:
321 self
.to_screen('[download] The file has already been downloaded')
323 def report_unable_to_resume(self
):
324 """Report it was impossible to resume download."""
325 self
.to_screen('[download] Unable to resume')
327 def download(self
, filename
, info_dict
):
328 """Download to a filename using the info from info_dict
329 Return True on success and False otherwise
332 nooverwrites_and_exists
= (
333 self
.params
.get('nooverwrites', False) and
334 os
.path
.exists(encodeFilename(filename
))
337 if not hasattr(filename
, 'write'):
338 continuedl_and_exists
= (
339 self
.params
.get('continuedl', True) and
340 os
.path
.isfile(encodeFilename(filename
)) and
341 not self
.params
.get('nopart', False)
344 # Check file already present
345 if filename
!= '-' and (nooverwrites_and_exists
or continuedl_and_exists
):
346 self
.report_file_already_downloaded(filename
)
347 self
._hook
_progress
({
348 'filename': filename
,
349 'status': 'finished',
350 'total_bytes': os
.path
.getsize(encodeFilename(filename
)),
354 min_sleep_interval
= self
.params
.get('sleep_interval')
355 if min_sleep_interval
:
356 max_sleep_interval
= self
.params
.get('max_sleep_interval', min_sleep_interval
)
357 sleep_interval
= random
.uniform(min_sleep_interval
, max_sleep_interval
)
359 '[download] Sleeping %s seconds...' % (
360 int(sleep_interval
) if sleep_interval
.is_integer()
361 else '%.2f' % sleep_interval
))
362 time
.sleep(sleep_interval
)
364 return self
.real_download(filename
, info_dict
)
366 def real_download(self
, filename
, info_dict
):
367 """Real download process. Redefine in subclasses."""
368 raise NotImplementedError('This method must be implemented by subclasses')
370 def _hook_progress(self
, status
):
371 for ph
in self
._progress
_hooks
:
374 def add_progress_hook(self
, ph
):
375 # See YoutubeDl.py (search for progress_hooks) for a description of
377 self
._progress
_hooks
.append(ph
)
379 def _debug_cmd(self
, args
, exe
=None):
380 if not self
.params
.get('verbose', False):
383 str_args
= [decodeArgument(a
) for a
in args
]
386 exe
= os
.path
.basename(str_args
[0])
388 self
.to_screen('[debug] %s command line: %s' % (
389 exe
, shell_quote(str_args
)))