+from __future__ import division, unicode_literals
+
import os
import re
import sys
import time
+import random
+from ..compat import compat_os_name
from ..utils import (
+ decodeArgument,
encodeFilename,
- timeconvert,
+ error_to_compat_str,
format_bytes,
+ shell_quote,
+ timeconvert,
)
Available options:
- verbose: Print additional info to stdout.
- quiet: Do not print messages to stdout.
- ratelimit: Download speed limit, in bytes/sec.
- retries: Number of times to retry for HTTP error 5xx
- buffersize: Size of download buffer in bytes.
- noresizebuffer: Do not automatically resize the download buffer.
- continuedl: Try to continue downloads if possible.
- noprogress: Do not print the progress bar.
- logtostderr: Log messages to stderr instead of stdout.
- consoletitle: Display progress in console window's titlebar.
- nopart: Do not use temporary .part files.
- updatetime: Use the Last-modified header to set output file timestamps.
- test: Download only first bytes to test the downloader.
- min_filesize: Skip files smaller than this size
- max_filesize: Skip files larger than this size
+ verbose: Print additional info to stdout.
+ quiet: Do not print messages to stdout.
+ ratelimit: Download speed limit, in bytes/sec.
+ retries: Number of times to retry for HTTP error 5xx
+ buffersize: Size of download buffer in bytes.
+ noresizebuffer: Do not automatically resize the download buffer.
+ continuedl: Try to continue downloads if possible.
+ noprogress: Do not print the progress bar.
+ logtostderr: Log messages to stderr instead of stdout.
+ consoletitle: Display progress in console window's titlebar.
+ nopart: Do not use temporary .part files.
+ updatetime: Use the Last-modified header to set output file timestamps.
+ test: Download only first bytes to test the downloader.
+ min_filesize: Skip files smaller than this size
+ max_filesize: Skip files larger than this size
+ xattr_set_filesize: Set ytdl.filesize user xattribute with expected size.
+ external_downloader_args: A list of additional command-line arguments for the
+ external downloader.
+ hls_use_mpegts: Use the mpegts container for HLS videos.
+ http_chunk_size: Size of a chunk for chunk-based HTTP downloading. May be
+ useful for bypassing bandwidth throttling imposed by
+ a webserver (experimental)
Subclasses of this one must re-define the real_download method.
"""
+ _TEST_FILE_SIZE = 10241
params = None
def __init__(self, ydl, params):
self.ydl = ydl
self._progress_hooks = []
self.params = params
+ self.add_progress_hook(self.report_progress)
@staticmethod
def format_seconds(seconds):
def calc_eta(start, now, total, current):
if total is None:
return None
+ if now is None:
+ now = time.time()
dif = now - start
- if current == 0 or dif < 0.001: # One millisecond
+ if current == 0 or dif < 0.001: # One millisecond
return None
rate = float(current) / dif
return int((float(total) - float(current)) / rate)
@staticmethod
def calc_speed(start, now, bytes):
dif = now - start
- if bytes == 0 or dif < 0.001: # One millisecond
+ if bytes == 0 or dif < 0.001: # One millisecond
return None
return float(bytes) / dif
return '%10s' % '---b/s'
return '%10s' % ('%s/s' % format_bytes(speed))
+ @staticmethod
+ def format_retries(retries):
+ return 'inf' if retries == float('inf') else '%.0f' % retries
+
@staticmethod
def best_block_size(elapsed_time, bytes):
new_min = max(bytes / 2.0, 1.0)
- new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
+ new_max = min(max(bytes * 2.0, 1.0), 4194304) # Do not surpass 4 MB
if elapsed_time < 0.001:
return int(new_max)
rate = bytes / elapsed_time
def report_error(self, *args, **kargs):
self.ydl.report_error(*args, **kargs)
- def slow_down(self, start_time, byte_counter):
+ def slow_down(self, start_time, now, byte_counter):
"""Sleep if the download speed is over the rate limit."""
- rate_limit = self.params.get('ratelimit', None)
+ rate_limit = self.params.get('ratelimit')
if rate_limit is None or byte_counter == 0:
return
- now = time.time()
+ if now is None:
+ now = time.time()
elapsed = now - start_time
if elapsed <= 0.0:
return
speed = float(byte_counter) / elapsed
if speed > rate_limit:
- time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit)
+ sleep_time = float(byte_counter) / rate_limit - elapsed
+ if sleep_time > 0:
+ time.sleep(sleep_time)
def temp_name(self, filename):
"""Returns a temporary filename for the given filename."""
- if self.params.get('nopart', False) or filename == u'-' or \
+ if self.params.get('nopart', False) or filename == '-' or \
(os.path.exists(encodeFilename(filename)) and not os.path.isfile(encodeFilename(filename))):
return filename
- return filename + u'.part'
+ return filename + '.part'
def undo_temp_name(self, filename):
- if filename.endswith(u'.part'):
- return filename[:-len(u'.part')]
+ if filename.endswith('.part'):
+ return filename[:-len('.part')]
return filename
+ def ytdl_filename(self, filename):
+ return filename + '.ytdl'
+
def try_rename(self, old_filename, new_filename):
try:
if old_filename == new_filename:
return
os.rename(encodeFilename(old_filename), encodeFilename(new_filename))
except (IOError, OSError) as err:
- self.report_error(u'unable to rename file: %s' % str(err))
+ self.report_error('unable to rename file: %s' % error_to_compat_str(err))
def try_utime(self, filename, last_modified_hdr):
"""Try to set the last-modified time of the given file."""
return
try:
os.utime(filename, (time.time(), filetime))
- except:
+ except Exception:
pass
return filetime
def report_destination(self, filename):
"""Report destination filename."""
- self.to_screen(u'[download] Destination: ' + filename)
+ self.to_screen('[download] Destination: ' + filename)
def _report_progress_status(self, msg, is_last_line=False):
- fullmsg = u'[download] ' + msg
+ fullmsg = '[download] ' + msg
if self.params.get('progress_with_newline', False):
self.to_screen(fullmsg)
else:
- if os.name == 'nt':
+ if compat_os_name == 'nt':
prev_len = getattr(self, '_report_progress_prev_line_length',
0)
if prev_len > len(fullmsg):
- fullmsg += u' ' * (prev_len - len(fullmsg))
+ fullmsg += ' ' * (prev_len - len(fullmsg))
self._report_progress_prev_line_length = len(fullmsg)
- clear_line = u'\r'
+ clear_line = '\r'
else:
- clear_line = (u'\r\x1b[K' if sys.stderr.isatty() else u'\r')
+ clear_line = ('\r\x1b[K' if sys.stderr.isatty() else '\r')
self.to_screen(clear_line + fullmsg, skip_eol=not is_last_line)
- self.to_console_title(u'youtube-dl ' + msg)
+ self.to_console_title('youtube-dl ' + msg)
- def report_progress(self, percent, data_len_str, speed, eta):
- """Report download progress."""
- if self.params.get('noprogress', False):
+ def report_progress(self, s):
+ if s['status'] == 'finished':
+ if self.params.get('noprogress', False):
+ self.to_screen('[download] Download completed')
+ else:
+ msg_template = '100%%'
+ if s.get('total_bytes') is not None:
+ s['_total_bytes_str'] = format_bytes(s['total_bytes'])
+ msg_template += ' of %(_total_bytes_str)s'
+ if s.get('elapsed') is not None:
+ s['_elapsed_str'] = self.format_seconds(s['elapsed'])
+ msg_template += ' in %(_elapsed_str)s'
+ self._report_progress_status(
+ msg_template % s, is_last_line=True)
+
+ if self.params.get('noprogress'):
return
- if eta is not None:
- eta_str = self.format_eta(eta)
- else:
- eta_str = 'Unknown ETA'
- if percent is not None:
- percent_str = self.format_percent(percent)
+
+ if s['status'] != 'downloading':
+ return
+
+ if s.get('eta') is not None:
+ s['_eta_str'] = self.format_eta(s['eta'])
else:
- percent_str = 'Unknown %'
- speed_str = self.format_speed(speed)
+ s['_eta_str'] = 'Unknown ETA'
- msg = (u'%s of %s at %s ETA %s' %
- (percent_str, data_len_str, speed_str, eta_str))
- self._report_progress_status(msg)
+ if s.get('total_bytes') and s.get('downloaded_bytes') is not None:
+ s['_percent_str'] = self.format_percent(100 * s['downloaded_bytes'] / s['total_bytes'])
+ elif s.get('total_bytes_estimate') and s.get('downloaded_bytes') is not None:
+ s['_percent_str'] = self.format_percent(100 * s['downloaded_bytes'] / s['total_bytes_estimate'])
+ else:
+ if s.get('downloaded_bytes') == 0:
+ s['_percent_str'] = self.format_percent(0)
+ else:
+ s['_percent_str'] = 'Unknown %'
- def report_progress_live_stream(self, downloaded_data_len, speed, elapsed):
- if self.params.get('noprogress', False):
- return
- downloaded_str = format_bytes(downloaded_data_len)
- speed_str = self.format_speed(speed)
- elapsed_str = FileDownloader.format_seconds(elapsed)
- msg = u'%s at %s (%s)' % (downloaded_str, speed_str, elapsed_str)
- self._report_progress_status(msg)
-
- def report_finish(self, data_len_str, tot_time):
- """Report download finished."""
- if self.params.get('noprogress', False):
- self.to_screen(u'[download] Download completed')
+ if s.get('speed') is not None:
+ s['_speed_str'] = self.format_speed(s['speed'])
+ else:
+ s['_speed_str'] = 'Unknown speed'
+
+ if s.get('total_bytes') is not None:
+ s['_total_bytes_str'] = format_bytes(s['total_bytes'])
+ msg_template = '%(_percent_str)s of %(_total_bytes_str)s at %(_speed_str)s ETA %(_eta_str)s'
+ elif s.get('total_bytes_estimate') is not None:
+ s['_total_bytes_estimate_str'] = format_bytes(s['total_bytes_estimate'])
+ msg_template = '%(_percent_str)s of ~%(_total_bytes_estimate_str)s at %(_speed_str)s ETA %(_eta_str)s'
else:
- self._report_progress_status(
- (u'100%% of %s in %s' %
- (data_len_str, self.format_seconds(tot_time))),
- is_last_line=True)
+ if s.get('downloaded_bytes') is not None:
+ s['_downloaded_bytes_str'] = format_bytes(s['downloaded_bytes'])
+ if s.get('elapsed'):
+ s['_elapsed_str'] = self.format_seconds(s['elapsed'])
+ msg_template = '%(_downloaded_bytes_str)s at %(_speed_str)s (%(_elapsed_str)s)'
+ else:
+ msg_template = '%(_downloaded_bytes_str)s at %(_speed_str)s'
+ else:
+ msg_template = '%(_percent_str)s % at %(_speed_str)s ETA %(_eta_str)s'
+
+ self._report_progress_status(msg_template % s)
def report_resuming_byte(self, resume_len):
"""Report attempt to resume at given byte."""
- self.to_screen(u'[download] Resuming download at byte %s' % resume_len)
+ self.to_screen('[download] Resuming download at byte %s' % resume_len)
- def report_retry(self, count, retries):
+ def report_retry(self, err, count, retries):
"""Report retry in case of HTTP error 5xx"""
- self.to_screen(u'[download] Got server HTTP error. Retrying (attempt %d of %d)...' % (count, retries))
+ self.to_screen(
+ '[download] Got server HTTP error: %s. Retrying (attempt %d of %s)...'
+ % (error_to_compat_str(err), count, self.format_retries(retries)))
def report_file_already_downloaded(self, file_name):
"""Report file has already been fully downloaded."""
try:
- self.to_screen(u'[download] %s has already been downloaded' % file_name)
+ self.to_screen('[download] %s has already been downloaded' % file_name)
except UnicodeEncodeError:
- self.to_screen(u'[download] The file has already been downloaded')
+ self.to_screen('[download] The file has already been downloaded')
def report_unable_to_resume(self):
"""Report it was impossible to resume download."""
- self.to_screen(u'[download] Unable to resume')
+ self.to_screen('[download] Unable to resume')
def download(self, filename, info_dict):
"""Download to a filename using the info from info_dict
Return True on success and False otherwise
"""
- # Check file already present
- if self.params.get('continuedl', False) and os.path.isfile(encodeFilename(filename)) and not self.params.get('nopart', False):
- self.report_file_already_downloaded(filename)
- self._hook_progress({
- 'filename': filename,
- 'status': 'finished',
- 'total_bytes': os.path.getsize(encodeFilename(filename)),
- })
- return True
+
+ nooverwrites_and_exists = (
+ self.params.get('nooverwrites', False)
+ and os.path.exists(encodeFilename(filename))
+ )
+
+ if not hasattr(filename, 'write'):
+ continuedl_and_exists = (
+ self.params.get('continuedl', True)
+ and os.path.isfile(encodeFilename(filename))
+ and not self.params.get('nopart', False)
+ )
+
+ # Check file already present
+ if filename != '-' and (nooverwrites_and_exists or continuedl_and_exists):
+ self.report_file_already_downloaded(filename)
+ self._hook_progress({
+ 'filename': filename,
+ 'status': 'finished',
+ 'total_bytes': os.path.getsize(encodeFilename(filename)),
+ })
+ return True
+
+ min_sleep_interval = self.params.get('sleep_interval')
+ if min_sleep_interval:
+ max_sleep_interval = self.params.get('max_sleep_interval', min_sleep_interval)
+ sleep_interval = random.uniform(min_sleep_interval, max_sleep_interval)
+ self.to_screen(
+ '[download] Sleeping %s seconds...' % (
+ int(sleep_interval) if sleep_interval.is_integer()
+ else '%.2f' % sleep_interval))
+ time.sleep(sleep_interval)
return self.real_download(filename, info_dict)
def real_download(self, filename, info_dict):
"""Real download process. Redefine in subclasses."""
- raise NotImplementedError(u'This method must be implemented by sublcasses')
+ raise NotImplementedError('This method must be implemented by subclasses')
def _hook_progress(self, status):
for ph in self._progress_hooks:
ph(status)
def add_progress_hook(self, ph):
- """ ph gets called on download progress, with a dictionary with the entries
- * filename: The final filename
- * status: One of "downloading" and "finished"
+ # See YoutubeDl.py (search for progress_hooks) for a description of
+ # this interface
+ self._progress_hooks.append(ph)
- It can also have some of the following entries:
+ def _debug_cmd(self, args, exe=None):
+ if not self.params.get('verbose', False):
+ return
- * downloaded_bytes: Bytes on disks
- * total_bytes: Total bytes, None if unknown
- * tmpfilename: The filename we're currently writing to
- * eta: The estimated time in seconds, None if unknown
- * speed: The download speed in bytes/second, None if unknown
+ str_args = [decodeArgument(a) for a in args]
- Hooks are guaranteed to be called at least once (with status "finished")
- if the download is successful.
- """
- self._progress_hooks.append(ph)
+ if exe is None:
+ exe = os.path.basename(str_args[0])
+
+ self.to_screen('[debug] %s command line: %s' % (
+ exe, shell_quote(str_args)))
Rapsys Git / youtubedl / blobdiff