Mirror of youtube-dl with not newest code that I had https://www.opennet.ru/opennews/art.shtml?num=53952
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

251 lines
10 KiB

from __future__ import unicode_literals
import errno
import os
import socket
import time
import re
from .common import FileDownloader
from ..compat import compat_urllib_error
from ..utils import (
class HttpFD(FileDownloader):
def real_download(self, filename, info_dict):
url = info_dict['url']
tmpfilename = self.temp_name(filename)
stream = None
# Do not include the Accept-Encoding header
headers = {'Youtubedl-no-compression': 'True'}
add_headers = info_dict.get('http_headers')
if add_headers:
basic_request = sanitized_Request(url, None, headers)
request = sanitized_Request(url, None, headers)
is_test = self.params.get('test', False)
if is_test:
request.add_header('Range', 'bytes=0-%s' % str(self._TEST_FILE_SIZE - 1))
# Establish possible resume length
if os.path.isfile(encodeFilename(tmpfilename)):
resume_len = os.path.getsize(encodeFilename(tmpfilename))
resume_len = 0
open_mode = 'wb'
if resume_len != 0:
if self.params.get('continuedl', True):
request.add_header('Range', 'bytes=%d-' % resume_len)
open_mode = 'ab'
resume_len = 0
count = 0
retries = self.params.get('retries', 0)
while count <= retries:
# Establish connection
data = self.ydl.urlopen(request)
# When trying to resume, Content-Range HTTP header of response has to be checked
# to match the value of requested Range HTTP header. This is due to a webservers
# that don't support resuming and serve a whole file with no Content-Range
# set in response despite of requested Range (see
# https://github.com/rg3/youtube-dl/issues/6057#issuecomment-126129799)
if resume_len > 0:
content_range = data.headers.get('Content-Range')
if content_range:
content_range_m = re.search(r'bytes (\d+)-', content_range)
# Content-Range is present and matches requested Range, resume is possible
if content_range_m and resume_len == int(content_range_m.group(1)):
# Content-Range is either not present or invalid. Assuming remote webserver is
# trying to send the whole file, resume is not possible, so wiping the local file
# and performing entire redownload
resume_len = 0
open_mode = 'wb'
except (compat_urllib_error.HTTPError, ) as err:
if (err.code < 500 or err.code >= 600) and err.code != 416:
# Unexpected HTTP error
elif err.code == 416:
# Unable to resume (requested range not satisfiable)
# Open the connection again without the range header
data = self.ydl.urlopen(basic_request)
content_length = data.info()['Content-Length']
except (compat_urllib_error.HTTPError, ) as err:
if err.code < 500 or err.code >= 600:
# Examine the reported length
if (content_length is not None and
(resume_len - 100 < int(content_length) < resume_len + 100)):
# The file had already been fully downloaded.
# Explanation to the above condition: in issue #175 it was revealed that
# YouTube sometimes adds or removes a few bytes from the end of the file,
# changing the file size slightly and causing problems for some users. So
# I decided to implement a suggested change and consider the file
# completely downloaded if the file size differs less than 100 bytes from
# the one in the hard drive.
self.try_rename(tmpfilename, filename)
'filename': filename,
'status': 'finished',
'downloaded_bytes': resume_len,
'total_bytes': resume_len,
return True
# The length does not match, we start the download over
resume_len = 0
open_mode = 'wb'
except socket.error as e:
if e.errno != errno.ECONNRESET:
# Connection reset is no problem, just retry
# Retry
count += 1
if count <= retries:
self.report_retry(count, retries)
if count > retries:
self.report_error('giving up after %s retries' % retries)
return False
data_len = data.info().get('Content-length', None)
# Range HTTP header may be ignored/unsupported by a webserver
# (e.g. extractor/scivee.py, extractor/bambuser.py).
# However, for a test we still would like to download just a piece of a file.
# To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
# block size when downloading a file.
if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
data_len = self._TEST_FILE_SIZE
if data_len is not None:
data_len = int(data_len) + resume_len
min_data_len = self.params.get('min_filesize')
max_data_len = self.params.get('max_filesize')
if min_data_len is not None and data_len < min_data_len:
self.to_screen('\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
return False
if max_data_len is not None and data_len > max_data_len:
self.to_screen('\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
return False
byte_counter = 0 + resume_len
block_size = self.params.get('buffersize', 1024)
start = time.time()
# measure time over whole while-loop, so slow_down() and best_block_size() work together properly
now = None # needed for slow_down() in the first loop run
before = start # start measuring
while True:
# Download and write
data_block = data.read(block_size if not is_test else min(block_size, data_len - byte_counter))
byte_counter += len(data_block)
# exit loop when download is finished
if len(data_block) == 0:
# Open destination file just in time
if stream is None:
(stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
assert stream is not None
filename = self.undo_temp_name(tmpfilename)
except (OSError, IOError) as err:
self.report_error('unable to open for writing: %s' % str(err))
return False
if self.params.get('xattr_set_filesize', False) and data_len is not None:
import xattr
xattr.setxattr(tmpfilename, 'user.ytdl.filesize', str(data_len))
except(OSError, IOError, ImportError) as err:
self.report_error('unable to set filesize xattr: %s' % str(err))
except (IOError, OSError) as err:
self.report_error('unable to write data: %s' % str(err))
return False
# Apply rate limit
self.slow_down(start, now, byte_counter - resume_len)
# end measuring of one loop run
now = time.time()
after = now
# Adjust block size
if not self.params.get('noresizebuffer', False):
block_size = self.best_block_size(after - before, len(data_block))
before = after
# Progress message
speed = self.calc_speed(start, now, byte_counter - resume_len)
if data_len is None:
eta = None
eta = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
'status': 'downloading',
'downloaded_bytes': byte_counter,
'total_bytes': data_len,
'tmpfilename': tmpfilename,
'filename': filename,
'eta': eta,
'speed': speed,
'elapsed': now - start,
if is_test and byte_counter == data_len:
if stream is None:
self.report_error('Did not get any data blocks')
return False
if tmpfilename != '-':
if data_len is not None and byte_counter != data_len:
raise ContentTooShortError(byte_counter, int(data_len))
self.try_rename(tmpfilename, filename)
# Update file modification time
if self.params.get('updatetime', True):
info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
'downloaded_bytes': byte_counter,
'total_bytes': byte_counter,
'filename': filename,
'status': 'finished',
'elapsed': time.time() - start,
return True