+from __future__ import unicode_literals
+
+import errno
import os
-import re
-import subprocess
-import sys
+import socket
import time
from .common import FileDownloader
-from ..utils import (
+from .dash import DashSegmentsFD
+from ..compat import (
compat_urllib_request,
compat_urllib_error,
+)
+from ..utils import (
ContentTooShortError,
-
encodeFilename,
sanitize_open,
- format_bytes,
)
class HttpFD(FileDownloader):
def real_download(self, filename, info_dict):
+ if info_dict.get('initialization_url') and list(filter(None, info_dict.get('segment_urls', []))):
+ return DashSegmentsFD(self.ydl, self.params).real_download(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'}
- if 'user_agent' in info_dict:
- headers['Youtubedl-user-agent'] = info_dict['user_agent']
+ add_headers = info_dict.get('http_headers')
+ if add_headers:
+ headers.update(add_headers)
basic_request = compat_urllib_request.Request(url, None, headers)
request = compat_urllib_request.Request(url, None, headers)
- if self.params.get('test', False):
- request.add_header('Range','bytes=0-10240')
+ 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)):
open_mode = 'wb'
if resume_len != 0:
- if self.params.get('continuedl', False):
+ if self.params.get('continuedl', True):
self.report_resuming_byte(resume_len)
- request.add_header('Range','bytes=%d-' % resume_len)
+ request.add_header('Range', 'bytes=%d-' % resume_len)
open_mode = 'ab'
else:
resume_len = 0
while count <= retries:
# Establish connection
try:
- if count == 0 and 'urlhandle' in info_dict:
- data = info_dict['urlhandle']
- data = compat_urllib_request.urlopen(request)
+ data = self.ydl.urlopen(request)
break
except (compat_urllib_error.HTTPError, ) as err:
if (err.code < 500 or err.code >= 600) and err.code != 416:
# Unable to resume (requested range not satisfiable)
try:
# Open the connection again without the range header
- data = compat_urllib_request.urlopen(basic_request)
+ 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:
self._hook_progress({
'filename': filename,
'status': 'finished',
+ 'downloaded_bytes': resume_len,
+ 'total_bytes': resume_len,
})
return True
else:
# The length does not match, we start the download over
self.report_unable_to_resume()
+ resume_len = 0
open_mode = 'wb'
break
+ except socket.error as e:
+ if e.errno != errno.ECONNRESET:
+ # Connection reset is no problem, just retry
+ raise
+
# Retry
count += 1
if count <= retries:
self.report_retry(count, retries)
if count > retries:
- self.report_error(u'giving up after %s retries' % 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", None)
- max_data_len = self.params.get("max_filesize", None)
+ max_data_len = self.params.get("max_filesize", None)
if min_data_len is not None and data_len < min_data_len:
- self.to_screen(u'\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (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(u'\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (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
- data_len_str = format_bytes(data_len)
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
- before = time.time()
- data_block = data.read(block_size)
- after = time.time()
+ 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:
break
- byte_counter += len(data_block)
- # Open file just in time
+ # Open destination file just in time
if stream is None:
try:
(stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
filename = self.undo_temp_name(tmpfilename)
self.report_destination(filename)
except (OSError, IOError) as err:
- self.report_error(u'unable to open for writing: %s' % str(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:
+ try:
+ 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))
+
try:
stream.write(data_block)
- except (IOError, OSError):
- self.to_stderr(u"\n")
- self.report_error(u'unable to write data: %s' % str(err))
+ except (IOError, OSError) as err:
+ self.to_stderr('\n')
+ 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, time.time(), byte_counter - resume_len)
+ speed = self.calc_speed(start, now, byte_counter - resume_len)
if data_len is None:
- eta = percent = None
+ eta = None
else:
- percent = self.calc_percent(byte_counter, data_len)
eta = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
- self.report_progress(percent, data_len_str, speed, eta)
self._hook_progress({
+ 'status': 'downloading',
'downloaded_bytes': byte_counter,
'total_bytes': data_len,
'tmpfilename': tmpfilename,
'filename': filename,
- 'status': 'downloading',
'eta': eta,
'speed': speed,
+ 'elapsed': now - start,
})
- # Apply rate limit
- self.slow_down(start, byte_counter - resume_len)
+ if is_test and byte_counter == data_len:
+ break
if stream is None:
- self.to_stderr(u"\n")
- self.report_error(u'Did not get any data blocks')
+ self.to_stderr('\n')
+ self.report_error('Did not get any data blocks')
return False
- stream.close()
- self.report_finish(data_len_str, (time.time() - start))
+ if tmpfilename != '-':
+ stream.close()
+
if data_len is not None and byte_counter != data_len:
raise ContentTooShortError(byte_counter, int(data_len))
self.try_rename(tmpfilename, filename)
'total_bytes': byte_counter,
'filename': filename,
'status': 'finished',
+ 'elapsed': time.time() - start,
})
return True
projects / youtube-dl / blobdiff