You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
youtube-dl/youtube_dl/downloader/dash.py

84 lines
3.3 KiB
Python

from __future__ import unicode_literals
import itertools
from .fragment import FragmentFD
from ..compat import compat_urllib_error
from ..utils import (
DownloadError,
urljoin,
)
class DashSegmentsFD(FragmentFD):
"""
Download segments in a DASH manifest
"""
FD_NAME = 'dashsegments'
def real_download(self, filename, info_dict):
fragment_base_url = info_dict.get('fragment_base_url')
fragments = info_dict['fragments'][:1] if self.params.get(
'test', False) else info_dict['fragments']
ctx = {
'filename': filename,
'total_frags': len(fragments),
}
self._prepare_and_start_frag_download(ctx)
fragment_retries = self.params.get('fragment_retries', 0)
skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
for frag_index, fragment in enumerate(fragments, 1):
if frag_index <= ctx['fragment_index']:
continue
success = False
# In DASH, the first segment contains necessary headers to
# generate a valid MP4 file, so always abort for the first segment
fatal = frag_index == 1 or not skip_unavailable_fragments
fragment_url = fragment.get('url')
if not fragment_url:
assert fragment_base_url
fragment_url = urljoin(fragment_base_url, fragment['path'])
headers = info_dict.get('http_headers')
fragment_range = fragment.get('range')
if fragment_range:
headers = headers.copy() if headers else {}
headers['Range'] = 'bytes=%s' % (fragment_range,)
for count in itertools.count():
try:
success, frag_content = self._download_fragment(ctx, fragment_url, info_dict, headers)
if not success:
return False
self._append_fragment(ctx, frag_content)
except compat_urllib_error.HTTPError as err:
# YouTube may often return 404 HTTP error for a fragment causing the
# whole download to fail. However if the same fragment is immediately
# retried with the same request data this usually succeeds (1-2 attempts
# is usually enough) thus allowing to download the whole file successfully.
# To be future-proof we will retry all fragments that fail with any
# HTTP error.
if count < fragment_retries:
self.report_retry_fragment(err, frag_index, count + 1, fragment_retries)
continue
except DownloadError:
# Don't retry fragment if error occurred during HTTP downloading
# itself since it has its own retry settings
if fatal:
raise
break
if not success:
if not fatal:
self.report_skip_fragment(frag_index)
continue
self.report_error('giving up after %s fragment retries' % count)
return False
self._finish_frag_download(ctx)
return True