from __future__ import unicode_literals
-import re
+from .fragment import FragmentFD
+from ..compat import compat_urllib_error
+from ..utils import (
+ DownloadError,
+ urljoin,
+)
-from .common import FileDownloader
-from ..compat import compat_urllib_request
-
-class DashSegmentsFD(FileDownloader):
+class DashSegmentsFD(FragmentFD):
"""
Download segments in a DASH manifest
"""
+
+ FD_NAME = 'dashsegments'
+
def real_download(self, filename, info_dict):
- self.report_destination(filename)
- tmpfilename = self.temp_name(filename)
- base_url = info_dict['url']
- segment_urls = info_dict['segment_urls']
-
- self.byte_counter = 0
-
- def append_url_to_file(outf, target_url, target_name):
- self.to_screen('[DashSegments] %s: Downloading %s' % (info_dict['id'], target_name))
- req = compat_urllib_request.Request(target_url)
- data = self.ydl.urlopen(req).read()
- outf.write(data)
- self.byte_counter += len(data)
-
- def combine_url(base_url, target_url):
- if re.match(r'^https?://', target_url):
- return target_url
- return '%s/%s' % (base_url, target_url)
-
- with open(tmpfilename, 'wb') as outf:
- append_url_to_file(
- outf, combine_url(base_url, info_dict['initialization_url']),
- 'initialization segment')
- for i, segment_url in enumerate(segment_urls):
- append_url_to_file(
- outf, combine_url(base_url, segment_url),
- 'segment %d / %d' % (i + 1, len(segment_urls)))
-
- self.try_rename(tmpfilename, filename)
-
- self._hook_progress({
- 'downloaded_bytes': self.byte_counter,
- 'total_bytes': self.byte_counter,
+ 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,
- 'status': 'finished',
- })
+ '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)
+
+ frag_index = 0
+ for i, fragment in enumerate(fragments):
+ frag_index += 1
+ if frag_index <= ctx['fragment_index']:
+ continue
+ # In DASH, the first segment contains necessary headers to
+ # generate a valid MP4 file, so always abort for the first segment
+ fatal = i == 0 or not skip_unavailable_fragments
+ count = 0
+ while count <= fragment_retries:
+ try:
+ fragment_url = fragment.get('url')
+ if not fragment_url:
+ assert fragment_base_url
+ fragment_url = urljoin(fragment_base_url, fragment['path'])
+ success, frag_content = self._download_fragment(ctx, fragment_url, info_dict)
+ if not success:
+ return False
+ self._append_fragment(ctx, frag_content)
+ break
+ 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.
+ count += 1
+ if count <= fragment_retries:
+ self.report_retry_fragment(err, frag_index, count, fragment_retries)
+ except DownloadError:
+ # Don't retry fragment if error occurred during HTTP downloading
+ # itself since it has own retry settings
+ if not fatal:
+ self.report_skip_fragment(frag_index)
+ break
+ raise
+
+ if count > fragment_retries:
+ if not fatal:
+ self.report_skip_fragment(frag_index)
+ continue
+ self.report_error('giving up after %s fragment retries' % fragment_retries)
+ return False
+
+ self._finish_frag_download(ctx)
return True
projects / youtube-dl / blobdiff