import os.path
import re
+import binascii
+try:
+ from Crypto.Cipher import AES
+ can_decrypt_frag = True
+except ImportError:
+ can_decrypt_frag = False
from .fragment import FragmentFD
from .external import FFmpegFD
-from ..compat import compat_urlparse
+from ..compat import (
+ compat_urllib_error,
+ compat_urlparse,
+ compat_struct_pack,
+)
from ..utils import (
encodeFilename,
sanitize_open,
+ parse_m3u8_attributes,
+ update_url_query,
)
FD_NAME = 'hlsnative'
@staticmethod
- def can_download(manifest):
+ def can_download(manifest, info_dict):
UNSUPPORTED_FEATURES = (
- r'#EXT-X-KEY:METHOD=(?!NONE)', # encrypted streams [1]
+ r'#EXT-X-KEY:METHOD=(?!NONE|AES-128)', # encrypted streams [1]
r'#EXT-X-BYTERANGE', # playlists composed of byte ranges of media files [2]
# Live streams heuristic does not always work (e.g. geo restricted to Germany
# r'#EXT-X-MEDIA-SEQUENCE:(?!0$)', # live streams [3]
# This heuristic also is not correct since segments may not be appended as well.
- # Twitch vods have EXT-X-PLAYLIST-TYPE:EVENT despite no segments will definitely
- # be appended to the end of the playlist.
+ # Twitch vods of finished streams have EXT-X-PLAYLIST-TYPE:EVENT despite
+ # no segments will definitely be appended to the end of the playlist.
# r'#EXT-X-PLAYLIST-TYPE:EVENT', # media segments may be appended to the end of
- # event media playlists [4]
+ # # event media playlists [4]
# 1. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.4
# 2. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.2.2
# 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.2
# 4. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.5
)
- return all(not re.search(feature, manifest) for feature in UNSUPPORTED_FEATURES)
+ check_results = [not re.search(feature, manifest) for feature in UNSUPPORTED_FEATURES]
+ check_results.append(can_decrypt_frag or '#EXT-X-KEY:METHOD=AES-128' not in manifest)
+ check_results.append(not info_dict.get('is_live'))
+ return all(check_results)
def real_download(self, filename, info_dict):
man_url = info_dict['url']
self.to_screen('[%s] Downloading m3u8 manifest' % self.FD_NAME)
- manifest = self.ydl.urlopen(man_url).read()
+
+ manifest = self.ydl.urlopen(self._prepare_url(info_dict, man_url)).read()
s = manifest.decode('utf-8', 'ignore')
- if not self.can_download(s):
+ if not self.can_download(s, info_dict):
+ if info_dict.get('extra_param_to_segment_url'):
+ self.report_error('pycrypto not found. Please install it.')
+ return False
self.report_warning(
'hlsnative has detected features it does not support, '
'extraction will be delegated to ffmpeg')
fd.add_progress_hook(ph)
return fd.real_download(filename, info_dict)
- fragment_urls = []
+ total_frags = 0
for line in s.splitlines():
line = line.strip()
if line and not line.startswith('#'):
- segment_url = (
- line
- if re.match(r'^https?://', line)
- else compat_urlparse.urljoin(man_url, line))
- fragment_urls.append(segment_url)
- # We only download the first fragment during the test
- if self.params.get('test', False):
- break
+ total_frags += 1
ctx = {
'filename': filename,
- 'total_frags': len(fragment_urls),
+ 'total_frags': total_frags,
}
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)
+ test = self.params.get('test', False)
+
+ extra_query = None
+ extra_param_to_segment_url = info_dict.get('extra_param_to_segment_url')
+ if extra_param_to_segment_url:
+ extra_query = compat_urlparse.parse_qs(extra_param_to_segment_url)
+ i = 0
+ media_sequence = 0
+ decrypt_info = {'METHOD': 'NONE'}
frags_filenames = []
- for i, frag_url in enumerate(fragment_urls):
- frag_filename = '%s-Frag%d' % (ctx['tmpfilename'], i)
- success = ctx['dl'].download(frag_filename, {'url': frag_url})
- if not success:
- return False
- down, frag_sanitized = sanitize_open(frag_filename, 'rb')
- ctx['dest_stream'].write(down.read())
- down.close()
- frags_filenames.append(frag_sanitized)
+ for line in s.splitlines():
+ line = line.strip()
+ if line:
+ if not line.startswith('#'):
+ frag_url = (
+ line
+ if re.match(r'^https?://', line)
+ else compat_urlparse.urljoin(man_url, line))
+ frag_name = 'Frag%d' % i
+ frag_filename = '%s-%s' % (ctx['tmpfilename'], frag_name)
+ if extra_query:
+ frag_url = update_url_query(frag_url, extra_query)
+ count = 0
+ while count <= fragment_retries:
+ try:
+ success = ctx['dl'].download(frag_filename, {
+ 'url': frag_url,
+ 'http_headers': info_dict.get('http_headers'),
+ })
+ if not success:
+ return False
+ down, frag_sanitized = sanitize_open(frag_filename, 'rb')
+ frag_content = down.read()
+ down.close()
+ break
+ except compat_urllib_error.HTTPError as err:
+ # Unavailable (possibly temporary) fragments may be served.
+ # First we try to retry then either skip or abort.
+ # See https://github.com/rg3/youtube-dl/issues/10165,
+ # https://github.com/rg3/youtube-dl/issues/10448).
+ count += 1
+ if count <= fragment_retries:
+ self.report_retry_fragment(err, frag_name, count, fragment_retries)
+ if count > fragment_retries:
+ if skip_unavailable_fragments:
+ i += 1
+ media_sequence += 1
+ self.report_skip_fragment(frag_name)
+ continue
+ self.report_error(
+ 'giving up after %s fragment retries' % fragment_retries)
+ return False
+ if decrypt_info['METHOD'] == 'AES-128':
+ iv = decrypt_info.get('IV') or compat_struct_pack('>8xq', media_sequence)
+ frag_content = AES.new(
+ decrypt_info['KEY'], AES.MODE_CBC, iv).decrypt(frag_content)
+ ctx['dest_stream'].write(frag_content)
+ frags_filenames.append(frag_sanitized)
+ # We only download the first fragment during the test
+ if test:
+ break
+ i += 1
+ media_sequence += 1
+ elif line.startswith('#EXT-X-KEY'):
+ decrypt_info = parse_m3u8_attributes(line[11:])
+ if decrypt_info['METHOD'] == 'AES-128':
+ if 'IV' in decrypt_info:
+ decrypt_info['IV'] = binascii.unhexlify(decrypt_info['IV'][2:].zfill(32))
+ if not re.match(r'^https?://', decrypt_info['URI']):
+ decrypt_info['URI'] = compat_urlparse.urljoin(
+ man_url, decrypt_info['URI'])
+ if extra_query:
+ decrypt_info['URI'] = update_url_query(decrypt_info['URI'], extra_query)
+ decrypt_info['KEY'] = self.ydl.urlopen(decrypt_info['URI']).read()
+ elif line.startswith('#EXT-X-MEDIA-SEQUENCE'):
+ media_sequence = int(line[22:])
self._finish_frag_download(ctx)
projects / youtube-dl / blobdiff