Merge pull request #8201 from remitamine/hls-aes

[downloader/hls] Add support for AES-128 encrypted segments in hlsnative downloader
This commit is contained in:
Yen Chi Hsuan 2016-06-19 13:25:08 +08:00 committed by GitHub
commit 6aeb64b673
3 changed files with 68 additions and 36 deletions

View file

@ -2,14 +2,24 @@ from __future__ import unicode_literals
import os.path import os.path
import re 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 .fragment import FragmentFD
from .external import FFmpegFD from .external import FFmpegFD
from ..compat import compat_urlparse from ..compat import (
compat_urlparse,
compat_struct_pack,
)
from ..utils import ( from ..utils import (
encodeFilename, encodeFilename,
sanitize_open, sanitize_open,
parse_m3u8_attributes,
) )
@ -21,7 +31,7 @@ class HlsFD(FragmentFD):
@staticmethod @staticmethod
def can_download(manifest): def can_download(manifest):
UNSUPPORTED_FEATURES = ( 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] 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 # Live streams heuristic does not always work (e.g. geo restricted to Germany
@ -39,7 +49,9 @@ class HlsFD(FragmentFD):
# 3. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.2 # 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 # 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(not (re.search(r'#EXT-X-KEY:METHOD=AES-128', manifest) and not can_decrypt_frag))
return all(check_results)
def real_download(self, filename, info_dict): def real_download(self, filename, info_dict):
man_url = info_dict['url'] man_url = info_dict['url']
@ -57,36 +69,58 @@ class HlsFD(FragmentFD):
fd.add_progress_hook(ph) fd.add_progress_hook(ph)
return fd.real_download(filename, info_dict) return fd.real_download(filename, info_dict)
fragment_urls = [] total_frags = 0
for line in s.splitlines(): for line in s.splitlines():
line = line.strip() line = line.strip()
if line and not line.startswith('#'): if line and not line.startswith('#'):
segment_url = ( total_frags += 1
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
ctx = { ctx = {
'filename': filename, 'filename': filename,
'total_frags': len(fragment_urls), 'total_frags': total_frags,
} }
self._prepare_and_start_frag_download(ctx) self._prepare_and_start_frag_download(ctx)
i = 0
media_sequence = 0
decrypt_info = {'METHOD': 'NONE'}
frags_filenames = [] frags_filenames = []
for i, frag_url in enumerate(fragment_urls): 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_filename = '%s-Frag%d' % (ctx['tmpfilename'], i) frag_filename = '%s-Frag%d' % (ctx['tmpfilename'], i)
success = ctx['dl'].download(frag_filename, {'url': frag_url}) success = ctx['dl'].download(frag_filename, {'url': frag_url})
if not success: if not success:
return False return False
down, frag_sanitized = sanitize_open(frag_filename, 'rb') down, frag_sanitized = sanitize_open(frag_filename, 'rb')
ctx['dest_stream'].write(down.read()) frag_content = down.read()
down.close() down.close()
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) frags_filenames.append(frag_sanitized)
# We only download the first fragment during the test
if self.params.get('test', False):
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:])
if not re.match(r'^https?://', decrypt_info['URI']):
decrypt_info['URI'] = compat_urlparse.urljoin(man_url, decrypt_info['URI'])
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) self._finish_frag_download(ctx)

View file

@ -53,6 +53,7 @@ from ..utils import (
mimetype2ext, mimetype2ext,
update_Request, update_Request,
update_url_query, update_url_query,
parse_m3u8_attributes,
) )
@ -1150,23 +1151,11 @@ class InfoExtractor(object):
}] }]
last_info = None last_info = None
last_media = None last_media = None
kv_rex = re.compile(
r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
for line in m3u8_doc.splitlines(): for line in m3u8_doc.splitlines():
if line.startswith('#EXT-X-STREAM-INF:'): if line.startswith('#EXT-X-STREAM-INF:'):
last_info = {} last_info = parse_m3u8_attributes(line)
for m in kv_rex.finditer(line):
v = m.group('val')
if v.startswith('"'):
v = v[1:-1]
last_info[m.group('key')] = v
elif line.startswith('#EXT-X-MEDIA:'): elif line.startswith('#EXT-X-MEDIA:'):
last_media = {} last_media = parse_m3u8_attributes(line)
for m in kv_rex.finditer(line):
v = m.group('val')
if v.startswith('"'):
v = v[1:-1]
last_media[m.group('key')] = v
elif line.startswith('#') or not line.strip(): elif line.startswith('#') or not line.strip():
continue continue
else: else:

View file

@ -2852,3 +2852,12 @@ def decode_packed_codes(code):
return re.sub( return re.sub(
r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)], r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
obfucasted_code) obfucasted_code)
def parse_m3u8_attributes(attrib):
info = {}
for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):
if val.startswith('"'):
val = val[1:-1]
info[key] = val
return info