Initiate new release.

[youtubedl] / youtube_dl / downloader / hls.py

diff --git a/youtube_dl/downloader/hls.py b/youtube_dl/downloader/hls.py
index fa983462babeb6a6bccc2f22992411dd6aea8a5b..8d7971e5d7042e5917deaf589cdebac7d488cd10 100644 (file)
--- a/youtube_dl/downloader/hls.py
+++ b/youtube_dl/downloader/hls.py
@@ -1,44 +1,138 @@
-import os
-import subprocess
+from __future__ import unicode_literals

 
 

-from .common import FileDownloader
+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,
+    compat_struct_pack,
+)

 from ..utils import (
     encodeFilename,
 from ..utils import (
     encodeFilename,

+    sanitize_open,
+    parse_m3u8_attributes,
+    update_url_query,

 )
 
 
 )
 
 

-class HlsFD(FileDownloader):
+class HlsFD(FragmentFD):
+    """ A limited implementation that does not require ffmpeg """
+
+    FD_NAME = 'hlsnative'
+
+    @staticmethod
+    def can_download(manifest):
+        UNSUPPORTED_FEATURES = (
+            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
+            # http://hls-geo.daserste.de/i/videoportal/Film/c_620000/622873/format,716451,716457,716450,716458,716459,.mp4.csmil/index_4_av.m3u8?null=0)
+            # 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 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]
+
+            # 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
+        )
+        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)
+        return all(check_results)
+

     def real_download(self, filename, info_dict):
     def real_download(self, filename, info_dict):

-        url = info_dict['url']
-        self.report_destination(filename)
-        tmpfilename = self.temp_name(filename)
-
-        args = ['-y', '-i', url, '-f', 'mp4', '-c', 'copy',
-            '-bsf:a', 'aac_adtstoasc', tmpfilename]
-
-        for program in ['avconv', 'ffmpeg']:
-            try:
-                subprocess.call([program, '-version'], stdout=(open(os.path.devnull, 'w')), stderr=subprocess.STDOUT)
-                break
-            except (OSError, IOError):
-                pass
-        else:
-            self.report_error(u'm3u8 download detected but ffmpeg or avconv could not be found')
-        cmd = [program] + args
-
-        retval = subprocess.call(cmd)
-        if retval == 0:
-            fsize = os.path.getsize(encodeFilename(tmpfilename))
-            self.to_screen(u'\r[%s] %s bytes' % (cmd[0], fsize))
-            self.try_rename(tmpfilename, filename)
-            self._hook_progress({
-                'downloaded_bytes': fsize,
-                'total_bytes': fsize,
-                'filename': filename,
-                'status': 'finished',
-            })
-            return True
-        else:
-            self.to_stderr(u"\n")
-            self.report_error(u'ffmpeg exited with code %d' % retval)
-            return False
+        man_url = info_dict['url']
+        self.to_screen('[%s] Downloading m3u8 manifest' % self.FD_NAME)
+        manifest = self.ydl.urlopen(man_url).read()
+
+        s = manifest.decode('utf-8', 'ignore')
+
+        if not self.can_download(s):
+            self.report_warning(
+                'hlsnative has detected features it does not support, '
+                'extraction will be delegated to ffmpeg')
+            fd = FFmpegFD(self.ydl, self.params)
+            for ph in self._progress_hooks:
+                fd.add_progress_hook(ph)
+            return fd.real_download(filename, info_dict)
+
+        total_frags = 0
+        for line in s.splitlines():
+            line = line.strip()
+            if line and not line.startswith('#'):
+                total_frags += 1
+
+        ctx = {
+            'filename': filename,
+            'total_frags': total_frags,
+        }
+
+        self._prepare_and_start_frag_download(ctx)
+
+        extra_param_to_segment_url = info_dict.get('extra_param_to_segment_url')
+        i = 0
+        media_sequence = 0
+        decrypt_info = {'METHOD': 'NONE'}
+        frags_filenames = []
+        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)
+                    if extra_param_to_segment_url:
+                        frag_url = update_url_query(frag_url, extra_param_to_segment_url)
+                    success = ctx['dl'].download(frag_filename, {'url': frag_url})
+                    if not success:
+                        return False
+                    down, frag_sanitized = sanitize_open(frag_filename, 'rb')
+                    frag_content = down.read()
+                    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)
+                    # 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'])
+                        if extra_param_to_segment_url:
+                            decrypt_info['URI'] = update_url_query(decrypt_info['URI'], extra_param_to_segment_url)
+                        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)
+
+        for frag_file in frags_filenames:
+            os.remove(encodeFilename(frag_file))
+
+        return True