youtube-dl

commit b8079a40bc61326b17a672b073dce6cdfa791fb5
parent 5bc8a73af69f4aac8b2df6f7c23ecfb4ee72e518
Author: Remita Amine <remitamine@gmail.com>
Date:   Sun, 28 Aug 2016 17:51:53 +0100

[turner] fix secure m3u8 formats downloading

Diffstat:
M
youtube_dl/downloader/hls.py
 | 
11
+++++++----
M
youtube_dl/extractor/turner.py
 | 
15
+++++++++++++--
M
youtube_dl/extractor/uplynk.py
 | 
4
+---

3 files changed, 21 insertions(+), 9 deletions(-)
diff --git a/youtube_dl/downloader/hls.py b/youtube_dl/downloader/hls.py
@@ -83,7 +83,10 @@ class HlsFD(FragmentFD):
 
         self._prepare_and_start_frag_download(ctx)
 
+        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'}
@@ -97,8 +100,8 @@ class HlsFD(FragmentFD):
                         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)
+                    if extra_query:
+                        frag_url = update_url_query(frag_url, extra_query)
                     success = ctx['dl'].download(frag_filename, {'url': frag_url})
                     if not success:
                         return False
@@ -124,8 +127,8 @@ class HlsFD(FragmentFD):
                         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)
+                        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:])
diff --git a/youtube_dl/extractor/turner.py b/youtube_dl/extractor/turner.py
@@ -11,6 +11,7 @@ from ..utils import (
     parse_duration,
     xpath_attr,
     update_url_query,
+    compat_urlparse,
 )
 
 
@@ -87,8 +88,18 @@ class TurnerBaseIE(InfoExtractor):
             if ext == 'smil':
                 formats.extend(self._extract_smil_formats(video_url, video_id, fatal=False))
             elif ext == 'm3u8':
-                formats.extend(self._extract_m3u8_formats(
-                    video_url, video_id, 'mp4', m3u8_id=format_id, fatal=False))
+                m3u8_formats = self._extract_m3u8_formats(
+                    video_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)
+                if m3u8_formats:
+                    # Sometimes final URLs inside m3u8 are unsigned, let's fix this
+                    # ourselves
+                    qs = compat_urlparse.urlparse(video_url).query
+                    if qs:
+                        query = compat_urlparse.parse_qs(qs)
+                        for m3u8_format in m3u8_formats:
+                            m3u8_format['url'] = update_url_query(m3u8_format['url'], query)
+                            m3u8_format['extra_param_to_segment_url'] = qs
+                    formats.extend(m3u8_formats)
             elif ext == 'f4m':
                 formats.extend(self._extract_f4m_formats(
                     update_url_query(video_url, {'hdcore': '3.7.0'}),
diff --git a/youtube_dl/extractor/uplynk.py b/youtube_dl/extractor/uplynk.py
@@ -33,9 +33,7 @@ class UplynkIE(InfoExtractor):
         formats = self._extract_m3u8_formats('http://content.uplynk.com/%s.m3u8' % path, display_id, 'mp4')
         if session_id:
             for f in formats:
-                f['extra_param_to_segment_url'] = {
-                    'pbs': session_id,
-                }
+                f['extra_param_to_segment_url'] = 'pbs=' + session_id
         self._sort_formats(formats)
         asset = self._download_json('http://content.uplynk.com/player/assetinfo/%s.json' % path, display_id)
         if asset.get('error') == 1: