Merge branch 'vlive' of https://github.com/ping/youtube-dl into ping-vlive

[youtube-dl] / youtube_dl / extractor / vevo.py

diff --git a/youtube_dl/extractor/vevo.py b/youtube_dl/extractor/vevo.py
index ebab8b86c6f5727795d54b234ee4adda7258cf39..c17094f8193f7678cc3d0a912c3d970f38e6bf7c 100644 (file)
--- a/youtube_dl/extractor/vevo.py
+++ b/youtube_dl/extractor/vevo.py
@@ -4,17 +4,19 @@ import re
 import xml.etree.ElementTree
 
 from .common import InfoExtractor
-from ..utils import (
-    compat_HTTPError,
+from ..compat import (
     compat_urllib_request,
+)
+from ..utils import (
     ExtractorError,
+    int_or_none,
 )
 
 
 class VevoIE(InfoExtractor):
     """
     Accepts urls from vevo.com or in the format 'vevo:{id}'
-    (currently used by MTVIE)
+    (currently used by MTVIE and MySpaceIE)
     """
     _VALID_URL = r'''(?x)
         (?:https?://www\.vevo\.com/watch/(?:[^/]+/(?:[^/]+/)?)?|
@@ -191,9 +193,29 @@ class VevoIE(InfoExtractor):
         # Download via HLS API
         formats.extend(self._download_api_formats(video_id))
 
+        # Download SMIL
+        smil_blocks = sorted((
+            f for f in video_info['videoVersions']
+            if f['sourceType'] == 13),
+            key=lambda f: f['version'])
+        smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
+            self._SMIL_BASE_URL, video_id, video_id.lower())
+        if smil_blocks:
+            smil_url_m = self._search_regex(
+                r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
+                default=None)
+            if smil_url_m is not None:
+                smil_url = smil_url_m
+        if smil_url:
+            smil_xml = self._download_webpage(
+                smil_url, video_id, 'Downloading SMIL info', fatal=False)
+            if smil_xml:
+                formats.extend(self._formats_from_smil(smil_xml))
+
         self._sort_formats(formats)
-        timestamp_ms = int(self._search_regex(
-            r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
+        timestamp_ms = int_or_none(self._search_regex(
+            r'/Date\((\d+)\)/',
+            video_info['launchDate'], 'launch date', fatal=False))
 
         return {
             'id': video_id,