[youtube] Do not crash if caption info is missing altogether (Fixes #1610)

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

diff --git a/youtube_dl/extractor/vimeo.py b/youtube_dl/extractor/vimeo.py
index 512e06e2a620161704c8b1d02854c1279276179e..2de56ac814462e3c3536ccac34b980b3e9a8bfb5 100644 (file)
--- a/youtube_dl/extractor/vimeo.py
+++ b/youtube_dl/extractor/vimeo.py
@@ -11,13 +11,14 @@ from ..utils import (
     get_element_by_attribute,
     ExtractorError,
     std_headers,
+    unsmuggle_url,
 )
 
 class VimeoIE(InfoExtractor):
     """Information extractor for vimeo.com."""
 
     # _VALID_URL matches Vimeo URLs
-    _VALID_URL = r'(?P<proto>https?://)?(?:(?:www|player)\.)?vimeo(?P<pro>pro)?\.com/(?:(?:(?:groups|album)/[^/]+)|(?:.*?)/)?(?P<direct_link>play_redirect_hls\?clip_id=)?(?:videos?/)?(?P<id>[0-9]+)(?:[?].*)?$'
+    _VALID_URL = r'(?P<proto>https?://)?(?:(?:www|player)\.)?vimeo(?P<pro>pro)?\.com/(?:(?:(?:groups|album)/[^/]+)|(?:.*?)/)?(?P<direct_link>play_redirect_hls\?clip_id=)?(?:videos?/)?(?P<id>[0-9]+)/?(?:[?].*)?$'
     _NETRC_MACHINE = 'vimeo'
     IE_NAME = u'vimeo'
     _TESTS = [
@@ -44,6 +45,16 @@ class VimeoIE(InfoExtractor):
                 u'title': u'Andy Allan - Putting the Carto into OpenStreetMap Cartography',
             },
         },
+        {
+            u'url': u'http://player.vimeo.com/video/54469442',
+            u'file': u'54469442.mp4',
+            u'md5': u'619b811a4417aa4abe78dc653becf511',
+            u'note': u'Videos that embed the url in the player page',
+            u'info_dict': {
+                u'title': u'Kathy Sierra: Building the minimum Badass User, Business of Software',
+                u'uploader': u'The BLN & Business of Software',
+            },
+        }
     ]
 
     def _login(self):
@@ -88,6 +99,12 @@ class VimeoIE(InfoExtractor):
         self._login()
 
     def _real_extract(self, url, new_video=True):
+        url, data = unsmuggle_url(url)
+        headers = std_headers
+        if data is not None:
+            headers = headers.copy()
+            headers.update(data)
+
         # Extract ID from URL
         mobj = re.match(self._VALID_URL, url)
         if mobj is None:
@@ -102,7 +119,7 @@ class VimeoIE(InfoExtractor):
             url = 'https://vimeo.com/' + video_id
 
         # Retrieve video webpage to extract further information
-        request = compat_urllib_request.Request(url, None, std_headers)
+        request = compat_urllib_request.Request(url, None, headers)
         webpage = self._download_webpage(request, video_id)
 
         # Now we begin extracting as much information as we can from what we
@@ -112,7 +129,8 @@ class VimeoIE(InfoExtractor):
 
         # Extract the config JSON
         try:
-            config = webpage.split(' = {config:')[1].split(',assets:')[0]
+            config = self._search_regex([r' = {config:({.+?}),assets:', r'c=({.+?);'],
+                webpage, u'info section', flags=re.DOTALL)
             config = json.loads(config)
         except:
             if re.search('The creator of this video has not given you permission to embed it on this domain.', webpage):
@@ -132,12 +150,22 @@ class VimeoIE(InfoExtractor):
         video_uploader_id = config["video"]["owner"]["url"].split('/')[-1] if config["video"]["owner"]["url"] else None
 
         # Extract video thumbnail
-        video_thumbnail = config["video"]["thumbnail"]
+        video_thumbnail = config["video"].get("thumbnail")
+        if video_thumbnail is None:
+            _, video_thumbnail = sorted((int(width), t_url) for (width, t_url) in config["video"]["thumbs"].items())[-1]
 
         # Extract video description
-        video_description = get_element_by_attribute("itemprop", "description", webpage)
-        if video_description: video_description = clean_html(video_description)
-        else: video_description = u''
+        video_description = None
+        try:
+            video_description = get_element_by_attribute("itemprop", "description", webpage)
+            if video_description: video_description = clean_html(video_description)
+        except AssertionError as err:
+            # On some pages like (http://player.vimeo.com/video/54469442) the
+            # html tags are not closed, python 2.6 cannot handle it
+            if err.args[0] == 'we should not get here!':
+                pass
+            else:
+                raise
 
         # Extract upload date
         video_upload_date = None
@@ -154,14 +182,15 @@ class VimeoIE(InfoExtractor):
         # TODO bind to format param
         codecs = [('h264', 'mp4'), ('vp8', 'flv'), ('vp6', 'flv')]
         files = { 'hd': [], 'sd': [], 'other': []}
+        config_files = config["video"].get("files") or config["request"].get("files")
         for codec_name, codec_extension in codecs:
-            if codec_name in config["video"]["files"]:
-                if 'hd' in config["video"]["files"][codec_name]:
+            if codec_name in config_files:
+                if 'hd' in config_files[codec_name]:
                     files['hd'].append((codec_name, codec_extension, 'hd'))
-                elif 'sd' in config["video"]["files"][codec_name]:
+                elif 'sd' in config_files[codec_name]:
                     files['sd'].append((codec_name, codec_extension, 'sd'))
                 else:
-                    files['other'].append((codec_name, codec_extension, config["video"]["files"][codec_name][0]))
+                    files['other'].append((codec_name, codec_extension, config_files[codec_name][0]))
 
         for quality in ('hd', 'sd', 'other'):
             if len(files[quality]) > 0:
@@ -173,8 +202,12 @@ class VimeoIE(InfoExtractor):
         else:
             raise ExtractorError(u'No known codec found')
 
-        video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
-                    %(video_id, sig, timestamp, video_quality, video_codec.upper())
+        video_url = None
+        if isinstance(config_files[video_codec], dict):
+            video_url = config_files[video_codec][video_quality].get("url")
+        if video_url is None:
+            video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
+                        %(video_id, sig, timestamp, video_quality, video_codec.upper())
 
         return [{
             'id':       video_id,