[wistia] Add extractor

youtube_dl/YoutubeDL.py

@@ -488,7 +488,8 @@ def make_result(embedded_info):
                 new_result = ie_result.copy()
                 for f in ('_type', 'url', 'ext', 'player_url', 'formats',
                           'entries', 'urlhandle', 'ie_key', 'duration',
-                          'subtitles', 'annotations', 'format'):
+                          'subtitles', 'annotations', 'format',
+                          'thumbnail', 'thumbnails'):
                     if f in new_result:
                         del new_result[f]
                     if f in embedded_info:

youtube_dl/extractor/__init__.py

@@ -178,6 +178,7 @@
 from .websurg import WeBSurgIE
 from .weibo import WeiboIE
 from .wimp import WimpIE
+from .wistia import WistiaIE
 from .worldstarhiphop import WorldStarHipHopIE
 from .xhamster import XHamsterIE
 from .xnxx import XNXXIE

youtube_dl/extractor/generic.py

@@ -169,8 +169,13 @@ def _real_extract(self, url):
         #   Site Name | Video Title
         #   Video Title - Tagline | Site Name
         # and so on and so forth; it's just not practical
-        video_title = self._html_search_regex(r'<title>(.*)</title>',
-            webpage, u'video title', default=u'video', flags=re.DOTALL)
+        video_title = self._html_search_regex(
+            r'(?s)<title>(.*?)</title>', webpage, u'video title',
+            default=u'video')
+
+        # video uploader is domain name
+        video_uploader = self._search_regex(
+            r'^(?:https?://)?([^/]*)/.*', url, u'video uploader')
 
         # Look for BrightCove:
         bc_url = BrightcoveIE._extract_brightcove_url(webpage)
@@ -188,7 +193,7 @@ def _real_extract(self, url):
 
         # Look for embedded YouTube player
         matches = re.findall(
-            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube.com/embed/.+?)\1', webpage)
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/embed/.+?)\1', webpage)
         if matches:
             urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
                      for tuppl in matches]
@@ -197,13 +202,26 @@ def _real_extract(self, url):
 
         # Look for embedded Dailymotion player
         matches = re.findall(
-            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion.com/embed/video/.+?)\1', webpage)
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
         if matches:
             urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
                      for tuppl in matches]
             return self.playlist_result(
                 urlrs, playlist_id=video_id, playlist_title=video_title)
 
+        # Look for embedded Wistia player
+        match = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
+        if match:
+            return {
+                '_type': 'url_transparent',
+                'url': unescapeHTML(match.group('url')),
+                'ie_key': 'Wistia',
+                'uploader': video_uploader,
+                'title': video_title,
+                'id': video_id,
+            }
+
         # Look for Bandcamp pages with custom domain
         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
         if mobj is not None:
@@ -247,14 +265,9 @@ def _real_extract(self, url):
         # here's a fun little line of code for you:
         video_id = os.path.splitext(video_id)[0]
 
-        # video uploader is domain name
-        video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
-            url, u'video uploader')
-
         return {
             'id':       video_id,
             'url':      video_url,
             'uploader': video_uploader,
-            'upload_date':  None,
             'title':    video_title,
         }

youtube_dl/extractor/wistia.py

@@ -0,0 +1,55 @@
+import json
+import re
+
+from .common import InfoExtractor
+
+
+class WistiaIE(InfoExtractor):
+    _VALID_URL = r'^https?://(?:fast\.)?wistia\.net/embed/iframe/(?P<id>[a-z0-9]+)'
+
+    _TEST = {
+        u"url": u"http://fast.wistia.net/embed/iframe/sh7fpupwlt",
+        u"file": u"sh7fpupwlt.mov",
+        u"md5": u"cafeb56ec0c53c18c97405eecb3133df",
+        u"info_dict": {
+            u"title": u"cfh_resourceful_zdkh_final_1"
+        },
+    }
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+
+        webpage = self._download_webpage(url, video_id)
+        data_json = self._html_search_regex(
+            r'Wistia.iframeInit\((.*?), {}\);', webpage, u'video data')
+
+        data = json.loads(data_json)
+
+        formats = []
+        thumbnails = []
+        for atype, a in data['assets'].items():
+            if atype == 'still':
+                thumbnails.append({
+                    'url': a['url'],
+                    'resolution': '%dx%d' % (a['width'], a['height']),
+                })
+                continue
+            if atype == 'preview':
+                continue
+            formats.append({
+                'format_id': atype,
+                'url': a['url'],
+                'width': a['width'],
+                'height': a['height'],
+                'filesize': a['size'],
+                'ext': a['ext'],
+            })
+        formats.sort(key=lambda a: a['filesize'])
+
+        return {
+            'id': video_id,
+            'title': data['name'],
+            'formats': formats,
+            'thumbnails': thumbnails,
+        }