[youtube:user] Preserve user/c path segment

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

diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py
index 23a6fcd88b6ffef90f3ce0fcdf7fe6abd770ca90..0d43acc4ac7ff0162bdcc95471ccdb2376c0f7f2 100644 (file)
--- a/youtube_dl/extractor/facebook.py
+++ b/youtube_dl/extractor/facebook.py
@@ -38,7 +38,8 @@ class FacebookIE(InfoExtractor):
                                 story\.php
                             )\?(?:.*?)(?:v|video_id|story_fbid)=|
                             [^/]+/videos/(?:[^/]+/)?|
-                            [^/]+/posts/
+                            [^/]+/posts/|
+                            groups/[^/]+/permalink/
                         )|
                     facebook:
                 )
@@ -93,6 +94,24 @@ class FacebookIE(InfoExtractor):
             'title': '"What are you doing running in the snow?"',
             'uploader': 'FailArmy',
         }
+    }, {
+        'url': 'https://m.facebook.com/story.php?story_fbid=1035862816472149&id=116132035111903',
+        'md5': '1deb90b6ac27f7efcf6d747c8a27f5e3',
+        'info_dict': {
+            'id': '1035862816472149',
+            'ext': 'mp4',
+            'title': 'What the Flock Is Going On In New Zealand  Credit: ViralHog',
+            'uploader': 'S. Saint',
+        },
+    }, {
+        'note': 'swf params escaped',
+        'url': 'https://www.facebook.com/barackobama/posts/10153664894881749',
+        'md5': '97ba073838964d12c70566e0085c2b91',
+        'info_dict': {
+            'id': '10153664894881749',
+            'ext': 'mp4',
+            'title': 'Facebook video #10153664894881749',
+        },
     }, {
         'url': 'https://www.facebook.com/video.php?v=10204634152394104',
         'only_matching': True,
@@ -106,10 +125,25 @@ class FacebookIE(InfoExtractor):
         'url': 'facebook:544765982287235',
         'only_matching': True,
     }, {
-        'url': 'https://m.facebook.com/story.php?story_fbid=1035862816472149&id=116132035111903',
+        'url': 'https://www.facebook.com/groups/164828000315060/permalink/764967300301124/',
         'only_matching': True,
     }]
 
+    @staticmethod
+    def _extract_url(webpage):
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
+        if mobj is not None:
+            return mobj.group('url')
+
+        # Facebook API embed
+        # see https://developers.facebook.com/docs/plugins/embedded-video-player
+        mobj = re.search(r'''(?x)<div[^>]+
+                class=(?P<q1>[\'"])[^\'"]*\bfb-(?:video|post)\b[^\'"]*(?P=q1)[^>]+
+                data-href=(?P<q2>[\'"])(?P<url>(?:https?:)?//(?:www\.)?facebook.com/.+?)(?P=q2)''', webpage)
+        if mobj is not None:
+            return mobj.group('url')
+
     def _login(self):
         (useremail, password) = self._get_login_info()
         if useremail is None:
@@ -183,13 +217,25 @@ class FacebookIE(InfoExtractor):
 
         video_data = None
 
-        BEFORE = '{swf.addParam(param[0], param[1]);});\n'
+        BEFORE = '{swf.addParam(param[0], param[1]);});'
         AFTER = '.forEach(function(variable) {swf.addVariable(variable[0], variable[1]);});'
-        m = re.search(re.escape(BEFORE) + '(.*?)' + re.escape(AFTER), webpage)
-        if m:
-            data = dict(json.loads(m.group(1)))
+        PATTERN = re.escape(BEFORE) + '(?:\n|\\\\n)(.*?)' + re.escape(AFTER)
+
+        for m in re.findall(PATTERN, webpage):
+            swf_params = m.replace('\\\\', '\\').replace('\\"', '"')
+            data = dict(json.loads(swf_params))
             params_raw = compat_urllib_parse_unquote(data['params'])
-            video_data = json.loads(params_raw)['video_data']
+            video_data_candidate = json.loads(params_raw)['video_data']
+            for _, f in video_data_candidate.items():
+                if not f:
+                    continue
+                if isinstance(f, dict):
+                    f = [f]
+                if isinstance(f, list):
+                    continue
+                if f[0].get('video_id') == video_id:
+                    video_data = video_data_candidate
+                    break
 
         def video_data_list2dict(video_data):
             ret = {}
@@ -200,7 +246,7 @@ class FacebookIE(InfoExtractor):
 
         if not video_data:
             server_js_data = self._parse_json(self._search_regex(
-                r'handleServerJS\(({.+})\);', webpage, 'server js data'), video_id)
+                r'handleServerJS\(({.+})\);', webpage, 'server js data', default='{}'), video_id)
             for item in server_js_data.get('instances', []):
                 if item[1][0] == 'VideoConfig':
                     video_data = video_data_list2dict(item[2][0]['videoData'])
@@ -219,6 +265,8 @@ class FacebookIE(InfoExtractor):
 
         formats = []
         for format_id, f in video_data.items():
+            if f and isinstance(f, dict):
+                f = [f]
             if not f or not isinstance(f, list):
                 continue
             for quality in ('sd', 'hd'):
@@ -274,8 +322,8 @@ class FacebookIE(InfoExtractor):
 
         if '/posts/' in url:
             entries = [
-                self.url_result('facebook:%s' % video_id, FacebookIE.ie_key())
-                for video_id in self._parse_json(
+                self.url_result('facebook:%s' % vid, FacebookIE.ie_key())
+                for vid in self._parse_json(
                     self._search_regex(
                         r'(["\'])video_ids\1\s*:\s*(?P<ids>\[.+?\])',
                         webpage, 'video ids', group='ids'),