Update the changelog.

[youtubedl] / youtube_dl / extractor / facebook.py

diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py
index 789dd79d5474ea8f7c9d18df244c686076e3e646..ce64e26831fdafceb97b6d8ae919c00a78f0f90f 100644 (file)
--- a/youtube_dl/extractor/facebook.py
+++ b/youtube_dl/extractor/facebook.py
@@ -334,7 +334,7 @@ class FacebookIE(InfoExtractor):
         if not video_data:
             server_js_data = self._parse_json(
                 self._search_regex(
         if not video_data:
             server_js_data = self._parse_json(
                 self._search_regex(

-                    r'bigPipe\.onPageletArrive\(({.+?})\)\s*;\s*}\s*\)\s*,\s*["\']onPageletArrive\s+(?:stream_pagelet|pagelet_group_mall|permalink_video_pagelet)',
+                    r'bigPipe\.onPageletArrive\(({.+?})\)\s*;\s*}\s*\)\s*,\s*["\']onPageletArrive\s+(?:pagelet_group_mall|permalink_video_pagelet|hyperfeed_story_id_\d+)',

                     webpage, 'js data', default='{}'),
                 video_id, transform_source=js_to_json, fatal=False)
             video_data = extract_from_jsmods_instances(server_js_data)
                     webpage, 'js data', default='{}'),
                 video_id, transform_source=js_to_json, fatal=False)
             video_data = extract_from_jsmods_instances(server_js_data)


@@ -379,6 +379,7 @@ class FacebookIE(InfoExtractor):
         if not video_data:
             raise ExtractorError('Cannot parse data')
 
         if not video_data:
             raise ExtractorError('Cannot parse data')
 

+        subtitles = {}

         formats = []
         for f in video_data:
             format_id = f['stream_type']
         formats = []
         for f in video_data:
             format_id = f['stream_type']


@@ -402,9 +403,17 @@ class FacebookIE(InfoExtractor):
             if dash_manifest:
                 formats.extend(self._parse_mpd_formats(
                     compat_etree_fromstring(compat_urllib_parse_unquote_plus(dash_manifest))))
             if dash_manifest:
                 formats.extend(self._parse_mpd_formats(
                     compat_etree_fromstring(compat_urllib_parse_unquote_plus(dash_manifest))))

+            subtitles_src = f[0].get('subtitles_src')
+            if subtitles_src:
+                subtitles.setdefault('en', []).append({'url': subtitles_src})

         if not formats:
             raise ExtractorError('Cannot find video formats')
 
         if not formats:
             raise ExtractorError('Cannot find video formats')
 

+        # Downloads with browser's User-Agent are rate limited. Working around
+        # with non-browser User-Agent.
+        for f in formats:
+            f.setdefault('http_headers', {})['User-Agent'] = 'facebookexternalhit/1.1'
+

         self._sort_formats(formats)
 
         video_title = self._html_search_regex(
         self._sort_formats(formats)
 
         video_title = self._html_search_regex(


@@ -428,7 +437,7 @@ class FacebookIE(InfoExtractor):
         timestamp = int_or_none(self._search_regex(
             r'<abbr[^>]+data-utime=["\'](\d+)', webpage,
             'timestamp', default=None))
         timestamp = int_or_none(self._search_regex(
             r'<abbr[^>]+data-utime=["\'](\d+)', webpage,
             'timestamp', default=None))

-        thumbnail = self._og_search_thumbnail(webpage)
+        thumbnail = self._html_search_meta(['og:image', 'twitter:image'], webpage)

 
         view_count = parse_count(self._search_regex(
             r'\bviewCount\s*:\s*["\']([\d,.]+)', webpage, 'view count',
 
         view_count = parse_count(self._search_regex(
             r'\bviewCount\s*:\s*["\']([\d,.]+)', webpage, 'view count',


@@ -442,6 +451,7 @@ class FacebookIE(InfoExtractor):
             'timestamp': timestamp,
             'thumbnail': thumbnail,
             'view_count': view_count,
             'timestamp': timestamp,
             'thumbnail': thumbnail,
             'view_count': view_count,

+            'subtitles': subtitles,

         }
 
         return webpage, info_dict
         }
 
         return webpage, info_dict