Update upstream source from tag 'upstream/2020.06.16.1'

[youtubedl] / youtube_dl / extractor / facebook.py

diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py
index 97cfe0fc38ed53f32fd868fd8faa6276a3362dfe..610d6674592384922f9df7af4da5958592ce56bd 100644 (file)
--- a/youtube_dl/extractor/facebook.py
+++ b/youtube_dl/extractor/facebook.py
@@ -57,7 +57,7 @@ class FacebookIE(InfoExtractor):
     _CHROME_USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.97 Safari/537.36'
 
     _VIDEO_PAGE_TEMPLATE = 'https://www.facebook.com/video/video.php?v=%s'
     _CHROME_USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.97 Safari/537.36'
 
     _VIDEO_PAGE_TEMPLATE = 'https://www.facebook.com/video/video.php?v=%s'

-    _VIDEO_PAGE_TAHOE_TEMPLATE = 'https://www.facebook.com/video/tahoe/async/%s/?chain=true&isvideo=true'
+    _VIDEO_PAGE_TAHOE_TEMPLATE = 'https://www.facebook.com/video/tahoe/async/%s/?chain=true&isvideo=true&payloadtype=primary'

 
     _TESTS = [{
         'url': 'https://www.facebook.com/video.php?v=637842556329505&fref=nf',
 
     _TESTS = [{
         'url': 'https://www.facebook.com/video.php?v=637842556329505&fref=nf',


@@ -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(


@@ -424,11 +433,11 @@ class FacebookIE(InfoExtractor):
         uploader = clean_html(get_element_by_id(
             'fbPhotoPageAuthorName', webpage)) or self._search_regex(
             r'ownerName\s*:\s*"([^"]+)"', webpage, 'uploader',
         uploader = clean_html(get_element_by_id(
             'fbPhotoPageAuthorName', webpage)) or self._search_regex(
             r'ownerName\s*:\s*"([^"]+)"', webpage, 'uploader',

-            fatal=False) or self._og_search_title(webpage, fatal=False)
+            default=None) or self._og_search_title(webpage, fatal=False)

         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


@@ -456,15 +466,18 @@ class FacebookIE(InfoExtractor):
             return info_dict
 
         if '/posts/' in url:
             return info_dict
 
         if '/posts/' in url:

-            entries = [
-                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'),
-                    video_id)]
-
-            return self.playlist_result(entries, video_id)
+            video_id_json = self._search_regex(
+                r'(["\'])video_ids\1\s*:\s*(?P<ids>\[.+?\])', webpage, 'video ids', group='ids',
+                default='')
+            if video_id_json:
+                entries = [
+                    self.url_result('facebook:%s' % vid, FacebookIE.ie_key())
+                    for vid in self._parse_json(video_id_json, video_id)]
+                return self.playlist_result(entries, video_id)
+
+            # Single Video?
+            video_id = self._search_regex(r'video_id:\s*"([0-9]+)"', webpage, 'single video id')
+            return self.url_result('facebook:%s' % video_id, FacebookIE.ie_key())

         else:
             _, info_dict = self._extract_from_url(
                 self._VIDEO_PAGE_TEMPLATE % video_id,
         else:
             _, info_dict = self._extract_from_url(
                 self._VIDEO_PAGE_TEMPLATE % video_id,