diff options
author | Sergey M․ <dstftw@gmail.com> | 2017-02-01 23:15:38 +0700 |
---|---|---|
committer | Sergey M․ <dstftw@gmail.com> | 2017-02-01 23:19:20 +0700 |
commit | b83ef507b457e6ea8c52265ea42b6c5d2c500a7e (patch) | |
tree | 8554cf9dad9f3f4b3772cd58e383f27ae1f5ef9a | |
parent | 000f207944e277e63dbec5a60007c30e3187d3fd (diff) |
[facebook] Fix extraction (closes #11926)
-rw-r--r-- | youtube_dl/extractor/facebook.py | 36 |
1 files changed, 27 insertions, 9 deletions
diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py index c0a7fc7d8..47bcc0dbc 100644 --- a/youtube_dl/extractor/facebook.py +++ b/youtube_dl/extractor/facebook.py @@ -12,14 +12,16 @@ from ..compat import ( compat_urllib_parse_unquote_plus, ) from ..utils import ( + clean_html, error_to_compat_str, ExtractorError, + get_element_by_id, int_or_none, + js_to_json, limit_length, sanitized_Request, + try_get, urlencode_postdata, - get_element_by_id, - clean_html, ) @@ -243,14 +245,30 @@ class FacebookIE(InfoExtractor): video_data = None + def extract_video_data(instances): + for item in instances: + if item[1][0] == 'VideoConfig': + video_item = item[2][0] + if video_item.get('video_id') == video_id: + return video_item['videoData'] + server_js_data = self._parse_json(self._search_regex( - r'handleServerJS\(({.+})(?:\);|,")', webpage, 'server js data', default='{}'), video_id) - for item in server_js_data.get('instances', []): - if item[1][0] == 'VideoConfig': - video_item = item[2][0] - if video_item.get('video_id') == video_id: - video_data = video_item['videoData'] - break + r'handleServerJS\(({.+})(?:\);|,")', webpage, + 'server js data', default='{}'), video_id, fatal=False) + + if server_js_data: + video_data = extract_video_data(server_js_data.get('instances', [])) + + 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', + webpage, 'js data', default='{}'), + video_id, transform_source=js_to_json, fatal=False) + if server_js_data: + video_data = extract_video_data(try_get( + server_js_data, lambda x: x['jsmods']['instances'], + list) or []) if not video_data: if not fatal_if_no_video: |