From fa4bc6e71261613cf530437a2407ff7b61ea6cb5 Mon Sep 17 00:00:00 2001 From: Remita Amine Date: Tue, 17 Oct 2017 10:07:37 +0000 Subject: [PATCH] [youtube] replace youtube redirect urls in description(fixes #14517) --- youtube_dl/extractor/youtube.py | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index 4e8db240d..5aef555fb 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -1622,6 +1622,17 @@ def extract_view_count(v_info): # description description_original = video_description = get_element_by_id("eow-description", video_webpage) if video_description: + + def replace_url(m): + redir_url = compat_urlparse.urljoin(url, m.group(1)) + parsed_redir_url = compat_urllib_parse_urlparse(redir_url) + if re.search(r'^(?:www\.)?(?:youtube(?:-nocookie)?\.com|youtu\.be)$', parsed_redir_url.netloc) and parsed_redir_url.path == '/redirect': + qs = compat_parse_qs(parsed_redir_url.query) + q = qs.get('q') + if q and q[0]: + return q[0] + return redir_url + description_original = video_description = re.sub(r'''(?x) ]*> [^<]+\.{3}\s* - ''', lambda m: compat_urlparse.urljoin(url, m.group(1)), video_description) + ''', replace_url, video_description) video_description = clean_html(video_description) else: fd_mobj = re.search(r'