diff options
| author | Philipp Hagemeister <[email protected]> | 2014-08-24 02:24:47 +0200 |
|---|---|---|
| committer | Philipp Hagemeister <[email protected]> | 2014-08-24 02:24:49 +0200 |
| commit | fa8deaf38b0d576d693a6565dcdb3b29877a4c94 (patch) | |
| tree | 589dcd2f2262463e4ebe24449377172f940408be | |
| parent | 685759005919d004dd82acfc4ed88ea60de6111a (diff) | |
| download | youtube-dl-fa8deaf38b0d576d693a6565dcdb3b29877a4c94.tar.gz youtube-dl-fa8deaf38b0d576d693a6565dcdb3b29877a4c94.zip | |
[generic] Prevent from downloading a .swf as a video
We're seeing quite a number of people who do not put a video file in the og:video field, but the player URL. Try to detect some of these and filter them out.
| -rw-r--r-- | youtube_dl/extractor/generic.py | 7 |
1 files changed, 6 insertions, 1 deletions
diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py index 2ba86878b..5bd315051 100644 --- a/youtube_dl/extractor/generic.py +++ b/youtube_dl/extractor/generic.py @@ -831,7 +831,12 @@ class GenericIE(InfoExtractor): m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage) # We only look in og:video if the MIME type is a video, don't try if it's a Flash player: if m_video_type is not None: - found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage) + def check_video(vurl): + vpath = compat_urlparse.urlparse(vurl).path + return not vpath.endswith('.swf') + found = list(filter( + check_video, + re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))) if not found: # HTML5 video found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage) |