Bug #364
open
TVDB scraper stopped working
Added by dandar3 about 4 years ago.
Updated about 4 years ago.
Category:
Scraper (Music/Video Metadata
Description
TVDB scraper stopped working again.
Issue reported here:
https://www.xbmc4xbox.org.uk/forum/viewtopic.php?f=6&t=14364
I've run a quick test myself and there's two things that I found.
1) thetvdb.com will redirect the HTTP into an HTTPS
2) it fails when unpacking the zip file.
Does XBMC4Xbox support HTTPS/SSL or does it fail silently providing the scraper with an invalid ZIP file that will fail to decompress?
10:41:56 M: 32382976 DEBUG: VideoInfoScanner: Found episode match f:\tv shows\the mentalist\the.mentalist.s02e01.640x360.avi (s2e1) [[Ss]([0-9]+)[][ ._-]*[Ee]([0-9]+)([^\\/]*)$]
10:41:56 M: 32387072 DEBUG: CurlFile::Open(D00E52AC) http://www.thetvdb.com/api/1D62F2F90030C444/series/82459/all/en.zip
...
10:41:56 M: 31993856 DEBUG: advapi32.dll fake function dllCryptAcquireContext() called
10:41:56 M: 31993856 DEBUG: advapi32.dll fake function dllCryptReleaseContext() called
10:41:56 M: 31993856 DEBUG: advapi32.dll fake function dllCryptAcquireContext() called
...
10:43:21 M: 33951744 ERROR: FileZip: extended local header, not supported!
10:34:21 M: 33951744 DEBUG: scraper: GetEpisodeList returned <episodeguide></episodeguide>
The HTTPS might be the issue cause I've download the ZIP archive, uploaded into into Z: drive, navigated from XBMC4Xbox > File Manager and it was able to open it. dammit.
Our libCurl should work ok with https but would have to debug it further. Did you try using https directly bypassing the redirect ?
buzz wrote:
Our libCurl should work ok with https but would have to debug it further. Did you try using https directly bypassing the redirect ?
Thanks Buzz, no I haven't, noticed accross the scraper files we were using HTTP as opposed to HTTPS and thought that might be issue. I will try the HTTPS directly and see how that goes, at least it will rule out the redirect as an issue. (Y)
This bugs the hell out of me. I've change the protocol from HTTP to HTTPS in tvdb.xml and it completely ignores it, it still goes HTTP and drives me nuts cause I don't know where it reads it from. I've even added some aditional log statements to prove it.
19:46:53 M: 35512320 DEBUG: CScraperParser::Load() - strXMLFile = special://xbmc/system/scrapers/video/
19:46:53 M: 35512320 DEBUG: CScraperParser::Load() - strXMLFile = special://xbmc/system/scrapers/video/tvdb.xml
19:46:53 M: 35442688 DEBUG: CFileItemList::Sort, sorting took 0 millis
19:46:53 M: 35442688 DEBUG: VideoInfoScanner: Found episode match f:\tv shows\the mentalist\the.mentalist.s02e02.640x360.avi (s2e2) [[Ss]([0-9]+)[][ ._-]*[Ee]([0-9]+)([^\\/]*)$]
19:46:53 M: 35373056 DEBUG: CScraperParser::Load() - strXMLFile = special://xbmc/system/scrapers/video/tvdb.xml
19:46:53 M: 35373056 DEBUG: CScraperUrl::Get() - scrURL.m_cache = (null)
19:46:53 M: 35373056 DEBUG: CScraperUrl::Get() - scrURL.m_url = http://www.thetvdb.com/api/1D62F2F90030C444/series/82459/all/en.zip
Definitely the right scraper file (Q:/system/scrapers/video/tvdb.xml) cause if I rename it or remove it it won't print out at all. I also renamed not only the protocol thinking there might somewhere a string conversion from one to the other, even changed the last part [language].zip and added something stupid, same thing logged and same thing I see going through the proxy.
Also if I go again to try and Refresh the same episode info in the attempt to ask me and scan again, it will come back very quickly. I have to choose another episode to see it kicking CURL and then it drives me nuts with the old URL.
Any ideas?
Also available in: Atom
PDF
Updated by 
Updated by