When importing cvinfo w/o full URL list index out of range causes exception

[boshuda]

When importing a folder containing a cvinfo file the following appeared in the log: ERROR :: mylar.scanLibrary.620 : LibraryScan : [IMPORT] Unable to complete the scan: list index out of range

I traced it through to libraryScan in librarysync.py, around line 254:

            try:
                with open(os.path.join(os.path.dirname(comlocation), 'cvinfo')) as f:
                    urllink = f.readline()

                if urllink:
                    cid = urllink.split('/')
                    if '4050-' in cid[-2]:
                        cvinfo_CID = re.sub('4050-', '', cid[-2]).strip()
                        logger.info('CVINFO file located within directory. Attaching everything in directory that is valid to ComicID: ' + str(cvinfo_CID))
                        #store the location of the cvinfo so it's applied to the correct directory (since we're scanning multile direcorties usually)
                        cvscanned_loc = os.path.dirname(comlocation)
                else:
                    logger.error("Could not read cvinfo file properly (or it does not contain any data)")
            except (OSError, IOError):
                logger.error("Could not read cvinfo file properly (or it does not contain any data)")

You can see it splits the urllink by the '/' character. However, a cvinfo file as described in Comicvine Scraper doesn't require the full url. Just the 4050-xxxxx part. So when it isn't able to create a list long enough to look for cid[-2], the index out of range error occurs. Seems like it would be reasonable to do a regex to look for the digits following the 4050- portion, and that would take care of any format a person might use.

While filing this, I did a fork and a fix. I'll commit the change and make a pull request.