Added code to prevent parsed search results with UNKNOWN quality from being cached

Revert "Revert "Added feature that stores and retrieves provider cache results from sickrage's cache server""

This reverts commit 9575d5fd

Revert "Added feature that stores and retrieves provider cache results from sickrage's cache server"

This reverts commit b024c058

Fixed issues with conflicting database index class names

fix: Added integrity check for databases that will auto-delete corrupt indexes, resolves marshal errors
fix: Properly sort next episode dates when checking for next air date of new episodes

fix: Misc cleanups, Resolved torrent provider seed ratio issue

fix: Fixed code for source updating

HOTFIX: Removed code in regards to auto-installation of requirements/depends due to inconsistent operational results

FIX: Conversion of old-style logging calls to new-style logging calls per new logging module code reqs.
FIX: Using unicode_literals to futureproof py2>py3 conversion.
FIX: Converted old string index lookups to new bytestring index lookups to comply with unicode_literals.
RESTRUCTURE: Gave tests overhaul to making testing of the app code easier and more efficiant.

This version of SickBeard uses both TVDB and TVRage to search and gather it's series data from allowing you to now have access to and download shows that you couldn't before because of being locked into only what TheTVDB had to offer.

Also this edition is based off the code we used in our XEM editon so it does come with scene numbering support as well as all the other features our XEM edition has to offer.

Please before using this with your existing database (sickbeard.db) please make a backup copy of it and delete any other database files such as cache.db and failed.db if present, we HIGHLY recommend starting out with no database files at all to make this a fresh start but the choice is at your own risk!

Enjoy!