Explicit failed status and increased error handling

[drags]

Currently if the ClamAV exec fails for any reason the script was simply raising an exception and dying. This causes confusion for both users and admins and could lead to endless re-scanning for files that would never complete.

This PR does a few things:

• Introduces an AV_STATUS_FAILED status variable to explicitly denote that a scan failed to complete
• Sends metrics for failed scans
• Catches disk space exhaustion exceptions when downloading files from s3
• Refactors scan.pys lambda_handler method to reduce size and so that pieces of it can be re-used. The new methods are remove_file and download_clamav_databases

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}