Wrap deleting error files in query from database into an esgpull command

[AtefBN]

For the time being, if a query runs into a non-responding datanode (404 or read timeout) or metadata issue (size or checksum) the files will remain stuck in a loop of errors for as long as you retry downloading the query.

It is easy to bypass this with 5 lines of esgpull python client. But it would be much cleaner/convenient to wrap this into a subcommand and call it from command line.
Something like:

esgpull flush-query <######>

Expected behaviour: flushes the db, optionally asks if query is to be deleted