resolves #200 rewrite main loop in aggregateContent to operate on unique URLs

• rename openOrCloneRepository to loadRepo; don't propogate unnecessary data
• group content sources by URL
• iterate over unique URLs
• extract repository processing logic into function (collectComponentVersions)
• extract ref processing logic into function (populateComponentVersion)
• add test for multiple content sources with same URL
• isolate preparation of cache dir into an await call
• remove superfluous async keyword in certain tests
• rearrange functions in content aggregator
• interpret string branches and tags pattern defined in playbook as CSV
• use pattern -.git for name of cloned folder
• use consistent style for requiring path.posix