Build Babel-output-wide testing platform

[gaurav]

This repository contains a Scala application that can load a single compendium or synonyms file and run various tests on it, but there are some tests that require us to search across the entire Babel output:

• Look for the same preferred labels being used in multiple cliques, possibly in different compendia files
• Look for the same ID being used in multiple cliques

Jim suggested that Souffle might be an efficient way of doing this for some tests (e.g. looking for the exact same preferred label shared by multiple identifiers), but at some point the size of the Babel outputs will be too much. In this case, the only way to run these tests would be to either provide a low-level endpoint on NodeNorm/NameRes or to write tests directly against the Redis/Solr databases.