Better datasource

[pirj]

Redis?

Neo4j (or another graph database)?
Pros:

• Would allow things like levels of proximity between files and affected specs, as well as reduce redundancy of spec file data
  Cons:
• harder to set up and to implement the changes