Investigate performance of parsing & mutating large files

[hugo-vrijswijk]

I've noticed on large files such as in fs2 Stryker4s will be very slow to parse and mutate the file. We should investigate where this performance issue comes from. Ideas off the top of my head:

• ScalaMeta parsing is just slow
• More traversing of the AST than needed
• Adding more mutants in a file increases the time needed to add the next mutant (because the tree gets larger each time)
• Mutation switching adds duplicate code that will never be executed, exponentially increasing the size of a file with each mutant

Note: if anyone reading this is running into the same issue (Stryker4s taking a long time to start) and you can share a large file on which this occurs, that would be very helpful!

[hugo-vrijswijk]

Fixed in 0.15.0. Parsing and instrumenting now seems to happen in about 600ms for Stream.scala