Stable test output in logs

[Mingun]

Use ListMap to represent expected test results so tests always run in the same order.

This will result the same order of failed tests in the logs which makes it much more easy to compare.

[GreyCat]

I love the idea, but I would suggest a slightly different implementation. Rather than copy-pasting long formula ListMap[LanguageCompilerStatic, String] everywhere, we already have an alias:

type ResultMap = Map[LanguageCompilerStatic, String]

And it's actually exactly how all test functions are defined:

  def full(src: String, srcType: DataType, expType: DataType, expOut: ResultMap)
  def full(src: String, tp: TypeProvider, expType: DataType, expOut: ResultMap)
  def everybody(src: String, expOut: String, expType: DataType = CalcIntType)
  def everybodyExcept(src: String, commonExpOut: String, rm: ResultMap, expType: DataType = CalcIntType)

Shall we just swap that to use ListMap + then swap all the test invocations to use ResultMap instead of lengthy Map[LanguageCompilerStatic, String]?

[GreyCat]

@Mingun Any objections to merge it this way, or we're good to go?

[Mingun]

I have no objections, please merge.