Improve Unit Tests

[marip8]

Problems

• No real expectations on the results of the tool path planner in any test, just that a valid optional data type is returned
• Surface walk planner does not produce expected results (see Rasters do not extend all the way across the mesh when rotating them #34)
  • 
  • 
  • 
  • Non-deterministic results from surface walk planner(see Inconsistent generated toolpaths using surface_raster_planner_demo.launch #103, Tool Path Planner unit tests sometimes don't pass #47)
    • 
    • 
• Eigen value edge path planner does not produce expected results
  • 
  • 
  • 
• Extra rasters feature also produces (one?) extra waypoints
  • 

Solutions

• Revise unit tests to use test fixtures to reduce code duplication
• Add meaningful expectations to the results of the tool path generation (e.g. number of strokes, number of points per stroke, normal direction
• Fix issues with surface walk and eigen value edge planners
• Transform test meshes by non-identity transformation to ensure results are correct and not just coincidentally correct (see Fix plane slice planner #136)