Langrange Quadratic FieldShape Incorrect for Triangles

[JaredCrean2]

The hasNodesIn function for LagrangeQuadratic returns results that are inconsistent with the EntityShape it returns from getEntityShape. The problem is hasNodesIn says there are nodes on dimension 2 entities, but getEntityShape only special cases Quads. For triangles, it calls the QuadraticBase function, which returns a serendipity (only 6 nodes) triangle. This leads to problems with numberings/fields, because using hasNodesIn to skip over entities in loops doesn't work.

For my case, I can work around it by using SerendipityQuadratic.