Fix equality definition to take bag semantics into account

[jeromesimeon]

Leibnitz equality (eq) equates two collections only if they have the same order.
Since we assume collections are bags, we should be using bag equality (equivalence modulo permutation).

Note that we avoided this issue for records by enforcing that well-typed records are sorted.
In this case, sorting would be more painful, since we would need a <= relation on all data (not just on strings).

• Add the right equivalence relation for data, and show that it is an equivalence relation and that it is decidable.
• Show that it is a morphism for the data_type relation.
• Show that it is a morphism for the evaluation relations.
  • CAMP
  • NRA*
  • NNRC*
• Change our untyped and typed equivalence definitions
  • CAMP
  • NRA*
  • NNRC*

In addition, operators (e.g., ADistinct), should be reviewed to check that they rely on the proper equality relation).