Build move-set comparison and perft validation tools

[binarylab2022-del]

Objective

Create reusable tools for comparing the legal moves produced by the current
readable move generator with moves produced by future optimized generators.

This issue does not implement new move generation and does not implement
perft. Perft is handled separately in Issue #7.

The current WinDama move generator remains the reference implementation.

Why this is needed

Comparing only the number of generated moves is insufficient.

For example:

Reference generator:
A, B, C

Candidate generator:
A, B, D