Malte Langkabel
59067a9c49
Every one of the 33 standalone tests rebuilt the same seven-line preamble and the same four-lambda BuildingSystem construction, while PlacementFixture sat alongside doing exactly that. Each session of this refactor made those 33 blocks a line longer, which is what made it worth fixing now. They were not quite identical: 25 used the configured belt speed and 8 a fast belt of one tile per tick, marked by comments rather than by anything the code said. The fixture now takes an optional belt speed and kFastBeltSpeed_tps names the concept, so the difference is visible at the call site instead of buried in a static_cast. The per-test comments that explained a choice — the fast belt, the RNG seed — are kept; the fixture uses the same seed 0 those tests set by hand. 452 test cases and 3431 assertions before and after, so nothing was dropped in the conversion. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YHcUerKAZKWNvSKJxYKbnG