Explicitly specify float to int conversions

Always explicitly specify how the conversion from float to integer should be done. That is, always specify one of floor, ceil, trunc, round, rint, or nearbyint. Or one of the long or long long variants, lround, llround, etc.

static_cast<int>(double) does the equivalent of static_cast<int>std::trunc(double), but that is not always the best choice. Often floor, or one of the round variants, is preferable. Explicitly specifying encourages us to make a thoughtful choice. If std::trunc is what you want, say that explicitly, too.