Add support for float inputs in trimesh.grouping.boolean_rows

[lucagrementieri]

The function trimesh.grouping.boolean_rows is described as a very general function to find matching rows in arbitrary matrices but internally it converts the arrays to the type np.int64.

Therefore, if the input arrays have type np.float32 or np.float64 the decimal digits are discarded leading to a completely different behaviour.

Would it be possible to generalise the function to compare the binary representation of rows of whatever datatype?

Example to reproduce

Trimesh version: 4.0.7

>>> import trimesh
>>> import numpy as np
>>> v1 = np.linspace(0,1,6, endpoint=False).reshape(2,3)
>>> v2 = np.linspace(0,2,12, endpoint=False).reshape(4,3)
>>> trimesh.grouping.boolean_rows(v1, v2)
array([[0, 0, 0]])

[lucagrementieri]

A simple solution should be to just remove the parameter dtype inside the call to np.asanyarray. https://github.com/mikedh/trimesh/blob/7c90a94a427e7af1152bdc2dfb3e6ab6cfd2826c/trimesh/grouping.py#L604