Check for overflow when casting non-int/float types in Numpy arrays

[Zac-HD]

In this StackOverflow question, the lack of overflow warnings for complex numbers led to some confusion.

We should add a check for overflow in arrays of dtype complex64 (complex128 and larger can't overflow, because we only generate 64-bit floats). Here's the code that currently handles this - it's probably sufficient to apply the floats check to both real and imaginary parts:

hypothesis/hypothesis-python/src/hypothesis/extra/numpy.py

Lines 135 to 148 in b4a064f

	# Used by self.insert_element to check that the value can be stored
	# in the array without e.g. overflowing. See issue #1385.
	if dtype.kind in (u'i', u'u'):
	self.check_cast = lambda x: np.can_cast(x, self.dtype, 'safe')
	elif dtype.kind == u'f' and dtype.itemsize == 2:
	max_f2 = (2. - 2 ** -10) * 2 ** 15
	self.check_cast = lambda x: \
	(not np.isfinite(x)) or (-max_f2 <= x <= max_f2)
	elif dtype.kind == u'f' and dtype.itemsize == 4:
	max_f4 = (2. - 2 ** -23) * 2 ** 127
	self.check_cast = lambda x: \
	(not np.isfinite(x)) or (-max_f4 <= x <= max_f4)
	else:
	self.check_cast = lambda x: True

If possible we should also check for truncation in string arrays.