easier typecheck and TypeError thrown

pyfqmr/Simplify.pyx

@@ -109,9 +109,9 @@ cdef class Simplify :
         self.normals_cpp.clear()
 
         # Here we will need some checks, just to make sure the right objets are passed
-        # Only checks on the faces is possible, as they are necessarily integers.
         # Maybe other checks like max(faces)==len(vertices) could be added
-        assert is_integer_array(faces), "Faces must contain only integers, check function argument order"
+        if not np.issubdtype(faces.dtype, np.integer):
+            raise TypeError("faces array must be of integer type")
 
         self.faces_mv = faces.astype(dtype="int32", subok=False, copy=False)
         self.vertices_mv = vertices.astype(dtype="float64", subok=False, copy=False)
@@ -244,25 +244,3 @@ cdef vector[vector[int]] setFacesNogil(int[:,:] faces, vector[vector[int]] vecto
             triangle.push_back(faces[i,j])
         vector_faces.push_back(triangle)
     return vector_faces
-
-
-# For faster type checking
-def is_integer_array(arr):
-    # Check if the dtype is integer before calling Cython function
-    if np.issubdtype(arr.dtype, np.integer):
-        return True
-    return is_integer_array_cython(arr)
-
-@cython.boundscheck(False)
-@cython.wraparound(False)
-@cython.nonecheck(False)
-def is_integer_array_cython(np.ndarray arr):
-    cdef Py_ssize_t i, j
-    cdef Py_ssize_t n_rows = arr.shape[0]
-    cdef Py_ssize_t n_cols = arr.shape[1]
-
-    for i in range(n_rows):
-        for j in range(n_cols):
-            if arr[i, j] % 1 != 0:
-                return False
-    return True