How to join two dtypes and two arrays (numpy)

Question

I'm trying to join two arrays into one: The first one is the easiest being just:

["str".....]

with the dtype:

dtype('S22')

The second one being a bit more complicated:

[([(int, "str"),(int, "str")],)...]

with the dtype:

dtype([('table_pointer', [('table', '<i2'), ('pointer', 'S22')], (2,))])

And when i try to join them with np.array:

pTable = np.array([table1, table2], dtype = [('keys', table1), ('vals', table2)])

I get the following error:

invalid literal for int() with base 10: b'str'

I've tried using np.stack, np.vstack and np.concatenate, but they all refused to accept dtypes with subarrays

Here is the full code:

    dtTable1 = np.dtype("<S22")
    table1 = np.fromiter(pointerTable.keys(), dtTable1)
    dt = np.dtype([('pointer', np.int16),('table',"<S22")])
    dtTable2 = np.dtype([('table_pointer', dt, (2))])
    table2 = np.fromiter(pointerTable.values(), dtype = dtTable2)
    pTable = np.array([table1, table2 ], dtype = [('keys', dtTable1 ), ('vals', dtTable2)])

It's supposed to create a array containing the two like:

array(array(Table1),array(Table2)

The arrays are both the same length