Decoding bwt is too slow

[Wittline]

I was doing some tests with your bwt decoding, and this method is too slow, this decoding approach is not recommended for large alphabets.

def reconstruct_bwm(bwt: str) -> List[str]:
        bwm = []
        # first loop to create seeds for lines O(n)
        for _ in range(0, len(bwt), 1):
            bwm.append('')

        for _ in range(0, len(bwt), 1):

            for i in range(0, len(bwt), 1):
                bwm[i] = bwt[i] + bwm[i]
 
            yield [line for line in bwm]
            bwm.sort()
            yield [line for line in bwm]