soundex.py

def soundex(name):
    code = {
        'b': '1',
        'f': '1',
        'p': '1',
        'v': '1',
        'c': '2',
        'g': '2',
        'j': '2',
        'k': '2',
        'q': '2',
        's': '2',
        'x': '2',
        'z': '2',
        'd': '3',
        't': '3',
        'l': '4',
        'm': '5',
        'n': '5',
        'r': '6'
    }
    vowels = ('a', 'e', 'i', 'o', 'u')

    # Generate a basic soundex encoding
    raw_sndx = [code[c] if c in code else -1 if c in vowels else -2 for c in name.lower()]

    # Prune the encoding
    pruned_sndx = []
    for i in range(2, len(raw_sndx)):
        if (raw_sndx[i] != raw_sndx[i-1] and
                not (i < len(raw_sndx) - 2 and raw_sndx[i+1] == -1 and raw_sndx[i] == raw_sndx[i+2])
                and int(raw_sndx[i]) > 0):
            pruned_sndx.append(raw_sndx[i])

    # Format the encoding
    sndx = [name[0], '-'] + (pruned_sndx[1:] if int(pruned_sndx[0]) < 0 else pruned_sndx)

    # Cut or pad the encoding to the right length
    while len(sndx) > 5:
        del sndx[-1]
    while len(sndx) < 5:
        sndx.append('0')

    # Return formatted soundex
    return ''.join(sndx)

if __name__ == '__main__':
    print(soundex('Gutierrez'))
    print(soundex('Pfister'))
    print(soundex('Jackson'))
    print(soundex('Tymczak'))
    print(soundex('VanDeusen'))
    print(soundex('Ashcraft'))