shitty_encoder_decoder.py

'''text encoder'''

'''
encoding first histogram
then lopsided tree
then encode

'''

sentence = 'abcdabac'
sigma = ['a','b','c','d']
alpha_dict = {
    'a': [0],
    'b': [1,0],
    'c': [1,1,0],
    'd': [1,1,1]
}

reverse_alpha_dict = {tuple(v):k for k,v in alpha_dict.items()}

def encode(sentence, sigma, alpha_dict, encode=[]):
    for s in sentence:
        bits = alpha_dict[s]
        encode.extend(bits)
    return encode

encoded = encode(sentence, sigma, alpha_dict)
print(encoded)
# [0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0]

def decode(message, reverse_alpha_dict):
    ret_message = []
    piv = 0
    leng = 1
    while(True):
        pos = tuple(message[piv: piv+leng])
        ret = reverse_alpha_dict.get(pos, None)
        if not ret:
            leng+=1
        else:
            ret_message.append(ret)
            piv += leng
            leng = 1
        if piv+leng > len(message):
            break

    return ''.join(ret_message)

message = [0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0]
returned = decode(message, reverse_alpha_dict)
print(returned)