BIP39 校验和有时会失败

BIP39 Checksum fails sometimes

我有这段代码,将用于根据BIP39 Spec实现助记词生成器。 问题是大多数时候校验和不正确,但在某些情况下它有效,这取决于给定的熵。 (我使用 iancoleman.io bip39 来测试我的校验和)。

观察到以下情况:

使用了 128 位的熵。

Correct
Entropy:  10111101100010110111100011101111111110100010000101111110100101100000001100111111001100010010010011110110011010010010001011011000
Checksum: 1110

Incorrect
Entropy:  01011010000000110011001001001001001110100011100101010001001100111001111111000110000000011011110111011000011001010111001101111100
My checksum: 1010
Iancoleman checksum:1110

第一个成功了,第二个失败了。您可以在下面找到我的功能。

我错过了什么?

def fill_bits(binary, bits):
    if len(binary) < bits:
        return "0" * (bits - len(binary)) + binary
    return binary


# generate a given number of entropy bits
def generate_entropy(bits=256):
    if bits < 128 or bits > 256:
        raise EntropyRangeExceeded

    entropybits = bin(int.from_bytes(os.urandom(bits // 8), byteorder=sys.byteorder))[2:]
    return fill_bits(entropybits, bits)


# returns the sha256 hash of the given input
def sha256(_input):
    return hashlib.sha256(_input.encode("utf-8")).hexdigest()


# returns the checksum of the input hash
# checksum is given by the first (entropy length / 32)
# bits of the sha256 hash applied on entropy bits
def get_checksum(_entropy):
    entropy_length = len(_entropy) // 32
    return bin(int(sha256(_entropy), 16))[2:][:entropy_length]

sha256中哈希计算错误。可以不进行utf8编码。相反,熵必须表示为字节数组(参见 to_bytes)并且哈希必须由此生成:

import hashlib
def sha256(_entropy):
    entBytes = int(_entropy, 2).to_bytes(len(_entropy) // 8, byteorder='big')
    return hashlib.sha256(entBytes).hexdigest()

此外,哈希必须用前导 0 值填充到 256 位的长度(参见 zfill),以便在校验和中也考虑前导 0 值:

def get_checksum(_entropy):
    entropy_length = len(_entropy) // 32
    return bin(int(sha256(_entropy), 16))[2:].zfill(256)[:entropy_length];

示例 1,来自 here,第 4 步:

_entropy = '0011001010000101011111010000101111111111101000001001000001001010110100010101111001001011000100111100011110001001111011110111011010010100110011001110111001100010111011010010101101010011110100100110101111110001100101011001000110100010000110110001100101110001'
print(get_checksum(_entropy)) # 11110011

例子2,你的第二个例子:

_entropy = '01011010000000110011001001001001001110100011100101010001001100111001111111000110000000011011110111011000011001010111001101111100'
print(get_checksum(_entropy)) # 1110

示例 3,前导 0 值,与 here 的结果进行比较:

_entropy = '10111101100011110111100011101111111110100010000101111110100101100000001100111111001100010010010011110110011011010010001011011000'
print(get_checksum(_entropy)) # 0010