Migrating jwk_convert.py to Python 3

[szepeviktor]

@piraz Could you help @jsabater to convert his key?

https://github.com/szepeviktor/debian-server-tools/blob/master/security/jwk_convert.py

Hi, Viktor!

Just tried to use your tool to convert a private_key.json from the latest version of Certbot to PEM format using Python 3.9 under Debian 11 Bullseye and found this error:

Traceback (most recent call last):
  File "/home/ansible/bin/jwk_convert.py", line 31, in <module>
    PKEY[k] = enc(v.encode())
  File "/home/ansible/bin/jwk_convert.py", line 26, in enc
    return '0x'+binascii.hexlify(base64.b64decode(data, b'-_')).upper()
TypeError: can only concatenate str (not "bytes") to str

I had converted the script to Python 3 using 2to3 but handling strings of bytes seems to need specific knowledge, which I don't have. I think the problem is here:

data += b'=' * missing_padding

I've tried to research this topic without success. Would you feel like upgrading the script to Python 3? I think that the need to convert JWK to PEM is going to be there for a while for all us Ansible users.

Incidentally, I'd love to create an Ansible Playbook to do the whole process, maybe through a module.

Thanks in advance and happy new year.

[piraz]

I think that was fixed here #109. Need to release it.

I'll do it today.

[piraz]

@szepeviktor now I understand what is going on. Let me see if I can help.

By the way, 0.9.9 is working right now. Automatoes will work fine with python 3.11.

[piraz]

The error is on the data's format and at the concatenation from the return. If data is string, than you will have an error:

def enc(data):
    missing_padding = 4 - len(data) % 4
    if missing_padding:
        data += b'=' * missing_padding #  <==== right here, because you are concatenating strings with bytes

    return '0x'+binascii.hexlify(base64.b64decode(data, b'-_')).upper()  # <=== and here, becuase '0x' is string and binascii.hexlify will return bytes

You can check if data is string and convert it to bytes and convert the '0x' to b'0x', which is the byte form.

Here is your function working:

def enc(data):
    if isinstance(data, str):
        data = data.encode()
    missing_padding = 4 - len(data) % 4
    if missing_padding:
        data += b"=" * missing_padding
    return b"0x" + binascii.hexlify(base64.b64decode(data, b"-_")).upper()

I like double quotes for strings, you can change it if you want. ;)