gocept.webtoken 4.0

The CryptographicKey utility

gocept.webtoken uses a global utility of the class gocept.webtoken.CryptographicKeys, which provides cryptographic keys for different purposes. It loads a set of public and private keys from disk. It takes the filesystem path to your key files and a list of key names:

>>> import gocept.webtoken
>>> import pkg_resources
>>> path_to_keys = pkg_resources.resource_filename(
...     'gocept.webtoken', 'testing/keys')
>>> keys = gocept.webtoken.CryptographicKeys(
...     path_to_keys, ['key1'])

For each of the names, a private key file of the same name and a public key file (with a .pub suffix) must reside inside the keys_dir.

The utility needs to be registered at the ZCA, either via a zcml file or via:

>>> import zope.component
>>> zope.component.provideUtility(keys)

Creating a token

Create a signed web token with the function create_web_token. You will need the private key name, which was registered at the CryptographycKey utility. It is referenced by its name and the suffix -private:

>>> expires_in = 300  # The token is valid for 300 seconds
>>> payload = {'your': 'data'}
>>> result = gocept.webtoken.create_web_token(
...     'key1-private', 'issuer', 'subject', expires_in, payload)
>>> sorted(result.keys())
['data', 'token']

The token is available under the key token, while the data encoded in the token is placed under the key data.

Creating a Bearer Authorization header

You can create an Bearer Authorization header either from a token_dict as returned by create_web_token or from a token directly:

>>> gocept.webtoken.create_authorization_header(b'<TOKEN>')
('Authorization', 'Bearer <TOKEN>')

Extracting a token from a Bearer Authorization header

Extract the token from a dict containing the headers of you request or from the value of the HTTP Authorization header itself:

>>> request_headers = dict(Authorization='Bearer <TOKEN>')
>>> b'<TOKEN>' == gocept.webtoken.extract_token(request_headers)
True

Decoding a token

Decode a signed web token with the function decode_web_token. You will need the public key name, which was registered at the CryptographycKey utility. It is referenced by its name and the suffix -public:

>>> result = gocept.webtoken.decode_web_token(
...     result['token'], 'key1-public', 'subject')

Note that the subject must match the subject given when the token was created.

The result contains all data encoded in the token. You can find the payload under the key data:

>>> {'your': 'data'} == result['data']
True

4.0 (2023-08-21)

• Drop support for Python 2.7, 3.5, 3.6.

• Add support for Python 3.9, 3.10, 3.11.

3.1.post1 (2020-04-08)

• Drop faulty python_requires in setup.py.

3.1 (2020-04-08)

• Migrate to Github.

• Test with Python 3.8 and PyPy3.

• Stop testing with Python 3.5.

3.0 (2018-11-14)

• Change license from ZPL to MIT.

• Add support for Python 3.7.

• Drop support for Python 3.4.

• Make subject check optional as some systems like Keycloak use a random uuid as the subject which is unknown for the decoder.

• Add audience parameter which is required to decode tokens generated by Keycloak.

2.0 (2018-01-08)

• Drop support for Python 3.3 but add it for 3.6.

• Make setup.py compatible with newer setuptools versions by no longer using absolute paths.

1.2.1 (2015-10-08)

• Fix extract_token to accept any collections.Mapping derived object.

1.2 (2015-10-08)

• Added helper functions to create a Bearer Authorization header and extract a token from it.

• Officially support Python 3.5.

1.1 (2015-10-01)

• Shortened imports for CryptographicKeys, create_web_token and decode_web_token, which are now importable directly from gocept.webtoken.

• Added documentation.

1.0 (2015-10-01)

• Add support for Python 3.3 and 3.4.

• Initial release, extracted from internally used package.