Usage Patterns

Below are suggested patterns for using Betamax efficiently.

Configuring Betamax in py.test’s

Betamax and (the project which instigated the creation of Betamax) both utilize py.test and its feature of configuring how the tests run with [1]. One pattern that I have found useful is to include this in your file:

import betamax

with betamax.Betamax.configure() as config:
    config.cassette_library_dir = 'tests/cassettes/'

This configures your cassette directory for all of your tests. If you do not check your cassettes into your version control system, then you can also add:

import os

if not os.path.exists('tests/cassettes'):

An Example from

You can configure other aspects of Betamax via the file. For example, in, I do the following:

import os

record_mode = 'none' if os.environ.get('TRAVIS_GH3') else 'once'

with betamax.Betamax.configure() as config:
    config.cassette_library_dir = 'tests/cassettes/'
    config.default_cassette_options['record_mode'] = record_mode
        os.environ.get('GH_AUTH', 'x' * 20)

In essence, if the tests are being run on Travis CI, then we want to make sure to not try to record new cassettes or interactions. We also, want to ensure we’re authenticated when possible but that we do not leave our placeholder in the cassettes when they’re replayed.

Using Human Readble JSON Cassettes

Using the PrettyJSONSerializer provided by the betamax_serializers package provides human readable JSON cassettes. Cassettes output in this way make it easy to compare modifications to cassettes to ensure only expected changes are introduced.

While you can use the serialize_with option when creating each individual cassette, it is simpler to provide this setting globally. The following example demonstrates how to configure Betamax to use the PrettyJSONSerializer for all newly created cassettes:

from betamax_serializers import pretty_json
# ...
config.default_cassette_options['serialize_with'] = 'prettyjson'

Updating Existing Betamax Cassettes to be Human Readable

If you already have a library of cassettes when applying the previous configuration update, then you will probably want to also update all your existing cassettes into the new human readable format. The following script will help you transform your existing cassettes:

import os
import glob
import json
import sys

    cassette_dir = sys.argv[1]
    cassettes = glob.glob(os.path.join(cassette_dir, '*.json'))
    print('Usage: {0} CASSETTE_DIRECTORY'.format(sys.argv[0]))

for cassette_path in cassettes:
    with open(cassette_path, 'r') as fp:
        data = json.load(fp)
    with open(cassette_path, 'w') as fp:
        json.dump(data, fp, sort_keys=True, indent=2,
                  separators=(',', ': '))
print('Updated {0} cassette{1}.'.format(
    len(cassettes), '' if len(cassettes) == 1 else 's'))

Copy and save the above script as and then run it like:


If you’re not already using a version control system (e.g., git, svn) then it is recommended you make a backup of your cassettes first in the event something goes wrong.