pytest-vcrpandas
Combine vcrpy with pandas to not only mock your HTTP interactions, but also tests your post-processing work to get a pandas DataFrame.
Usage
# test_meteo_client.py
def test_meteo_get_wind_speed(vcrpandas, client):
start = pd.Timestamp('2018-01-01', tz='CET')
end = pd.Timestamp('2018-02-01', tz='CET')
with vcrpandas("meteo_get_wind_speed") as recorder:
df = client.get_wind_speed(start, end)
recorder(df)# call pytest for the first time to generate samples
> pytest --vcr-record="new_episodes"
# you can now replay it:
> pytestBehind the hood, 2 files are generated by the first run:
- a meteo_get_wind_speed.yaml, which is the reponse obtained by the first HTTP interaction.
- a meteo_get_wind_speed.pickle that serve to compare the formatting from the mocked response to the pandas DataFrame, types included.
Installation
pip install pytest-vcrpandasAdvanced Usage
To configure VCR, here is how to procee:
@pytest.fixture
def my_vcrpandas(vcrpandas):
custom_vcr_config = dict(
serializer='json',
cassette_library_dir='fixtures/myclient/cassettes',
match_on=['uri', 'method'],
filter_headers=['authorization'], # filter information from HTTP Headers
filter_query_parameters=['api_key'], # filter information from HTTP query string
filter_post_data_parameters=['client_secret'], # filter information from HTTP post data
)
vcrpandas.config.update(custom_vcr_config)
return vcrpandas
def test_mysupertest(my_vcrpandas):
with my_vcrpandas("random_bucket_name") as recorder:
recorder(CL.get_stuff())Refere to `VCR configuration <https://vcrpy.readthedocs.io/en/latest/configuration.html`_ and VCR Advanced Features to get the full list of options.
