better acceptance test metadata feedback

[mikeAdamss]

What is this

We have an acceptance test that compares the metadata we're generating with a fixture of expected metadata.

It works well, but the feedback on a failure just prints the contens of a json file, which in some cases will be quite big, making it hard to drill down to where the failure is.

This task is to see if we can get more specific feedback from it.

What to do

so this line:

dp-data-pipelines/features/data_ingress_v1.feature

Line 21 in c134415

And the metadata should match 'fixtures/correct_metadata.json'

calls this function:

dp-data-pipelines/features/steps/data.py

Line 111 in c134415

def step_impl(context, correct_metadata):

We need more specific feedback than just printing the dict.

Take some time to investigate the options. One (possible) option I stumbeld accross is https://github.com/inveniosoftware/dictdiffer, but there might be others.

Note - use make feature to run the. acceptance tests.

Acceptance Criteria

• Comparing two dictionaries with the acceptance tests results in some sort of feedback telling you what is different.