velvet citrus
Hey folks, trying to be specific as I can about my issue, and have not been able to find resources online specifically about content=, as opposed to json=, text=, etc
My question relates to the requests-mock library, and why content= works differently depending on if you simply call get(...) vs register_uri(...). I want to use register_uri(...) to (eventually) use the feature response lists allowing the ability to provide multiple data sets per api call. For now, I cannot get even this more simple setup to work, seen below:
For example, this is how to do a mock for a request:
json_blob = json.dumps([{'foo': 'bar'},{'fizz': 'buzz'}]).encode()
# m is a Mocker object
m.get(
"https://my-api-endpoint.com/path/to/endpoint",
content=json_blob # this is a JSON encoded Byte String
)
^ this code works just fine
However, if you want to register the URI so you dont have to pass in content:
m.register_uri(
"GET",
"https://my-api-endpoint.com/path/to/endpoint",
content=json_blob
)
m.get("https://my-api-endpoint.com/path/to/endpoint")
you actually get an error when it gets called:
response.json()
^^^^^^^^^^^^^^^
FAILED tests/test_my_api_endpoint.py::test_endpoint - requests.exceptions.JSONDecodeError: Expecting value: line 1 column 1 (char 0)
I have printed out the response object I got from the mock, and it shows that the content on the response object = b'', in other words, it's empty, whereas the simple get(URL, content=json_blob) version provides an actual object. Can anyone help me figure this one out? Thank you!
vale mason
