I want to upload a single file to Workiva and associate it with an existing table using this endpoint in Python.

My code:

import requests


headers = {
    'Authorization': f'Bearer {access_token}',  # I've got an access_token from the Workiva authentication API
    # 'Content-Type': 'multipart/form-data'  # If I set this, the API will respond '{"body":"No multipart boundary parameter in Content-Type","code":400,"details":null}'
}

file_name = 'test.csv'
table_id = '643ff5b61d0742769d25fe579b53521a'

url = 'https://h.eu.wdesk.com/s/wdata/prep/api/v1/file'
params = {
    'tableId': table_id
}

with open(file_name, 'rb') as file_to_upload:
    # print(file_to_upload.readline()) -> this works fine
    files = {
        'file': (file_name, file_to_upload)
    }
    r = requests.post(url, params=params, headers=headers, files=files)

    print(r.status_code)  # -> prints 400
    print(r.request.headers)  # shows that the Content-Type is set correctly, including the boundary
    print(r.request.body)  # shows the file content I'm sending
    print(r.content)

The API responds with '{"body":["contentType must not be null"],"code":400,"details":null}'. However, this is misleading, because in the header I can clearly see that the Content-Type is set correctly, including the boundary.

If I set the Content-Type manually, the API complains that no boundary is set. If I set the boundary manually, I get an EOF reached error, i.e. I can't set the boundary manually (the boundary must be calculated, not set manually).

I can successfully upload my csv-file via Postman, so the csv-file should be fine, as well as my understanding of how to use the API.

Am I doing something wrong or is this a bug in the Workiva API code?