Going to calculate the sparse field before calling the schema constructor rather than calling _init_fields again because they've labelled it explicitly as a private API.
Also, now if both and include and field[primary_type] parameters are passed, the included relationship must be present in the fields parameter because relationships are fields
Fix for https://github.com/miLibris/flask-rest-jsonapi/issues/209
Going to calculate the sparse field before calling the schema constructor rather than calling
_init_fieldsagain because they've labelled it explicitly as a private API.Also, now if both and
includeandfield[primary_type]parameters are passed, the included relationship must be present in the fields parameter because relationships are fields