[BUG] Job status inconsistencies

[fmigneault]

When a submitted process is just picked up by celery, it should report running status. Sometimes it does so for a brief moment when calling GET /processes/{id}/job/{id}, but doing a subsequent calls, the process returns to accepted status. This is probably due to the status mapper trying to be compliant to some status names that do not include running, which falls back to accepted.

Once the process has been correctly picked up and set to running, it must remain as such until it transits to failed/succeeded/dismissed accordingly.

As per OGC reference, the succeeded status is named successful. This should be adjusted to respect the standard. Internally we should still verify all variations to make sure.
https://raw.githubusercontent.com/opengeospatial/wps-rest-binding/master/core/openapi/schemas/statusInfo.yaml

The same above reference also shows that OGC now considers dismissed as a valid status. It should be added to the OGC status mapping, and mapping function should be updated accordingly.

weaver/weaver/status.py

Line 45 in c7fd7fd

STATUS_COMPLIANT_OGC: frozenset([STATUS_ACCEPTED, STATUS_RUNNING, STATUS_SUCCEEDED, STATUS_FAILED]), # noqa: E241, E501

Must had a test that validates this transition of status more explicitly.
Similar procedure as for following test should be used, but calling get-job-status at fast intervals (no delays) to attempt catching brief/sporadic status transitions.

weaver/tests/functional/test_builtin.py

Line 77 in f555d3e

def test_jsonarray2netcdf_execute(self):