[Core] Fix type of job_id when querying job status

[cblmemo]

Previously, our common_utils.encode_payload used json.dumps to handle a dictionary with integer and None type as key, which will not be restored as the original type in default json.loads implementation; also, our click treat job_id as string, causing a lot of problems. This PR fixes these problems.

$ sky logs --status g 1
Getting job status...
{'1': None, 1: 'SUCCEEDED'}
<sky-payload>{"1": null, "1": "SUCCEEDED"}</sky-payload>

{'1': <JobStatus.SUCCEEDED: 'SUCCEEDED'>}
Job 1: SUCCEEDED

Tested (run the relevant ones):

• Code formatting: bash format.sh
• Any manual or new tests for this PR (please specify below)
  • sky logs --status cluster-with-no-job-submitted
  • sky logs --status cluster-with-job

>>> import sky
>>> sky.core.job_status('cluster-with-no-job-submitted', None)
Getting job status...
{None: None}
>>> sky.core.job_status('cluster-with-job', None)
Getting job status...
{1: <JobStatus.SUCCEEDED: 'SUCCEEDED'>}

• All smoke tests: pytest tests/test_smoke.py
• Relevant individual smoke tests: pytest tests/test_smoke.py::test_fill_in_the_name
• Backward compatibility tests: bash tests/backward_comaptibility_tests.sh

[Michaelvll]

Thanks for fixing the issue @cblmemo! Left several comments.

[Michaelvll]

Can we add some docstr for this function to list all the possible cases for the dict? For example:
{None: None} there is no job on the cluster
{1: RUNNING, 2: None} job 2 does not exist

[cblmemo]

It is already included in the docstring of core.py::job_status. I added a reference to it 🫡

[Michaelvll]

Thanks for the fix @cblmemo! LGTM.