[RayJob]Refactor the sidecar container failure detection logic and extract common functions

[zhuyanhuazhuyanhua]

Why are these changes needed?

In SidecarMode, the ray-job-submitter runs as a container inside the head Pod with the head Pod's RestartPolicy set to Never. If the submitter exits non-zero before the RayCluster reaches Ready (e.g. the job fails while workers are still being scheduled), the terminated submitter container keeps the head Pod Ready=False, so RayCluster.Status.State never becomes Ready.

The RayJob Initializing -> Running transition is gated on RayCluster.Status.State == Ready, and the only submitter-failure detector (checkSubmitterAndUpdateStatusIfNeeded) runs only in the Running state. As a result the RayJob is wedged in Initializing forever: it never transitions to Failed, shutdownAfterJobFinishes never runs, and the autoscaler keeps retrying workers — even though the submitter failure is already observable.

This PR detects a terminal submitter failure during Initializing (SidecarMode) and fails fast. The feature-gate-aware detection (SidecarSubmitterRestart) is extracted into checkSidecarSubmitterFailedAndUpdateStatus and shared by the Initializing and Running states; the Initializing call skips when the head Pod does not exist yet.

Related issue number

Closes #4945 #4946

Checks

• I've made sure the tests are passing.

Testing:

• go build ./controllers/ray/..., go vet ./controllers/ray/, gofumpt -l — clean
• New unit test TestCheckSidecarSubmitterFailedAndUpdateStatus: submitter Terminated exit != 0 → Failed/SubmissionFailed; same with JobStatus already Failed → AppFailed; exit 0 → no-op; still Running → no-op; SidecarSubmitterRestart enabled → routes to the restart-count check
• go test ./controllers/ray/ -run 'TestCheckSidecarSubmitterFailedAndUpdateStatus|TestCheckIsRestartCountExceeded' — pass