Add index_sbom_occurrence_refs task to seed ES index on setup

What does this MR do and why?

This MR adds an index_sbom_occurrence_refs task to Search::RakeTaskExecutorService, adds :sbom_occurrence_refs to Search::Elastic::TriggerIndexingWorker::TASKS so initial setup seeds it, and exposes the rake task gitlab:elastic:index_sbom_occurrence_refs. The task is gated on Search::Elastic::SbomOccurrenceRefIndexHelper.indexing_allowed? and logs a skip message when indexing is off.

It reuses Sbom::BulkEsOperationService rather than calling track! directly, since Sbom::OccurrenceRef#es_parent resolves project.namespace.root_ancestor and that service already owns the required preloading. That service gains an optional bookkeeping_service: keyword so initial setup enqueues onto Elastic::ProcessInitialBookkeepingService instead of the incremental queue; existing callers keep the default.

References

1. Relates to - #612844 (closed)
2. Reference MR - !197147 (merged)

How to set up and validate locally

1. Empty the index:

   curl -XPOST "http://localhost:9200/gitlab-development-sbom_occurrence_refs/_delete_by_query?refresh=true" -H "Content-Type: application/json" -d '{"query":{"match_all":{}}}'
   curl -s "http://localhost:9200/gitlab-development-sbom_occurrence_refs/_count"

2. In bundle exec rails console, run the task:

   Search::RakeTaskExecutorService.new(logger: Logger.new($stdout)).execute(:index_sbom_occurrence_refs)

   Expect Indexing SBOM occurrence refs... then ... done, and Elastic::ProcessInitialBookkeepingService.queue_size to equal Sbom::OccurrenceRef.count.

3. Drain the queue, repeating until it returns [0, 0]. The second number is the failure count and must stay 0:

   Elastic::ProcessInitialBookkeepingService.new.execute

4. Refresh and count. The document count should equal Sbom::OccurrenceRef.count:

   curl -XPOST "http://localhost:9200/gitlab-development-sbom_occurrence_refs/_refresh"
   curl -s "http://localhost:9200/gitlab-development-sbom_occurrence_refs/_count"

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.