Discovery

Discovery is the second service in the .proto, alongside AllocationBooking. It is read-only: the consumer browses — within their permissions — which Repositories exist and the availability inside them.

Grouper / Group trees

Segmentation is a tree (parent / is_parent + a metadata map), at two levels:

• Repository level — segments events (competition, stadium, date, round…).
• Bucket level — segments availability inside an event (sector, type…). The leaf bucket-group is the Bucket itself.

A symmetric trio at each level: Grouper → Group → Repository ‖ Grouper → Group → Bucket.

The RPCs

RPC	Returns
ListRepositoryGroupers	grouping dimensions for events (competition, stadium…).
ListRepositoryGroups	values within a grouper (e.g. grouper Competition → Brasileirão).
ListRepositories	the event summaries themselves.
ListBucketGroupers	grouping dimensions for availability within a Repository.
ListBuckets	the leaf buckets + their availability count (the browse-side “Check-per-Bucket”).
ListUnits	the individual seats of one Bucket — for reserved seating / a seat map.

ListUnits is open (read), scoped to one Bucket at a time (bounds the payload). A QUANTITATIVE Bucket has no addressable Unit, so it is browsed by count only.

Identity comes from the session

The consumer’s identity (Subject / Channel), used to filter by permission, comes from the session — not from these messages. The grouping dimensions are populated at ingestion (the product). Discovery stays a clean, open read surface.