fn group_ranges_by_primary_end(
ranges: &[PartitionRange],
descending: bool,
) -> Vec<(Timestamp, usize, usize)>Expand description
Group consecutive ranges by their primary end value.
Returns a vector of (primary_end, start_idx_inclusive, end_idx_exclusive) tuples. Ranges with the same primary end MUST be processed together because they may overlap and contain values that belong to the same “top-k” result.