Examples of IdService

• com.urbanairship.datacube.IdService
  A service for translating byte[]s consistently into unique IDs. If the given byte[] has been seen before, the same identifier will be returned as the previous call. If this is the first time byte[] has been seen, a new identifier will be assigned. Different identifier lengths can be used to save space. For example, if you have a dimension that stores a boolean, it can only have two distinct values, so you only need one byte to encode datapoints in that dimension. On the other hand, to store the name of a city, you would need more bytes to encode it since there are many cities. Implementations can assume that numIdBytes will never change over the life of a cube for a given dimension. Implementations are required to be thread safe. Also, implementations must assume that other JVMs will be sharing the underlying database, so some kind of external concurrency control is required when assigning IDs. It's not OK for the same input to be translated differently; if a particular input translates into a particular unique ID once, it must translate in the same way everywhere else forever. Also, a unqiue ID must never be reused for different inputs. This means you must use locks, transactions, or compare-and-swap on your database when assigning IDs.

Examples of com.urbanairship.datacube.IdService