Storage keys are arrays because they are hierarchical and they allow the storage subsystem to do
range queries for all keys that have a particular prefix. For example, incremental changes for a
given document might be stored under [<documentId>, "incremental", <SHA256>].
automerge-repo mostly uses keys in the following form:
[documentId, "snapshot", hash] // OR
[documentId, "incremental", hash]
However, the storage adapter implementation should be agnostic to the meaning of the key and should not assume any particular structure.
A storage key is an array of strings that represents a path to a value in a StorageAdapter.