NameKeySet implementation suitable for storing small set of NameKey
values (generally 8 or less). Uses linear search, and is thus the
most compact presentation for a set
Notes about usage:
- All Strings contained in
NameKey
instances are assumed
interned, so that equality comparison can be done (both for values
stored and keys used)
- It is assumed that sets are never empty, ie. always contain at
least one entry.
- It is assumed that caller has ensured that there are no duplicates
in the set -- this data structure does no further validation.