It would be interesting if there would support for Judy based atom string tables.
At the moment it's possible to create a JudySL or JudyHS array that maps a string or array-of-bytes to a word.
However, given that "word" as key does not allow me to recover the original string or array-of-bytes without creating another JudyL array that has to contain another copy of the string or array-of-bytes.
It would be great if Judy would have builtin support for atom tables, which allows a string -> value lookup where the value by itself is a key to lookup the string again (without having to store the string twice).
Atom tables are used very frequently to minimize storage and they require fast bidirectional lookups. It sounds like an ideal candidate to implement this using Judy arrays, but as far as I can see, the current implementation doesn't allow me to create such an atom table efficiently. Or am I missing something?