I am looking for a map-like data structure with the following properties:
- it uses subsets of some set S as keys. The size of S is potentially unbounded, but does not change during the runtime
- the values associated with the keys are generic objects - poiters/references. Their properties are not important for the data structure
- the size of the subsets, used as keys, can vary between keys, i.e. one key can be a 3 element subset, while another might be a 7 element subset
- it must support insertion of new entries (key - value pairs) into the "map"
- it does not need to support removal of entries from the "map"
- if I have some subset B of S and want to get a value from the map using B, I want the following:
- find the largest key (the key with the larges cardinality) in the map that is a subset of B, and return it along with the value associated with it
- if B itself is a key in the map, I want to return the entry for B
- if there are multiple keys that satisfy this condition, I don't care which one gets returned
- if there are multiple keys that satisfy this condition, I don't care if a different one is returned for consecutive read operations
- if neccessary, I can guarantee that the entries with the atoms (one element subsets of S) as keys are always present in the map
My understanding is, that the keys form a partially ordered set based on set inclusion, and the read opperation is to find some sort of an infimum.
I looked at the disjoint-set (union-find) data strcutrue, but as far as I understand, this is not what I am looking for.
I have also looked at other questions around here with similar keywords, but could not find anything that would fit my requirements:
Any help is greatly appriciated. Doubly so, if I would be pointed in the dirrection of a javascript/typescript implementation of the data structure.
