I have some code where I read an object off a JMS topic (unfortunately I cannot share the actual code, so I will use a hypothetical example) - let's assume the object is Employee (firstName, lastName, postCode, email, etc), and I insert the employee object into a Map where the key is the email address for quick retrieval. However, my requirements are likely to change and I may also need to look up by firstName. So now I have to maintain two maps, one keyed by email and one keyed by firstName. Then if I get a new requirement, I have to maintain an additional map, and so on and so forth - nightmare code ensues.
I had a look at the Guava BiMap, which is great if you only have 2 fields and both of them are unique, but I was wondering if there is a Util out there (Apache Commons Collections, Guava Collections, any other) that will give me this out of the box - essentially, I want something like a List which I can very simply insert to and remove from, and it will handle several backing maps for me so that finding elements by a given field is roughly O(1). My current implementation is to use Java 8 parallelStream and filter, but this won't scale once my list grows (which it potentially could up to a few million entries).
The only other option I have come up with so far is to use an in-memory database such as HSQLDB, index every column, and use some simple SQL to retrieve my values. I believe this will work but has a fair amount of overhead maintaining a schema, writing queries etc.
TL;DR: Is there a collection type in a lilbrary that looks like List but that can give me O(1) retrieval based on any field in the objects?
