This is a small, barely working, incomplete implementation of a Directed Acyclic Word Graph, a.k.a. Minimal Acyclic Finite State Automaton, a data structure closely related to a Trie.
Its purpose is to efficiently store a large word dataset and allow for quick lookups, which is typically used in spell checkers, autocomplete engines or environments with limited memory.
The work here was largely inspired by Steve Hanov's article. The main difference is that this implementation should work with unsorted data, and allow value retrieval.
Use at your own risk. Patches welcome.
npm install udawg