3 ms·

Less than a naive dictionary; it's a very compact and natural representation of ordered lists of unique words.

• pmjordan 16y ago
  I realise that, I'm interested in the per-object overhead. This approaches zero for the single giant string.
  • rudiger 16y ago
    Take a look at Saurabh Odhyan's JavaScript prefix trie implementation; https://github.com/odhyan/trie https://github.com/odhyan/trie