efrt

efrt

a prefix/suffix trie-based data-structure optimised for compression of english words

based on mckoss/lookups by Mike Koss and bits.js by Steve Hanov it can compress a wordlist/dictionary into a very compact form, so that filesize/http/bandwidth is low.

clients though can query from the compressed form ultra-quick, with performance that's comparable to a straight-up javascript obj.

By doing the fancy stuff ahead-of-time, efrt lets you ship much bigger word-lists to the client-side, while ensuring there's no big unpacking step - so that users are always on the critical path.

var efrt = require('efrt')//
var words = [
  'coolage', //must boring, lowercase, non-unicode
  'cool',
  'cool cat',
  'cool.com',
  'coolamungo'
];
//pack these words as tightly as possible
var compressed = efrt.pack(words);
//cool0;! cat,.com,a0;ge,mungo

//pull it apart into a lookup-trie
var trie = efrt.unpack(compressed);

//hit it!
console.log(trie.has('cool'));//true
console.log(trie.has('miles davis'));//false

Demo

if you're doing the second step in the client, you can load only the unpack-half of the library(~3k):

<script src="./builds/efrt-unpack.min.js"></script>
<script>
  $.get('./compressedStuff.txt', (str)=>{
    var trie=unpack(str);
    trie.has('miles davis');
  })
</script>

the words you input should be pretty normalized. Spaces and unicode are good, but numbers, case-sensitivity, and some punctuation are not (yet) supported.