A javascript standard data structure library which benchmark against C++ STL.
Note that our official version starts from 2.0.0. In order to avoid unnecessary trouble, please select the latest version when downloading (>= 2.0.0).
Unfortunately, I try my best to improve the performance of hash table but failed. Ideally, the hash table should be O(1) and in my project when the size of each bucket is 1, its time is much longer than Set and Map.
After analysis, I find the operator new
costs nearly 2s when the size is 1e6, This seems to be the bottleneck of js.
You can try the following code to verify my statement.
console.time("run");
const arr = new Array(1000000);
arr[100000] = new Array();
console.timeEnd("run");
You can imagine what happens when I repeat the above operation many times.
If you want to use hash container in js, please use set or map in ES6.
The official Set and Map are implemented using hash table instead of RBTree.
For more information about js hash table, please refer to https://stackoverflow.com/questions/368280/javascript-hashmap-equivalent.
- Vector
- Stack
- Queue
- LinkList
- Deque
- PriorityQueue
- Set (using RBTree)
- Map (using RBTree)
- HashSet (for reference only)
- HashMap (for reference only)
- node.js (using commonjs)
- react/vue (using es5)
- browser (support most browsers including IE8+)
Download directly
- js-sdsl.js (for development)
- js-sdsl.min.js (for production)
Or install js-sdsl using npm
npm install js-sdsl
To help you have a better use, we provide this API document.
<!-- you can download the file locally and import it or import it dynamically by using url. -->
<script src="https://zly201.github.io/js-sdsl/js-sdsl.min.js"></script>
<script>
const { Vector } = sdsl;
const myVector = new Vector();
// you code here...
</script>
Just like other packages.
If you want to get more help, viewing src/test.ts
may help.
You can pull this repository and run yarn build
to rebuild this library.
Before publishing, we conducted strict unit tests on each function, you can see testResult.txt
to find our test
results or run yarn test
to reproduce it.
This project is MIT licensed.