This project addresses the issues associated with performance compromise as the dataset grows in client side search operations using some inbuilt libraries and some search alogrithms.
This project offers a unique solution using vanilla javascript and Web Worker for parallel processing.
You know often time we face a problem that we want to build a search functionality for large dataset. But it is little difficult to build, even using the libraries like fuse.js.
We use various search algorithms, different hacky solutions but initially all performs well untill your data grows to a large extent. Then the performance of these solutions starts to become slow.
Now, in this case you are reluctant to not go for server side and build your search functionality from scratch. Well there could be many reasons to that. Such as, it can make UX bad. Also, building this from scratch in server side is hell lot of task. You might introduce some additional complexties to it.
Instead you seek the solution which can be implemented in a client side. It should be efficient enough to not block the main thread of javascript and at the same time maintain the good UX.
But, we can achieve this using vanilla javascript and using Web Worker in client side only. This project offers the search functionality for large dataset using web worker and dividing the data into different chunks based on the logical cores of a system. This approach allows for faster search performance without blocking the main thread and without compromising search capabilities.
Installation of this project is very easy.
-
Fork the repo.
-
Clone the project
https://github.com/<YOUR_USERNAME>/warp-search.git
- Run npm install.
npm i
- Run the project.
npm start