Fast Vue SSR with Rust and QuickJS

栏目: IT技术 · 发布时间: 5年前

内容简介:An ongoing experiment usingSo far using a thread pool and channels to communicate with the Warp route handler. The goal is to get a full Node-likeNode outperforms QuickJS by a wide margin. Especially with enough cores and memory. However, QuickJS is very s

Fast Vue SSR with Rust and QuickJS

An ongoing experiment using Rust , Warp and QuickJS to server-side render Vue.js applications.

use renderer::RendererPool;
#[tokio::main]
pub async fn main() -> io::Result<()> {
    let pool = Arc::new(Mutex::new(RendererPool::new(64)));
    let renderer = warp::path::full().map(move |path: FullPath| {
        let renderer = Arc::clone(&pool);
        let s = path.as_str().to_string();
        // Currently only passing path to renderer is possible
        // Full Request object is a WIP
        let result = renderer.lock().unwrap().render(s);
        result
    });
    let routes = warp::path::full()
        .and(renderer)
        .map(|_, result| reply::html(result));

So far using a thread pool and channels to communicate with the Warp route handler. The goal is to get a full Node-like IncomingMessage object available as $ssrContext.req . It already includes a /static handler and serves a code-splitted build on the client via Rollup.

Node outperforms QuickJS by a wide margin. Especially with enough cores and memory. However, QuickJS is very small and has very low memory consumption, so running it threaded in a Rust shell makes it possible to have very high throughput using very few resources in comparison.

Inspired by Xinjiang Shao's experiment .

Running

  1. Install Rust .
  2. npm install
  3. npm test

Or npm run build for generating the Rust binary.


以上就是本文的全部内容,希望本文的内容对大家的学习或者工作能带来一定的帮助,也希望大家多多支持 码农网

查看所有标签

猜你喜欢:

本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们

深入理解计算机系统(原书第3版)

深入理解计算机系统(原书第3版)

Randal E.Bryant、David O'Hallaron / 龚奕利、贺莲 / 机械工业出版社 / 2016-11 / 139.00元

和第2版相比,本版内容上*大的变化是,从以IA32和x86-64为基础转变为完全以x86-64为基础。主要更新如下: 基于x86-64,大量地重写代码,首次介绍对处理浮点数据的程序的机器级支持。 处理器体系结构修改为支持64位字和操作的设计。 引入更多的功能单元和更复杂的控制逻辑,使基于程序数据流表示的程序性能模型预测更加可靠。 扩充关于用GOT和PLT创建与位置无关代码的......一起来看看 《深入理解计算机系统(原书第3版)》 这本书的介绍吧!

RGB转16进制工具
RGB转16进制工具

RGB HEX 互转工具

SHA 加密
SHA 加密

SHA 加密工具

UNIX 时间戳转换
UNIX 时间戳转换

UNIX 时间戳转换