SPA应用性能优化（懒加载）

前提：

如今开发方式都是采用前后台分离的方式，前台采用的方式则是单页面应用开发简称SPA，这种开发模式最大的一个特点就是将有所代码打包成了一个文件，

这会导致了一个问题就是如果这个应用过大，打出来的这个文件也会越大，性能也就随之降低了，那么如何优化这个问题呢，今天就记录下懒加载策略吧。

那么什么是懒加载呢~

懒加载概念：

懒加载也就是按需加载，只要有需要才会加载的一种策略，常用于性能优化上（个人理解）通过懒加载可以避免加载不必要的资源。

工程级别：

在微服务架构里，前台也需要同步，这就需要每个模块都需要build出属于自己的js跟css，当跳转到目标模块时，我们不需要加载其他模块的文件，于是我们就

可以利用懒加载策略进行优化

利用第三方lazyload.js对目标模块进行文件读写

第三方lazyload文件

/*jslint browser: true, eqeqeq: true, bitwise: true, newcap: true, immed: true, regexp: false */

/**

LazyLoad makes it easy and painless to lazily load one or more external

JavaScript or CSS files on demand either during or after the rendering of a web

page.

Supported browsers include Firefox 2+, IE6+, Safari 3+ (including Mobile

Safari), Google Chrome, and Opera 9+. Other browsers may or may not work and

are not officially supported.

Visit https://github.com/rgrove/lazyload/ for more info.

Copyright (c) 2011 Ryan Grove <ryan@wonko.com>

All rights reserved.

Permission is hereby granted, free of charge, to any person obtaining a copy of

this software and associated documentation files (the 'Software'), to deal in

the Software without restriction, including without limitation the rights to

use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of

the Software, and to permit persons to whom the Software is furnished to do so,

subject to the following conditions:

The above copyright notice and this permission notice shall be included in all

copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR

IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS

FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR

COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER

IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN

CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

@module lazyload

@class LazyLoad

@static

*/

LazyLoad = (function (doc) {

    // -- Private Variables ------------------------------------------------------

    // User agent and feature test information.

    var env,

        // Reference to the <head> element (populated lazily).

        head,

        // Requests currently in progress, if any.

        pending = {},

        // Number of times we've polled to check whether a pending stylesheet has

        // finished loading. If this gets too high, we're probably stalled.

        pollCount = 0,

        // Queued requests.

        queue = { css: [], js: [] },

        // Reference to the browser's list of stylesheets.

        styleSheets = doc.styleSheets;

    // -- Private Methods --------------------------------------------------------

    /**

    Creates and returns an HTML element with the specified name and attributes.

    @method createNode

    @param {String} name element name

    @param {Object} attrs name/value mapping of element attributes

    @return {HTMLElement}

    @private

    */

    function createNode(name, attrs) {

        var node = doc.createElement(name), attr;

        for (attr in attrs) {

            if (attrs.hasOwnProperty(attr)) {

                node.setAttribute(attr, attrs[attr]);

            }

        }

        return node;

    }

    /**

    Called when the current pending resource of the specified type has finished

    loading. Executes the associated callback (if any) and loads the next

    resource in the queue.

    @method finish

    @param {String} type resource type ('css' or 'js')

    @private

    */

    function finish(type) {

        var p = pending[type],

            callback,

            urls;

        if (p) {

            callback = p.callback;

            urls = p.urls;

            urls.shift();

            pollCount = 0;

            // If this is the last of the pending URLs, execute the callback and

            // start the next request in the queue (if any).

            if (!urls.length) {

                callback && callback.call(p.context, p.obj);

                pending[type] = null;

                queue[type].length && load(type);

            }

        }

    }

    /**

    Populates the <code>env</code> variable with user agent and feature test

    information.

    @method getEnv

    @private

    */

    function getEnv() {

        var ua = navigator.userAgent;

        env = {

            // True if this browser supports disabling async mode on dynamically

            // created script nodes. See

            // http://wiki.whatwg.org/wiki/Dynamic_Script_Execution_Order

            async: doc.createElement('script').async === true

        };

        (env.webkit = /AppleWebKit\//.test(ua))

            || (env.ie = /MSIE|Trident/.test(ua))

            || (env.opera = /Opera/.test(ua))

            || (env.gecko = /Gecko\//.test(ua))

            || (env.unknown = true);

    }

    /**

    Loads the specified resources, or the next resource of the specified type

    in the queue if no resources are specified. If a resource of the specified

    type is already being loaded, the new request will be queued until the

    first request has been finished.

    When an array of resource URLs is specified, those URLs will be loaded in

    parallel if it is possible to do so while preserving execution order. All

    browsers support parallel loading of CSS, but only Firefox and Opera

    support parallel loading of scripts. In other browsers, scripts will be

    queued and loaded one at a time to ensure correct execution order.

    @method load

    @param {String} type resource type ('css' or 'js')

    @param {String|Array} urls (optional) URL or array of URLs to load

    @param {Function} callback (optional) callback function to execute when the

      resource is loaded

    @param {Object} obj (optional) object to pass to the callback function

    @param {Object} context (optional) if provided, the callback function will

      be executed in this object's context

    @private

    */

    function load(type, urls, callback, obj, context) {

        var _finish = function () { finish(type); },

            isCSS = type === 'css',

            nodes = [],

            i, len, node, p, pendingUrls, url;

        env || getEnv();

        if (urls) {

            // If urls is a string, wrap it in an array. Otherwise assume it's an

            // array and create a copy of it so modifications won't be made to the

            // original.

            urls = typeof urls === 'string' ? [urls] : urls.concat();

            // Create a request object for each URL. If multiple URLs are specified,

            // the callback will only be executed after all URLs have been loaded.

            //

            // Sadly, Firefox and Opera are the only browsers capable of loading

            // scripts in parallel while preserving execution order. In all other

            // browsers, scripts must be loaded sequentially.

            //

            // All browsers respect CSS specificity based on the order of the link

            // elements in the DOM, regardless of the order in which the stylesheets

            // are actually downloaded.

            if (isCSS || env.async || env.gecko || env.opera) {

                // Load in parallel.

                queue[type].push({

                    urls: urls,

                    callback: callback,

                    obj: obj,

                    context: context

                });

            } else {

                // Load sequentially.

                for (i = 0, len = urls.length; i < len; ++i) {

                    queue[type].push({

                        urls: [urls[i]],

                        callback: i === len - 1 ? callback : null, // callback is only added to the last URL

                        obj: obj,

                        context: context

                    });

                }

            }

        }

        // If a previous load request of this type is currently in progress, we'll

        // wait our turn. Otherwise, grab the next item in the queue.

        if (pending[type] || !(p = pending[type] = queue[type].shift())) {

            return;

        }

        head || (head = doc.head || doc.getElementsByTagName('head')[0]);

        pendingUrls = p.urls.concat();

        for (i = 0, len = pendingUrls.length; i < len; ++i) {

            url = pendingUrls[i];

            if (isCSS) {

                node = env.gecko ? createNode('style') : createNode('link', {

                    href: url,

                    rel: 'stylesheet'

                });

            } else {

                node = createNode('script', { src: url });

                node.async = false;

            }

            node.className = 'lazyload';

            node.setAttribute('charset', 'utf-8');

            if (env.ie && !isCSS && 'onreadystatechange' in node && !('draggable' in node)) {

                node.onreadystatechange = function () {

                    if (/loaded|complete/.test(node.readyState)) {

                        node.onreadystatechange = null;

                        _finish();

                    }

                };

            } else if (isCSS && (env.gecko || env.webkit)) {

                // Gecko and WebKit don't support the onload event on link nodes.

                if (env.webkit) {

                    // In WebKit, we can poll for changes to document.styleSheets to

                    // figure out when stylesheets have loaded.

                    p.urls[i] = node.href; // resolve relative URLs (or polling won't work)

                    pollWebKit();

                } else {

                    // In Gecko, we can import the requested URL into a <style> node and

                    // poll for the existence of node.sheet.cssRules. Props to Zach

                    // Leatherman for calling my attention to this technique.

                    node.innerHTML = '@import "' + url + '";';

                    pollGecko(node);

                }

            } else {

                node.onload = node.onerror = _finish;

            }

            nodes.push(node);

        }

        for (i = 0, len = nodes.length; i < len; ++i) {

            head.appendChild(nodes[i]);

        }

    }

    /**

    Begins polling to determine when the specified stylesheet has finished loading

    in Gecko. Polling stops when all pending stylesheets have loaded or after 10

    seconds (to prevent stalls).

    Thanks to Zach Leatherman for calling my attention to the @import-based

    cross-domain technique used here, and to Oleg Slobodskoi for an earlier

    same-domain implementation. See Zach's blog for more details:

    http://www.zachleat.com/web/2010/07/29/load-css-dynamically/

    @method pollGecko

    @param {HTMLElement} node Style node to poll.

    @private

    */

    function pollGecko(node) {

        var hasRules;

        try {

            // We don't really need to store this value or ever refer to it again, but

            // if we don't store it, Closure Compiler assumes the code is useless and

            // removes it.

            hasRules = !!node.sheet.cssRules;

        } catch (ex) {

            // An exception means the stylesheet is still loading.

            pollCount += 1;

            if (pollCount < 200) {

                setTimeout(function () { pollGecko(node); }, 50);

            } else {

                // We've been polling for 10 seconds and nothing's happened. Stop

                // polling and finish the pending requests to avoid blocking further

                // requests.

                hasRules && finish('css');

            }

            return;

        }

        // If we get here, the stylesheet has loaded.

        finish('css');

    }

    /**

    Begins polling to determine when pending stylesheets have finished loading

    in WebKit. Polling stops when all pending stylesheets have loaded or after 10

    seconds (to prevent stalls).

    @method pollWebKit

    @private

    */

    function pollWebKit() {

        var css = pending.css, i;

        if (css) {

            i = styleSheets.length;

            // Look for a stylesheet matching the pending URL.

            while (--i >= 0) {

                if (styleSheets[i].href === css.urls[0]) {

                    finish('css');

                    break;

                }

            }

            pollCount += 1;

            if (css) {

                if (pollCount < 200) {

                    setTimeout(pollWebKit, 50);

                } else {

                    // We've been polling for 10 seconds and nothing's happened, which may

                    // indicate that the stylesheet has been removed from the document

                    // before it had a chance to load. Stop polling and finish the pending

                    // request to prevent blocking further requests.

                    finish('css');

                }

            }

        }

    }

    return {

        /**

        Requests the specified CSS URL or URLs and executes the specified

        callback (if any) when they have finished loading. If an array of URLs is

        specified, the stylesheets will be loaded in parallel and the callback

        will be executed after all stylesheets have finished loading.

        @method css

        @param {String|Array} urls CSS URL or array of CSS URLs to load

        @param {Function} callback (optional) callback function to execute when

          the specified stylesheets are loaded

        @param {Object} obj (optional) object to pass to the callback function

        @param {Object} context (optional) if provided, the callback function

          will be executed in this object's context

        @static

        */

        css: function (urls, callback, obj, context) {

            load('css', urls, callback, obj, context);

        },

        /**

        Requests the specified JavaScript URL or URLs and executes the specified

        callback (if any) when they have finished loading. If an array of URLs is

        specified and the browser supports it, the scripts will be loaded in

        parallel and the callback will be executed after all scripts have

        finished loading.

        Currently, only Firefox and Opera support parallel loading of scripts while

        preserving execution order. In other browsers, scripts will be

        queued and loaded one at a time to ensure correct execution order.

        @method js

        @param {String|Array} urls JS URL or array of JS URLs to load

        @param {Function} callback (optional) callback function to execute when

          the specified scripts are loaded

        @param {Object} obj (optional) object to pass to the callback function

        @param {Object} context (optional) if provided, the callback function

          will be executed in this object's context

        @static

        */

        js: function (urls, callback, obj, context) {

            load('js', urls, callback, obj, context);

        }

    };

})(this.document);

封装路由

class LazyLoader extends React.Component {

    constructor(props) {

        super(props);

        this.state = {

            module: null

        }

    }

    componentWillMount() {

        this.load(this.props.module, this.props.path);

    }

    componentWillReceiveProps(newProps) {

        this.load(newProps.module, newProps.path);

    }

    load(module, path) {

        let cssPath = path + '',

            jsPath = path + '';

        LazyLoad.css(cssPath);

        LazyLoad.js(jsPath, () => {

            this.state.module = Modules[module];

        })

    }

    render() {

        if (!this.state.module) return null;

        else this.props.children(this.state.module);

    }

}

const HTMLModule = ({ name, path }) => {

    let param = location.param;

    location.param = null;

    let ModuleComponent = Modules[name];

    return (

        <React.Fragment>

            {

                ModuleComponent ?

                    <ModuleComponent param={param} /> :

                    <LazyLoader path={path} module={name}>

                        {(Item) => <Item param={param} />}

                    </LazyLoader>

            }

        </React.Fragment>

    );

};

class ModulesRouter extends React.Component {

    constructor(props) {

        super(props);

    }

    render() {

        return (

            <Switch>

                <Route path={'/Module/Target'} render={() => <HTMLModule name='Target' path={'./Module/Target'} />} />

                <Redirect from='*' to='/403' />

            </Switch>

        );

    }

}

export default ModulesRouter;

路由与组件级别：

这里我们利用React16.6.X以后的版本新特性React.lazy跟React.Suspense进行路由以及组件部分的优化

路由设计

root路由

const RootRouter = () => {

    return <HashRouter>

        <React.Suspense fallback={<div>loading...</div>}>

            <Switch>

                <Route exact path="/" component={Home} />

                <Route path="/oAuthPromisition" component={oAuthPromisition} />

                <Route path="/WeChatPromisitionCheck" component={WeChatPromisitionCheck} />

                <Route path="/403" component={Forbidden} />

                <Route path="/sub" component={ModuleRouter} />

            </Switch>

        </React.Suspense>

    </HashRouter>

}

子路由设计

import {

    Route,

    Switch

} from 'react-router-dom';

function lazyWithPreload(factory) {

    const Component = React.lazy(factory);

    Component.preload = factory;

    return Component;

}

const LeaveMsg = lazyWithPreload(() => { return import(/* webpackChunkName: "leavemessage" */ '../JSX/Components/LeaveMessage/LeaveMessage') });

const Blog = lazyWithPreload(() => { return import(/* webpackChunkName: "blog" */ '../JSX/Components/Blog/Blog') });

const Article = lazyWithPreload(() => { return import(/* webpackChunkName: "article" */ '../JSX/Components/Blog/Article') });

const Production = lazyWithPreload(() => { return import(/* webpackChunkName: "production" */ '../JSX/Components/Production/index') });

const About = lazyWithPreload(() => { return import(/* webpackChunkName: "about" */ '../JSX/Components/About/About') });

const Write = lazyWithPreload(() => { return import(/* webpackChunkName: "write" */ '../JSX/Components/Blog/Write') });

const Overview = lazyWithPreload(() => { return import(/* webpackChunkName: "overview" */ '../JSX/Components/Overview/Overview') });

function moduleHtml(component) {

    component.preload()

    return component;

}

const SubRouter = () => {

    return <Switch>

        <Route path="/sub/leavemessage" component={moduleHtml(LeaveMsg)} />

        <Route path="/sub/blog" component={moduleHtml(Blog)} />

        <Route path="/sub/article" component={moduleHtml(Article)} />

        <Route path="/sub/production" component={moduleHtml(Production)} />

        <Route path="/sub/about" component={moduleHtml(About)} />

        <Route path="/sub/write" component={moduleHtml(Write)} />

        <Route path="/sub/overview" component={moduleHtml(Overview)} />

    </Switch>

}

export default SubRouter;

路由设计总结：

利用import动态加载组件，以达到按需加载的目的，但是在动态加载的时候webpack会发生代码切割，所有我们同时也要进行webpack配置

首先配置分割规则

output: {

  path: path.resolve(__dirname, "./public/"),

  filename: '[name].bundle.js',

  chunkFilename: '[name].module.js', //非入口文件命名规则

},

配置chunkFilename即非入口文件命名规则，就是这些由webpack分割的文件命名规范

由于webpack分割之后的路径查找方式是通过build输出路径查找的，因此如果我们单纯的将静态文件(index.html)当做根文件来加载，是找不到分割之后的代码路径的，

因此需要配置HtmlWebpackPlugin

new HtmlWebpackPlugin({

  template: 'Index.html',

  filename: 'index.html',

  chunks: [

    'index'

  ]
 }),

然后将资源文件，以及第三方文件拷贝到build输出文件夹下

new CopyWebpackPlugin([

      {

        from: __dirname + '/Lib/editor',

        to: __dirname + '/public/editor',

        // toType: 'directory', //file 或者 dir  可选，默认是文件

        // ignore: ['.*']

      },{

        from: __dirname + '/picSrc',

        to: __dirname + '/public/picSrc',

      }

    ]),

这样webpack配置修改完成了

组件级别跟路由级别基本类似，但是我们要考虑到，如果这么拆分代码，会产生很多文件，根据实际情况使用。

SPA应用性能优化（懒加载）的更多相关文章

[转]JavaScript 的性能优化：加载和执行
原文链接:http://www.ibm.com/developerworks/cn/web/1308_caiys_jsload/index.html?ca=drs- JavaScript 的性能优化: ...
【转】js JavaScript 的性能优化：加载和执行
JavaScript 的性能优化:加载和执行转自:https://www.ibm.com/developerworks/cn/web/1308_caiys_jsload/ 随着 Web2.0 技术的 ...
vue性能优化1--懒加载
懒加载也叫延迟加载,即在需要的时候进行加载.随用随载.为什么需要懒加载?像vue这种单页面应用,如果没有应用懒加载,运用webpack打包后的文件将会异常的大,造成进入首页时,需要加载的内容过多,时间 ...
JavaScript 的性能优化：加载和执行
随着 Web2.0 技术的不断推广,越来越多的应用使用 javascript 技术在客户端进行处理,从而使 JavaScript 在浏览器中的性能成为开发者所面临的最重要的可用性问题.而这个问题又因 ...
JavaScript的性能优化：加载和执行
随着 Web2.0 技术的不断推广,越来越多的应用使用 javascript 技术在客户端进行处理,从而使 JavaScript 在浏览器中的性能成为开发者所面临的最重要的可用性问题.而这个问题又因 ...
Android进阶：ListView性能优化异步加载图片使滑动效果流畅
ListView 是一种可以显示一系列项目并能进行滚动显示的 View,每一行的Item可能包含复杂的结构,可能会从网络上获取icon等的一些图标信息,就现在的网络速度要想保持ListView运行的很 ...
【摘要】JavaScript 的性能优化：加载和执行
1.浏览器遇到js代码会暂停页面的下载和渲染,谁晓得js代码会不会把html给强奸(改变)了: 2.延迟脚本加载:defer 属性 <html> <head> <titl ...
前端性能优化_css加载会造成哪些阻塞现象？
css的加载是不会阻塞DOM的解析,但是会阻塞DOM的渲染,会阻塞link后面js语句的执行.这是由于浏览器为了防止html页面的重复渲染而降低性能,所以浏览器只会在加载的时候去解析dom树,然后等在 ...
移动端性能优化动态加载JS、CSS
JS CODE (function() { /** * update: * 1.0 */ var version = "insure 1.1.0"; var Zepto = Zep ...
web前端如何性能优化提高加载速度
前端优化有以下几种途径: 一.减少HTTP请求数量和次数: 二.使用CDN: 三.添加Expires头: 四.压缩组件: 五.将样式表放在头部: 六.将脚本放在底部: 七.避免CSS表达式: 八.使用 ...

随机推荐

Codeforces1203F2. Complete the Projects (hard version) （贪心+贪心+01背包）
题目链接:传送门思路: 对于对rating有提升的项目,肯定做越多越好,所以把$b_{i} >= 0$的项目按rating要求从小到大贪心地都做掉,得到最高的rating记为r. 对于剩余的$ ...
windows环境下 RabbitMQ 安装时创建用户流程命令
https://blog.csdn.net/xiaojieblog/article/details/70332469
【HDU4034】Graph
题目大意:给定一个图的最短路,求原图中至少存在多少条边. 题解:利用 Floyd 的性质,枚举边 d[i][j],若存在一个不是两端点的点,使得 d[i][j]=d[i][k]+d[k][j] 成立, ...
JDK的动态代理深入解析（Proxy,InvocationHandler）（转）
JDK的动态代理深入解析(Proxy,InvocationHandler)(转) 一.什么是动态代理动态代理可以提供对另一个对象的访问,同时隐藏实际对象的具体事实.代理一般会实现它所表示的实际对象的 ...
java数据结构2--集合总论
集合类 0.1.为什么出现集合类? 面向对象语言对事物的体现都是以对象的形式,所以为了方便对多个对象的操作,就对对象进行存储,集合就是存储对象最常用的一种方式(容器)Java中集合也是类,真正用来存储 ...
HDU-4185-Oil Skimming（最大匹配）
链接: https://vjudge.net/problem/HDU-4185 题意: Thanks to a certain "green" resources company, ...
forEach、map、filter、reduce的区别
1.相同点: 都会循环遍历数组中的每一项: map().forEach()和filter()方法里每次执行匿名函数都支持3个参数,参数分别是:当前元素.当前元素的索引.当前元素所属的数组: 匿名函数中 ...
Linux基础教程 linux下查询history操作时间的方法
要在linux操作系统中查看history记录的操作时间,可以按如下步骤实现: 学习linux 1,修改/etc/profile文件,在末尾添加:exporthisttimeformat=”%f %t ...
BigDecimal除法问题
BigDecimal类的主要功能是进行小数的大数计算,而且最重要的是可以精确到指定的四舍五入位数. 如果要进行四舍五入的操作,则必须依靠以下的方法:public BigDecimal divide(B ...
BZOJ3438 小M的作物（和拓展）
题目链接:戳我我们如果要选择一种种植情况的话,一定是其他的选择都不可行了.这种决策问题用最小割来处理最好不过. 建图方式--A为源点,B为汇点.然后将每个点分别向A,B连边,边权为种植它的价值.组合 ...

SPA应用性能优化（懒加载）

SPA应用性能优化（懒加载）的更多相关文章

随机推荐

热门专题