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

PAT Basic 1042 字符统计 (20 分)
请编写程序,找出一段给定文字中出现最频繁的那个英文字母. 输入格式: 输入在一行中给出一个长度不超过 1000 的字符串.字符串由 ASCII 码表中任意可见字符及空格组成,至少包含 1 个英文字母, ...
跳跃表-原理及Java实现
跳跃表-原理及Java实现引言: 上周现场面试阿里巴巴研发工程师终面,被问到如何让链表的元素查询接近线性时间.笔者苦思良久,缴械投降.面试官告知回去可以看一下跳跃表,遂出此文. 跳跃表的引入我们知 ...
【POJ2486】Apple Tree
题目大意:给定一棵 N 个节点的有根树,点有点权,边权均为1.现允许从根节点出发走 K 步,求可以经过的点权之和最大是多少. 题解:可以将点权看作是价值,将可以走的步数看作是重量,则转化成了一个树上背 ...
java基础语法3 方法
方法的定义-方法的特点 -方法的应用-方法的重载-数组定义-数组初始化-二维数组-Java中参数传递的特点:值传递 7.方法 7.1 方法的定义什么是方法?Method方法就是定义在类中的,具有特定 ...
encode()和decode()两个函数
编码可以将抽象字符以二进制数据的形式表示,有很多编码方法,如utf-8.gbk等,可以使用encode()函数对字符串进行编码,转换成二进制字节数据,也可用decode()函数将字节解码成字符串:用d ...
Kafka(华为FusionInsight )操作命令
华为大数据kafka操作web界面创建角色.用户.用户管理角色进入服务器环境,进入客户端目录/opt/hadoopclient,导入环境变量source bigdata_env.切换用户kinit k ...
jvm——class类文件的结构
class类文件并不一定以磁盘的形式存在,也可以是由类加载器直接生成的二进制流,他其实是一种数据结构,类似于c语言结构体,这种数据结构只有两种数据类型:无符号数和表. 1.魔数:类似于文件拓展名,CA ...
VS2017 + EF + MySQL 环境配置
我使用过程中遇到的坑(血泪啊) 安装环境VS2017MVC+WIN10+EF6+MySQL8.0.12 1.安装MySQL connector一定要6.10.8,8.0以上全是坑,会闪退!!! 2.安 ...
Vue3.0 Function API---------引用
1.了解 Vue 3.0 是否有 break change,就像 Python 3 / Angular 2 一样? 不,100% 兼容 Vue 2.0,且暂未打算废弃任何 API(未来也不).之前有草 ...
ASP.NET如何实现断点续传的上传、下载功能？
1 背景用户本地有一份txt或者csv文件,无论是从业务数据库导出.还是其他途径获取,当需要使用蚂蚁的大数据分析工具进行数据加工.挖掘和共创应用的时候,首先要将本地文件上传至ODPS,普通的小文件通 ...

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