Vue3相关源码-Vue Router源码解析(一)

本文基于vue-router 4.1.6版本源码进行分析

本文重点分析Vue Router的WebHashHistory模式，不会对WebHistory和MemoryHistory模式过多分析

文章内容

从Vue Router的初始化代码入手，逐步分析对应的代码流程和涉及到的操作方法（push、replace、pop）

本文将着重于：

• Vue Router是如何利用routes数组建立路由映射，路由权重是如何初始化
• Vue Router的push、replace、pop流程具体执行了什么？是如何找到对应的路由数据
• Vue Router提供的RouterView和RouterLink源码分析
• Vue Router提供的多个组合式API的源码分析，包括onBeforeRouteLeave、onBeforeRouteLeave、useRouter、useRoute、useLink
• Vue Router源码注释中涉及到的issues的问题讲解和对应的修复代码分析

本文的最后将基于多个问题进行Vue Router源码的系统性总结

前置知识

Vue Router介绍

摘录于Vue Router官方文档

Vue Router 是 Vue.js 的官方路由。它与 Vue.js 核心深度集成，让用 Vue.js 构建单页应用变得轻而易举。功能包括：

• 嵌套路由映射
• 动态路由选择
• 模块化、基于组件的路由配置
• 路由参数、查询、通配符
• 展示由 Vue.js 的过渡系统提供的过渡效果
• 细致的导航控制
• 自动激活 CSS 类的链接
• HTML5 history 模式或 hash 模式
• 可定制的滚动行为
• URL 的正确编码

window.location属性

属性	描述
href	https://test.example.com:8090/vue/router#test?search=1
protocol	https:
host/hostname	test.example.com
port	8090
pathname	/vue/router
search	?search=1
hash	#test

：文档根 URL 元素

摘录于https://developer.mozilla.org/en-US/docs/Web/HTML/Element/base

• HTML 元素 指定用于一个文档中包含的所有相对 URL 的根 URL
• 只能有一个 元素
• 一个文档的基本 URL，可以通过使用 document.baseURI(en-US) 的 JS 脚本查询，如果文档不包含 元素，baseURI 默认为 document.location.href

前置问题

• 外部定义的路由，是如何在Vue Router内部建立联系的
• Vue Router是如何实现push、replace、pop操作的
• Vue Router是如何命中多层嵌套路由，比如/parent/child/child1需要加载多个组件，是如何实现的
• Vue Router有什么导航？触发的流程是怎样的
• Vue Router的导航守卫是如何做到链式调用的
• Vue Router的beforeRouteEnter和beforeRouteUpdate的触发时机
• Vue Router中route和router的区别
• hash模式跟h5 history模式在Vue Router中有什么区别
• Vue Router的hash模式重定向后还会保留浏览记录吗？比如重定向后再使用router.go(-1)会返回重定向之前的页面吗
• Vue Router的hash模式什么地方最容易导致路由切换失败

上面所有问题将在第二篇文章Vue3相关源码-Vue Router源码解析(二)中进行解答

示例代码

代码来自于Vue Router官方文档

// 1. 定义路由组件.
// 也可以从其他文件导入
const Home = { template: '
Home
' }
const About = { template: '
About
' }

// 2. 定义一些路由
// 每个路由都需要映射到一个组件。
// 我们后面再讨论嵌套路由。
const routes = [
  { path: '/', component: Home },
  { path: '/about', component: About },
]

// 3. 创建路由实例并传递 `routes` 配置
// 你可以在这里输入更多的配置，但我们在这里
// 暂时保持简单
const router = VueRouter.createRouter({
  // 4. 内部提供了 history 模式的实现。为了简单起见，我们在这里使用 hash 模式。
  history: VueRouter.createWebHashHistory(),
  routes, // `routes: routes` 的缩写
})

// 5. 创建并挂载根实例
const app = Vue.createApp({})
//确保 _use_ 路由实例使
//整个应用支持路由。
app.use(router)

app.mount('#app')

分析的核心示例代码

const router = VueRouter.createRouter({
  history: VueRouter.createWebHashHistory(),
  routes
})
const app = Vue.createApp({})
app.use(router)

下面将按照创建history、初始化router、Vue使用router的顺序进行源码分析

createWebHashHistory创建history

hash模式是用createWebHashHistory() 创建的历史记录模式

从下面代码可以知道，处理了base数据：如果没有传入url，则拼接location.pathname + location.search，然后加上#
最后还是调用了createWebHistory(base)的方法进行history的创建

function createWebHashHistory(base) {
    // Make sure this implementation is fine in terms of encoding, specially for IE11
    // for `file://`, directly use the pathname and ignore the base
    // href="https://example.com"的location.pathname也是"/"
    base = location.host ? (base || location.pathname + location.search) : '';
    // allow the user to provide a `#` in the middle: `/base/#/app`
    if (!base.includes('#'))
        base += '#';
    return createWebHistory(base);
}

直接看上面的代码会有点懵，直接看官方文档会比较好理解点，下面内容参考自https://router.vuejs.org/zh/api/#createwebhashhistory，是createWebHashHistory(base)的参数说明

Parameter	Type	Description
base	string	提供可选的 base。默认是 location.pathname + location.search。如果 head 中有一个 ，它的值将被忽略，而采用这个参数。但请注意它会影响所有的 history.pushState() 调用，这意味着如果你使用一个 标签，它的 href 值必须与这个参数相匹配 (请忽略 # 后面的所有内容)

// at https://example.com/folder
createWebHashHistory() // 给出的网址为 `https://example.com/folder#`
createWebHashHistory('/folder/') // 给出的网址为 `https://example.com/folder/#`
// 如果在 base 中提供了 `#`，则它不会被 `createWebHashHistory` 添加
createWebHashHistory('/folder/#/app/') // 给出的网址为 `https://example.com/folder/#/app/`
// 你应该避免这样做，因为它会更改原始 url 并打断正在复制的 url
createWebHashHistory('/other-folder/') // 给出的网址为 `https://example.com/other-folder/#`

// at file:///usr/etc/folder/index.html
// 对于没有 `host` 的位置，base被忽略
createWebHashHistory('/iAmIgnored') // 给出的网址为 `file:///usr/etc/folder/index.html#`

从上面例子可以看出，如果传入base字符串，会以传入base优先级最高去拼接，然后才是location.pathname + location.search
同时会检测传入base是否含有"#"，如果没有，则在后面添加"#"

当然，也会存在传入base跟location.pathname不一样的情况，如上面例子/other-folder/所示，会直接更改网址去除location.pathname，改为base

createWebHistory创建history

整体概述

routerHistory对标的是原生的history对象，routerHistory在原生的history对象的API基础上，增添了一些逻辑处理，实现Vue的路由切换、路由映射组件、组件切换的功能

function createWebHistory(base) {
    // 步骤1: normalizeBase整理url
    base = normalizeBase(base);
    // 步骤2: useHistoryStateNavigation
    const historyNavigation = useHistoryStateNavigation(base);
    // 步骤3: useHistoryListeners
    const historyListeners = useHistoryListeners(base, 
      historyNavigation.state, historyNavigation.location, historyNavigation.replace);
    // 步骤4: 合并historyNavigation和historyListeners，整合为routerHistory
    const routerHistory = assign({
        // it's overridden right after
        location: '',
        base,
        go,
        createHref: createHref.bind(null, base),
    }, historyNavigation, historyListeners);
    Object.defineProperty(routerHistory, 'location', {
        enumerable: true,
        get: () => historyNavigation.location.value,
    });
    Object.defineProperty(routerHistory, 'state', {
        enumerable: true,
        get: () => historyNavigation.state.value,
    });
    return routerHistory;
}
function go(delta, triggerListeners = true) {
    if (!triggerListeners)
        historyListeners.pauseListeners();
    history.go(delta);
}

步骤1: normalizeBase

从上面分析可以知道，如果有传入base 或者 本身存在location.pathname，那么到达normalizeBase(base)的base就不可能为空

但是也有可能存在

• 没有传入base+本身不存在location.pathname
• 没有传入base+本身不存在location.host(file://xxx)

当没有base时，我们就会检测是否具有标签，标签的含义可以参考https://developer.mozilla.org/en-US/docs/Web/HTML/Element/base，如果没有标签，则使用location.href作为基础的链接，如果有标签，则使用该标签作为基础链接

For example, given and this link: #anchor>To anchor. The link points to https://example.com/#anchor

然后替换域名部分的字符串，最终返回base="/xxxx"或者"#xxxxx"

function normalizeBase(base) {
    if (!base) {
        if (isBrowser) {
            // respect  tag
            const baseEl = document.querySelector('base');
            base = (baseEl && baseEl.getAttribute('href')) || '/';
            // 剔除https://xx.xxxx.com这部分的字母
            base = base.replace(/^\w+:\/\/[^\/]+/, '');
        } else {
            base = '/';
        }
    }
    if (base[0] !== '/' && base[0] !== '#')
        base = '/' + base;
    // 删除尾部的斜杠
    return removeTrailingSlash(base);
}
const TRAILING_SLASH_RE = /\/$/;
const removeTrailingSlash = (path) => path.replace(TRAILING_SLASH_RE, '');

一般情况下，经过normalizeBase()得到的base的结构为location.pathname+location.search+"#"，具体结构如下面所示：

下面链接为测试链接，仅仅表示不考虑特殊情况下的base结构，没有其它含义

base = "/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#"

步骤2: useHistoryStateNavigation

整体概述

该方法初始化的historyNavigation包含了当前的路由、当前的堆栈信息以及对应的push加操作和replace覆盖操作，同时检测history.state是否为空，如果为空，需要压入一个初始化的currentLocation

除了push、replace等常规操作，我们知道一个路由还需要具备路由后退的监听，Vue Router将后退操作的监听放在下一节要分析的useHistoryListeners中

// const historyNavigation = useHistoryStateNavigation(base);
function useHistoryStateNavigation(base) {
    const { history, location } = window;
    // currentLocation.value=除去"#"后的字符串
    const currentLocation = {
        value: createCurrentLocation(base, location),
    };
    const historyState = { value: history.state };
    // build current history entry as this is a fresh navigation
    if (!historyState.value) {
        changeLocation(currentLocation.value, {
            back: null,
            current: currentLocation.value,
            forward: null,
            // the length is off by one, we need to decrease it
            position: history.length - 1,
            replaced: true,
            // don't add a scroll as the user may have an anchor, and we want
            // scrollBehavior to be triggered without a saved position
            scroll: null,
        }, true);
    }

    function changeLocation(to, state, replace) {...}
    function replace(to, data) {...}
    function push(to, data) {...}

    return {
        location: currentLocation,
        state: historyState,
        push,
        replace,
    };
}

createCurrentLocation()

从 window.location对象中创建一个规范化的数据作为当前的路径，可以简单认为最后得到的路径就是上面注册的routes中的其中一个路由路径

具体逻辑分析请看下面的例子分析

function createCurrentLocation(base, location) {
  const { pathname, search, hash } = location;
  // allows hash bases like #, /#, #/, #!, #!/, /#!/, or even /folder#end
  const hashPos = base.indexOf('#');
  if (hashPos > -1) {
    let slicePos = hash.includes(base.slice(hashPos))
      ? base.slice(hashPos).length
      : 1;
    let pathFromHash = hash.slice(slicePos);
    // prepend the starting slash to hash so the url starts with /#
    if (pathFromHash[0] !== '/')
      pathFromHash = '/' + pathFromHash;
    return stripBase(pathFromHash, '');
  }
  const path = stripBase(pathname, base);
  return path + search + hash;
}
function stripBase(pathname, base) {
    // no base or base is not found at the beginning
    if (!base || !pathname.toLowerCase().startsWith(base.toLowerCase()))
        return pathname;
    return pathname.slice(base.length) || '/';
}

延续上面例子所用的文件路径，我们可以知道：

location.href = "http://localhost:63342/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#/about";
base = "/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#";
location.hash = "#/about";

location.pathname = "/Frontend-Articles/vue3-debugger/router/vue-router.html";

经过hash.slice(slicePos)得到的pathFromHash为

pathFromHash = "/about"

最终经过stripBase(pathFromHash, '')，直接返回pathFromHash = "/about"

!!!!上面的举例是比较正常一点的路径，但是现实中肯定会存在各种各样的路径=_=后面遇到奇怪的路径再回来补充例子，暂时跳过这部分规范化路径逻辑的思考

最终得到目前的路由路径为

const currentLocation = {
    value: "/about"
};

changeLocation()

function changeLocation(to, state, replace) {
    const hashIndex = base.indexOf('#');
    const url = hashIndex > -1
        ? (location.host && document.querySelector('base')
            ? base
            : base.slice(hashIndex)) + to
        : createBaseLocation() + base + to;
    try {
        // BROWSER QUIRK
        // NOTE: Safari throws a SecurityError when calling this function 100 times in 30 seconds
        history[replace ? 'replaceState' : 'pushState'](state, '', url);
        historyState.value = state;
    } catch (err) {
        // Force the navigation, this also resets the call count
        location[replace ? 'replace' : 'assign'](url);
    }
}

传入要跳转的路径to，然后拼接要跳转的路径url
在上面的例子中，我们知道

base = "/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#";
url = base.slice(hashIndex)) + to = "#/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#/about"

changeLocation()本质就是拼接了跳转路径，然后

• 调用浏览器原生API提供的window.history.pushState/window.history.replaceState方法
• 更新historyState.value

一开始调用changeLocation()传入的state数据如下所示

{
    back: null,
    current: currentLocation.value,
    forward: null,
    // the length is off by one, we need to decrease it
    position: history.length - 1,
    replaced: true,
    // don't add a scroll as the user may have an anchor, and we want
    // scrollBehavior to be triggered without a saved position
    scroll: null,
}

changeLocation()的过程中如果发生错误，则强制跳转，调用原生的window.location.assign(url)：使窗口加载并显示指定url处的document

导航发生后，用户可以通过按“后退”按钮导航回到location.assign替换的页面，即这个方法会产生浏览器历史记录，不是直接覆盖当前页面

window.location.assign可以参考文档assign，跟window.location.replace不同点在于，replace不会保留目前要替换的页面记录

push()

代码逻辑也比较清晰和简单，主要分为5个步骤

• 使用forward: to记录要跳转的路由，并且放入currentState中
• 使用changeLocation()进行history.replaceState()替换当前的浏览记录
• 使用buildState()构建出新的路由对象，形成新的路由state
• 使用changeLocation()进行history.pushState()新增一条新的浏览记录
• 更新多个方法都使用的变量currentLocation为当前的新的路由地址currentLocation.value=to

function push(to, data) {
    const currentState = assign({},
        // use current history state to gracefully handle a wrong call to
        // history.replaceState
        // https://github.com/vuejs/router/issues/366
        historyState.value, history.state, {
        forward: to,
        scroll: computeScrollPosition(),
    });
    if (!history.state) {
        // 如果当前没有history.state，报错提示：
        // 如果您手动调用 history.replaceState，请确保保留现有的历史状态
    }
    changeLocation(currentState.current, currentState, true);
    const state = assign({}, buildState(currentLocation.value, to, null), { position: currentState.position + 1 }, data);
    changeLocation(to, state, false);
    currentLocation.value = to;
}
function buildState(back, current, forward, replaced = false, computeScroll = false) {
    return {
        back,
        current,
        forward,
        replaced,
        position: window.history.length,
        scroll: computeScroll ? computeScrollPosition() : null,
    };
}

replace()

构建出新的路由对象，形成新的路由state，然后使用changeLocation()进行history.replaceState()替换当前的浏览记录

在changeLocation()中也进行historyState.value=state

更新多个方法都使用的变量currentLocation为当前的新的路由地址currentLocation.value=to

function replace(to, data) {
    const state = assign({}, history.state, buildState(historyState.value.back,
        // keep back and forward entries but override current position
        to, historyState.value.forward, true), data, { position: historyState.value.position });
    changeLocation(to, state, true);
    currentLocation.value = to;
}

---

小结

从上面的代码可以知道，historyNavigation最终是一个对象，具有location、state、push、replace属性，其中

• location：代表除去"#"之后的路径内容，管理当前页面的路径数据，比如location="/about"

• state：保存原生的window.history.state数据，比如下面的histroyState就是代表window.history.state数据

  // history.state格式如下
  const histroyState = {
    back: null,
    current: "/Frontend-Articles/vue3-debugger/router/vue-router.html?_ijt=q70k46ba94eddqp8fu39ta28fh&_ij_reload#/about",
    forward: null,
    position: 14,
    replaced: true,
    scroll: null
  }
  const state = { value: histroyState };

• push()：新增路由时调用的方法
• replace()：替换路由时调用的方法

步骤3: useHistoryListeners

整体概述

传入初始化路径以及historyNavigation维护的状态和方法，进行historyListeners的初始化

const historyListeners = useHistoryListeners(base, 
      historyNavigation.state, historyNavigation.location, historyNavigation.replace);

function useHistoryListeners(base, historyState, currentLocation, replace) {
    let listeners = [];
    let teardowns = [];
    let pauseState = null;
    const popStateHandler = ({ state, }) => {...};
    function pauseListeners() {...}
    function listen(callback) {...}
    function beforeUnloadListener() {...}
    function destroy() {...}
    
    // set up the listeners and prepare teardown callbacks
    window.addEventListener('popstate', popStateHandler);
    window.addEventListener('beforeunload', beforeUnloadListener);
    return {
        pauseListeners,
        listen,
        destroy,
    };
}

popStateHandler()

监听后退事件，这个监听方法比较复杂并且非常重要！

简单示例

window.addEventListener("popstate", (event) => {
  console.log(
    `location: ${document.location}, state: ${JSON.stringify(event.state)}`
  );
});
history.pushState({ page: 1 }, "title 1", "?page=1");
history.pushState({ page: 2 }, "title 2", "?page=2");
history.replaceState({ page: 3 }, "title 3", "?page=3");
history.back(); // Logs "location: http://example.com/example.html?page=1, state: {"page":1}"
history.back(); // Logs "location: http://example.com/example.html, state: null"
history.go(2); // Logs "location: http://example.com/example.html?page=3, state: {"page":3}"

触发时机

• history.pushState() 或者history.replaceState()不会触发popStateHandler()
• 用户手动点击浏览器的后退按钮、history.back() 、 history.forward()都会触发popStateHandler()
• history.go(数字) 本质就是history.back() / history.forward()调用1次或者多次，也会触发popStateHandler()

触发时携带的参数state

摘录于https://developer.mozilla.org/en-US/docs/Web/API/Window/popstate_event#the_history_stack

如果被激活的历史条目是通过调用 history.pushState() 创建的，或者受到调用 history.replaceState() 的影响，则 popstate 事件的状态属性包含历史条目状态对象的副本

换句话说，popStateHandler({state})中state不为空的前提是我们一直都有使用pushState/replaceState

源码分析

从上面的简单示例可以知道，携带的参数state是后退后当前的路由参数，如果存在，直接把它当作目前最新的historyState.value=state，然后根据后退前缓存的数据fromState计算出对应的delta，最终调用listeners进行监听方法的触发

const popStateHandler = ({ state, }) => {
    const to = createCurrentLocation(base, location);
    const from = currentLocation.value;
    const fromState = historyState.value;
    let delta = 0;
    if (state) {
        currentLocation.value = to;
        historyState.value = state;
        // ignore the popstate and reset the pauseState
        if (pauseState && pauseState === from) {
            pauseState = null;
            return;
        }
        delta = fromState ? state.position - fromState.position : 0;
    } else {
        replace(to);
    }
    listeners.forEach(listener => {
        listener(currentLocation.value, from, {
            delta,
            type: NavigationType.pop,
            direction: delta
                ? delta > 0
                    ? NavigationDirection.forward
                    : NavigationDirection.back
                : NavigationDirection.unknown,
        });
    });
};
window.addEventListener('popstate', popStateHandler);

如果没有携带参数state，我们就无法知道目前的状态，因此会直接使用const to = createCurrentLocation(base, location)构建出新的路由对象，形成新的路由state，然后使用changeLocation()进行history.replaceState()替换当前的浏览记录

没有携带参数state时，listeners进行监听方法触发的状态为NavigationDirection.unknown

function replace(to, data) {
    const state = assign({}, history.state, buildState(historyState.value.back,
        // keep back and forward entries but override current position
        to, historyState.value.forward, true), data, { position: historyState.value.position });
    changeLocation(to, state, true);
    currentLocation.value = to;
}

listen() and pauseListeners()

提供给外部注册监听的方法，并且返回停止该监听的方法teardown

function listen(callback) {
    // set up the listener and prepare teardown callbacks
    listeners.push(callback);
    const teardown = () => {
        const index = listeners.indexOf(callback);
        if (index > -1)
            listeners.splice(index, 1);
    };
    teardowns.push(teardown);
    return teardown;
}
function pauseListeners() {
    pauseState = currentLocation.value;
}

---

问题：

1. 什么时候触发listen()注册监听方法？
2. listen()注册的监听方法有什么用处？
3. 为什么要使用pauseListeners()暂停监听？
4. pauseListeners()是如何暂停方法执行的？

这些问题将在第二篇文章Vue3相关源码-Vue Router源码解析(二)的总结Vue Router的hash模式什么地方最容易导致路由切换失败进行解答

---

beforeUnloadListener()注册监听

beforeunload 事件在即将离开当前页面（刷新或关闭）时触发，触发时会重置history.replaceState保留当前的scroll（来自Vue Router的代码git提交记录描述）

该事件可用于弹出对话框，提示用户是继续浏览页面还是离开当前页面，比如“确定要离开此页吗？”

function beforeUnloadListener() {
    const { history } = window;
    if (!history.state)
        return;
    history.replaceState(assign({}, history.state, { scroll: computeScrollPosition() }), '');
}
window.addEventListener('beforeunload', beforeUnloadListener);

destroy()

清除所有listeners监听方法，移除全局注册的事件

function destroy() {
    for (const teardown of teardowns)
        teardown();
    teardowns = [];
    window.removeEventListener('popstate', popStateHandler);
    window.removeEventListener('beforeunload', beforeUnloadListener);
}

步骤4: go()和routerHistory初始化

除了上述的3个步骤，最后一个步骤就是将上面步骤得到的historyNavigation、historyListeners以及对应的其它基础数据进行合并成为routerHistory

其中go()事件除了第一个参数回退的层级，还有第二个参数triggerListeners，阻止监听器触发执行，该方法的分析将放在第二篇文章Vue3相关源码-Vue Router源码解析(二)的总结中进行解答

function go(delta, triggerListeners = true) {
    if (!triggerListeners)
        historyListeners.pauseListeners();
    history.go(delta);
}
const routerHistory = assign({
    // it's overridden right after
    location: '',
    base,
    go,
    createHref: createHref.bind(null, base),
}, historyNavigation, historyListeners);
Object.defineProperty(routerHistory, 'location', {
    enumerable: true,
    get: () => historyNavigation.location.value,
});
Object.defineProperty(routerHistory, 'state', {
    enumerable: true,
    get: () => historyNavigation.state.value,
});
return routerHistory;

createMemoryHistory创建history

createMemoryHistory会创建一个基于内存的history，主要目的是为了处理SSR
与前面两个方法不同的是，createMemoryHistory维护一个队列queue和一个position，来保证历史记录存储的正确性

这里不展开详细分析，请读者自行研究

createRouter创建VueRouter对象

整体概述

createRouter()对VueRouter.createWebHashHistory()和routes的数据进行合并和拼接，组合成router对象

// Vue初始化Vue Router时的示例代码
const router = VueRouter.createRouter({
    // 4. 内部提供了 history 模式的实现。为了简单起见，我们在这里使用 hash 模式。
    history: VueRouter.createWebHashHistory(),
    routes, // `routes: routes` 的缩写
})

function createRouter(options) {
    const matcher = createRouterMatcher(options.routes, options);
    //...多个方法，基本都是为下面初始化router的属性
    //比如addRoute()、removeRoute()、hasRoute()
    const router = {
        currentRoute,
        listening: true,
        addRoute,
        removeRoute,
        hasRoute,
        getRoutes,
        resolve,
        options,
        push,
        replace,
        go,
        back: () => go(-1),
        forward: () => go(1),
        beforeEach: beforeGuards.add,
        beforeResolve: beforeResolveGuards.add,
        afterEach: afterGuards.add,
        onError: errorHandlers.add,
        isReady,
        install(app) {
            //...省略，外部app.use()时调用
        },
    };
    return router;
}

这个方法的逻辑非常繁杂，本文不会具体展开讲，后面会讲一些重点的方法以及根据一些业务去分析createRouter里面的方法，我们只需要明白，这个方法创建的数据就是Vue Router提供给外部使用的对象即可

createRouterMatcher: 根据外部传入的options.routes初始化路由配置

整体概述

根据外部传入的options.routes初始化路由配置，建立matchers数组

注意：createRouterMatcher()内部有addRoute()方法，上面createRouter()内部也有addRoute()方法，不要搞混

// globalOptions={history, routes}
function createRouterMatcher(routes, globalOptions) {
    // normalized ordered array of matchers
    const matchers = [];
    const matcherMap = new Map();
    globalOptions = mergeOptions({ strict: false, end: true, sensitive: false }, globalOptions);
    function getRecordMatcher(name)
    function addRoute(record, parent, originalRecord) {...}
    function removeRoute(matcherRef) {...}
    function getRoutes() {...}
    function insertMatcher(matcher) {...}
    function resolve(location, currentLocation) {...}
    // add initial routes
    routes.forEach(route => addRoute(route));
    return { addRoute, resolve, removeRoute, getRoutes, getRecordMatcher };
}

从上面的代码，我们可以知道，除了一些方法的初始化之外，主要就执行了两个步骤：mergeOptions()和routers.forEach(route=>addRoute(route))

mergeOptions

合并传递过来的参数

// partialOptions={history, routes}
// defaults={ strict: false, end: true, sensitive: false }
function mergeOptions(defaults, partialOptions) {
    const options = {};
    for (const key in defaults) {
        options[key] = key in partialOptions ? partialOptions[key] : defaults[key];
    }
    return options;
}

routers.forEach(route=>addRoute(route))

function addRoute(record, parent, originalRecord) {
    //...
    const mainNormalizedRecord = normalizeRouteRecord(record);
    const normalizedRecords = [
        mainNormalizedRecord,
    ];
    if ('alias' in record) {
        const aliases = typeof record.alias === 'string' ? [record.alias] : record.alias;
        for (const alias of aliases) {
            // ...省略处理别名的逻辑
            normalizedRecords.push(assign({}, mainNormalizedRecord, {...}));
        }
    }
    let matcher;
    let originalMatcher;
    for (const normalizedRecord of normalizedRecords) {
        //...
        // 步骤1: 创建matcher对象
        matcher = createRouteRecordMatcher(normalizedRecord, parent, options);
        //...
        // 步骤2: 创建matcher对象和它的children-matcher之间的关系
        if (mainNormalizedRecord.children) {
            const children = mainNormalizedRecord.children;
            for (let i = 0; i < children.length; i++) {
                addRoute(children[i], matcher, originalRecord && originalRecord.children[i]);
            }
        }
        //...
        // 步骤3: 将创建好的matcher插入到matchers数组中
        insertMatcher(matcher);
    }
    return originalMatcher
        ? () => fcd
            // since other matchers are aliases, they should be removed by the original matcher
            removeRoute(originalMatcher);
        }
        : noop;
}

步骤1: createRouteRecordMatcher创建matcher对象

在路由对象record的基础上进行路由权重的计算以及正则表达式的构建，为后面路径的映射提供对应的matcher对象

function createRouteRecordMatcher(record, parent, options) {
    const parser = tokensToParser(tokenizePath(record.path), options);
    // warn against params with the same name

    const existingKeys = new Set();
    for (const key of parser.keys) {
        existingKeys.add(key.name);
    }

    const matcher = assign(parser, {
        record,
        parent,
        // these needs to be populated by the parent
        children: [],
        alias: [],
    });
    if (parent) {
        // both are aliases or both are not aliases
        // we don't want to mix them because the order is used when
        // passing originalRecord in Matcher.addRoute
        if (!matcher.record.aliasOf === !parent.record.aliasOf)
            parent.children.push(matcher);
    }
    return matcher;
}

tokenizePath()和tokensToParser()解析多种模式下的路由路径，下面将着重分析Vue Router4中的路由权重计算逻辑

多种类型的路由介绍

下面内容参考自Vue Router官方文档

静态路由

const routes = [
  { path: '/', component: Home },
  { path: '/about', component: About },
]

带参数的动态路由匹配
像 /users/johnny 和 /users/jolyn 这样的 URL 都会映射到同一个路由

const routes = [
    // 动态字段以冒号开始
    { path: '/users/:id', component: User },
]

正则表达式路由
常规参数只匹配url 片段之间的字符，用/分隔。如果我们想匹配任意路径，我们可以使用自定义的路径参数正则表达式，在路径参数后面的括号中加入 正则表达式

const routes = [
    // 将匹配所有内容并将其放在 `$route.params.pathMatch` 下
    { path: '/:pathMatch(.*)*', name: 'NotFound', component: NotFound }
]

const routes = [
    // 将匹配以 `/user-` 开头的所有内容，并将其放在 `$route.params.afterUser` 下
    { path: '/user-:afterUser(.*)', component: UserGeneric }
]

const routes = [
  // /:orderId -> 仅匹配数字
  { path: '/:orderId(\\d+)' },
  // /:productName -> 匹配其他任何内容
  { path: '/:productName' },
]

const routes = [
  // /:chapters ->  匹配 /one, /one/two, /one/two/three, 等
  { path: '/:chapters+' },
  // /:chapters -> 匹配 /, /one, /one/two, /one/two/three, 等
  { path: '/:chapters*' },
]

具体例子分析

const routes = [
    {path: '/', component: Home},
    {path: '/child', component: Child1}, // 静态路由
    {path: '/child/:id', component: Child2}, // 动态路由
    {path: '/child/:id?', component: Child3}, // 动态路由（可选）
    {path: '/:child1(\\d+)', component: Child4}, // 动态路由（限制数字）TokenizerState.ParamRegExp
    {path: '/:child2+', component: Child5}, // 动态路由（可重复）
]

具体例子使用Vue Router官方文档提供的在线测试链接Path Ranker进行测试，结果为下图所示

下面将分析如何得到上面图中的分数以及对应的路由匹配逻辑

tokenizePath()

该方法就是直接根据不同的const char=path[i++]进行不同状态的赋值，然后形成对应的数据
下面就是Vue Router源码中标记的几种状态

const enum TokenizerState {
  Static, // 静态数据，比如/child这种类型中的"child"
  Param, // 动态路由以及常见的正则表达式，比如"/:child1(\\d+)"
  ParamRegExp, // custom re for a param
  ParamRegExpEnd, // check if there is any ? + *
  EscapeNext,
}

直接使用debugger断点调试该方法即可清晰明白整个流程

取一个比较复杂的正则表达式例子为：path: "/c\\hil\\d3/new:c\\hild1(\\d+)?"，这个例子中有静态数据/child3，有动态数据:child1，也有一些正则表达式\d+以及?，还有一些中间乱入的\\
tokenizePath()核心方法为下面代码块，根据判断目前item是"/"，还是":"，还是"("，然后进行不同的状态的赋值

function tokenizePath(path) {
    while (i < path.length) {
        char = path[i++];
        if (char === '\\' && state !== 2 /* TokenizerState.ParamRegExp */) {
            previousState = state;
            state = 4 /* TokenizerState.EscapeNext */;
            continue;
        }
        switch (state) {
            case 0 /* TokenizerState.Static */:
                break;
            case 4 /* TokenizerState.EscapeNext */:
                break;
            case 1 /* TokenizerState.Param */:
                break;
            case 2 /* TokenizerState.ParamRegExp */:
                break;
            case 3 /* TokenizerState.ParamRegExpEnd */:
                break;
            default:
                crash('Unknown state');
                break;
        }
    }
}

• TokenizerState.Static: 遇到"/"，进入该状态，处理静态路径，比如"/child3"
• TokenizerState.Param: 遇到":"，进入该状态，处理动态路径，比如":child1"
• TokenizerState.ParamRegExp: 遇到"("，进入该状态，开始处理正则表达式
• TokenizerState.ParamRegExpEnd: 遇到")"，进入该状态，结束处理正则表达式，然后重新回到TokenizerState.Static状态
• TokenizerState.EscapeNext: 在不是TokenizerState.ParamRegExp(处理正则表达式)的状态下遇到了"\\"，直接跳过"\\"，比如/c\\hil\\d3->/child3

最终形成segment数组为：

segment.push({
    type: 1 /* TokenType.Param */,
    value: buffer, // path的部分内容
    regexp: customRe, // 正则表达式的内容
    repeatable: char === '*' || char === '+', // 是否允许重复，代表"+"或者"*"这些表示重复的正则表达式
    optional: char === '*' || char === '?', // 是否可选，代表"?"或者"*"这些表示可选的正则表达式
});

segment代表每一个片段的值，在path: "/c\\hil\\d3/new:c\\hild1(\\d+)?"中，可以分为2个片段

• c\\hil\\d3
• new:c\\hild1(\\d+)?

最终要根据segment再形成一个总体的数组tokens

function finalizeSegment() {
    if (segment)
        tokens.push(segment);
    segment = [];
}

path: "/c\\hil\\d3/new:c\\hild1(\\d+)?"形成tokens如下所示

• c\\hil\\d3: 静态路径
• new:c\\hild1(\\d+)?: 静态路径+动态路径（包含正则表达式+可选）

[
  [
    {
      "type": 0,
      "value": "child3"
    }
  ],
  [
    {
      "type": 0,
      "value": "new"
    },
    {
      "type": 1,
      "value": "child1",
      "regexp": "\\d+",
      "repeatable": false,
      "optional": true
    }
  ]
]

tokensToParser()

遍历tokenizePath()拿到的tokens进行权重得分、正则表达式的计算

function tokensToParser(segments, extraOptions) {
    //...
    for (const segment of segments) {
        //...
    }
    return {
        re,
        score,
        keys,
        parse,
        stringify,
    };
}

下面分析正则表达式re和权重分数score的构建流程

正则表达式re的构建

TokenType.Static: 静态路径，比如path="/child"，使用原来的路径值，会进行一些特殊字符的转译，最终形成re=/^\/child$/i

const REGEX_CHARS_RE = /[.+*?^${}()[\]/\\]/g;
if (!tokenIndex)
    pattern += '/';
pattern += token.value.replace(REGEX_CHARS_RE, '\\$&');

TokenType.Param: 动态匹配路径，包含各种正则表达式
当动态匹配路径不包含正则表达式时，直接使用BASE_PARAM_PATTERN = '[^/]+?'作为最终re的值
比如path="/child/:id"，最终形成re=/^\/child\/([^/]+?)$/i

由于path包含静态路径+动态路径，因此re= 静态路径+动态路径 = ^\/child\/ + ([^/]+?)+i

当动态匹配路径包含正则表达式时，直接路径中的正则表达式，即segment.regexp去构建最终的re的值
比如path="/:child1(\\d+)?"，得到的segment.regexp="\\d+"，最终形成re=/^(?:\/(\d+))?$/i

const BASE_PARAM_PATTERN = '[^/]+?';
const re = regexp ? regexp : BASE_PARAM_PATTERN;
let subPattern = repeatable ? `((?:${re})(?:/(?:${re}))*)` : `(${re})`;
// prepend the slash if we are starting a new segment
if (!tokenIndex)
    subPattern =
        // avoid an optional / if there are more segments e.g. /:p?-static
        // or /:p?-:p2
        optional && segment.length < 2
            ? `(?:/${subPattern})`
            : '/' + subPattern;
if (optional)
    subPattern += '?';

如果path中包含可重复的正则表达式，比如path="/:child1(\\d+)+"，得到的segment.regexp="\\d+"，由于最后有个+，因此repeatable=true，最终形成re=/^\/((?:\d+)(?:\/(?:\d+))*)$/i

---

权重分数score的构建

const enum PathScore {
  _multiplier = 10,
  Root = 9 * _multiplier, // just /
  Segment = 4 * _multiplier, // /a-segment
  SubSegment = 3 * _multiplier, // /multiple-:things-in-one-:segment
  Static = 4 * _multiplier, // /static
  Dynamic = 2 * _multiplier, // /:someId
  BonusCustomRegExp = 1 * _multiplier, // /:someId(\\d+)
  BonusWildcard = -4 * _multiplier - BonusCustomRegExp, // /:namedWildcard(.*) we remove the bonus added by the custom regexp
  BonusRepeatable = -2 * _multiplier, // /:w+ or /:w*
  BonusOptional = -0.8 * _multiplier, // /:w? or /:w*
  // these two have to be under 0.1 so a strict /:page is still lower than /:a-:b
  BonusStrict = 0.07 * _multiplier, // when options strict: true is passed, as the regex omits \/?
  BonusCaseSensitive = 0.025 * _multiplier, // when options strict: true is passed, as the regex omits \/?
}

原始分数为PathScore.Segment=40

• 如果对大小写敏感（options.sensitive=true），则增加分数PathScore.BonusCaseSensitive=0.25，即使40+0.25=40.25
• 如果最后一个segment设置了options.strict=true，额外得到PathScore.BonusStrict的0.7

上面是初始化的分数，下面将针对各种状态进行分析

TokenType.Static

静态路径会额外得到PathScore.Static的+40分，如/child的分数是40+40+0.7=80.7

TokenType.Param

动态匹配路径会额外得到PathScore.Dynamic的+20分，如/:child的分数是40+20+0.7=60.7

• 如果包含正则表达式，正则表达式会额外得到PathScore.BonusCustomRegExp的+10分，如/:child1(\\d+)的分数是60+10+0.7=70.7
• 如果包含可选optional符号，额外得到PathScore.BonusOptional的-10分，如/:child1?的分数是60-10+0.7=60.7
• 如果包含匹配所有字符.*符号，额外得到PathScore.BonusWildcard的-50分，如/:child1(.*)的分数是60+10-50+0.7=20.7
• 如果包含重复repeatable符号，额外得到PathScore.BonusRepeatable的-20分，如/:child1+的分数是60-20+0.7=40.7

/:child1+的正则表达式为空，即TokenizePath [[{"type":1,"value":"child1","regexp":"","repeatable":true,"optional":false}]]

/:child1?同理，正则表达式也为空，因此没有PathScore.BonusCustomRegExp的+10分

---

如果存在多段segment

• /child/:child1，分数为child+:child1=[[80], [60.7]]
• /child/pre-:child1，分数为child+pre-+:child1=[[80],[80,60.7]]

---

小结

通过tokenizePath()拿到routes解析后的路由数据

{
  "type": 1,
  "value": "child1",
  "regexp": "\\d+",
  "repeatable": false,
  "optional": true
}

根据tokenizePath()解析后的数据，进行路由权重的计算，通过tokensToParser()拿到权重以及拼接对应的正则表达式，形成一个更加完善的路由数据

parse()和stringify()会在使用RouterMatcher匹配路径时用到

{
  re, //正则表达式
  score, //权重分数
  keys, //{name,repeatable,optional}
  parse, //动态路由匹配，匹配出动态路由对应那个静态路径的方法
  stringify //格式化params得到path的方法
}

最终拼接所有数据形成matcher，createRouteRecordMatcher()返回数据如下所示

function createRouteRecordMatcher(record, parent, options) {
    const parser = tokensToParser(tokenizePath(record.path), options);
    const matcher = assign(parser, {
        record,
        parent,
        // these needs to be populated by the parent
        children: [],
        alias: [],
    });
    return matcher;
}

步骤2: 创建matcher对象和它的children-matcher之间的关系

遍历目前route的children，将当前route创建的matcher作为parent传入addRoute()，递归调用addRoute()创建新的matcher

if (mainNormalizedRecord.children) {
    const children = mainNormalizedRecord.children;
    for (let i = 0; i < children.length; i++) {
        addRoute(children[i], matcher, originalRecord && originalRecord.children[i]);
    }
}

步骤3: insertMatcher()将创建好的matcher插入到matchers数组和matcherMap对象中

• 使用comparePathParserScore()对matchers进行排序，每次从头开始遍历let i=0，如果目前matchers[i]权重较大，则i++，否则直接调用matchers.splice(i, 0, matcher)插入matcher进去
• matcherMap以路由对象record的名称作为key，matcher作为value

在使用matchers查找路径时，会使用matcher=matchers.find(m => m.re.test(path))，权重越大的元素放在越前面，会最先被找到，因此查找路径时会先找到权重最大的那个matcher

function insertMatcher(matcher) {
    let i = 0;
    while (i < matchers.length &&
        comparePathParserScore(matcher, matchers[i]) >= 0 &&
        // Adding children with empty path should still appear before the parent
        // https://github.com/vuejs/router/issues/1124
        (matcher.record.path !== matchers[i].record.path ||
            !isRecordChildOf(matcher, matchers[i])))
        i++;
    matchers.splice(i, 0, matcher);
    // only add the original record to the name map
    if (matcher.record.name && !isAliasRecord(matcher))
        matcherMap.set(matcher.record.name, matcher);
}

而权重的比较也比较简单，不是按照总分计算权重，而是根据数组中的每一项从头到尾进行比较，如果a<b，则返回大于0的数字

• 如果相同index比较能出结果，直接返回结果

  • 长度相同时，直接比较相同index，直接返回差值
  • 如果a的长度只有1，并且a[0]=PathScore.Static+PathScore.Segment，表示a是一个静态路径，并且只有一个元素，那么a的权重较大，返回-1，反之返回1
  • 如果b的长度只有1，并且b[0]=PathScore.Static+PathScore.Segment，表示b是一个静态路径，并且只有一个元素，那么b的权重较大，返回-1，反之返回1
  • 如果a和b的长度相同，每一个值比较都等于0，证明a===b，返回0，继续下面流程
• 如果相同index的结果都一样，并且长度相差1，则比较最后一位是否为负数，哪个为负数，就比较小
• 如果相同index的结果都一样，并且长度不止1，则比较谁的长度大，谁大权重就大

function comparePathParserScore(a, b) {
    let i = 0;
    const aScore = a.score;
    const bScore = b.score;
    while (i < aScore.length && i < bScore.length) {
        const comp = compareScoreArray(aScore[i], bScore[i]);
        // do not return if both are equal
        if (comp)
            return comp;
        i++;
    }
    if (Math.abs(bScore.length - aScore.length) === 1) {
        if (isLastScoreNegative(aScore))
            return 1;
        if (isLastScoreNegative(bScore))
            return -1;
    }
    return bScore.length - aScore.length;
}
function compareScoreArray(a, b) {
    let i = 0;
    while (i < a.length && i < b.length) {
        const diff = b[i] - a[i];
        // only keep going if diff === 0
        if (diff)
            return diff;
        i++;
    }
    if (a.length < b.length) {
        return a.length === 1 && a[0] === 40 /* PathScore.Static */ + 40 /* PathScore.Segment */
            ? -1
            : 1;
    }else if (a.length > b.length) {
        return b.length === 1 && b[0] === 40 /* PathScore.Static */ + 40 /* PathScore.Segment */
            ? 1
            : -1;
    }
    return 0;
}

使用RouterMatcher匹配路由

在上面createRouterMatcher()中，我们知道了怎么初始化匹配路由，在这个小节中，我们将分析如何利用matcher进行路由的匹配

// 只保留matcher匹配的代码逻辑
function createRouter(options) {
    const matcher = createRouterMatcher(options.routes, options);
    //...多个方法，基本都是为下面初始化router的属性
    //比如addRoute()、removeRoute()、hasRoute()
    const router = {
        currentRoute,
        listening: true,
        addRoute,
        removeRoute,
        hasRoute,
        getRoutes,
        resolve,
        options,
        push,
        replace,
        go,
        back: () => go(-1),
        forward: () => go(1),
        beforeEach: beforeGuards.add,
        beforeResolve: beforeResolveGuards.add,
        afterEach: afterGuards.add,
        onError: errorHandlers.add,
        isReady,
        install(app) {
            //...省略，外部app.use()时调用
        },
    };
    return router;
}

push(routerHistory.location)整体概述

这个push()方法是在createRouter()创建的局部方法，不是上面useHistoryStateNavigation()创建的router.push()方法

当外部调用router.push()跳转到新的路由时，实际调用的是pushWithRedirect()，而在pushWithRedirect()中第一行代码，使用resolve(to)进行当前要跳转的路由的计算

function push(to) {
    return pushWithRedirect(to);
}
function pushWithRedirect(to, redirectedFrom) {
    const targetLocation = (pendingLocation = resolve(to));
    const from = currentRoute.value;
    const data = to.state;
    const force = to.force;
    // to could be a string where `replace` is a function
    const replace = to.replace === true;
    const shouldRedirect = handleRedirectRecord(targetLocation);
    if (shouldRedirect) {
        //...处理重定向的逻辑
        return pushWithRedirect(...)
    }
    // if it was a redirect we already called `pushWithRedirect` above
    const toLocation = targetLocation;
    toLocation.redirectedFrom = redirectedFrom;
    //...处理SameRouteLocation的情况

    // ...去除failure的处理，默认都成功
    return navigate(toLocation, from)
        .then((failure) => {
            failure = finalizeNavigation(toLocation, from, true, replace, data);
            triggerAfterEach(toLocation, from, failure);
            return failure;
        });
}

resolve(rawLocation, currentLocation)解析跳转路径

传递的rawLocation主要分为2种情况进行分析

• rawLocation为字符串，比如rawLocation="./child1"，然后调用
  parseURL()->matcher.resolve()->routerHistory.createHref()
• rawLocation为Object数据，处理它携带的path、params，然后调用matcher.resolve()
  ->routerHistory.createHref()

下面将先对parseURL()、matcher.resolve()、routerHistory.createHref()展开分析，然后再分析resolve()的整体流程

parseURL()解析路径拿到fullPath、path、query、hash

传入参数

• parseQuery()是一个方法，可以解析链接中?A=xx&B=xx的部分，返回一个key-value数据，开发者可在初始化传入自定义的解析方法
• location代表即将要跳转的路由路径
• currentLocation代表目前的路由路径

在这个方法中，我们通过"#"以及"?"拿到对应的字符串片段，塞入到hash和query字段中，将删除掉"#"和"?"的部分塞入到path字段中

function parseURL(parseQuery, location, currentLocation = '/') {
    let path, query = {}, searchString = '', hash = '';
    // Could use URL and URLSearchParams but IE 11 doesn't support it
    // TODO: move to new URL()
    const hashPos = location.indexOf('#');
    let searchPos = location.indexOf('?');
    // the hash appears before the search, so it's not part of the search string
    if (hashPos < searchPos && hashPos >= 0) {
        searchPos = -1;
    }
    if (searchPos > -1) {
        path = location.slice(0, searchPos);
        searchString = location.slice(searchPos + 1, hashPos > -1 ? hashPos : location.length);
        query = parseQuery(searchString);
    }
    if (hashPos > -1) {
        path = path || location.slice(0, hashPos);
        // keep the # character
        hash = location.slice(hashPos, location.length);
    }
    // no search and no query
    path = resolveRelativePath(path != null ? path : location, currentLocation);
    // empty path means a relative query or hash `?foo=f`, `#thing`
    return {
        fullPath: path + (searchString && '?') + searchString + hash,
        path,
        query,
        hash,
    };
}

然后触发resolveRelativePath()方法，代码逻辑也非常简单，就是判断toPosition是否是相对路径

• 如果是"/"开头，不处理直接返回
• 如果是"."开头，则代表跟fromSegments同级目录，不进行position的处理
• 如果是".."开头，代表是fromSegments的上一级目录，进行position--

最终拼接绝对路径进行返回

function resolveRelativePath(to, from) {
    if (to.startsWith('/'))
        return to;
    if (!from.startsWith('/')) {
        warn(`Cannot resolve a relative location without an absolute path. Trying to resolve "${to}" from "${from}". It should look like "/${from}".`);
        return to;
    }
    if (!to)
        return from;
    const fromSegments = from.split('/');
    const toSegments = to.split('/');
    let position = fromSegments.length - 1;
    let toPosition;
    let segment;
    for (toPosition = 0; toPosition < toSegments.length; toPosition++) {
        segment = toSegments[toPosition];
        // we stay on the same position
        if (segment === '.')
            continue;
        // go up in the from array
        if (segment === '..') {
            // we can't go below zero, but we still need to increment toPosition
            if (position > 1)
                position--;
            // continue
        }
        // we reached a non-relative path, we stop here
        else
            break;
    }
    return (fromSegments.slice(0, position).join('/') +
        '/' +
        toSegments
            // ensure we use at least the last element in the toSegments
            .slice(toPosition - (toPosition === toSegments.length ? 1 : 0))
            .join('/'));
}

最终返回绝对路径下的全路径fullPath（包括query和hash）以及对应的path、query、hash

function parseURL(parseQuery, location, currentLocation = '/') {
    //...
    return {
        fullPath: path + (searchString && '?') + searchString + hash,
        path,
        query,
        hash,
    };
}

matcher.resolve(location, currentLocation)拿到路径上所有对应的matcher数组

注意，matcher是createRouterMatcher()返回的对象，具有多个方法属性
createRouterMatcher()内部会进行addRoute()创建对应的matcherMap对象，提供给matcher.resolve()使用

// globalOptions={history, routes}
function createRouterMatcher(routes, globalOptions) {
    //...
    function resolve(location, currentLocation) {...}
    return { addRoute, resolve, removeRoute, getRoutes, getRecordMatcher };
}

matcher.resolve()方法，主要分为3个条件进行查找

function resolve(location, currentLocation) {
    let matcher;
    let params = {};
    let path;
    let name;
    if ('name' in location && location.name) {
        //...传递name查找路由
    } else if ('path' in location) {
       //...传递path查找路由
    } else {
       //...其它
    }
    const matched = [];
    let parentMatcher = matcher;
    while (parentMatcher) {
        matched.unshift(parentMatcher.record);
        parentMatcher = parentMatcher.parent;
    }
    return {
        name,
        path,
        params,
        matched,
        meta: mergeMetaFields(matched),
    };
}

function insertMatcher(matcher) {
    //...
    if (matcher.record.name && !isAliasRecord(matcher))
        matcherMap.set(matcher.record.name, matcher);
}

下面将针对上面这个代码进行分解

传递name查找对应的路由

router.push({ name: 'user', params: { username: 'eduardo' } })

直接根据name找到对应的matcher，然后进行params的合并，这里直接利用location.params覆盖currentLocation.params重复的key

paramsFromLocation(): 筛选第一个传入Object，筛选出key存在于第二个参数，即第二个参数存在的key，才能保留下来第一个参数对应的key-value

matcher = matcherMap.get(location.name);
name = matcher.record.name;
params = assign(
    paramsFromLocation(currentLocation.params,
        matcher.keys.filter(k => !k.optional).map(k => k.name)),
    location.params &&
    paramsFromLocation(location.params, matcher.keys.map(k => k.name)));
path = matcher.stringify(params);

然后根据params进行对应路径的拼接，如果是静态路径，直接拼接静态路径的值，如果是动态路径，则拼接传递的params值

如果动态匹配路由没有传递对应的参数，并且是不可选optional=false，则会报错

function stringify(params) {
    let path = '';
    let avoidDuplicatedSlash = false;
    for (const segment of segments) {
        if (!avoidDuplicatedSlash || !path.endsWith('/'))
            path += '/';
        avoidDuplicatedSlash = false;
        for (const token of segment) {
            if (token.type === 0 /* TokenType.Static */) {
                path += token.value;
            } else if (token.type === 1 /* TokenType.Param */) {
                const { value, repeatable, optional } = token;
                const param = value in params ? params[value] : '';
                const text = isArray(param)
                    ? param.join('/')
                    : param;
                if (!text) {
                    if (optional) {
                        // 可选条件下，如果path.endsWith('/')，则去掉最后面的'/'
                    } else { throw new Error(`Missing required param "${value}"`); }
                }
                path += text;
            }
        }
    }
    return path || '/';
}

传递path查找对应的路由

通过正则表达式匹配路径找到对应的matcher，通过matcher.parse()解析路径，拿到路由动态匹配的字符串
路由配置为: /:child1->router.push("/test")->params: ["child1": ["test"]]

path = location.path;
matcher = matchers.find(m => m.re.test(path));
if (matcher) {
    //params: ["child1": ["test"]]
    params = matcher.parse(path);
    name = matcher.record.name;
}

没有传递name也没有传递path

获取当前路由的matcher，然后合并传递的params，使用params形成新的路径

matcher = currentLocation.name
    ? matcherMap.get(currentLocation.name)
    : matchers.find(m => m.re.test(currentLocation.path));
name = matcher.record.name;
params = assign({}, currentLocation.params, location.params);
path = matcher.stringify(params);

---

寻找到路由后，遍历这条路由所有segment，拿到所有的matcher，涉及路径上的多个Component

我们加载子Component的同时也会加载父Component

const matched = [];
let parentMatcher = matcher;
while (parentMatcher) {
    matched.unshift(parentMatcher.record);
    parentMatcher = parentMatcher.parent;
}
return {
    name,
    path,
    params,
    matched,
    meta: mergeMetaFields(matched),
};

routerHistory.createHref()

将base(除去域名后的剩余路径)，使用正则表达式替换为"#"，比如"/router/vue-router-path-ranker.html?a=1#test#xixi"->"#test#xixi"

此时location为目前的路径，比如/one/two/three，那么组合起来就是"#test#xixi/one/two/three"

一般来说应该是"#/one/two/three"，不会出现两个#

const href = routerHistory.createHref(fullPath);

const BEFORE_HASH_RE = /^[^#]+#/;
function createHref(base, location) {
    return base.replace(BEFORE_HASH_RE, '#') + location;
}

resolve()详细分析

parseURL(): 解析得到fullPath、path、query、hash
matcher.resolve(): 解析得到matchedRouted，初始化传入routes得到的数组对象，包含该路由所映射的组件，对应的参数以及一系列的路由守卫，具体的数据结构如下面代码块所示
routerHistory.createHref(): 解析得到href

// matcher.resolve()
function resolve(location, currentLocation) {
    return {
        name,
        path,
        params,
        matched,
        meta: mergeMetaFields(matched),
    };
}
// matched: RouteRecordNormalized[]
interface RouteRecordNormalized {
    path: _RouteRecordBase['path']
    redirect: _RouteRecordBase['redirect'] | undefined
    name: _RouteRecordBase['name']
    components: RouteRecordMultipleViews['components'] | null | undefined
    children: RouteRecordRaw[]
    meta: Exclude<_RouteRecordBase['meta'], void>
    props: Record
    beforeEnter: _RouteRecordBase['beforeEnter']
    leaveGuards: Set
    updateGuards: Set
    enterCallbacks: Record
    instances: Record
    aliasOf: RouteRecordNormalized | undefined
}

从上面的分析，我们可以知道parseURL()、matcher.resolve()、routerHistory.createHref()的返回值，现在我们可以对resolve()方法进行具体的分析

function resolve(rawLocation, currentLocation) {
    currentLocation = assign({}, currentLocation || currentRoute.value);
    if (typeof rawLocation === 'string') {
        const locationNormalized = parseURL(parseQuery$1, rawLocation, currentLocation.path);
        const matchedRoute = matcher.resolve({ path: locationNormalized.path }, currentLocation);
        const href = routerHistory.createHref(locationNormalized.fullPath);
        return assign(locationNormalized, matchedRoute, {
            params: decodeParams(matchedRoute.params),
            hash: decode(locationNormalized.hash),
            redirectedFrom: undefined,
            href,
        });
    } else {
        let matcherLocation;
        if ('path' in rawLocation) {
            matcherLocation = assign({}, rawLocation, {
                path: parseURL(parseQuery$1, rawLocation.path, currentLocation.path).path,
            });
        } else {
            const targetParams = assign({}, rawLocation.params);
            for (const key in targetParams) {
                if (targetParams[key] == null) {
                    delete targetParams[key];
                }
            }
            matcherLocation = assign({}, rawLocation, {
                params: encodeParams(rawLocation.params),
            });
            currentLocation.params = encodeParams(currentLocation.params);
        }
        const matchedRoute = matcher.resolve(matcherLocation, currentLocation);
        const hash = rawLocation.hash || '';
        matchedRoute.params = normalizeParams(decodeParams(matchedRoute.params));
        const fullPath = stringifyURL(stringifyQuery$1, assign({}, rawLocation, {
            hash: encodeHash(hash),
            path: matchedRoute.path,
        }));
        const href = routerHistory.createHref(fullPath);
        return assign({
            fullPath,
            hash,
            query: stringifyQuery$1 === stringifyQuery
                ? normalizeQuery(rawLocation.query)
                : (rawLocation.query || {}),
        }, matchedRoute, {
            redirectedFrom: undefined,
            href,
        });
    }
}

parseURL(): 解析得到fullPath、path、query、hash

matcher.resolve(): 解析得到matchedRouted，初始化传入routes得到的Array，包含该路由所映射的组件，对应的参数以及一系列的路由守卫

routerHistory.createHref(): 解析得到href

typeof rawLocation === 'string'时，字符串代表路由path，通过matcher.resolve({path: xxx})获取对应的路由对象

const locationNormalized = parseURL(parseQuery$1, rawLocation, currentLocation.path);
const matchedRoute = matcher.resolve({ path: locationNormalized.path }, currentLocation);
const href = routerHistory.createHref(locationNormalized.fullPath);
// locationNormalized is always a new object
return assign(locationNormalized, matchedRoute, {
    params: decodeParams(matchedRoute.params),
    hash: decode(locationNormalized.hash),
    redirectedFrom: undefined,
    href,
});

typeof rawLocation !== 'string'时，也是同样的逻辑

• 如果有path属性，则跟上面的逻辑一致，通过matcher.resolve({path: xxx})获取对应的路由对象
• 如果没有path属性，处理params数据（从path中提取的已解码参数字典），然后再通过matcher.resolve({path: xxx})获取对应的路由对象

fullPath: path + (searchString && '?') + searchString + hash

let matcherLocation;
//======== 第1部分：整理参数 ========
if ('path' in rawLocation) {
    matcherLocation = assign({}, rawLocation, {
        path: parseURL(parseQuery$1, rawLocation.path, currentLocation.path).path,
    });
} else {
    //...
    matcherLocation = assign({}, rawLocation, {
        params: encodeParams(rawLocation.params),
    });
    currentLocation.params = encodeParams(currentLocation.params);
}
//======== 第2部分：matcher.resolve ========
const matchedRoute = matcher.resolve(matcherLocation, currentLocation);
const hash = rawLocation.hash || '';
matchedRoute.params = normalizeParams(decodeParams(matchedRoute.params));
const fullPath = stringifyURL(stringifyQuery$1, assign({}, rawLocation, {
    hash: encodeHash(hash),
    path: matchedRoute.path,
}));
//======== 第3部分：routerHistory.createHref ========
const href = routerHistory.createHref(fullPath);
// ======== 第4部分：返回对象数据 ========
return assign({
    fullPath,
    hash,
    query: stringifyQuery$1 === stringifyQuery
        ? normalizeQuery(rawLocation.query)
        : (rawLocation.query || {}),
}, matchedRoute, {
    redirectedFrom: undefined,
    href,
});

resolve()最终返回值对象也就是下文中targetLocation的值和toLocation的值

---

通过讲解上面的一系列方法，现在我们可以开始解析push()->pushWithRedirect()

pushWithRedirect()真正push逻辑

我们通过resolve(to)拿到了目前匹配的路径对象，然后处理重定向的逻辑，然后相同Route对象逻辑，然后触发navigate()方法

function pushWithRedirect(to, redirectedFrom) {
    const targetLocation = (pendingLocation = resolve(to));
    const from = currentRoute.value;
    const data = to.state;
    const force = to.force;
    // to could be a string where `replace` is a function
    const replace = to.replace === true;
    const shouldRedirect = handleRedirectRecord(targetLocation);
    if (shouldRedirect) {
        //...处理重定向的逻辑
        return pushWithRedirect(...)
    }
    // if it was a redirect we already called `pushWithRedirect` above
    const toLocation = targetLocation;
    toLocation.redirectedFrom = redirectedFrom;
    //...处理SameRouteLocation的情况

    // ...去除failure的处理，默认都成功
    return navigate(toLocation, from)
        .then((failure) => {
            failure = finalizeNavigation(toLocation, from, true, replace, data);
            triggerAfterEach(toLocation, from, failure);
            return failure;
        });
}

navigate(toLocation, from)

function navigate(to, from) {
    let guards;
    // 将旧的路由的beforerRouteLeave守卫函数放入guards
  
    return (runGuardQueue(guards)// 先执行beforerRouteLeave守卫函数
        .then(() => {
            // 处理全局的beforeEach守卫函数
        })
        .then(() => {
            // 处理该路由的beforeRouteUpdate守卫函数
        })
        .then(() => {
            // 处理该路由的beforeEnter守卫函数
        }).then(() => {
            // 处理该路由的beforeRouteEnter守卫函数
        }).then(() => {
            // 处理全局的beforeResolve守卫函数
        })
        // catch any navigation canceled
        .catch(err => isNavigationFailure(err, 8 /* ErrorTypes.NAVIGATION_CANCELLED */)
            ? err
            : Promise.reject(err)));
}
function runGuardQueue(guards) {
    return guards.reduce((promise, guard) => promise.then(() => guard()), Promise.resolve());
}

路由切换失败情况分析

除了执行一系列的导航守卫需要关注外，我们还需要关注下error发生时的处理情况，主要的错误情况分为

export const enum ErrorTypes {
  // they must be literals to be used as values, so we can't write
  // 1 << 2
  MATCHER_NOT_FOUND = 1,
  NAVIGATION_GUARD_REDIRECT = 2,
  NAVIGATION_ABORTED = 4,
  NAVIGATION_CANCELLED = 8,
  NAVIGATION_DUPLICATED = 16,
}

ErrorTypes.MATCHER_NOT_FOUND

当使用matcher.solve()寻找对应的matched数据时，如果传入的参数是路由的name，但是我们却无法根据name找到对应的matcher时，我们返回ErrorTypes.MATCHER_NOT_FOUND的错误

因为对于一个路由，name是唯一的标识，如果传入name，会根据matcherMap去找对应存储过的matcher

function createRouterMatcher(routes, globalOptions) {
    function resolve(location, currentLocation) {
        let matcher;
        let params = {};
        let path;
        let name;
        if ('name' in location && location.name) {
            matcher = matcherMap.get(location.name);
            if (!matcher)
                throw createRouterError(1 /* ErrorTypes.MATCHER_NOT_FOUND */, {
                    location,
                });
            //...
        }

        return {
            name,
            path,
            params,
            matched,
            meta: mergeMetaFields(matched),
        };
    }
}

ErrorTypes.NAVIGATION_ABORTED

如下面代码块所示，我们导航守卫会传递对应的next()提供给外部使用，比如开发者在外部的next()返回false，则触发NAVIGATION_ABORTED，表示取消导航

ErrorTypes.NAVIGATION_GUARD_REDIRECT

如下面代码块所示，我们导航守卫会传递对应的next()提供给外部使用，比如开发者在外部的next()返回"/login"或者{"name": "login"}，则isRouteLocation=true，从而返回NAVIGATION_GUARD_REDIRECT，表示导航重定向到其它路由

// guards.push(guardToPromiseFn(guard, to, from));
function guardToPromiseFn(guard, to, from, record, name) {
    return () => new Promise((resolve, reject) => {
        const next = (valid) => {
            if (valid === false) {
                reject(createRouterError(4 /* ErrorTypes.NAVIGATION_ABORTED */, {
                    from,
                    to,
                }));
            } else if (valid instanceof Error) {
                reject(valid);
            } else if (isRouteLocation(valid)) {
                reject(createRouterError(2 /* ErrorTypes.NAVIGATION_GUARD_REDIRECT */, {
                    from: to,
                    to: valid,
                }));
            } else {
                resolve();
            }
        };
    }
}

ErrorTypes.NAVIGATION_CANCELLED

如下面代码块所示，当使用push()->pushWithRedirect()->navigate()->finalizeNavigation()时，会进行checkCanceledNavigation()的检测，如果当前要跳转的路由跟pushWithRedirect()的路由不同时，说明又有新的导航已经发生，之前的导航取消

ErrorTypes.NAVIGATION_DUPLICATED

如下面代码块所示，如果没有使用force，当检测到相同路由时，会产生NAVIGATION_DUPLICATED错误，阻止继续调用navigate()->finalizeNavigation()

isSameRouteLocation: matched、params、query、hash都相同

function pushWithRedirect(to, redirectedFrom) {
    const targetLocation = (pendingLocation = resolve(to));
    if (!force && isSameRouteLocation(stringifyQuery$1, from, targetLocation)) {
        failure = createRouterError(16 /* ErrorTypes.NAVIGATION_DUPLICATED */, { to: toLocation, from });
    }
      if(failure) return Promise.resolve(failure);
    return navigate(toLocation, from).then((failure) => {
        if (!failure) {
            failure = finalizeNavigation(toLocation, from, true, replace, data);
        }
    });
}
function finalizeNavigation(toLocation, from, isPush, replace, data) {
    // a more recent navigation took place
    const error = checkCanceledNavigation(toLocation, from);
    if (error)
        return error;
}
function checkCanceledNavigation(to, from) {
    if (pendingLocation !== to) {
        return createRouterError(8 /* ErrorTypes.NAVIGATION_CANCELLED */, {
            from,
            to,
        });
    }
}

finalizeNavigation()

从下面代码块可以知道，finalizeNavigation()的步骤为：

• 触发routerHistory.replace/routerHistory.push更新
• 更新currentRoute.value为目前的路由路径
• 检测是否是初始化阶段，如果是初始化阶段，则触发setupListeners()方法

function finalizeNavigation(toLocation, from, isPush, replace, data) {
    //...处理错误
    if (isPush) {
        if (replace || isFirstNavigation)
            routerHistory.replace(toLocation.fullPath, assign({
                scroll: isFirstNavigation && state && state.scroll,
            }, data));
        else
            routerHistory.push(toLocation.fullPath, data);
    }
    // accept current navigation
    currentRoute.value = toLocation;
    handleScroll(toLocation, from, isPush, isFirstNavigation);
    markAsReady();
}
function markAsReady(err) {
    if (!ready) {
        // still not ready if an error happened
        ready = !err;
        setupListeners();
        readyHandlers
            .list()
            .forEach(([resolve, reject]) => (err ? reject(err) : resolve()));
        readyHandlers.reset();
    }
    return err;
}

setupListeners()注册pop操作相关监听方法

初始化会进行一次push()操作，此时就是初始化阶段

!ready，push()->navigate()->finalizeNavigation()

初始化阶段会进行routerHistory.listen()的方法注册

function finalizeNavigation(toLocation, from, isPush, replace, data) {
    markAsReady();
}
function markAsReady(err) {
    if (!ready) {
        ready = !err;
        setupListeners();
    }
    return err;
}
function setupListeners() {
    // avoid setting up listeners twice due to an invalid first navigation
    if (removeHistoryListener)
        return;
    removeHistoryListener = routerHistory.listen((to, _from, info) => {
        if (!router.listening)
            return;
        // cannot be a redirect route because it was in history
        const toLocation = resolve(to);
        //...处理重定向的逻辑

        pendingLocation = toLocation;
        const from = currentRoute.value;
        // TODO: should be moved to web history?
        if (isBrowser) {
            saveScrollPosition(getScrollKey(from.fullPath, info.delta), computeScrollPosition());
        }
        //...去除错误处理
        navigate(toLocation, from)
            .then((failure) => {
                finalizeNavigation(
                    // after navigation, all matched components are resolved
                    toLocation, from, false);
                triggerAfterEach(toLocation, from, failure);
            })
            .catch(noop);
    });
}

那么初始化就注册的监听在什么时候会触发呢？这个监听又有什么作用呢？

在上面我们说到useHistoryListeners()初始化的时候，我们提供了listen()方法进行事件的注册，然后在popStateHandler()触发时，进行listeners注册方法的调用

因为我们可以知道，这个监听事件本质就是为了在用户进行浏览器后退按钮点击时，能够正常监听到路由变化并且自动实现Component切换

function useHistoryListeners(base, historyState, currentLocation, replace) {
    let listeners = [];
    let teardowns = [];
    // TODO: should it be a stack? a Dict. Check if the popstate listener
    // can trigger twice
    let pauseState = null;
    const popStateHandler = ({ state, }) => {...};
    function pauseListeners() {...}
    function listen(callback) {...}
    function beforeUnloadListener() {...}
    function destroy() {...}
    
    // set up the listeners and prepare teardown callbacks
    window.addEventListener('popstate', popStateHandler);
    window.addEventListener('beforeunload', beforeUnloadListener);
    return {
        pauseListeners,
        listen,
        destroy,
    };
}
const popStateHandler = ({ state, }) => {
    //...
    listeners.forEach(listener => {
        listener(currentLocation.value, from, {
            delta,
            type: NavigationType.pop,
            direction: delta
                ? delta > 0
                    ? NavigationDirection.forward
                    : NavigationDirection.back
                : NavigationDirection.unknown,
        });
    });
};

app.use(router)使用VueRouter

use(plugin, ...options) {
    if (plugin && isFunction(plugin.install)) {
        installedPlugins.add(plugin);
        plugin.install(app, ...options);
    }
    else if (isFunction(plugin)) {
        installedPlugins.add(plugin);
        plugin(app, ...options);
    }
    return app;
}

router.install(app)

从上面Vue3的源码可以知道，最终会触发Vue Router的install()方法

const START_LOCATION_NORMALIZED = {
    path: '/',
    name: undefined,
    params: {},
    query: {},
    hash: '',
    fullPath: '/',
    matched: [],
    meta: {},
    redirectedFrom: undefined,
};
const currentRoute = vue.shallowRef(START_LOCATION_NORMALIZED);
const router = {
    //....
    addRoute,
    removeRoute,
    push,
    replace,
    beforeEach: beforeGuards.add,
    isReady,
    install(app) {
        //...省略，外部app.use()时调用
    },
};

由于篇幅原因，接下来的分析请看下一篇文章Vue3相关源码-Vue Router源码解析(二)

参考文章

1. 7张图，从零实现一个简易版Vue-Router，太通俗易懂了！
2. VueRouter4路由权重

Vue系列其它文章

1. Vue2源码-响应式原理浅析
2. Vue2源码-整体流程浅析
3. Vue2源码-双端比较diff算法 patchVNode流程浅析
4. Vue3源码-响应式系统-依赖收集和派发更新流程浅析
5. Vue3源码-响应式系统-Object、Array数据响应式总结
6. Vue3源码-响应式系统-Set、Map数据响应式总结
7. Vue3源码-响应式系统-ref、shallow、readonly相关浅析
8. Vue3源码-整体流程浅析
9. Vue3源码-diff算法-patchKeyChildren流程浅析