Koa是Node应用广泛的后端框架,它的“洋葱”模型被人津津乐道,Koa源码实现也非常简练,那就一起来看下Koa的源码吧。
工程化
# .editorconfig
# editorconfig.org
root = true
[*]
indent_style = space
indent_size = 2
end_of_line = lf
charset = utf-8
trim_trailing_whitespace = true
insert_final_newline = true
[*.md]
trim_trailing_whitespace = false
[Makefile]
indent_style = tab
# .travis.yml
language: node_js
node_js:
- 8
- 10
- 12
cache:
directories:
- wrk/bin
- node_modules
before_script:
- npm prune
- "[ ! -f wrk/bin/wrk ] && rm -rf wrk && git clone https://github.com/wg/wrk.git && make -C wrk && mkdir wrk/bin && mv wrk/wrk wrk/bin || true"
- export PATH=$PATH:$PWD/wrk/bin/
script:
- npm run lint
- npm run test-cov
- npm run bench
after_script:
# only upload the coverage.json file
- bash <(curl -s https://codecov.io/bash) -f coverage/coverage-final.json
npm prune会移除无关的包,所谓无关指的是没有在父包的依赖关系列表中列出的包。参考npm prune
wrk是一个http测试工具
wrk -t12 -c400 -d30s http://127.0.0.1:8080/index.html
意思就是看看30s,用12个线程,开400个http连接能处理多少个http请求。
再看package.json
{
"name": "koa",
"version": "2.12.0",
"description": "Koa web app framework",
"main": "lib/application.js",
"scripts": {
"test": "egg-bin test test",
"test-cov": "egg-bin cov test",
"lint": "eslint benchmarks lib test",
"bench": "make -C benchmarks",
"authors": "git log --format='%aN <%aE>' | sort -u > AUTHORS"
},
"repository": "koajs/koa",
"keywords": [
"web",
"app",
"http",
"application",
"framework",
"middleware",
"rack"
],
"license": "MIT",
"dependencies": {
"accepts": "^1.3.5",
"cache-content-type": "^1.0.0",
"content-disposition": "~0.5.2",
"content-type": "^1.0.4",
"cookies": "~0.8.0",
"debug": "~3.1.0",
"delegates": "^1.0.0",
"depd": "^1.1.2",
"destroy": "^1.0.4",
"encodeurl": "^1.0.2",
"escape-html": "^1.0.3",
"fresh": "~0.5.2",
"http-assert": "^1.3.0",
"http-errors": "^1.6.3",
"is-generator-function": "^1.0.7",
"koa-compose": "^4.1.0",
"koa-convert": "^1.2.0",
"on-finished": "^2.3.0",
"only": "~0.0.2",
"parseurl": "^1.3.2",
"statuses": "^1.5.0",
"type-is": "^1.6.16",
"vary": "^1.1.2"
},
"devDependencies": {
"egg-bin": "^4.13.0",
"eslint": "^6.5.1",
"eslint-config-koa": "^2.0.0",
"eslint-config-standard": "^14.1.0",
"eslint-plugin-import": "^2.18.2",
"eslint-plugin-node": "^10.0.0",
"eslint-plugin-promise": "^4.2.1",
"eslint-plugin-standard": "^4.0.1",
"mm": "^2.5.0",
"supertest": "^3.1.0"
},
"engines": {
"node": "^4.8.4 || ^6.10.1 || ^7.10.1 || >= 8.1.4"
},
"files": [
"lib"
]
}
这里需要留意的是有一个authors
命令, git log --format='%aN <%aE>' | sort -u > AUTHORS
这条命令是把git 提交记录的姓名和邮件排序输出到AUTHORS文件。
核心代码
入口
先看构造函数
module.exports = class Application extends Emitter {
/**
* Initialize a new `Application`.
*
* @api public
*/
/**
*
* @param {object} [options] Application options
* @param {string} [options.env='development'] Environment
* @param {string[]} [options.keys] Signed cookie keys
* @param {boolean} [options.proxy] Trust proxy headers
* @param {number} [options.subdomainOffset] Subdomain offset
* @param {boolean} [options.proxyIpHeader] proxy ip header, default to X-Forwarded-For
* @param {boolean} [options.maxIpsCount] max ips read from proxy ip header, default to 0 (means infinity)
*
*/
constructor(options) {
super();
options = options || {};
this.proxy = options.proxy || false;
this.subdomainOffset = options.subdomainOffset || 2;
this.proxyIpHeader = options.proxyIpHeader || 'X-Forwarded-For';
this.maxIpsCount = options.maxIpsCount || 0;
this.env = options.env || process.env.NODE_ENV || 'development';
if (options.keys) this.keys = options.keys;
this.middleware = [];
this.context = Object.create(context);
this.request = Object.create(request);
this.response = Object.create(response);
if (util.inspect.custom) {
this[util.inspect.custom] = this.inspect;
}
}
}
就是一些初始化流程,把options的一些选项挂在到this上。这里有一个要留意的是代理相关的配置。
listen(...args) {
debug('debug')
const server = http.createServer(this.callback())
return server.listen(...args)
}
这里http.createServer接收的是一个function(req, res) {}
函数,这里调用了calllback生成这样一个函数。这样的设计使得可以将callback生成的处理器放到任何能够接受监听器函数中用于启动一个服务,比如说http.createServer或者express。
callback() {
const fn = compose(this.middleware)
if (!this.listenerCount('error)) this.on('error', this.onerror)
const handleRequest = (req, res) => {
const ctx = this.createContext(req, res)
return this.handleRequest(ctx, fn)
}
}
这里主要就是构造出一个上下文对象,然后在handleRequest里去处理这个请求。
先看下创建上下文的函数createContext
吧
createContext(req, res) {
const context = Object.create(this.context);
const request = context.request = Object.create(this.request);
const response = context.response = Object.create(this.response);
context.app = request.app = response.app = this;
context.req = request.req = response.req = req;
context.res = request.res = response.res = res;
request.ctx = response.ctx = context;
request.response = response;
response.request = request;
context.originalUrl = request.originalUrl = req.url;
context.state = {};
return context;
}
前面构造函数中对this.context
, this.request
, this.response
进行了初始化
this.context = Object.create(context);
this.request = Object.create(request);
this.response = Object.create(response);
createContext
做的就是给context变量里的req,res,ctx进行赋值,吧res和req不仅挂到了context上,还挂到了contex.request
, context.response
上。
接下来看一下koa为人津津乐道的插件机制。在koa github主页上有这么一段话介绍koa
- Expressive HTTP middleware framework for node.js to make web applications and APIs more enjoyable to write. Koa’s middleware stack flows in a stack-like manner, allowing you to perform actions downstream then filter and manipulate the response upstream
koa提供了一个web框架的核心部分,不包含路由处理,静态文件处理(这些都是通过中间件来实现),我们经常会把这种叫做洋葱模型。
实际使用中使用use来给koa应用注册中间件,中间件按照注册顺序按序调用,前面的中间件可以调用next()来控制下一个中间件的执行。
const Koa = require('koa');
const app = new Koa();
// logger
app.use(async (ctx, next) => {
await next();
const rt = ctx.response.get('X-Response-Time');
console.log(`${ctx.method} ${ctx.url} - ${rt}`);
});
// x-response-time
app.use(async (ctx, next) => {
const start = Date.now();
await next();
const ms = Date.now() - start;
ctx.set('X-Response-Time', `${ms}ms`);
});
// response
app.use(async ctx => {
ctx.body = 'Hello World';
});
app.listen(3000);
通过中间件可以做到调用下游,控制流回上游
那这个是怎么实现的呢?先来看use的实现
use(fn) {
if (typeof fn !== 'function') throw new TypeError('middleware must be a function!');
if (isGeneratorFunction(fn)) {
deprecate('Support for generators will be removed in v3. ' +
'See the documentation for examples of how to convert old middleware ' +
'https://github.com/koajs/koa/blob/master/docs/migration.md');
fn = convert(fn);
}
debug('use %s', fn._name || fn.name || '-');
this.middleware.push(fn);
return this;
}
use函数实现非常简单,就是把注册进来中间件函数(context, next) ⇒ {}
存放到实例的middleware里面。那么这个middleware是怎么起作用的呢?
细心的你可能在之前的callback
函数中发现有这样一行
const fn = compose(this.middleware)
这行代码就是实现中间件的精髓。compose将所有的中间件组合成一个函数。
先不看官方的实现,我们可以自己实现一个这样的compose函数
const compose = (middlewares) => (ctx) => {
const dispatch = (i) => {
const fn = middlewares[i]
// 最后一个中间件调用next不能报错
if (!fn) {
return Promise.resolve()
}
return Promise.resolve(fn(ctx, () => dispatch(i + 1)))
}
dispatch(0)
}
可以看到,compose生成了一个新的函数,在这个函数中我们把ctx传入进去,所有的中间件都会执行,执行的顺序通过给中间件函数传递的第二个参数来控制,也就是说把在第n+1个中间件函数的执行权交给第n个中间件函数。
koa实际是使用了koa/compose。这里一并把代码贴出来
module.exports = compose
/**
* Compose `middleware` returning
* a fully valid middleware comprised
* of all those which are passed.
*
* @param {Array} middleware
* @return {Function}
* @api public
*/
function compose (middleware) {
if (!Array.isArray(middleware)) throw new TypeError('Middleware stack must be an array!')
for (const fn of middleware) {
if (typeof fn !== 'function') throw new TypeError('Middleware must be composed of functions!')
}
/**
* @param {Object} context
* @return {Promise}
* @api public
*/
return function (context, next) {
// last called middleware #
let index = -1
return dispatch(0)
function dispatch (i) {
if (i <= index) return Promise.reject(new Error('next() called multiple times'))
index = i
let fn = middleware[i]
if (i === middleware.length) fn = next
if (!fn) return Promise.resolve()
try {
return Promise.resolve(fn(context, dispatch.bind(null, i + 1)));
} catch (err) {
return Promise.reject(err)
}
}
}
}
相对于简陋版本的compose来说,这里的compose考虑了不能在一个中间件中调用多次next,还处理了最后一个中间件不调用next的情况。
Koa构造函数还有两个函数,toJSON, inspect,这俩函数就是返回指定的几个字段。
/**
* Return JSON representation.
* We only bother showing settings.
*
* @return {Object}
* @api public
*/
toJSON() {
return only(this, [
'subdomainOffset',
'proxy',
'env'
]);
}
/**
* Inspect implementation.
*
* @return {Object}
* @api public
*/
inspect() {
return this.toJSON();
}
再看下处理请求的部分,处理请求包括错误处理,数据响应。这里有一个onFinished(res, onerror)是在响应结束或者关闭或者错误的时候执行一下onerror方法。on-finished
handleRequest(ctx, fnMiddleware) {
const res = ctx.res
res.statusCode = 404
const onerror = err => ctx.onerror(err)
const handleResponse = () => respond(ctx)
onFinished(res, onerror)
return fnMiddleware(ctx).then(handleResponse).catch(onerror)
}
function respond(ctx) {
// allow bypassing koa
if (false === ctx.respond) return
if (!ctx.writable) return
const res = ctx.res
let body = ctx.body
const code = ctx.status;
// statuses.empty(code) returns true if a status code expects an empty body
if (statuses.empty[code]) {
// strip headers
ctx.body = null
return res.end()
}
if ('HEAD' === ctx.method) {
if (!res.headersSent && !ctx.response.has('Content-Length')) {
const { length } = ctx
if (Number.isInteger(length)) ctx.length = length
}
return res.end()
}
// status body
if (null === body) {
if (ctx.reponse._expliciNullBody) {
ctx.response.remove('Content-Type')
ctx.response.remove('Transfer-Encoding')
return res.end()
}
if (ctx.req.httpVersionMajaor >= 2) {
body = String(code)
} else {
body = ctx.message || String(code)
}
if (!res.headersSend) {
ctx.type = 'text'
ctx.length = Buffer.byteLength(body)
}
return res.end(body)
}
// responses
if (Buffer.isBuffer(body)) return res.end(body)
if ('string' === typeof body) return res.end(body)
if (body instanceof Stream) return body.pipe(res)
// body: json
body = JSON.stringify(body)
if (!res.headersSent) {
ctx.length = Buffer.byteLength(body)
}
res.end(body)
}
这里处理了不同类型的body,这里可以看到如果res.headersSent
为false的话,会写一些header。
context, request, response
context
const COOKIES = Symbol('context#cookies')
const proto = module.exports = {
inspect() {
if (this === proto) return this
return this.toJSON()
},
// 返回对象的所有json表示
toJSON() {
return {
request: this.request.toJSON(),
response: this.response.toJSON(),
app: this.app.toJSON(),
originalUrl: this.originalUrl,
req: '<original node req>',
res: '<original node res>',
socket: '<original node socket>'
}
},
// assertion
assert: httpAssert,
// throw error
throw(...arg) {
throw createError(...args)
},
// default error handling
onerror(err) {
//
},
// cookies
get cookies() {
if (!this[COOKIES]) {
this[COOKIES] = new Cookie(this.req, this.res, {
keys: this.app.keys,
secure: this.request.secure
})
}
},
set cookies(_cookies) {
this[COOKIES] = _cookies
}
}
/**
* Response delegation.
*/
delegate(proto, 'response')
.method('attachment')
.method('redirect')
.method('remove')
.method('vary')
.method('has')
.method('set')
.method('append')
.method('flushHeaders')
.access('status')
.access('message')
.access('body')
.access('length')
.access('type')
.access('lastModified')
.access('etag')
.getter('headerSent')
.getter('writable');
/**
* Request delegation.
*/
delegate(proto, 'request')
.method('acceptsLanguages')
.method('acceptsEncodings')
.method('acceptsCharsets')
.method('accepts')
.method('get')
.method('is')
.access('querystring')
.access('idempotent')
.access('socket')
.access('search')
.access('method')
.access('query')
.access('path')
.access('url')
.access('accept')
.getter('origin')
.getter('href')
.getter('subdomains')
.getter('protocol')
.getter('host')
.getter('hostname')
.getter('URL')
.getter('header')
.getter('headers')
.getter('secure')
.getter('stale')
.getter('fresh')
.getter('ips')
.getter('ip');
context里写了很多setter,getter,并且使用delegate把request,response的东西挂到委托到context上。
request
const IP = Symbol('context#ip')
module.exports = {
get header() {
return this.req.headers
},
set header(val) {
this.req.headers = val
},
get url() {
return this.req.url
},
set url(val) {
this.req.url = val
},
get method() {
return this.req.method
},
set method(val) {
this.req.method = val
},
get query() {
const str = this.querystring;
const c = this._querycache = this._querycache || {};
return c[str] || (c[str] = qs.parse(str))
},
set query(obj) {
this.querystring = q.stringify(obj)
},
get querystring() {
if (!this.req) return ''
return parse(this.req).query || ''
},
set querystring() {
const url = parse(this.req)
if (url.search === `?${str}`) return;
url.search = str
url.patch = null
this.url = stringify(url)
},
get socket() {
return this.socket
},
get length() {
const len = this.get('Content-Length')
if (len === '') return
return ~~len
},
/**
* Return request's remote address
* When `app.proxy` is `true`, parse
* the "X-Forwarded-For" ip address list and return the first one
*
* @return {String}
* @api public
*/
get ip() {
if (!this[IP]) {
this[IP] = this.ips[0] || this.socket.remoteAddress || '';
}
return this[IP];
},
set ip(_ip) {
this[IP] = _ip;
},
get(field) {
const req = this.req;
switch (field = field.toLowerCase()) {
case 'referer':
case 'referrer':
return req.headers.referrer || req.headers.referer || '';
default:
return req.headers[field] || '';
}
},
/**
* Return JSON representation.
*
* @return {Object}
* @api public
*/
toJSON() {
return only(this, [
'method',
'url',
'header'
]);
}
}
可以看到这里也都是一些setter,getter,还有一些封装的方法比如get。
不过从代码中可以学习到一些编码技巧。
const c = this._querycache = this._querycache || {}; // 存只避免多次访问
return ~~len // 位运算,这里相当于parseInt
位运算参考文章 位运算符在JS中的妙用
上面代码还有一个比较意思的是,cookies和ip的访问都是用了Symbol。原因参考
response
与request 类似,不再贴代码。
测试
在package.json里script中
"scripts": {
"test": "egg-bin test test",
"test-cov": "egg-bin cov test",
"lint": "eslint benchmarks lib test",
"bench": "make -C benchmarks",
"authors": "git log --format='%aN <%aE>' | sort -u > AUTHORS"
},
可以看到这里功能测试使用的是egg-bin(mocha)
基准测试
make -C benchmarks
,首先可以了解一下make。-C是改变Makefile的读取目录,这里意思就是去benchmark目录下找Makefile文件。
// benchmark/Makefile
all: middleware
middleware:
@./run 1 false $@
@./run 5 false $@
@./run 10 false $@
@./run 15 false $@
@./run 20 false $@
@./run 30 false $@
@./run 50 false $@
@./run 100 false $@
@./run 1 true $@
@./run 5 true $@
@./run 10 true $@
@./run 15 true $@
@./run 20 true $@
@./run 30 true $@
@./run 50 true $@
@./run 100 true $@
@echo
.PHONY: all middleware
// benchmark/middleware.js
'use strict';
const Koa = require('..');
const app = new Koa();
// number of middleware
let n = parseInt(process.env.MW || '1', 10);
const useAsync = process.env.USE_ASYNC === 'true';
console.log(` ${n}${useAsync ? ' async' : ''} middleware`);
while (n--) {
if (useAsync) {
app.use(async(ctx, next) => next());
} else {
app.use((ctx, next) => next());
}
}
const body = Buffer.from('Hello World');
if (useAsync) {
app.use(async(ctx, next) => { await next(); ctx.body = body; });
} else {
app.use((ctx, next) => next().then(() => ctx.body = body));
}
app.listen(3333);
功能测试
测试代码就不看了,有兴趣自己去github上看看。