[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8},"actionData","errors","root",{"_11":-5},"routes/user.$name.replies",{"_9":10,"_11":12,"_13":14,"_15":16,"_17":18,"_19":20},"replies",[36,37],"user",{"_21":22,"_13":14,"_23":24,"_25":26,"_27":28,"_29":30,"_31":32,"_33":16,"_34":35},"loginname","mad","page",1,"limit",50,"total",2,"id","55","avatar_url","http://www.gravatar.com/avatar/eeaf078094fc76f1fadc1a957bb18960?s=48","githubUsername","","create_at","2011-05-08 18:44:02","score",5,"topic_count",0,"reply_count","is_star",false,{"_21":58,"_39":59,"_41":-5,"_42":60,"_44":61,"_46":62,"_48":35,"_49":35,"_33":63,"_51":64,"_27":65,"_54":66},{"_21":38,"_39":40,"_41":-5,"_42":43,"_44":45,"_46":47,"_48":35,"_49":35,"_33":50,"_51":52,"_27":53,"_54":55},"111","author_id","3","tab","content","nodePaintChat - node.js开发的多人协作绘图程序\r\n
\r\n
作者: hpf1908 (CNODEJS社区)\r\n

\r\n

$\"\"$

\r\n

\r\n
\r\n
\r\n
演示地址：http://paintchat.cnodejs.net/\r\n

\r\n

\r\n
$\"\"$ \r\n
\r\n

\r\n

\r\n
","title","node协作绘图程序在线测试","last_reply_at","2013-02-19 08:15:36.508","good","top",11,"visit_count",7436,"2011-05-08 15:20:43","author",{"_13":56,"_23":57},"kongwu","http://www.gravatar.com/avatar/1bfc8def4d9f790b4157a17aeba5f55e?s=48","105","28","

题记：

\r\n
有一天去面试，面试官问我，nodejs中的require是阻塞还是非阻塞。我心想，模块的引用都是先require之后再使用，应该是阻塞，但是对require的实现机制并不了解，就坦诚说不知道了。回来之后，我就翻了一遍node的源码，就简单写一下自己对node中模块的实现理解，纰漏之处，望大家指正。\r\n

js模块的实现

\r\n
其实早先时候在cnodejs问答中已经有人提到，模块中module,exports是怎么实现的，什么含义。当时是我回答的说是一个开发的接口，其实这是不对的。node中模块的实现，其实是依赖于闭包的，也就是说，module,exports其实都是外部传入的参数，这个简化形式如下：\r\n

function NativeModule(id){\r\n
    this.id=id;\r\n
    this.filename=id+\".js\";\r\n
    this.exports={};\r\n
}\r\n
NativeModule.require=function(id){\r\n
    var module=new NativeModule(id);\r\n
    module.compile();\r\n
    return  module.exports();\r\n
};\r\n
NativeModule.prototype.compile = function() {  \r\n
   (function(exports, require, module, __filename, __dirname){\r\n
        exports=module.exports=function(){alert(\"leexiaosi\")};\r\n
   }(this.exports, NativeModule.require, this, this.filename));\r\n
};\r\n
NativeModule.require(\"leexiaosi\")

\r\n
(附:上面的代码可以复制出来直接运行)\r\n
通过上面的代码我们惊奇的发现，其实我们看见的模块文件的代码，一般而言都是类似\r\n

exports=module.exports=function(){alert(\"leexiaosi\")};

\r\n
同时我们也就明白，为什么在模块文件中用var定义的变量都是私有的了。这里关键点就是\r\n

NativeModule.prototype.compile

\r\n
这个函数的实现。为了方便理解，我们设\r\n

var fnImport=function(exports, require, module, __filename, __dirname)

\r\n
那么，根据闭包的原理，这个函数在执行时，其[[scope]]会记录fnImport这个函数定义的作用域的各个变量，\r\n

fnImport函数中的形参值，即this.exports, NativeModule.require, this, this.filename

fnImport函数中function声明的函数

fnImport函数中通过var声明的变量

fnImport函数的父函数的作用域[[scope]]

\r\n
fnImport这个函数对其fnImport.[[scope]]的操作都是有权限的。故this.exports会被更改，尽管在NativeModule的构造函数中this.exports的定义是空对象。\r\n
当然，事实上的NativeModule的实现要比这个复杂得多，尽管一般而言，非核心模块加载都是依赖于module这个模块，但是，module的实现模块加载的基本思想也是这样，只是module中增加了模块的检索功能。而模块文件中，require正是NativeModule.require这个函数。从return module.exports()来看，require初次加载模块时候必然是阻塞的（初次加载之后会被缓存，所以加载之后再require就不是阻塞的了）。","node中的require和exports","2016-07-19 15:18:11.995",8,22395,"2011-05-15 11:09:13",{"_13":67,"_23":68},"leexiaosi","http://www.gravatar.com/avatar/7322babbf6eccdd9c69b0929cee9afb1?s=48"]