如何寫 Lisp 解釋器？

關於代碼的話我搜到了兩個例子

http://blog.csdn.net/sedgewick/article/details/5928653

http://www.googies.info/articles/lispy.html

但具體說, 寫一個解釋器應該怎樣考慮呢?

function interpret(form, env, ctx){
if(form instanceof Array){
switch(form[0]){
case "lambda": {
var params = form[1];
var body = form[2];
return ctx(function(ctx){ return function() {
var e = Object.create(env);
for(var j = 0; j &< params.length; j++) e[params[j]] = arguments[j]; return interpret(body, e, ctx) }}) }; case "if": { var test = form[1], consequent = form[2], alternate = form[3]; return interpret(form[1], env, function(c){ if(c) return interpret(consequent, env, ctx) else return interpret(alternate, env, ctx) }) }; case "callcc": { return interpret(form[1], env, function(f){ var fctx = function(){ return function(x){ return ctx(x) }}; return f(fctx)(fctx) }) }; case "quote": { return ctx(form[1]) }; default: { return interpretCall(form, env, ctx); }; } } else if(typeof form === "string") { return ctx(env[form]) } else { return ctx(form) } } function interpretCall(form, env, ctx){ return interpret(form[0], env, function(callee){ return interpret$(form.slice(1), env, function(args){ return callee(ctx).apply(null, args) }) }) } function interpret$(form, env, ctx){ if(!form.length) return ctx(null); else return interpret(form[0], env, function(x0){ return interpret$(form.slice(1), env, function(x$){ if(x$) return ctx([x0]) else return ctx([x0].concat(x$)) }) }) } function id(x){ return x } var env0 = { trace: function(ctx){ return function(x) { return ctx(console.log(x)) }}}; interpret(["trace", ["callcc", ["lambda", ["return"], ["return", ["quote", 3]]]]], env0, id);

因為之前有人說我偷懶沒做 call/cc，今天我就做它一回。全文按照 Danvy 他們的抽象解釋模式寫成，顯式傳遞 Continuation。

function $(C,E,K){if(!(C instanceof Array))return K("string"==typeof C?
E[C]:C);switch(C[0]){case"lambda":return K(function(K){return function(
){for(var e=Object.create(E),u=0;u&