Lisp解释器的实现-默认操作符的实现(6)

在上文中，我们提到了 Evaluator 有默认的操作符，本文将实现这些操作符。

默认操作符的实现

def accumulate(proc,init,args):
    result = init
    for i in args:
        result = proc(result,i)
    return result

def add(*args):
    return accumulate(lambda a,b:a + b,0,args)
def subtract(*args):
    return accumulate(lambda a,b:a - b,0,args)
def multiply(*args):
    return accumulate(lambda a,b:a * b,1,args)
def divide(*args):
    return accumulate(lambda a,b:a / b,1,args)

def to_list(*args):
    return accumulate(lambda a,b:a + [b],[],args)

def cons(first,second):
    return [first] + [second]
def car(pair):
    return pair[0]
def cdr(pair):
    if len(pair) == 2:
        return pair[1]
    return pair[1:]
def null(seq):
    return seq == []

在 Evaluator 里设置默认操作符

class Evaluator:
    def __init__(self):
        self._exp_parser = ExpressionParser()
        self._init_primitives()
    def eval(self,exp,env):
        if exp == []: return

        if isinstance(exp,Expression) == False:
            exp = self._exp_parser.parse(exp)

        if isinstance(exp,Expression) == False:
            if len(exp) == 1:
                exp = VariableExpression(exp)
            else:
                args = list(map(lambda arg: self.eval(arg,env),exp[1:]))
                proc = env.get_variable(exp[0])
                if proc == None:
                    proc = self._primitives[exp[0]]
                
                if isinstance(proc,ProcedureExpression):
                    for i in range(len(proc.args)):
                        proc.env.define_variable(proc.args[i],args[i])
                return self.apply(proc,args)
        return exp.eval(self,env)
    def apply(self,proc,args):
        if proc in self._primitives.values():
            return proc(*args)
        return self.eval(proc.proc,proc.env)
    def _init_primitives(self):
        self._primitives = {
            "+": add,
            "-": subtract,
            "*": multiply,
            "/": divide,
            "list": to_list,
            "cons": cons,
            "car": car,
            "cdr": cdr,
            "null?": null
        }