Nim语言中如何声明相互调用的过程原型？

完全可以用**前置声明（forward declaration）**解决这个问题，不需要用到面向对象代码。

Nim编译器要求调用过程前必须先知道这个过程的存在，相互调用的情况下没法把所有过程都放在最前面，这时候只需要先写出每个过程的签名（名字、参数类型、返回值类型），告诉编译器这些过程会在后面定义，之后再写具体实现即可。

修改后的代码如下：

# 前置声明三个过程的签名
proc mydouble(num: int): int
proc mysum(firstnum, secondnum: int): int
proc mymultiply(firstnum, secondnum: int): int

# 具体实现部分保持不变
proc mydouble(num: int): int =
  echo "Warning, someone is trying to double a number, which is forbidden by law!"
  echo "I will allow it for now, but I filed it under the capital crimes register."
  return mymultiply(num, 2)

proc mysum(firstnum, secondnum: int): int =
  if firstnum==secondnum:
    return mydouble(firstnum)
  else:
    return firstnum + secondnum

proc mymultiply(firstnum, secondnum: int): int =
  if firstnum == 2:
    return mydouble(secondnum)
  elif secondnum == 2:
    return mydouble(firstnum)
  else:
    var res=0
    for x in 1..secondnum:
      res=mysum(res,firstnum)
    return res

var a,b,c: int
a = 3
b = 2
c = -1

c=mydouble(a)
echo c # expecting 6, with a warning about crime.
c=mysum(a,b)
echo c # expecting 5
c=mysum(a-1,b)
echo c # expecting 4, with a warning about crime.
c=mymultiply(a,a)
echo c # expecting 9
c=mymultiply(a,b)
echo c # expecting 6, with a warning about crime.

这样编译时就不会再报未声明标识符的错误了——前置声明相当于给编译器打了个招呼："这些过程我后面会写，先别报错"。

内容的提问来源于stack exchange，提问作者GiMS_Pang