花了一下午的时间完成了一个简单语言的解释器,我会在最后帖出所有代码,但是今天不打算详细解释程序的每一个步骤,最近考虑找实习、做论文,要花一些时间。有时间了我会解释每一部分,在这里只提醒一下读者,程序的写作过程和它呈现出来的不一样,总体来说我的写作过程是先写一个只包含一条指令的解释器,然后逐渐加入其他指令。ps:我是多么的想看看高手们写程序的过程,而不只是结果,但是就像graham说的“写的过程往往显得冗长”,所以这方面的书籍不多。我觉得比较接近的书包括《clean code》、《paip》、《software tools》。只可惜现在只看完了一本半。
想法的来源
昨天圣诞节,坐在不能上网的实验室,觉得好无聊。然后想起了学习《可计算性与计算复杂性》的时候,里面用到了一种含有5条指令的原语言,老师也要求我们用这个语言写程序,所以我就想写一个解释器,来执行用该语言,并把这个解释器作为自己的圣诞礼物。
原语言描述
书中说它是一个fortran式的简单语言,所以我就给它起了个简单的名字Metafor,表示meta fortran。下面来看一下它的具体描述,然后贴上我在代码里更详细的描述,我为每一种指令都起了个名字(注意其中把赋值语句命名为setf是由于我最近在写common lisp多一些)。
+-------------------------------------------------------------------------------------------+
指令 描述
x = x + 1 变量x的值加1
x = x - 1 变元x的值减1。若x的值为0,则结果仍为0
TO A IF x≠0 若x≠0,则转标号为A的指令;否则执行下一条指令
TO A 无条件转到标号为A的指令
y=x 把x的值赋给变元y,x值保持不变
+-------------------------------------------------------------------------------------------+
1. Metafor has five instructions: name instruction description inc: x = x + 1, Increase by 1 the value of the variable x. dec: x = x - 1, If the value of x is 0, leave it unchanged; otherwise decrease by 1 the value of x. con_goto: TO A IF x != 0, If the value of x is nonzero, perform the instruction with the label A next; otherwise proceed to the next instruction in the list. goto: TO A, Perform the instruction with the label A next. setf: y = x, Change the value variable y to the value of variable x. 2. Some specification: (1) Input variables are represented as: x1, x2, x3, ... (2) Local variables are represented as: z1, z2, z3, ... (3) Output variable is represented as: y note: the num 1 is often omitted(i.e., x stands for x1 and z stands for z1). 3. Labeled instructions: Instructions may or may not have labels. When an instruction is labeled, the label is written to its left in square brackets. For example, [B] Z = Z - l 4. A smaple program: "Program to compute y = x1+ x2" y = x1 [B] TO A IF x2 != 0 TO E [A] x2 = x2 - 1 y = y + 1 TO B 4. For more information, please refer to 《可计算性与计算复杂性》, 周长林、李占山著.
整体思路