Compiler Principles Study 1: Removing Comments from Code

Preface

I've started learning compiler principles～ For all compiler principles exercises, as usual, I'll still implement them in my favorite C# language, running on the .NetCore platform～ All the code for this series has already been uploaded to github, project homepage:

https://github.com/Deali-Axy/CompilerConstructionLearning

This Problem

Delete // comments and /…/ comments from source programs written in high-level programming languages such as C or C++, and retain the source program after deletion. It is required to save it in file form.

Approach Analysis

The main function of the program is to eliminate comments from already written source programs. In source programs, comments have two forms: one is single-line comments, denoted by “//”, and the other is multi-line comments, denoted by “/…/”. For these two forms, the program uses if..else.. statements to judge and handle them accordingly. There may also be another situation here: the above two kinds of comment symbols may appear inside quotation marks, and comment symbols that appear inside quotation marks do not have commenting functionality, so comment symbols that appear inside quotation marks should not be eliminated. Therefore, the program written this time will analyze three cases.

First case, single-line comments:

if (ch != temp)
{
    // 这里就是单行注释
    ofile.put(ch);
    ch = ifile.get();
}

or

if (ch != temp)
{
    /* 这里就是单行注释 */
    ofile.put(ch);
    ch = ifile.get();
}

Second case, block comments:

if (ifile.fail() || ofile.fail())
{
    cerr << "open file fail\n";
    return EXIT_FAILURE;
    /*返回值EXIT_FAILURE（在cstdlib库中定义）,用于向操作系统报*
    告打开文件失败*/
}

Third case, trailing comments:

ifile.close(); // 关闭文件
ofile.close();
cout << "/////*////ret/rtr////";
system("pause");
return 0;

There's also a key point to note

You can see this line

cout << "/////*////ret/rtr////";

This string of code wrapped in double quotes contains many slashes, so we need to avoid identifying these slashes as comments. The method I used here is to first replace the strings that contain comment symbols before processing comments, and then add them back after the comments have been deleted.

Implementation Code

The comments are written in great detail, so together with the approach analysis above, I won't continue analyzing the code～

var sReader = new StreamReader(filePath);
var newSource = "";
var inBlock = false;
var replaceFlag = false;
var tempLine = ""; // 用于保存被替换的特殊行代码
while (!sReader.EndOfStream)
{
    var line = sReader.ReadLine();
    if (line.Length == 0) continue; // 去除空行

    var quotationPattern = "^(.*?)\".*//.*\"";
    var quotationResult = Regex.Match(line, quotationPattern);
    if (quotationResult.Success)
    {
        System.Console.WriteLine("替换特殊代码，双引号中包裹注释斜杠");
        tempLine = quotationResult.Groups[0].Value;
        replaceFlag = true;
        line = Regex.Replace(line, quotationPattern, REPLACEMENT);
    }

    // 单行注释
    if (line.Trim().StartsWith(@"//"))
        continue;
    if (line.Trim().StartsWith(@"/*") && line.EndsWith(@"*/"))
        continue;

    // 注释块
    if (Regex.Match(line.Trim(), @"^/\*").Success)
        inBlock = true;
    if (Regex.Match(line.Trim(), @"\*/$").Success)
    {
        inBlock = false;
        continue;
    }

    // 行后注释
    // 使用非贪婪模式(.+?)匹配第一个//
    var pattern = @"^(.*?)//(.*)";
    // var pattern = @"[^(.*?)//(.*)]|[^(.*?)/\*(.*)\*/]";
    var result = Regex.Match(line, pattern);
    if (result.Success)
    {
        System.Console.WriteLine("发现行后注释：{0}", result.Groups[2]);
        line = result.Groups[1].Value;
    }

    // 还原被替换的代码
    if (replaceFlag)
    {
        System.Console.WriteLine("还原特殊代码");
        line = line.Replace(REPLACEMENT, tempLine);
        replaceFlag = false;
    }

    if (inBlock) continue;
    newSource += line + Environment.NewLine;
}

var outputPath = "output/exp1.src";
System.Console.WriteLine("去除注释完成，创建新文件。");
using (var sWriter = new StreamWriter(outputPath))
{
    sWriter.Write(newSource);
}
System.Console.WriteLine("操作完成！文件路径：{0}", outputPath);

Result Testing

Source file

#include <iostream>
#include <fstream>
#include <iomanip>
#include <cstdlib>
using namespace std;

int main()
{
    cout << '/';
    ifstream ifile; //建立文件流对象
    ofstream ofile;
    ifile.open("f:\\上机实验题\\C++\\ConsoleApplication2\\ConsoleApplication2\\源.cpp"); //打开F盘根目录下的fileIn.txt文件
    ofile.open("f:\\上机实验题\\C++\\ConsoleApplication2\\ConsoleApplication2\\源.obj");
    if (ifile.fail() || ofile.fail())
    { //测试打开操作是否成功
        cerr << "open file fail\n";
        return EXIT_FAILURE;
        /*返回值EXIT_FAILURE（在cstdlib库中定义）,用于向操作系统报*
    告打开文件失败*/
    }
    char ch;
    ch = ifile.get(); //进行读写操作
    while (!ifile.eof())
    {
        if (ch == 34)
        {                   //双引号中若出现“//”，双引号中的字符不消除
            char temp = ch; //第一个双引号
            ofile.put(ch);
            ch = ifile.get();
            while (!ifile.eof())
            {
                if (ch != temp)
                { //寻找下一个双引号
                    ofile.put(ch);
                    ch = ifile.get();
                }
                else
                {
                    ofile.put(ch);
                    break;
                }
            }
            ch = ifile.get();
            continue; //双引号情况结束，重新新一轮判断
        }
        if (ch == 47)
        { //出现第一个斜杠
            char temp2 = ch;
            ch = ifile.get();
            if (ch == 47)
            { //单行注释情况
                ch = ifile.get();
                while (!(ch == '\n'))
                    ch = ifile.get();
            }
            else if (ch == '*')
            { //多行注释情况
                while (1)
                {
                    ch = ifile.get();
                    while (!(ch == '*'))
                        ch = ifile.get();
                    ch = ifile.get();
                    if (ch == 47)
                        break;
                }
                ch = ifile.get();
            }
            else
            {
                ofile.put(temp2); //temp2保存第一个斜杠，当上述两种情况都没有时，将此斜杠输出
            }
            //ch = ifile.get();
        }
        //cout << ch << endl;
        ofile.put(ch);    //将字符写入文件流对象中
        ch = ifile.get(); //从输入文件对象流中读取一个字符
    }
    ifile.close(); //关闭文件
    ofile.close();
    cout << "/////*////ret/rtr////";
    system("pause");
    return 0;
}

Result after processing

#include <iostream>
#include <fstream>
#include <iomanip>
#include <cstdlib>
using namespace std;
int main()
{
    cout << '/';
    ifstream ifile; 
    ofstream ofile;
    ifile.open("f:\\上机实验题\\C++\\ConsoleApplication2\\ConsoleApplication2\\源.cpp"); 
    ofile.open("f:\\上机实验题\\C++\\ConsoleApplication2\\ConsoleApplication2\\源.obj");
    if (ifile.fail() || ofile.fail())
    { 
        cerr << "open file fail\n";
        return EXIT_FAILURE;
    }
    char ch;
    ch = ifile.get(); 
    while (!ifile.eof())
    {
        if (ch == 34)
        {                   
            char temp = ch; 
            ofile.put(ch);
            ch = ifile.get();
            while (!ifile.eof())
            {
                if (ch != temp)
                { 
                    ofile.put(ch);
                    ch = ifile.get();
                }
                else
                {
                    ofile.put(ch);
                    break;
                }
            }
            ch = ifile.get();
            continue; 
        }
        if (ch == 47)
        { 
            char temp2 = ch;
            ch = ifile.get();
            if (ch == 47)
            { 
                ch = ifile.get();
                while (!(ch == '\n'))
                    ch = ifile.get();
            }
            else if (ch == '*')
            { 
                while (1)
                {
                    ch = ifile.get();
                    while (!(ch == '*'))
                        ch = ifile.get();
                    ch = ifile.get();
                    if (ch == 47)
                        break;
                }
                ch = ifile.get();
            }
            else
            {
                ofile.put(temp2); 
            }
        }
        ofile.put(ch);    
        ch = ifile.get(); 
    }
    ifile.close(); 
    ofile.close();
    cout << "/////*////ret/rtr////";
    system("pause");
    return 0;
}

Complete Code

https://github.com/Deali-Axy/CompilerConstructionLearning/blob/master/code/exp/exp1/Exp1.cs

References:

• Regular expression metacharacters: https://www.runoob.com/regexp/regexp-metachar.html
• Removing comments in JavaScript: https://segmentfault.com/a/1190000015611632

Welcome to Discuss

For questions, please leave a message on the WeChat official account backend; I will reply to every message~

• WeChat official account: 画星星高手
• Coding livestream: https://live.bilibili.com/11883038
• Zhihu: https://www.zhihu.com/people/dealiaxy
• Column: https://zhuanlan.zhihu.com/deali