合肥工业大学编译原理实验一词法分析

基本思路

词法分析是对输入语句串中一个个单词符号进行分析，最后格式化输出种别码，类型，位置等信息。那么，就可以考虑一次读入一个字符将它们拼接成一个字符串，当碰到空格或者分界符（‘，’， ‘；’， ‘）’， ‘（’， ‘[’, ‘]’ ）时，就把前面已读的字符串格式化输出，再输出当前分界符，然后再往后读下一个字符，直到程序尾部。

看代码之前

这个程序写得并不是很好，当时时间有限，就随便写写应付验收，但效果还是不错的。还可在此基础上修改修改，提高效率，缩减代码量，比如用利用字典来索引种别码，判断逻辑修改为循环而不是多个if-else等。

完整代码

#include <iostream>
#include <ctype.h>
#include <cstdlib>
#include <cstdio>
using namespace std;
int typeOfWord(string word);//字符串的种别码
void printsub(string word, int row, int col); //格式化输出
bool isDivider(char c);//是否是分界符
bool isrela(char c);//是否是关系运算符
bool isAlop(char c);//算术运算符

string k[8] = {"do", "end", "for", "if", "printf", "scanf", "then", "while"};//关键字
string s[6] = {",", ";", "(", ")", "[", "]"}; //分界符
string aop[4] = {"+", "-", "*", "/"}; //算术运算符表
string rop[6] = {"<", "<=", "=", ">", ">=", "<>"};//关系运算符
string identi[10] = {"a", "i", "n"}; //标识符

int main()
{
    int row, col;
    char c;
    string inString = "", alString = "";
    row = col = 1; //从第一行第一列开始计数
    FILE *fp;
    fp = fopen("source.txt", "r");
    if(fp == NULL)//打开操作不成功
    {
        printf("file open failed.\n");
        return -1;
    }
    cout << "单词\t二元序列\t类型\t位置(行，列)\n";
    cout << "\t(单词种别，单词属性)\n";
    do
    {
        c = fgetc(fp); //读取一个字符
        if(c == EOF)
            break;
        if(isalpha(c))//如果是字母
        {
            inString += c;
            continue;
        }
        else if(isrela(c)) //关系符
        {
            inString += c;
            continue;
        }
        else if(isdigit(c)) //数字
        {
            inString += c;
            continue;
        }
        else if (c == ' ') //如果是空格，就执行输出程序
        {
            printsub(inString, row, col);
            inString = "";
            col += 1; //列加1
            continue;
        }
        else if (c == '\n') //换行符
        {
            row += 1;
            col = 1;
            continue;
        }
        else if(isAlop(c)) //算数运算符
        {
            if(inString.size() != 0)
            {
                printsub(inString, row, col); //先把前一个字符输出
                inString = "";
                col += 1; //列加1
            }
            alString += c;
        }
        else if(isDivider(c))
        {
            if(inString.size() != 0) //如果inString长度不为0，就先输出它
            {
                printsub(inString, row, col); //先把前一个字符输出
                inString = "";
                col += 1; //列加1
            }
            if(alString.size() != 0)//如果alString长度不为0，就再输出它
            {
                printsub(alString, row, col);
                col += 1;
                alString = "";
            }
            //然后再输出当前分界符
            inString = c;
            printsub(inString, row, col);
            inString = "";
            col += 1; //列加1
        }
        else //else啥都不是，就是error
        {
            printf("%6s\t%6s\t%9s\t(%d,%d)\n",inString.c_str(), "Error", "Error", row, col);
            col += 1;
        }
    }while(!feof(fp));
    return 0;
}

int typeOfWord(string word)
{
    int res = 0, i;
    bool found = false;
    if(!found)
        for(i = 0; i < 8; i++) //判断是否为关键字
        {
            if (k[i] == word)
            {
                res = 1;
                found = true;
                break;
            }
        }
    //判断是否为分界符
    if(!found)
        for (i = 0; i < 6; i++)
        {
            if (s[i] == word)
            {
                res = 2;
                found = true;
                break;
            }
        }
    //判断算术运算符
    if(!found)
        for (i = 0; i < 4; i++)
        {
            if (aop[i] == word)
            {
                res = 3;
                found = true;
                break;
            }
        }
    //关系运算符
    if(!found)
        for (i = 0; i < 6; i++)
        {
            if (rop[i] == word)
            {
                res = 4;
                found = true;
                break;
            }
        }
    //判断无符号数
    bool isDigit = true;
    if(!found)
    {
        for (i = 0; i < int(word.size()); i++)
        {
            if (!isdigit(word[i]))
            {
                isDigit = false;
                break;
            }
        }
        if (isDigit)
        {
            found = true;
            res = 5;
        }
        else
            res = -1;
    }
    //查标识符
    bool isIden = false;
    if(!found)
    {
        for (i = 0;i < 10; i++)
        {
            if (word == identi[i])
            {
                isIden = true;
                break;
            }
        }
        if(isIden)
            res = 6;
        else res = -1;
    }
    return res;
}
//格式化输出
void printsub(string word, int row, int col)
{
    int type = typeOfWord(word);
    switch (type)
    {
    case 1:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "关键字", row, col);
        break;

    case 2:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "分界符", row, col);
        break;

    case 3:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "算数运算符", row, col);
        break;

    case 4:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "关系运算符", row, col);
        break;

    case 5:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "常数", row, col);
        break;

    case 6:
        printf("%6s\t(%d,%3s)\t%10s\t(%d,%d)\n",word.c_str(), type, word.c_str(), "标识符", row, col);
        break;

    case -1:
        printf("%6s\t%6s\t%9s\t(%d,%d)\n",word.c_str(), "Error", "Error", row, col);
        break;
    }
}

bool isrela(char c)
{
    bool res = false;
    if (c == '>')
        res = true;
    else if(c == '<')
        res = true;
    else if(c == '=')
        res = true;
    return res;
}

bool isAlop(char c)//这里就可以修改成循环的形式
{
    bool res = false;
    if (c == '+')
        res = true;
    else if(c == '-')
        res = true;
    else if(c == '*')
        res = true;
    else if (c == '/')
        res = true;
    return res;
}

bool isDivider(char c)
{
    char s[] = {',', ';', '(', ')', '[', ']'};
    bool res = false;
    for(int i = 0; i < 6; i++)
    {
        if(s[i] == c)
        {
            res = true;
            break;
        }
    }
    return res;
}

敲重点！

这个实验最坑的是把++运算符也识别为错误，验收时老师可能会问这个怎么实现的，还有就是后面识别3b这种为错是怎么实现的，换成3b45之类的还能不能识别，就这两点需要注意。

如果可能的话

如果这篇文章对你有帮助的话，能否动动小手打个赏呢？下一篇是实验二LL1分析。886