PL0词法分析程序_PL0词法分析_

#include <stdio.h> #include "string.h" enum symbol { nul, ident, number, plus, minus, times, slash, oddsym, eql, neq, lss, leq, gtr, geq, lparen, rparen, comma, semicolon, period, becomes, beginsym, endsym, ifsym, thensym, whilesym, writesym, readsym, dosym, callsym, constsym, varsym, procsym, }; #define al 10 /* 符号的最大长度 */ #define norw 13 /* 关键字个数 */ #define nmax 14 /* number的最大位数 */ #define fctnum 8 char line[81]; /* 读取行缓冲区 */ #define getchdo if(-1 == getch()) return -1 FILE* fin; char fname[al]; enum symbol ssym[256]; /* 单字符的符号值 */ char word[norw][al]; /* 保留字 */ enum symbol wsym[norw]; /* 保留字对应的符号值 */ int err; /* 错误计数器 */ int cc, ll; /* getch使用的计数器，cc表示当前字符(ch)的位置 */ int cx; /* 虚拟机代码指针, 取值范围[0, cxmax-1]*/ char ch; /* 获取字符的缓冲区，getch 使用 */ char a[al+1]; /* 临时符号, 多出的一个字节用于存放0 */ char id[al+1]; /* 当前ident, 多出的一个字节用于存放0 */ enum symbol sym; /* 当前的符号 */ int num; /* 当前number */ void error(int n); int getch(); void init(); int getsym(); int main() { printf("Input pl/0 file? "); scanf("%s", fname); /* 输入文件名 */ fin = fopen(fname, "r"); if (fin) { init(); /* 初始化 */ err = 0; cc = cx = ll = 0; ch = ' '; do { getsym(); if(sym==1)//标识符 种别及其本身 { printf("(%d,'%s')", sym, a ); } else if(sym==2)// 无符号整数 种别及其本身 { printf("(%d,'%d')", sym, num ); } else { printf("(%d)", sym);//其他（保留字，分界符，操作符等）的种别 } } while (sym != 18);//"."的种别是18 printf("\n"); fclose(fin); } else { printf("Can't open file!\n"); } printf("\n"); return 0; } void error(int n) { char space[81]; memset(space,32,81); space[cc-1]=0; //出错时当前符号已经读完，所以cc-1 printf("****%s!%d\n", space, n); err++; } int getch() { if (cc == ll) { if (feof(fin)) { printf("program incomplete"); return -1; } ll=0; cc=0; //printf("%d ", cx); ch = ' '; while (ch != 10) { //fscanf(fin,"%c", &ch) //richard if (EOF == fscanf(fin,"%c", &ch)) { line[ll] = 0; break; } //end richard //printf("%c", ch); line[ll] = ch; ll++; } printf("\n");//每读取一行 } ch = line[cc]; cc++; return 0; } void init() { int i; /* 设置单字符符号 */ for (i=0; i<=255; i++) { ssym[i] = nul; } ssym['+'] = plus; ssym['-'] = minus; ssym['*'] = times; ssym['/'] = slash; ssym['('] = lparen; ssym[')'] = rparen; ssym['='] = eql; ssym[','] = comma; ssym['.'] = period; ssym['#'] = neq; ssym[';'] = semicolon; /* 设置保留字名字,按照字母顺序，便于折半查找 */ strcpy(&(word[0][0]), "begin"); strcpy(&(word[1][0]), "call"); strcpy(&(word[2][0]), "const"); strcpy(&(word[3][0]), "do"); strcpy(&(word[4][0]), "end"); strcpy(&(word[5][0]), "if"); strcpy(&(word[6][0]), "odd"); strcpy(&(word[7][0]), "procedure"); strcpy(&(word[8][0]), "read"); strcpy(&(word[9][0]), "then"); strcpy(&(word[10][0]), "var"); strcpy(&(word[11][0]), "while"); strcpy(&(word[12][0]), "write"); /* 设置保留字符号 */ wsym[0] = beginsym; wsym[1] = callsym; wsym[2] = constsym; wsym[3] = dosym; wsym[4] = endsym; wsym[5] = ifsym; wsym[6] = oddsym; wsym[7] = procsym; wsym[8] = readsym; wsym[9] = thensym; wsym[10] = varsym; wsym[11] = whilesym; wsym[12] = writesym; } int getsym() { int i,j,k; /* the original version lacks "\r", thanks to foolevery */ while (ch==' ' || ch==10 || ch==13 || ch==9) /* 忽略空格、换行、回车和TAB */ { getchdo; } if (ch>='a' && ch<='z') { /* 名字或保留字以a..z开头 */ k = 0; do { if(k<al) { a[k] = ch; k++; } getchdo; } while (ch>='a' && ch<='z' || ch>='0' && ch<='9'); a[k] = 0; strcpy(id, a); i = 0; j = norw-1; do { /* 搜索当前符号是否为保留字 */ k = (i+j)/2; if (strcmp(id,word[k]) <= 0) { j = k - 1; } if (strcmp(id,word[k]) >= 0) { i = k + 1; } } while (i <= j); if (i-1 > j) { sym = wsym[k]; } else { sym = ident; /* 搜索失败则，是名字或数字 */ } } else { if (ch>='0' && ch<='9') { /* 检测是否为数字：以0..9开头 */ k = 0; num = 0; sym = number; do { num = 10*num + ch - '0'; k++; getchdo; } while (ch>='0' && ch<='9'); /* 获取数字的值 */ k--; if (k > nmax) { error(30); } } else { if (ch == ':') /* 检测赋值符号 */ { getchdo; if (ch == '=') { sym = becomes; getchdo; } else { sym = nul; /* 不能识别的符号 */ } } else { if (ch == '<') /* 检测小于或小于等于符号 */ { getchdo; if (ch == '=') { sym = leq; getchdo; } else { sym = lss; } } else { if (ch=='>') /* 检测大于或大于等于符号 */ { getchdo; if (ch == '=') { sym = geq; getchdo; } else { sym = gtr; } } else { sym = ssym[ch]; /* 当符号不满足上述条件时，全部按照单字符符号处理 */ //getchdo; //richard if (sym != period) { getchdo; } //end richard } } } } } return 0; }