判断给定字符串是否匹配规则表达式

#include <stdio.h>
#include <regex.h>

void test_match(const char *pattern, const char *input, int maxmatches) {
    int i = 0;
    int regrc = 0;
    regex_t regex;

    printf("pattern=%s\n", pattern);
    printf("input  =%s\n", input);

    regrc = regcomp(&regex, pattern, REG_EXTENDED);
    if (regrc != 0) {
       printf("cannot regcomp pattern, return %d\n", regrc);
       return;
    }

    regmatch_t matches[maxmatches];
    regrc = regexec(&regex, input, maxmatches, matches, 0);
    if (regrc == 0) {
       int i;
       printf("regex match found\n");
       for (i = 0; i < maxmatches; i++) {
          if (matches[i].rm_so != -1) {
             printf("    matches[%d]=%.*s\n", i, matches[i].rm_eo - matches[i].rm_so, &input[matches[i].rm_so]);
           }
        }
    } else if (regrc == REG_NOMATCH) {
       printf("regex ok, but REG_NOMATCH\n");
    } else {
       char err[128] = {'\0'};
       printf("regex failed, return %d\n", regrc);
       regerror(regrc, &regex, err, sizeof(err));
    }

    regfree(&regex);
    return;
}


int main(int argc, char * argv[]) {
    const char * input   = "CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))";
    const char * pattern = "^CAST\\(.*\\)";
    int maxmatches = 1;
    test_match(pattern, input, maxmatches);
}

这个例子判断给定字符串是否符合一个CAST函数调用。
运行结果：

pattern=^CAST\(.* AS ([A-Z_]+).*\)
input  =CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
regex match found
    matches[0]=CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))

下面介绍第三个参数maxmatches的使用，她主要用在子匹配上。以上述为例，如果需要取出AS后面的类型值，那么我们就可以用：

    const char * input   = "CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))";
    const char * pattern = "^CAST\\(.* AS ([A-Z_]+).*\\)";
    int maxmatches = 2;

运行结果：

pattern=^CAST\(.* AS ([A-Z_]+).*\)
input  =CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
regex match found
    matches[0]=CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
    matches[1]=CHAR

需要注意的点是：

1. matches[0]是最长的匹配串。
2. matches[1..n]是子匹配串，根据需求可以多个。
3. 子匹配串在表达式中使用括号标识。
4. 如果需要匹配括号符合(即不作为子匹配,只是当作普通字符),那么需要转义(\)。
5. 子匹配串可以嵌套使用，例如：

    const char * input   = "CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))";
    const char * pattern = "^CAST(.* AS ([A-Z_]+).*)";
    int maxmatches = 3;
    test_match(pattern, input, maxmatches);

得到的输出结果就是：

pattern=^CAST(.* AS ([A-Z_]+).*)
input  =CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
regex match found
    matches[0]=CAST(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
    matches[1]=(TO_CHAR(TBL.FLD, '0000000000D') AS CHAR(12))
    matches[2]=CHAR

matches[0]: 是整个匹配的串
matches[1]: 是第一个子匹配的串，即函数CAST的参数内容
matches[1]: 是第二个子匹配的串，即嵌套表达式([A-Z_]+)的匹配子串。