posix 解析uevent导致错误的正则表达式

Question

我正在尝试使用下面的代码解析 uevent，但我认为我的正则表达式不正确导致 regcomp 函数失败。

有人可以帮忙吗？我正在尝试做类似 this.

的事情

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

int main ()
{
  char * source = "change@/devices/soc/799999.i2c/i2c-3/3-0015/power_supply/battery";
  char * regexString = "(?<action>[a-zA-Z]+)@\/(?<dev_path>.*)\/(?<subsystem>[a-zA-z]+)\/(?<name>[a-zA-z]+)";
  size_t maxGroups = 4;

  regex_t regexCompiled;
  regmatch_t groupArray[maxGroups];

  if (regcomp(&regexCompiled, regexString, REG_EXTENDED))
    {
      printf("Could not compile regular expression.\n");
      return 1;
    };
  regfree(&regexCompiled);

  return 0; 
}

我得到 "Could not compile regular expression."。这意味着 regcomp 无法识别正则表达式。

Answer 1

当我使用代码报错时：

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

int main(void)
{
  //char * source = "change@/devices/soc/799999.i2c/i2c-3/3-0015/power_supply/battery";
  char * regexString = "(?<action>[a-zA-Z]+)@\/(?<dev_path>.*)\/(?<subsystem>[a-zA-z]+)\/(?<name>[a-zA-z]+)";
  //size_t maxGroups = 4;

  regex_t regexCompiled;
  //regmatch_t groupArray[maxGroups];

  int rc;
  if ((rc = regcomp(&regexCompiled, regexString, REG_EXTENDED)) != 0)
    {
      char buffer[1024];
      regerror(rc, &regexCompiled, buffer, sizeof(buffer));
      printf("Could not compile regular expression (%d: %s).\n", rc, buffer);
      return 1;
    }
  regfree(&regexCompiled);

  return 0; 
}

我得到输出：

Could not compile regular expression (13: repetition-operator operand invalid).

问题出在您使用的符号 (? 中：

"(?<action>[a-zA-Z]+)@\/(?<dev_path>.*)\/(?<subsystem>[a-zA-z]+)\/(?<name>[a-zA-z]+)"

该符号适用于 PCRE 而不是 POSIX。 PCRE 在 ( 之后使用 ? 正是因为它在其他正则表达式系统（例如 POSIX）中无效。

因此，如果您想使用 PCRE 正则表达式，请安装并使用 PCRE 库。

否则，您需要使用：

"([a-zA-Z]+)@\/(.*)\/([a-zA-z]+)\/([a-zA-z]+)"

有了这个，并注意到你需要一个 regmatch_t 用于匹配的整个字符串加上 4 个捕获组（总共 5 个捕获），你可以写：

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

int main(void)
{
    char *source = "change@/devices/soc/799999.i2c/i2c-3/3-0015/power_supply/battery";
    // char * regexString = "(?<action>[a-zA-Z]+)@\/(?<dev_path>.*)\/(?<subsystem>[a-zA-z]+)\/(?<name>[a-zA-z]+)";
    size_t maxGroups = 5;
    char *regexString =  "([a-zA-Z]+)@\/(.*)\/([a-zA-z]+)\/([a-zA-z]+)";

    regex_t regexCompiled;
    regmatch_t groupArray[maxGroups];

    int rc;
    if ((rc = regcomp(&regexCompiled, regexString, REG_EXTENDED)) != 0)
    {
        char buffer[1024];
        regerror(rc, &regexCompiled, buffer, sizeof(buffer));
        printf("Could not compile regular expression (%d: %s).\n", rc, buffer);
        return 1;
    }
    if ((rc = regexec(&regexCompiled, source, maxGroups, groupArray, 0)) != 0)
    {
        char buffer[1024];
        regerror(rc, &regexCompiled, buffer, sizeof(buffer));
        printf("Could not execute regular expression (%d: %s).\n", rc, buffer);
        return 1;
    }

    printf("Match successful:\n");
    for (size_t i = 0; i < maxGroups; i++)
    {
        int so = groupArray[i].rm_so;
        int eo = groupArray[i].rm_eo;
        printf("%zu: %d..%d [%.*s]\n", i, so, eo, eo - so, &source[so]);
    }

    regfree(&regexCompiled);

    return 0;
}

输出为：

Match successful:
0: 0..64 [change@/devices/soc/799999.i2c/i2c-3/3-0015/power_supply/battery]
1: 0..6 [change]
2: 8..43 [devices/soc/799999.i2c/i2c-3/3-0015]
3: 44..56 [power_supply]
4: 57..64 [battery]

posix 解析uevent导致错误的正则表达式

posix regular expression for parsing uevent causing error

c

regex

posix-ere