metric.c

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

#define METRIC_BUFSIZ 8192

int
clean_metric(char *metric, char *allowed_chars) {
    char cleaned_metric[METRIC_BUFSIZ];
    char *p, *q, *lastnl;
    char *firstspace = NULL;

    q = cleaned_metric;
    for (p = metric; *p != '\0'; p++) {
        if (*p == '\n' || *p == '\r') {
            /* end of metric */
            lastnl = p;

            /* just a newline on its own? some random garbage? skip */
            if (q == cleaned_metric || firstspace == NULL) {
                q = cleaned_metric;
                firstspace = NULL;
                continue;
            }
            *q++ = '\n';
            *q = '\0'; /* can do this because we substract one from buf */

            return strcmp(cleaned_metric, metric);
        } else if (*p == ' ' || *p == '\t' || *p == '.') {
            /* separator */
            if (q == cleaned_metric) {
                /* make sure we skip this on next iteration to
                 * avoid an infinite loop, issues #8 and #51 */
                lastnl = p;
                continue;
            }
            if (*p == '\t')
                *p = ' ';
            if (*p == ' ' && firstspace == NULL) {
                if (*(q - 1) == '.')
                    q--;  /* strip trailing separator */
                firstspace = q;
                *q++ = ' ';
            } else {
                /* metric_path separator or space,
                 * - duplicate elimination
                 * - don't start with separator/space */
                if (*(q - 1) != *p && (q - 1) != firstspace)
                    *q++ = *p;
            }
        } else if (firstspace != NULL ||
                    (*p >= 'a' && *p <= 'z') ||
                    (*p >= 'A' && *p <= 'Z') ||
                    (*p >= '0' && *p <= '9') ||
                    strchr(allowed_chars, *p)) {
            /* copy char */
            *q++ = *p;
        } else {
            /* something barf, replace by underscore */
            *q++ = '_';
        }
    }
    return 1;
}

int
main(int argc, char **argv) {
    if (argc < 2) {
        printf("Usage: ./a.out <metric_file>\n");
        return 1;
    }
    char *metric_file = argv[1];
    char metric[METRIC_BUFSIZ];
    char metric_name[METRIC_BUFSIZ];
    char *allowed_chars = "-_:#";

    FILE *fp = fopen(metric_file, "r");
    while (fgets(metric_name, METRIC_BUFSIZ, fp) != NULL) {
        int size = strlen(metric_name);
        metric_name[size - 1] = '\0';  /* strip trailing '\n' */
        snprintf(metric, METRIC_BUFSIZ, "%s %d %d\n", metric_name, 123, 1444320007);
        if (clean_metric(metric, allowed_chars) != 0) {
            printf("metric=%s\n", metric);
        }
    }
    return 0;
}