pcre2

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

#define PCRE2_CODE_UNIT_WIDTH 8

#include <pcre2.h>

void odczyt_pliku_tekstowego(void);

char *checkLineWithRegex(char *pattern, char *subject, char *replacement);

FILE *inputFile;
FILE *outputFile;
char PLIK_TXT[] = "/home/rocket/CLionProjects/zpcre2/sample2.tex";
char patternPreserveTextInsideBrackets[] =
        "(.*)\\\\(?>title|author|date|enumsentence|section|subsection|emph|href|url).?\\{?(.*?)\\}(.*)?";
char patternDeleteAllOccurences[] = "";
char paternDeletWholeLine[] =
        "^(?>\\s*)\\\\(?>documentclass|usepackage|begin|maketitle|begin|end|includegraphics|nodeconnect|hspace).*$";
PCRE2_UCHAR *returnLine;


int main(int argc, char **argv) {
    returnLine = malloc(sizeof(PCRE2_UCHAR) * 1024);
    odczyt_pliku_tekstowego();

    return 0;
}

void odczyt_pliku_tekstowego(void) {
    char buffer[255];
    char *temp = malloc(sizeof(char) * 255);
    char outName[60];


    if ((inputFile = fopen(PLIK_TXT, "r")) != NULL) {
        fprintf(stdout, "Otwarto plik %s w trybie odczytu tekstowego.\n", PLIK_TXT);
        strcpy(outName, PLIK_TXT);
        outputFile = fopen(strcat(outName, "_cleaned.txt"), "w");
        printf("Utworzono plik tekstowy %s\n", outName);

        while (fgets(buffer, 255, (FILE *) inputFile)) {
            strcpy(temp, buffer);
            if (strcmp(temp, "\n") == 0) {
                continue;
            }
            checkLineWithRegex(paternDeletWholeLine, temp, "");
            checkLineWithRegex(patternPreserveTextInsideBrackets, returnLine, "${1} ${2} ${3}");

            fputs(returnLine, outputFile);
            printf("%s\n", returnLine);
        }


        fclose(outputFile);
        free(temp);
        fclose(inputFile);
    } else {
        printf("Nie moge odczytac pliku %s!\n", PLIK_TXT);
    }
}

char *checkLineWithRegex(char *patternIn, char *subjectIn, char *replacementIn) {
    int error;
    PCRE2_SIZE erroffset;

    const PCRE2_SPTR pattern = (PCRE2_SPTR) patternIn;
    const PCRE2_SPTR subject = (PCRE2_SPTR) subjectIn;
    const PCRE2_SPTR replacement = (PCRE2_SPTR) replacementIn;
    int subject_length = (PCRE2_SIZE) strlen((char *) subject);

    pcre2_code *re = pcre2_compile(pattern, PCRE2_ZERO_TERMINATED, 0, &error, &erroffset, 0);
    if (re == 0) {
        return "";
    }

    pcre2_jit_compile(re, PCRE2_JIT_COMPLETE);

    PCRE2_UCHAR output[1024] = "";
    PCRE2_SIZE outlen = sizeof(output) / sizeof(PCRE2_UCHAR);

    pcre2_match_data *match_data = pcre2_match_data_create_from_pattern(re, NULL);;


    int matched = pcre2_match(
            re,                   /* the compiled pattern */
            subject,              /* the subject string */
            subject_length,       /* the length of the subject */
            0,                    /* start at offset 0 in the subject */
            0,                    /* default options */
            match_data,           /* block for storing the result */
            NULL);
    if (matched == 1) {
        pcre2_code_free(re);
        printf("returnLine: %s, output: %s\n", returnLine, output);

        return strcpy((char *) returnLine, (char *) output);
    }

    int rc = pcre2_substitute(re, subject, PCRE2_ZERO_TERMINATED, 0,
                              PCRE2_SUBSTITUTE_GLOBAL | PCRE2_SUBSTITUTE_EXTENDED, 0, 0, replacement,
                              PCRE2_ZERO_TERMINATED, output, &outlen);

    if (rc > 0) {
        printf("returnLine: %s, output: %s\n", returnLine, output);

        return strcpy((char *) returnLine, (char *) output);

    } else {
        printf("returnLine: %s, subjectIn: %s\n", returnLine, subjectIn);

        return strcpy((char *) returnLine, subjectIn);
    }
}