[UPDATE] [Code] [Fastest String Search Algorithm]

[UPDATE] [Code] [Fastest String Search Algorithm]

[Amit Choudhary]

Last time I had sent old code. This is the latest code.

Again, I am not sending this code so that this code gets included in glibc.

I am sending this code because glibc mailing list is in public domain and
getting indexed by search engines. So, people outside of glibc can see this
algo if they search for fastest string search algorithm on web.

Code:

===================================================================================

// Choudhary string search algorithm
static char * choudhary_string_search_algorithm(char *text, char *pattern)
{

#define false 0
#define true 1
#define ALPHABET_SIZE 256

    int i = 0;
    int end_index = 0;
    int not_found = false;

    char pattern_char[ALPHABET_SIZE] = {0};

    int text_len = strlen(text);
    int pattern_len = strlen(pattern);

    int pi_44 = pattern_len - 1;
    int pi_34 = (3 * pattern_len) / 4;
    int pi_24 = pattern_len / 2;
    int pi_14 = pattern_len / 4;

    int last_failed_index = -1;

    // preprocessing
    for (i = 0; i < pattern_len; i++) {
        pattern_char[(int)(pattern[i])] = 1;
    }

    // now search
    for (i = 0; i < text_len; i++) {

        if ((text_len - i) < pattern_len) {
            return NULL;
            //return -1;
        }

        if (pattern[pi_44] != text[i + pi_44]) {

            last_failed_index = pi_44;

            // this character doesn't appear in pattern, so skip
            if (pattern_char[(int)(text[i + pi_44])] == 0) {
                i = i + pi_44;
            }

            continue;

        } else if (pattern[pi_34] != text[i + pi_34]) {

            last_failed_index = pi_34;

            // this character doesn't appear in pattern, so skip
            if (pattern_char[(int)(text[i + pi_34])] == 0) {
                i = i + pi_34;
            }

            continue;

        } else if (pattern[pi_24] != text[i + pi_24]) {

            last_failed_index = pi_24;

            // this character doesn't appear in pattern, so skip
            if (pattern_char[(int)(text[i + pi_24])] == 0) {
                i = i + pi_24;
            }

            continue;

        } else if (pattern[pi_14] != text[i + pi_14]) {

            last_failed_index = pi_14;

            // this character doesn't appear in pattern, so skip
            if (pattern_char[(int)(text[i + pi_14])] == 0) {
                i = i + pi_14;
            }

            continue;

        } // end of if-else.. block

        // compare with character at last failed index.
        if (last_failed_index >= 0) {

            if (pattern[last_failed_index] != text[i + last_failed_index]) {
                continue;
            }

        }

        if (pattern[0] == text[i]) {

            //full_pattern_search = full_pattern_search + 1;
            end_index = i + pi_44;
            not_found = false;
            int index = 0;

            for (index = i; index <= end_index; index++) {
                if (text[index] != pattern[index - i]) {
                    last_failed_index = index - i;
                    not_found = true;
                    break;
                }
            } // end of inner for loop

            if (not_found == false) { // match is found
                return (text + i);
                //return i;
            } else if (pattern_char[(int)(text[index])] == 0) {
                i = index;
            }
        } // end of if pattern[0]

    } // end of outer for loop

    return NULL;
    //return -1;

} // end of choudhary_string_search_algorithm

===================================================================================

Amit