strlen performance comparison

README

A small pair of C programs to answer pepe-ramollo's question at www.reddit.com/r/C_Programming/comments/u53xv/i_dont_get_the_implementation_of_strlen/

use make test to run

Makefile

all: mystrlen strlen
mystrlen: mystrlen.c makestrs.h
	gcc -O2 mystrlen.c -o mystrlen
strlen: strlen.c makestrs.h
	gcc -O2 strlen.c -o strlen
clean:
	rm mystrlen strlen
test: mystrlen strlen
	time ./mystrlen > /dev/null
	time ./strlen > /dev/null

makestrs.h

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

#define NUMSTRS 40000

char** makestrs()
{
	char **strlist = malloc(sizeof(char*) * NUMSTRS);

	int i;
	//create NUMSTRS strings pointed to by array elements of strlist
	//for simplicity, the strings are all filled with 'a' except for the final null byte
	//additionally, they increase in size from 1 to NUMSTRS+1
	for (i = 0; i < NUMSTRS; i++) {
		strlist[i] = malloc(sizeof(char) * (i+1));
		int j;
		for (j = 0; j < i+1; j++) {
			if (j == i) {
				strlist[i][j] = '\0';
			} else {
				strlist[i][j] = 'a';
			}
		}
		//printf("string %d is %s (size %d)\n", i, strlist[i], strlen(strlist[i]));
	}
	return strlist;
}

mystrlen.c

//test the performance of the naive implementation
#include <stdlib.h>

#include "makestrs.h"

size_t mystrlen (const char *str)
{
	size_t len = 0;
	while (str[len]!='\0'){
		len++;
	}
	return len;
}

int main()
{
	char **strings = makestrs();

	int i;
	for (i = 0; i < NUMSTRS; i++) {
		printf("%ld\n",mystrlen(strings[i]));
	}

	return 0;
}

strlen.c

//test the performance of the naive implementation
#include <string.h>

#include "makestrs.h"

int main()
{
	char **strings = makestrs();

	int i;
	for (i = 0; i < NUMSTRS; i++) {
		printf("%ld\n", strlen(strings[i]));
	}

	return 0;
}