Thread: To recreate strcpy

Yeah I kind of came to realise that the 'not' check was redundant, but dident say anything cos I'd look dumb :P I guess theres no sneaking my dodgy code past you guys tho

Anyway my first attempt sucked so I had a go at rewriting the function. This should be faster, and I managed to shave a few characters off it:

Code:

int StrCmp(char *a, char *b) 
{   
    while(*a || *b) 
    { 
        if(*a != *b) return(*a > *b) ? 1 : -1; 
        *a++; *b++;
    }                 
    return 0;         
}

Code:

  *a++; *b++;

You can get rid of the stars since you're not using the values of *a and *b anyway.

Originally Posted by mike_g

Code:

int StrCmp(char *a, char *b) 
{   
    while(*a || *b) 
    { 
        if(*a != *b) return(*a > *b) ? 1 : -1; 
        *a++; *b++;
    }                 
    return 0;         
}

This still compares against the null byte. Try writing it using while(*a && *b) instead (other stuff will change around too).

Originally Posted by robatino

You can get rid of the stars since you're not using the values of *a and *b anyway.

Nice one. Now when I remove the whitespace it fits snuggly into one line of code, which fits with the rest of the functions

Originally Posted by brewbuck

This still compares against the null byte.

I see I forgot to make the char pointer unsigned which would cause problems, but apart from that wouldent the null byte equate to less than whatever it is being compared against anyway?

Originally Posted by mike_g

Nice one. Now when I remove the whitespace it fits snuggly into one line of code, which fits with the rest of the functions


I see I forgot to make the char pointer unsigned which would cause problems, but apart from that wouldent the null byte equate to less than whatever it is being compared against anyway?

In this case, yes.

I wrote my own fmod function that only dealt with cases where the number was less than twice that of the number you want to divide by.

It was significantly faster.

I think that's what put me off relying on other people's code.

I wrote several functions in a basic language that were apparently faster than the equivalent functions that were built in. I think the difference was mainly down to function overheads (or something like that) rather than my code actually being faster.