Re: AW: octstr_compare

On Wed, 10 Mar 2004 11:19:16 +0100
JЖrg Pommnitz <[email protected]> wrote:

> I'm not sure, but I think a NULL octstr would already fail the
> sems_valid test and cause an assertion.
> 
> -----UrsprЭngliche Nachricht-----
> Von: Yury Mikhienko [mailto:[email protected]]
> Gesendet: Mittwoch, 10. MДrz 2004 11:16
> An: [email protected]
> Betreff: octstr_compare
> 
> 
> Hi All!
> 
> What are think about the add following code into octstr.c/octstr_compare?
> 
> int octstr_compare(const Octstr *ostr1, const Octstr *ostr2)
> {
>     int ret;
>     long len;
> 
>     seems_valid(ostr1);
>     seems_valid(ostr2);
> 
> +    if ((ostr1 == NULL) && (ostr2 == NULL))
> +       return 0;
>         
>     if (ostr1->len < ostr2->len)
>         len = ostr1->len;
>     else
>         len = ostr2->len;
> 
>     if (len == 0) {
>         if (ostr1->len == 0 && ostr2->len > 0)
>             return -1;
>         if (ostr1->len > 0 && ostr2->len == 0)
>             return 1;
>         return 0;
>     }
> 
>     ret = memcmp(ostr1->data, ostr2->data, len);
>     if (ret == 0) {
>         if (ostr1->len < ostr2->len)
>             ret = -1;
>         else if (ostr1->len > ostr2->len)
>             ret = 1;
>     }
>     return ret;
> }
> 

But if I use disable_assertion option?

-- 
 
Best regards,
Yury Mikhienko.
IT ERP group head, ZAO "Mobikom-Kavkaz"
tel: (8632) 704188