On 12/23/22 13:26, Alejandro Colomar wrote:

> Well, I do believe snprintf is also misdesigned, for the same reasons that the 
> strlcpy(3) manual page states that you should use strlcpy(3) for catenating 

typo fix:  s/should/shouldn't/

> strings, but rather strlcat(3):
> 
>         To detect truncation, perhaps while building a pathname, something like
>         the following might be used:
> 
>               char *dir, *file, pname[MAXPATHLEN];
> 
>               ...
> 
>               if (strlcpy(pname, dir, sizeof(pname)) >= sizeof(pname))
>                       goto toolong;
>               if (strlcat(pname, file, sizeof(pname)) >= sizeof(pname))
>                       goto toolong;
> 
>         Since it is known how many  characters  were  copied  the  first  time,
>         things can be sped up a bit by using a copy instead of an append:
> 
>               char *dir, *file, pname[MAXPATHLEN];
>               size_t n;
> 
>               ...
> 
>               n = strlcpy(pname, dir, sizeof(pname));
>               if (n >= sizeof(pname))
>                       goto toolong;
>               if (strlcpy(pname + n, file, sizeof(pname) ‐ n) >= sizeof(pname) ‐ n)
>                       goto toolong;
> 
>         However,  one  may question the validity of such optimizations, as they
>         defeat the whole purpose of strlcpy() and strlcat().  As  a  matter  of
>         fact, the first version of this manual page got it wrong.

-- 
<http://www.alejandro-colomar.es/>