Re: [PATCH v2] ntoab: Handle LONG_MIN case.

[Jan Nieuwenhuizen]

Danny Milosavljevic writes:

> * lib/mes/ntoab.c (ntoab): Handle LONG_MIN case.
> ---
>  lib/mes/ntoab.c | 25 ++++++++++++++++++++++++-
>  1 file changed, 24 insertions(+), 1 deletion(-)

Oh, great!

> diff --git a/lib/mes/ntoab.c b/lib/mes/ntoab.c
> index a83b4d31..25780652 100644
> --- a/lib/mes/ntoab.c
> +++ b/lib/mes/ntoab.c
> @@ -20,6 +20,10 @@
>  
>  #include <assert.h>
>  #include <mes/lib.h>
> +#include <stdint.h>
> +
> +#define STR(x) #x
> +#define XSTR(s) STR(s)
>  
>  char *
>  ntoab (long x, int base, int signed_p)
> @@ -34,7 +38,26 @@ ntoab (long x, int base, int signed_p)
>    if (signed_p && x < 0)
>      {
>        sign_p = 1;
> -      u = -x;
> +      if (x == LONG_MIN)
> +        {
> +           /* Cannot do u = (-x), so avoid it. */
> +           long i;
> +           if (base == 10)
> +             return XSTR(LONG_MIN);
> +           /* We cause the same result as
> +                i = (-x) % base
> +                u = (-x) / base
> +              would in mathematics if x and base were integers.
> +              It will be the case that -base < x <= 0 after the loop.
> +              Because base > 1, the quotient will definitely fit into u.
> +              i will contain the last digit to print. */
> +           for (u = 0; x <= -base; x += base, ++u)
> +             ;
> +           i = -x;
> +           *p-- = i > 9 ? 'a' + i - 10 : '0' + i;
> +        }
> +      else
> +        u = -x;
>      }
>    else
>      u = x;

I like the comments, that's very helpful.  Do you think we should have
add a test?

LGTM!

Thanks!
Janneke

-- 
Jan Nieuwenhuizen <janneke@gnu.org> | GNU LilyPond http://lilypond.org
Freelance IT http://JoyofSource.com | Avatar® http://AvatarAcademy.com