RE: [Dev-C++] itoa - atoi : another implementation

[Ioannis V. <iv...@em...>]

> -----Original Message-----
> From: Per Westermark [mailto:pw...@ia...]=20
> Sent: Wednesday, April 02, 2003 9:44 PM
> To: Ioannis Vranos
> Cc: 'Dev-C++'
> Subject: RE: [Dev-C++] itoa - atoi : another implementation=20
>=20
>=20
> The most common is one of the following declarations:
> void itoa(int num, char *buf, int base);
> void itoa(int num, char *buf);
>=20
> Neither have problem with lost memory, and both are reentrant.
>=20
> The first form is always useful, since it allows adjustable=20
> number base,
> i.e. binary, trinary, ...



I had written some time ago an unsigned to string conversion function in =
any
number system. But after some time I erased it because except of =
hexadecimal
and octadecimal systems, it is very rare to use trinary etc in a =
program.
And for hex and oct we can do the job with the printf() family.




> The second form is only needed in embedded programs, since it requires
> less RTL support than the big-big-big sprintf() - note that sprintf()
> doesn't know whet additional support that is required, so it=20
> has code for
> left-adjust, right-adjust, int, long, signed, unsigned,=20
> string, pointer,
> float, double, ...



I haven't played with embedded systems, but for usual ones I consider =
the
horsepower/memory enough so as to have the entire C90. I think that if a
system can't hold C90, it can only be programmed with assembly. Memory =
is
very cheap these days. For example one could place an intel processor =
with 1
mb ram easily in an embedded system (e.g. just a 286 or some of the =
intel
processors used for palmtops etc, or from other companies).



Ioannis Vranos
=20
* Programming pages: http://www.noicys.freeurl.com
* Alternative URL: http://run.to/noicys