[avr-gcc-list] Re: sprintf

[David Brown]

Daniel O'Connor wrote:
> On Saturday 28 February 2009 09:00:50 David Kelly wrote:
> > On Fri, Feb 27, 2009 at 10:37:30PM +0100, Vincent Trouilliez wrote:
> > > On Fri, 27 Feb 2009 22:10:16 +0100
> > >
> > > David Brown <address@hidden> wrote:
> > > >         sprintf((A_String + i), "%c", 0xff)
> > > >
> > > > has exactly the same effect as:
> > > >
> > > >         A_String[i] = 0xff;
> > > >         A_String[i + 1] = 0x00;
> > > >         delay_about_1000_processor_cycles();
> > > >         waste_about_2000_bytes_of_code_space();
> > > ROTF ! :-))))
> > sprintf() only costs 2kB? Sure enough, this little program is 2132 bytes
> > after strip:
>
> Depends which vprintf you ask for, see 
> http://www.nongnu.org/avr-libc/user-manual/group__avr__stdio.html#ga3b98c0d17b35642c0f3e4649092b9f1
>
> Also, while it IS a pig (relative to the micro flash size and CPU) it is
> very powerful and for a lot of applications the speed is not critical.

I didn't mean to imply that there is a problem with sprintf itself - 
I've had occasion to use it in my own programs.  The point is that it is 
totally inappropriate in this case - not only is direct access to the 
array hugely smaller and faster, but it is very much clearer to write 
and easier to understand.

When you want to use the features of sprintf (and friends) for text and 
number formatting, then it is much easier and faster than writing your 
own code.