Re: %Ps of t_REAL

[Bill Allombert <Bill.Allombert@math.u-bordeaux1.fr>]

On Thu, Oct 25, 2012 at 11:26:17PM +0200, Bill Allombert wrote:
> Hello PARI developers,
> 
> ? \p10000
>    realprecision = 10018 significant digits (10000 digits displayed)
> ? a=Pi;
> ? \p38
>    realprecision = 38 significant digits
> ? printf("%s",a)
> 3.1415926535897932384626433832795028842
> 
> I know this is the documented behaviour but I cannot say I ever liked it.
> I would much prefer to get 10018 digits. (If I wanted 38 digits, I would
> do the computation to 38 digits, or I would write %.38s).
> 
> It is worse in library mode where you do not have any control over
> realprecision. It is especially painful with err_printf which is 
> used for debugging, when the interesting digits are discarded.

We discussed about changing the behaviour, but there are two problems:

1) %g convert the data to t_REAL:
? printf("%g",12)
12.000000000000000000000000000000000000

2) "%s" does not allow to specify a precision. Instead:
   *  s:  Convert  to a character string.   If a precision is given,  no more
          than the specified number of characters are written.

? printf("%.8s",[Pi,23])
[3.14159

So we are stuck.

Cheers,
Bill.