Front page | perl.perl5.porters |
Postings from October 2000
On Thu, Oct 05, 2000 at 01:13:11PM +0100, Simon Cozens wrote:
> On Thu, Oct 05, 2000 at 07:15:28AM -0500, Jarkko Hietaniemi wrote:
> > Urque? By the time you can are able to say C<$a+3> the looks_like_number
> > thing has already been done and you have an IV/UV/NV called $a.
>
> Eh? But $a may be a PV. If it's a PV, you have to re-examine it, at
> runtime, to see whether the PV contains an NV or an IV/UV. (If we want
Ahhh, you were talking about PVs. I think in that case the looks_like_number
is used once and then numeric version is cached.
> to split IV and NV treatment of arithmetic operators) At which point,
> you might as well just use the NV method anyway.
perl -MDevel::Peek -e '$a="123";$b=$a+3;Dump($a)'
SV = PVNV(0x140007c80) at 0x140022140
REFCNT = 1
FLAGS = (NOK,POK,pNOK,pPOK)
IV = 0
NV = 123
PV = 0x140001030 "123"\0
CUR = 3
LEN = 4
Yup, seems like it becomes an NV. Grr. It must be using
some other routine than is being used in parse time since
perl -MDevel::Peek -e '$a=123;$b=$a+3;Dump($a)'
SV = PVNV(0x140007c80) at 0x140022140
REFCNT = 1
FLAGS = (IOK,NOK,pIOK,pNOK)
IV = 123
NV = 123
PV = 0
I must wonder what's the use of ever bothering to try to parse
numbers as IVs/UVs since as soon as they are used they become NVs.
--
$jhi++; # http://www.iki.fi/jhi/
# There is this special biologist word we use for 'stable'.
# It is 'dead'. -- Jack Cohen
Thread Previous
|
Thread Next