This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Hexadecimal float literals, for perl #122219
authorJarkko Hietaniemi <jhi@iki.fi>
Tue, 12 Aug 2014 12:39:40 +0000 (08:39 -0400)
committerJarkko Hietaniemi <jhi@iki.fi>
Thu, 14 Aug 2014 02:44:06 +0000 (22:44 -0400)
commit61e61fbc49c95894c46dd5128133731fc2092e07
tree000efc7c83f1e235f440163bca5f467b3299558b
parent40bca5ae9c72f416f0e0e056ecf8e205a03e5be3
Hexadecimal float literals, for perl #122219

Do not use system strtod/strtold since its C99 (which we do not require).

If we can fit the mantissa bits of a NV to a UV (which is the case for
the most common format, 64-bit IEEE 754 double, which has 52 mantissa bits)
we will use that, and only in the end convert to NV.  If we cannot fit
the bits, accumulate the bits incrementally to a NV.
MANIFEST
perl.h
pod/perldata.pod
pod/perldiag.pod
t/op/hexfp.t [new file with mode: 0644]
toke.c