develooper Front page | perl.perl5.porters | Postings from January 2003

Re: [PATCH] operation on `PL_na' may be undefined

Thread Previous | Thread Next
From:
Nick Ing-Simmons
Date:
January 17, 2003 08:17
Subject:
Re: [PATCH] operation on `PL_na' may be undefined
Message ID:
20030117161719.2651.5@bactrian.elixent.com
Nicholas Clark <nick@unfortu.net> writes:
>Sorry about the delay in replying to the original question.
>On Wed, Dec 11, 2002 at 11:28:48AM -0500, John Peacock wrote:
>> Rafael Garcia-Suarez wrote:
>> >%_ : placeholder for an SV that will get stringified. See sv_vcatpvfn() in 
>> >sv.c.
>> >I can't find doc on this in the perl sources, but there a small paragraph
>> >about it in the Cozens & Jenness.
>
>I can't find any docs either. In fact, a straight grep on %_ in the
>current sources is most unrewarding.
>
>> Neat!  Is it supported in 5.005???

IIRC Chip put it in way back.

>
>Well, I had a grep in the 5.004_04 sources and it's in there quite a lot,
>such as:
>
>    if (in_eval & 2)
>        warn("%_", msg);
>    else if (in_eval)
>        sv_catsv(GvSV(errgv), msg);
>
>so I wondered what had happened to them all. The same piece of code is now:
>
>    if (PL_in_eval & EVAL_WARNONLY)
>        Perl_warn(aTHX_ "%"SVf, msg);
>    else
>        qerror(msg);
>
>with this in perl.h
>
>#ifndef SVf
>#  ifdef CHECK_FORMAT
>#    define SVf "p"
>#  else
>#    define SVf "_"
>#  endif
>#endif
>
>That doesn't seem to be in Devel::PPort (yet - patches welcome I guess)
>
>I wonder how many parts of the perl core can be reduced slightly in size
>by changing %s SvPV_nolen(sv) to %_ sv.
>This would have the beneficial side effect of making the non-error code path
>more direct, which might have a slight speed gain.

Again IIRC I thought that %_ was not quite the same as %s-of-PV.
That is I seem to recall one motivation for put it in was to avoid stringifying
things that were SvNOK and do %g on those and %ld on SvIOK - thuse there
might be spots where changing to %_ would change what you got.


>
>It's not in 5.003_07's sources, but I don't think anyone really supports that
>any more.
>
>Nicholas Clark
-- 
Nick Ing-Simmons
http://www.ni-s.u-net.com/


Thread Previous | Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About