for lvalue_sub #6761
Comments
From @AbigailCreated by @Abigail#!/usr/bin/perl use strict; my @foo = qw /one two three/; print for foo; However, I expect it to print "onetwothree" (which is does in 5.6.0). This bug occurs from 5.6.1 up to and including bleadperl. Abigail Perl Info |
From kasei@cpan.orgpadav is leaving an arrayref on the stack when it told to return an |
|
The RT System itself - Status changed from 'new' to 'open' |
From voropaev.andrey@gmail.comCreated by voropaev.andrey@gmail.comI know, that lvalue subs are experimental. So this is not really an issue, just The code example ########################################## sub canmod : lvalue { @a; } Produces --------------------------------------------- Perl Info |
From kasei@cpan.orgAttached a fix for lvalue subs returning an array into the args of I've modified leavesublv in that case to convert the returned arrayref |
From kasei@cpan.org0001-fix-for-23790.patchFrom ba8f5d50c4a9a94867872ba5c341fa7e5b546e48 Mon Sep 17 00:00:00 2001
From: Marty Pauley <marty@martian.org>
Date: Sun, 24 Oct 2010 18:02:40 +0900
Subject: [PATCH] fix for #23790.
padav is leaving an arrayref on the stack when producing the return value for an
lvalue sub. But when this is in an argument list it really should be a array,
not a ref. So, in leavesublv I check for this case and expand the arrayref to
an array.
---
pp_hot.c | 23 +++++++++++++++++++++++
t/op/sub_lval.t | 12 +++++++++---
2 files changed, 32 insertions(+), 3 deletions(-)
diff --git a/pp_hot.c b/pp_hot.c
index 34542c2..2d6cba9 100644
--- a/pp_hot.c
+++ b/pp_hot.c
@@ -2598,6 +2598,29 @@ PP(pp_leavesublv)
if (gimme == G_SCALAR)
goto temporise;
if (gimme == G_ARRAY) {
+ mark = newsp + 1;
+ /* We want an array here, but padav will have left us an arrayref for an lvalue,
+ * so we need to expand it */
+ if(SvTYPE(*mark) == SVt_PVAV) {
+ AV *const av = MUTABLE_AV(*mark);
+ const I32 maxarg = AvFILL(av) + 1;
+ (void)POPs; /* get rid of the array ref */
+ EXTEND(SP, maxarg);
+ if (SvRMAGICAL(av)) {
+ U32 i;
+ for (i=0; i < (U32)maxarg; i++) {
+ SV ** const svp = av_fetch(av, i, FALSE);
+ SP[i+1] = svp
+ ? SvGMAGICAL(*svp) ? (mg_get(*svp), *svp) : *svp
+ : &PL_sv_undef;
+ }
+ }
+ else {
+ Copy(AvARRAY(av), SP+1, maxarg, SV*);
+ }
+ SP += maxarg;
+ PUTBACK;
+ }
if (!CvLVALUE(cx->blk_sub.cv))
goto temporise_array;
EXTEND_MORTAL(SP - newsp);
diff --git a/t/op/sub_lval.t b/t/op/sub_lval.t
index aedaba0..d0ba84a 100644
--- a/t/op/sub_lval.t
+++ b/t/op/sub_lval.t
@@ -3,7 +3,7 @@ BEGIN {
@INC = '../lib';
require './test.pl';
}
-plan tests=>74;
+plan tests=>76;
sub a : lvalue { my $a = 34; ${\(bless \$a)} } # Return a temporary
sub b : lvalue { ${\shift} }
@@ -527,8 +527,7 @@ TODO: {
is($blah, 8, "yada");
}
-TODO: {
- local $TODO = "bug #23790";
+{ # bug #23790
my @arr = qw /one two three/;
my $line = "zero";
sub lval_array () : lvalue {@arr}
@@ -538,6 +537,13 @@ TODO: {
}
is($line, "zeroonetwothree");
+
+ sub trythislval { scalar(@_)."x".join "", @_ }
+ is(trythislval(lval_array()), "3xonetwothree");
+
+ sub changeme { $_[2] = "free" }
+ changeme(lval_array);
+ is("@arr", "one two free");
}
{
--
1.7.2.3
|
From @cpansproutOn Mon Nov 01 08:46:20 2010, maokt wrote:
Thank you. Applied as 91e34d8. |
|
@cpansprout - Status changed from 'open' to 'resolved' |
From @cpansproutOn Thu Oct 28 02:23:36 2010, voropaev.andrey@gmail.com wrote:
Thank you for the report. This is the same as bug #23790, which has just |
|
The RT System itself - Status changed from 'new' to 'open' |
|
@cpansprout - Status changed from 'open' to 'resolved' |
From @cpansproutOn Thu Nov 04 00:35:16 2010, sprout wrote:
For the record: That patch only worked if the array were the first thing |
Migrated from rt.perl.org#23790 (status was 'resolved')
Searchable as RT23790$
The text was updated successfully, but these errors were encountered: