Front page | perl.perl5.porters |
Postings from June 2008
Rafael Garcia-Suarez schrieb:
> 2008/6/20 via RT Nicholas Clark <perlbug-followup@perl.org>:
>> $ ./perl -Ilib -e 'sub foo {return return}'
>> $
>>
>> How come this isn't a syntax error? It can't possibly do anything useful.
>> (Actually, it seems that anything can be passed the return value of return.
>> C<print return> works just fine).
>
> return is technically parsed as a list operator (LSTOP).
> That enables using it in compound expressions.
>
> Note that this syntax works also for other control flow statements :
>
> $ perl -wle '$i=3;while(1){print($i--||last)}'
> 3
> 2
> 1
>
> I don't really see a reason to change it.
Well, if we once want to emit a warning, here it is:
$ ./perl -e'sub foo {return return}; foo'
Ignored return return at -e line 1.
$ ./perl -e'sub foo {return 'xx', return}; foo'
Ignored return return at -e line 1.
$ ./perl -e'sub foo {return 1 || return}; foo'
$ ./perl -e'sub foo {return 0 || return}; foo'
Ignored return return at -e line 1.
I'm not entirely sure if optimizing it away is correct though.
The only way I found is in peep(), not in ck_return() and not in perly.y
See
http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2008-06/msg00650.html
diff -u op.c.orig op.c
--- op.c.orig 2008-06-17 11:25:08.000000000 +0000
+++ op.c 2008-06-24 16:40:24.359375000 +0000
@@ -8822,6 +8822,18 @@
assert (!cPMOP->op_pmstashstartu.op_pmreplstart);
}
break;
+
+ case OP_RETURN:
+ if ((o->op_type != OP_LEAVESUBLV)
+ && (o->op_next->op_type == OP_RETURN))
+ { /* [perl #56150] return return */
+ Perl_warner(aTHX_ packWARN(WARN_SYNTAX),
+ "Ignored return return");
+ o->op_type = OP_NULL;
+ /*o->op_ppaddr = PL_ppaddr[OP_NULL];*/
+ o->op_opt = 1;
+ }
+ break;
}
oldop = o;
}
--
Reini Urban
http://phpwiki.org/ http://murbreak.at/
Thread Previous
|
Thread Next