On 7 Nov 2011, at 11:55, Peter Haworth wrote:
> On Fri, 4 Nov 2011 21:49:49 +0000, Dave Hodgkinson wrote:
>> On 4 Nov 2011, at 20:16, Graham Barr wrote:
>>> On Nov 4, 2011, at 11:36 , Chisel wrote:
>>>>
>>>> # why no error?!
>>>> $ perl -M5.14.0 -wle 'for my $i (qw/foo/) { eval { $i.=q{}; next; }; } say
>>>> "done"'
>>>> done
>>>
>>> Because the next never happens, check $@ you will see
>>>
>>> Modification of a read-only value attempted
>>
>> A more enlightening error would be nice.
>>
>> I find myself tempted to next in the $@ check. What's a good
>> idiom for failing out of an eval then?
>
> Isn't that what die or return are for? Assuming that eval is the only thing
> in the body of the loop, of course.
True. I'm ingesting HoP at the moment, hopefully some of that will
rub off into better structured code.
