Re: [jruby-user] Re: JRuby 1.7.13: Why does [ "%0d" % nil ] not generate an error as it does under Ruby 1.9.3?

Hi Karol;

Thanks for your comments... I'm unlikely to venture into fixing it myself,
but will report it.

It is a subtle difference that was difficult to find, and one that I
*least* expected to be the source of some grief when I ported my MRI app to
JRuby.

Many thanks,
Tim


On Tue, Sep 16, 2014 at 8:12 AM, Karol Bucek <[email protected]> wrote:

>
>  Hi Tim, JRuby tries to match MRI to the "max" ... so this is likely a
> 1.9.3 / 2.1 compatibility issue.
>
> If you need it I would report it (even better try to fix it on my own) -
> it's likely just adding a nil check on the right method (Java involved).
> Esp. since this might end up as lower priority since it's probably not
> that crucial to max the error being raise ... might be wrong here.
>
> Hope that explains your concerns, K.
>
>
>
> On Tue, Sep 16, 2014 at 1:42 PM, Tim Griffin <[email protected]>
> wrote:
>
>> Hi again,
>>
>> Anyone with any thoughts on this?
>> Tim
>>
>>
>> On Wed, Sep 10, 2014 at 12:20 PM, Tim Griffin <[email protected]>
>> wrote:
>>
>>> Hi all;
>>>
>>> I've just bumped into a little anomoly between JRuby 1.7.13 and Ruby
>>> 1.9.3.
>>>
>>> *Under Ruby 1.9.3:*
>>> This statement generates an error:
>>> 3] pry(#<Package>)> *p "%0d" % nil*
>>> TypeError: can't convert nil into Integer
>>> from (pry):8:in `%'
>>> [4] pry(#<Package>)>
>>>
>>> *Under JRuby 1.7.13:*
>>> The same statement does *not* generate an error:
>>> [4] pry(#<Package>)> p "%0d" % nil.0.0.0:8080-1) [4] pry(#<Package>)> *p
>>> "%0d" % nil*
>>> 12:07:07,902 INFO  [stdout] (http-/0.0.0.0:8080-1) "0"
>>> 12:07:07,903 INFO  [stdout] (http-/0.0.0.0:8080-1) => "0"
>>>
>>>
>>> Can anyone account for the difference? Which is the "correct" response?
>>>
>>> Thanks,
>>> Tim
>>>
>>>
>>>
>>
>