Re: [chatter] [dylan-lang/opendylan] 7b3f54: Tabs-to-spaces

On Jul 16, 2013, at 11:03 AM, Hannes Mehnert <[email protected]> wrote:

> Hi,
> 
> Bruce fixed a regression of this commit already, I still have a question:
> 
> On 07/15/2013 03:42, GitHub wrote:
>>  Commit: 9f7256ae131c66e994f6a94ed4b2ee7a4ed8663f
>>      https://github.com/dylan-lang/opendylan/commit/9f7256ae131c66e994f6a94ed4b2ee7a4ed8663f
>>  Author: Dustin Voss <[email protected]>
>>  Date:   2013-07-14 (Sun, 14 Jul 2013)
>> 
>>  Changed paths:
>>    M sources/dfmc/modeling/objects.dylan
>>    M sources/dylan/collection.dylan
>>    M sources/io/print.dylan
> 
> (io/print.dylan, starting at line 853 -- modifications by the above
> commit partly stripped):
> define method write-class-name (obj-class :: <class>, stream :: <stream>)
>     => ();
>  let cname = obj-class.class-name;
>  if (cname)
>    write(stream, cname);
>   else
> -    write(stream, "<unnamed-class>");
> +    print(obj-class, stream);
>   end if;
> end method write-class-name;
> 
> Is this correct? I'm slightly worried that the call to print ends in
> non-termination (by recursively calling write-class-name somehow). Is
> there any test case for this? Is there any class whose debug-name is #f
> (class-name, implemented just above write-class-name, accesses debug-name)?
> 
> If not, we can also safely remove all these conditionals.


I will double-check to make sure print() terminates. But as to the other part, I do not know if there are any classes whose debug-name is #f.
_______________________________________________
hackers mailing list
[email protected]
https://lists.opendylan.org/mailman/listinfo/hackers