[rabbitmq-discuss] changes to rabbit_amqqueue:declare in trunk

[anders conbere]

On Mon, Sep 15, 2008 at 12:36 AM, Matthias Radestock
<matthias at lshift.net> wrote:
> Anders,
>
> anders conbere wrote:
>>
>> On Mon, Sep 15, 2008 at 12:13 AM, Matthias Radestock
>> <matthias at lshift.net> wrote:
>>>
>>> There's a type signature near the beginning of rabbit_amqqueue:
>>>
>>> -spec(declare/4 :: (queue_name(), bool(), bool(), amqp_table()) ->
>>>            amqqueue()).
>>
>> Yeah, I just didn't find "queue_name" to be particularly illuminating
>> about what the actual value should be.
>
> I got a bit carried away with the type abstractions ;) ...
>
> Take a look at rabbit.hrl:
>
> -type(queue_name() :: r('queue')).
> -type(r(Kind) ::
>      #resource{virtual_host :: vhost(),
>                kind         :: Kind,
>                name         :: resource_name()}).
>
> and then rabbit_framing_spec.hrl:
>
> -type(resource_name() :: binary()).
> -type(vhost() :: binary()).

Ah interesting, I haven't quite seen anything like that in erlang yet.

~ Anders

>
>
> Matthias.
>