Cheers Gordon,
It does indeed look very similar to the preceeding " else if (type ==
TYPE_EXCHANGE || type == TYPE_TOPIC) " block so it's almost certainly
just a bit of copy/paste.
Incidentally I noticed in the preceeding block the following:
"
try {
std::pair<boost::shared_ptr<Exchange>, bool> result =
createExchange(name, exchangeType, durable,
alternateExchange, arguments, userId, connectionId);
if (!result.second) {
throw ObjectAlreadyExists(name);
}
} catch (const UnknownExchangeTypeException&) {
throw Exception(QPID_MSG("Invalid exchange type: " <<
exchangeType));
}
"
but the binding just has
"
bind(binding.queue, binding.exchange, binding.key, arguments,
userId, connectionId);
"
Which looks like the reason adding duplicate bindings doesn't cause the
broker to barf, but trying to add queues/exchanges with the same name does.
Thanks again for your response, it helped my sanity a bit - which was
much needed considering the weirdness I came across this weekend trying
to map things between the C++ and Java Brokers, I definitely think
there's a lot of scope for "unification" there :-)
Frase
On 11/03/13 09:52, Gordon Sim wrote:
On 03/10/2013 07:25 PM, Fraser Adams wrote:
would anyone be able to confirm that "exchange-type" has never
*actually* needed to be set in a QMF create binding call?
Yes, I can confirm that is the case. The type of exchange never needs
to be specified as it is implied by the exchange instance which is
identified by name. I agree the code you quoted is a confusing in that
regard, I suspect it has just been a bit of sloppy cut-and-paste.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
