Github user gemmellr commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/2115#discussion_r191792711
--- Diff:
artemis-protocols/artemis-amqp-protocol/src/main/java/org/apache/activemq/artemis/protocol/amqp/broker/AMQPMessage.java
---
@@ -606,22 +606,36 @@ public String getAddress() {
@Override
public AMQPMessage setAddress(String address) {
- this.address = SimpleString.toSimpleString(address,
coreMessageObjectPools == null ? null :
coreMessageObjectPools.getAddressStringSimpleStringPool());
+ internalSetAddress(address);
+ setProtonAddress(address);
return this;
}
+ private void internalSetAddress(String address) {
+ this.address = SimpleString.toSimpleString(address,
coreMessageObjectPools == null ? null :
coreMessageObjectPools.getAddressStringSimpleStringPool());
+ }
+
@Override
public AMQPMessage setAddress(SimpleString address) {
this.address = address;
+ setProtonAddress(address.toString());
--- End diff --
I must be misinterpretting things then. The setAddress methods (again,
there are two) appear used already in several places of note within the broker.
For example, I interpret this bit as meaning every AMQP message received on
a link with a target address will have it called on it, is that the case?
https://github.com/apache/activemq-artemis/blob/2.6.0/artemis-protocols/artemis-amqp-protocol/src/main/java/org/apache/activemq/artemis/protocol/amqp/broker/AMQPSessionCallback.java#L448
If so, doesnt that mean that anyone reencoding the message for any reason,
even if they have changed nothing, might find the payload changes due to the
setAddress side effects?
---
