Fix routing loop of Service request when client and server are server by the same bridge

[JEnoch]

In case a Service Server and a Client for the same Service are running in the same domain and discovered by the same bridge, the Client can directly communicate with the Server. Nevertheless, the bridge is also routing the Client's request to the Server, leading to a duplicate request received by the Server.

It seems to not cause an issue for lot of ROS Services Server (the Client will drop the 2nd reply anyway).
But for Lifecycle Nodes that are based on Services for state transitions, this leads to some error messages

E.g. to reproduce with lifecycle demo on a single host:

• ros2 run lifecycle lifecycle_talker
• ros2 run lifecycle lifecycle_listener
• zenoh-bridge-ros2dds
• ros2 run lifecycle lifecycle_service_client

On Transition 3 this error message shows up for lifecycle_talker:

[WARN] [1703856397.595288020] [rcl_lifecycle]: No transition matching 3 found for current state active
[ERROR] [1703856397.595358561] []: Unable to start transition 3 from current state active: Transition is not registered., at /Users/runner/mambaforge/conda-bld/ros-humble-rcl-lifecycle-0_1675746578761/work/ros-humble-rcl-lifecycle/src/work/src/rcl_lifecycle.c:356

This PR fixes that making sure that the Client Request when routed by the bridge as a Zenoh query doesn't loop back to the same bridge (using .allowed_destination(Locality::Remote))
The PR also improve requests/replies routing logs and operation names for clarity.

[DariusIMP]

LGTM