O(1) fair-queueing in XREP implemented

Up to now the complexity of fair-queueing in XREP was O(n).

Signed-off-by: Martin Sustrik <sustrik@250bpm.com>

1 files changed, 8 insertions, 0 deletions

diff --git a/src/pipe.hpp b/src/pipe.hpp
index d3bf866..df85bc2 100644
--- a/src/pipe.hpp
+++ b/src/pipe.hpp
@@ -25,6 +25,7 @@
 #include "ypipe.hpp"
 #include "config.hpp"
 #include "object.hpp"
+#include "stdint.hpp"
 #include "array.hpp"
 
 namespace zmq
@@ -67,6 +68,10 @@ namespace zmq
         //  Specifies the object to send events to.
         void set_event_sink (i_pipe_events *sink_);
 
+        //  Pipe endpoint can store an opaque ID to be used by its clients.
+        void set_pipe_id (uint32_t id_);
+        uint32_t get_pipe_id ();
+
         //  Returns true if there is at least one message to read in the pipe.
         bool check_read ();
 
@@ -176,6 +181,9 @@ namespace zmq
         //  asks us to.
         bool delay;
 
+        //  Opaque ID. To be used by the clients, not the pipe itself.
+        uint32_t pipe_id;
+
         //  Returns true if the message is delimiter; false otherwise.
         static bool is_delimiter (msg_t &msg_);