Question

I am trying to understand how the following scenario could be implemented using reactive extensions. Other solutions that I have been looking at involve managing a series of timers… something that I am looking to avoid if possible.

Scenario
A stream of “ActionRequests” occurring. The ActionRequest’s streaming in are from different object sources. I am only interested in picking off the most recent ActionRequest once requests have stopped coming in from a specific source for a period of x seconds

So breaking that down.

Events are being continuously received from different object sources.

The different object sources are denoted by a guid – see the ActionRequests class

After x number of seconds of silence from the object source the latest event is pushed out.

The older messages can be discarded once a more recent ActionRequest of interest has been received

public class ActionRequest
{
    public ActionRequest(Guid quoteId, DateTime now)
    {
        QuoteId = quoteId;
        RequestTimestamp = now;
    }

    public Guid QuoteId { get; set; }
    public DateTime RequestTimestamp { get; set; }

}

Any thoughts on this would be appreciated.

Answer 1

Answer was provided in RX Forums by Richard Hein.

Short version of the answer is below. A more complete sample has been added to the thread in the RX forums.

var throttled = observable.GroupBy(tar => tar.Value.QuoteId).SelectMany(g => g.Throttle(TimeSpan.FromMilliseconds(1500)));