Use time.NewTimer() instead of time.After() (#56)

* Use `time.NewTimer()` instead of `time.After()`

For efficiency, cancel time.Timer after no longer needed.
See https://pkg.go.dev/time#After

* use defer timer.Stop() instead

---------

Co-authored-by: Andrey Kaipov <[email protected]>

Author: afriza

api/client.go

@@ -68,9 +68,12 @@ func (c *Client) SendRequest(requestBody Params, responseBody interface{}) error
 	}
 
 	var response *opcodes.RequestResponse
+
+	timer := time.NewTimer(c.ResponseTimeout * time.Millisecond)
+	defer timer.Stop()
 	select {
 	case response = <-c.IncomingResponses:
-	case <-time.After(c.ResponseTimeout * time.Millisecond):
+	case <-timer.C:
 		return fmt.Errorf("request %s: timeout waiting for response from server", name)
 	}
 

client.go

@@ -172,10 +172,12 @@ func (c *Client) connect() (err error) {
 	go c.handleRawServerMessages(authComplete)
 	go c.handleOpcodes(authComplete)
 
+	timer := time.NewTimer(c.ResponseTimeout * time.Millisecond)
+	defer timer.Stop()
 	select {
 	case a := <-authComplete:
 		return a
-	case <-time.After(c.ResponseTimeout * time.Millisecond):
+	case <-timer.C:
 		return fmt.Errorf("timeout waiting for authentication: %dms", c.ResponseTimeout)
 	}
 }
@@ -355,7 +357,7 @@ func (c *Client) writeEvent(event interface{}) {
 			// incoming events was full (but might not be by now),
 			// so safely read off the oldest, and write the latest
 			select {
-			case _ = <-c.IncomingEvents:
+			case <-c.IncomingEvents:
 			default:
 			}