Moved flushing execution to the last span in the context and improved…

… the documentation to give examples of flushing properly manually

integrations/langfuse/src/haystack_integrations/components/connectors/langfuse/langfuse_connector.py

@@ -23,18 +23,25 @@ class LangfuseConnector:
     Lastly, you may disable flushing the data after each component by setting the `HAYSTACK_LANGFUSE_ENFORCE_FLUSH` environent
     variable to `false`. By default, the data is flushed after each component and blocks the thread until the data is sent to
     Langfuse. **Caution**: Disabling this feature may result in data loss if the program crashes before the data is sent to Langfuse.
-    Make sure you will call langfuse.flush() explicitly before the program exits. e.g. by using: 
+    Make sure you will call langfuse.flush() explicitly before the program exits. e.g. by using tracer.actual_tracer.flush(): 
+    
     ```python
+    from haystack.tracing import tracer
+
     try:
         # your code here
     finally:
-        langfuse.flush()
+        tracer.actual_tracer.flush()
     ```
     or in FastAPI by defining a shutdown event handler:
     ```python
+    from haystack.tracing import tracer
+
+    # ...
+    
     @app.on_event("shutdown")
     async def shutdown_event():
-        langfuse.flush()
+        tracer.actual_tracer.flush()
     ```
 
     Here is an example of how to use it:

integrations/langfuse/src/haystack_integrations/tracing/langfuse/tracer.py

@@ -148,9 +148,12 @@ def trace(self, operation_name: str, tags: Optional[Dict[str, Any]] = None) -> I
         if len(self._context) == 1:
             # The root span has to be a trace, which need to be removed from the context after the pipeline run
             self._context.pop()
+
+            if self.enforce_flush:
+                self.flush()
 
-        if self.enforce_flush:
-            self._tracer.flush()
+    def flush(self):
+        self._tracer.flush()
 
     def current_span(self) -> Span:
         """