Immediate auto analysis and batched logging #48
Description
Describe the bug
I have been able to observe that with batched logging, there are instances where the exception/error logs for a failed test do not appear in ReportPortal until after the scenario is completed and when the launch has been finished. This makes immediate auto analysis unavailable for use here as it requires the exception logs at the time of the scenario finishing.
After diving into the source code of the behave agent, I see that in the client it flushes the log batcher when finishing the launch. I think something similar may need to happen before sending the request to finish a scenario so that the exception logs are available to the analyzer for immediate auto analysis.
self._log(self._log_batcher.flush())