When DownloadStatus is finished, it calls ProgressDisplay.stop(), which in turn calls ProgressDisplay._print_summary()
Here, in _print_summary(), we incorrectly assume that task.completed means observed steps, but in fact, when the download finishes, it is equal to task.total, which leads to issue with incorrect average download speed - no matter where we resume the download process, we always get task.total steps
Solution
Extend the ProgressDisplay class by adding a resumed_at attribute to keep track of the progress that has already been done.
Change the ProgressDisplay.stop() method to pass task.completed - self.resumed_at instead of task.completed
Problem
When
DownloadStatusis finished, it callsProgressDisplay.stop(), which in turn callsProgressDisplay._print_summary()Here, in
_print_summary(), we incorrectly assume thattask.completedmeans observed steps, but in fact, when the download finishes, it is equal totask.total, which leads to issue with incorrect average download speed - no matter where we resume the download process, we always gettask.totalstepsSolution
ProgressDisplayclass by adding aresumed_atattribute to keep track of the progress that has already been done.ProgressDisplay.stop()method to passtask.completed - self.resumed_atinstead oftask.completedRelated
Solves the issue #1516