Advance output of last points in continuous scan

[reszelaz]

#1651 points that output of last points in continuous scan often comes too late.

• Change logger for data events processing messages to GSF classes level (CTScan and TScan) instead of the macro level. These messages may be confusing if macro is run in debug mode (the messages are interleaved with the output fo the scan records)
• Delay overshoot correction until after ending all records.

Fixes #1651.

[reszelaz]

I have implemented the delayed info on overshoot correction as proposed in #1651 (comment).
The output now looks like this, in case the overshoot correction was done:

   99       9.9       0.1       0.2       0.3       0.4       11.9  
  100        10       nan       nan       nan       nan        12   
Overshoot was corrected
Operation saved in /tmp/test2d.h5 (HDF5::NXscan)
Scan #17 ended at Fri Oct 29 11:16:06 2021, taking 0:00:15.137578. Dead time 99.7% (setup time 0.3%, motion dead time 99.7%)

and in case it was disabled (senv ScanOvershootCorrection False):

   99       9.9       nan       nan       nan       nan       11.9  
  100        10       0.1       0.2       0.3       0.4        12   
Overshoot was not corrected
Operation saved in /tmp/test2d.h5 (HDF5::NXscan)
Scan #18 ended at Fri Oct 29 11:22:56 2021, taking 0:00:15.207367. Dead time 99.8% (setup time 0.2%, motion dead time 99.8%)

Before it could look like this:

   99       9.9       0.1       0.2       0.3       0.4       11.9  
Correcting overshoot...
  100        10       nan       nan       nan       nan        12   
Operation saved in /tmp/test2d.h5 (HDF5::NXscan)
Scan #16 ended at Fri Oct 29 11:06:15 2021, taking 0:00:15.607811. Dead time 99.9% (setup time 0.1%, motion dead time 99.9%)

In case of the ascanc macros the overshoot correction info messages are printed in the old way - there we are not processing the data events asynchronously so there is no risk to mix with the records output.

@johanfforsberg could you please review and eventually merge it? Thanks!