Add a script to parse the timing logs into performance reports

[paulfd]

This script uses the log files generated in earlier commits to output a report detailing the relative time spent in different part of the callback (filters, panning, ....) as well as timings for the background loading of files. An example report is attached here report.zip where a midi file is render through 3 different piano libraries. It is meant to be used to compare performance between patches or releases, or debug problems for users.

Usage:

usage: performance_report.py [-h] [--output OUTPUT] [--title TITLE] [-v]
                             files [files ...]

Plot performance summary and generate a detailed report on sfizz's performance

positional arguments:
  files            The csv log files to consider

optional arguments:
  -h, --help       show this help message and exit
  --output OUTPUT  The detailed output report file name
  --title TITLE    The report title
  -v, --verbose    Verbose console output