Native Filesystem Parser for SGA-V2 #74
Description
Assuming my findings in MAK-Relic-Tool/SGA-Core#26 are true; pyfilesystem is EXTREMELY slow because of all my abstractions.
I think at one point my design philosophy was 'map data structures to the file' and I went on a spree of finding solutions to solve that problem; going so far as to attempting to make a c-style struct library. That didn't end well.
The other takeaway, is that parallelization (while it might provide some benefits) isn't the solution; my latest tests show that serial extraction is just as good, leaving me to believe that IO is the biggest issue.
Perhaps more testing is in order; I have yet to add timings to the extraction strategies to get a better breakdown of where in the extraction process time is being taken