I have just "academic" question for Git experts.

I could not find anything about how Git actually internaly compose files from all the commits. I could not imagine that it iterate through all commits untill now (go from the very first commit and keep changing/adding/deleting etc. until it have current version).

Wouldn't it be slow for large codebases with MANY commits, many big files etc.? But maybe I am imagine it wrong. Maybe it do the work once in a while and save the "complete picture" after some commits to keep the "composing" fast, so it doesn't need to start from the very beginning, I dont know :) ... something like 'memoization'.

Has anybody here better understanding of what is going on internally in Git when it figuring out current files?