Why not having a txt file with indented tree, a bootstrap function that parses the file and creates the tree and a test for that function?
replyYou will have proper diff for the tree this way.
Well, partly because I already have a tool similar to that in mind as a separate project :) (The files also need to have certain content, so I need something a bit more complex that can specify that.)
replyBut mainly because, as you acknowledge, that process needs its own tests. Basic unzipping functionality is already tested for me.