Identical sub trees sounds like a pretty reasonable interview question, probably an hour or two of work (more because you want to actually do it without hand wa

Identical sub trees sounds like a pretty reasonable interview question, probably an hour or two of work (more because you want to actually do it without hand waiving!). Create a tree datastructure that mimics the file system and hash each file in place in the tree. Then create a new tree structure, where for each parent, you sort the children's hash's and hash them. Any duplicate hashes should be identical. I actually wouldn't be surprised if just straight: find $ROOT_OF_SEARCH -type f | xargs shasum | sort > file_hashes cat file_hashes | awk '{print $1}' | sort | uniq -c | sort -n > hash_freq vim -O file_hashes hash_freq was good enough > subtrees that are older versions of newer trees etc This is more complicated. How would you concretely define the relationship between older and newer versions?