On-the-fly while crawling,to avoid empty files.
You want options for skipping files over and/or under a certain file size?
One note though: Some sites still use e.g. a 1px x 1px transparent gif file (oldschool HTML trick for forcing browsers to render tables correctly) - but just to point out there may be legitimate very small files in a website - the example I have is just one.
I do understand skipping over large files. Adding todo list to add such an option to "Scan website | Output filters"