This looks awesome and exactly what I have been looking for.
One question about implementation just out of curiosity, is there any database? I'm worried that when it gets to hundreds or thousands of pages querying things becomes slow if it's just scanning files.