Do you think future devs on this tool can use a new fast method to find content within files?
Halfway because it is fast, but it's fast because it keeps the index entirely within RAM and thus you can't yet throw an arbitrarily-large disk of stuff at it to content-index.