That's intentional, they do it both to stop adblockers from working and to fight scrapers. Last I checked, the mangling algorithm switched up fairly regularly but I don't use Facebook much these days so it might have changed.
It might be easier to screen capture the page and extract text via OCR than deal with a constantly shifting DOM, since at least the former (screenshots) will remain more likely static.
Ahh, OK. I remember reading about that vaguely. Coincidence I chose FB. I know Google's DOM is equally unintelligible.