I wouldn’t disregard it simply due to its age - that is a bad habit I see all too often (“Old = bad!”). Just to make sure before posting this, I went to some random parts of the book and pasted example code into a utop REPL for the latest Ocaml version and it worked fine. So it’s not like the code has bit rotted. It’s still worth a read if someone is interested in languages. It is not worth reading if you are looking for a tutorial on spinning up some form of web app or server app that uses all the shiny things people are currently excited about.
As far as unicode goes, strings in OCaml are just bytes, so it won't mangle any unicode without you telling it to. If you need to process text that uses code points above the ASCII ones, there is camomile, which can be installed easily with opam and has documentation here: http://camomile.sourceforge.net/dochtml/CamomileLibrary.html . I haven't used it but it looks straightforward enough.
Note that sometimes people in forums will complain about an OCaml library not being updated in years. In the OCaml universe, this is normal. Many libraries don't get much churn and continue to work unmodified for years. As long as they've been modified since the switch to immutable strings, and as long as they don't depend on libraries that do churn (Base does a bit, as it's still in development, but they'll stop soon I think) they'll probably work. I use several such.
I don't mind writing dune files, but yeah, ocaml tooling is a bit lacking by modern standards...
dnf install ocaml ocaml-camomile-devel emacs-tuareg-mode