Introduction to the OCaml Toplevel

An OCaml toplevel is a chat between the user and OCaml. The user writes OCaml code, and UTop evaluates it. This is why it is also called a Read-Eval-Print-Loop (REPL). Several OCaml toplevels exist, like ocaml and utop. We recommend using UTop, which is part of the OCaml Platform toolchain.

To run UTop, we use the utop command, which looks like this:

$ utop
────────┬─────────────────────────────────────────────────────────────┬─────────
        │ Welcome to utop version 2.12.1 (using OCaml version 5.0.0)! │
        └─────────────────────────────────────────────────────────────┘

Type #utop_help for help about using utop.

─( 17:00:09 )─< command 0 >──────────────────────────────────────{ counter: 0 }─
utop #

Press Ctrl-D (end of file) or enter #quit;; to exit utop.

UTop displays a hash prompt #, similar to the $ in the CLI. This # means it is waiting for input, so you can start writing your code after the prompt. To evaluate it, add a double semicolon ;; to signal the end of the expression and press Enter.

For instance, consider the following code snippet:

# 2 + 2;;
- : int = 4

In the code snippet above, 2 + 2;; is the user's input, and - : int = 4 is the output of OCaml.

If you need to amend the code before hitting Enter, you can use your keyboard's right and left arrows to move inside the text. The up and down arrows allow navigation through previously evaluated expressions. Typing Enter without a double semicolon ;; will create a new line, so you can write multiple-line expressions this way.

Commands beginning with a hash character #, such as #quit or #help, are not evaluated by OCaml; they are interpreted as commands by UTop.

You're now ready to hack with UTop! If you hit any issue with the toplevel, don't hesitate to ask on Discuss.