Avatar of jef

jef's solution

to Forth in the OCaml Track

Published at Feb 22 2019 · 0 comments
Instructions
Test suite
Solution

Note:

This exercise has changed since this solution was written.

Implement an evaluator for a very simple subset of Forth.

Forth is a stack-based programming language. Implement a very basic evaluator for a small subset of Forth.

Your evaluator has to support the following words:

  • +, -, *, / (integer arithmetic)
  • DUP, DROP, SWAP, OVER (stack manipulation)

Your evaluator also has to support defining new words using the customary syntax: : word-name definition ;.

To keep things simple the only data type you need to support is signed integers of at least 16 bits size.

You should use the following rules for the syntax: a number is a sequence of one or more (ASCII) digits, a word is a sequence of one or more letters, digits, symbols or punctuation that is not a number. (Forth probably uses slightly different rules, but this is close enough.)

Words are case-insensitive.

Getting Started

For installation and learning resources, refer to the exercism help page.

Installation

To work on the exercises, you will need Opam and Base. Consult opam website for instructions on how to install opam for your OS. Once opam is installed open a terminal window and run the following command to install base:

opam install base

To run the tests you will need OUnit. Install it using opam:

opam install ounit

Running Tests

A Makefile is provided with a default target to compile your solution and run the tests. At the command line, type:

make

Interactive Shell

utop is a command line program which allows you to run Ocaml code interactively. The easiest way to install it is via opam:

opam install utop

Consult utop for more detail.

Feedback, Issues, Pull Requests

The exercism/ocaml repository on GitHub is the home for all of the Ocaml exercises.

If you have feedback about an exercise, or want to help implementing a new one, head over there and create an issue. We'll do our best to help you!

Submitting Incomplete Solutions

It's possible to submit an incomplete solution so you can see how others have completed the exercise.

test.ml

open Base
open OUnit2
open Forth

let print_int_list_option (xs: int list option) = match xs with
| None -> "None"
| Some xs -> "Some [" ^ String.concat ~sep:";" (List.map ~f:Int.to_string xs) ^ "]"
let ae exp got _test_ctxt = assert_equal ~printer:print_int_list_option exp got

let parsing_and_numbers_tests = [
   "numbers just get pushed onto the stack" >::
     ae (Some [1; 2; 3; 4; 5]) (evaluate ["1 2 3 4 5"]);
]


let addition_tests = [
   "can add two numbers" >::
     ae (Some [3]) (evaluate ["1 2 +"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["+"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 +"]);
]


let subtraction_tests = [
   "can subtract two numbers" >::
     ae (Some [-1]) (evaluate ["3 4 -"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["-"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 -"]);
]


let multiplication_tests = [
   "can multiply two numbers" >::
     ae (Some [8]) (evaluate ["2 4 *"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["*"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 *"]);
]


let division_tests = [
   "can divide two numbers" >::
     ae (Some [4]) (evaluate ["12 3 /"]);
   "performs integer division" >::
     ae (Some [2]) (evaluate ["8 3 /"]);
   "errors if dividing by zero" >::
     ae None (evaluate ["4 0 /"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["/"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 /"]);
]


let combined_arithmetic_tests = [
   "addition and subtraction" >::
     ae (Some [-1]) (evaluate ["1 2 + 4 -"]);
   "multiplication and division" >::
     ae (Some [2]) (evaluate ["2 4 * 3 /"]);
]


let dup_tests = [
   "copies the top value on the stack" >::
     ae (Some [1; 1]) (evaluate ["1 DUP"]);
   "is case-insensitive" >::
     ae (Some [1; 2; 2]) (evaluate ["1 2 Dup"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["dup"]);
]


let drop_tests = [
   "removes the top value on the stack if it is the only one" >::
     ae (Some []) (evaluate ["1 drop"]);
   "removes the top value on the stack if it is not the only one" >::
     ae (Some [1]) (evaluate ["1 2 drop"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["drop"]);
]


let swap_tests = [
   "swaps the top two values on the stack if they are the only ones" >::
     ae (Some [2; 1]) (evaluate ["1 2 swap"]);
   "swaps the top two values on the stack if they are not the only ones" >::
     ae (Some [1; 3; 2]) (evaluate ["1 2 3 swap"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["swap"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 swap"]);
]


let over_tests = [
   "copies the second element if there are only two" >::
     ae (Some [1; 2; 1]) (evaluate ["1 2 over"]);
   "copies the second element if there are more than two" >::
     ae (Some [1; 2; 3; 2]) (evaluate ["1 2 3 over"]);
   "errors if there is nothing on the stack" >::
     ae None (evaluate ["over"]);
   "errors if there is only one value on the stack" >::
     ae None (evaluate ["1 over"]);
]


let user_defined_words_tests = [
   "can consist of built-in words" >::
     ae (Some [1; 1; 1]) (evaluate [": dup-twice dup dup ;"; "1 dup-twice"]);
   "execute in the right order" >::
     ae (Some [1; 2; 3]) (evaluate [": countup 1 2 3 ;"; "countup"]);
   "can override other user-defined words" >::
     ae (Some [1; 1; 1]) (evaluate [": foo dup ;"; ": foo dup dup ;"; "1 foo"]);
   "can override built-in words" >::
     ae (Some [1; 1]) (evaluate [": swap dup ;"; "1 swap"]);
   "can override built-in operators" >::
     ae (Some [12]) (evaluate [": + * ;"; "3 4 +"]);
   "cannot redefine numbers" >::
     ae None (evaluate [": 1 2 ;"]);
   "errors if executing a non-existent word" >::
     ae None (evaluate ["foo"]);
]

let () =
  run_test_tt_main (
    "forth tests" >:::
      List.concat [
        parsing_and_numbers_tests; 
        addition_tests; 
        subtraction_tests; 
        multiplication_tests; 
        division_tests;
        combined_arithmetic_tests; 
        dup_tests; 
        drop_tests;
        swap_tests; 
        over_tests; 
        user_defined_words_tests
        ]
  )
open Base

type state_type =
  { stack : int list
  ; commands : (string * string) list
  ; has_error : bool }

let init_state = {stack = []; commands = []; has_error = false}

let is_number string =
  String.fold string ~init:true ~f:(fun result char ->
      if not (Char.is_digit char) then false else result )
;;

let is_command_definition string = Char.( = ) string.[0] ':'

let add_command commands string =
  let tokens = String.split string ~on:' ' |> List.to_array in
  let command_name = tokens.(1) in
  if is_number command_name
  then None
  else
    let command_body =
      String.concat_array
        ~sep:" "
        (Array.sub tokens ~pos:2 ~len:(Array.length tokens - 3))
    in
    Some ((command_name, command_body) :: commands)
;;

let math stack op = match stack with a :: b :: tl -> Some (op b a :: tl) | _ -> None
let divide stack = match stack with 0 :: _tl -> None | _ -> math stack ( / )
let dup stack = match stack with hd :: tl -> Some (hd :: hd :: tl) | _ -> None
let drop stack = match stack with _hd :: tl -> Some tl | _ -> None
let swap stack = match stack with a :: b :: tl -> Some (b :: a :: tl) | _ -> None
let over stack = match stack with a :: b :: tl -> Some (b :: a :: b :: tl) | _ -> None

let rec apply_token state token =
  match List.Assoc.find state.commands token ~equal:String.( = ) with
  | Some command ->
    let new_state = evaluate_single state command in
    if new_state.has_error then None else Some new_state.stack
  | _ ->
    (match String.lowercase token with
    | "+" -> math state.stack ( + )
    | "-" -> math state.stack ( - )
    | "*" -> math state.stack ( * )
    | "/" -> divide state.stack
    | "dup" -> dup state.stack
    | "drop" -> drop state.stack
    | "swap" -> swap state.stack
    | "over" -> over state.stack
    | number when is_number number -> Some (Int.of_string number :: state.stack)
    | _ -> None)

and evaluate_single state string =
  if is_command_definition string
  then
    match add_command state.commands string with
    | None -> {state with has_error = true}
    | Some new_commands -> {state with commands = new_commands}
  else
    String.split string ~on:' '
    |> List.fold_until
         ~init:state
         ~finish:(fun state -> state)
         ~f:(fun state token ->
           match apply_token state token with
           | None -> Stop {state with has_error = true}
           | Some new_stack -> Continue {state with stack = new_stack} )
;;

let evaluate strings =
  List.fold_until
    strings
    ~init:init_state
    ~finish:(fun state -> Some (List.rev state.stack))
    ~f:(fun state string ->
      let new_state = evaluate_single state string in
      if new_state.has_error then Stop None else Continue new_state )
;;

Community comments

Find this solution interesting? Ask the author a question to learn more.

What can you learn from this solution?

A huge amount can be learned from reading other people’s code. This is why we wanted to give exercism users the option of making their solutions public.

Here are some questions to help you reflect on this solution and learn the most from it.

  • What compromises have been made?
  • Are there new concepts here that you could read more about to improve your understanding?