Assign6 (Spring-2025.Specifications.Assign6)

Programming

The purpose of this assignment is to give you a taste of the mini-projects by building an S-expressions calculator. Here is a simple example of an expression in this language.

(?
  (< (+ 2 3) (+ 3 3))
  0
  (+ 1 1))

This expression has type $\texttt{int}$ and evaluates to $\texttt{0}$ .

In more familiar OCaml-like syntax, this would look like:

if 2 + 3 < 3 + 3 then 0 else 1 + 1

In the mini-projects, you'll be given three things: syntax, typing rules, and semantics. It will then be your task to implement an interpreter. Note that several of the parts below require little to no new code.

Here's the syntax of our toy language.

\begin{align*} \textcolor{blue}{\texttt{<expr>}} &::= \textcolor{blue}{\texttt{<int>}} \\ &\hspace{3mm}| \hspace{2.5mm} \textcolor{red}{\texttt{(}} \ \textcolor{red}{\texttt{+}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{red}{\texttt{)}} \\ &\hspace{3mm}| \hspace{2.5mm} \textcolor{red}{\texttt{(}} \ \textcolor{red}{\texttt{<}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{red}{\texttt{)}} \\ &\hspace{3mm}| \hspace{2.5mm} \textcolor{red}{\texttt{(}} \ \textcolor{red}{\texttt{?}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{blue}{\texttt{<expr>}} \ \textcolor{red}{\texttt{)}} \\ \end{align*}

Next, the typing rules. Expressions can either be integers or Boolean values. These should feel familiar. We'll elide names because we won't be writing derivations for this system. Note that there is no need for a context because there are no local variables!

\frac {n \text{ is an integer literal}} {n : \texttt{int}} \qquad \frac {e_1 : \texttt{int} \qquad e_2 : \texttt{int} } {\texttt{(} \ \texttt{+} \ e_1 \ e_2 \ \texttt{)} : \texttt{int}} \qquad \frac {e_1 : \texttt{int} \qquad e_2 : \texttt{int} } {\texttt{(} \ \texttt{<} \ e_1 \ e_2 \ \texttt{)} : \texttt{bool}} \qquad \frac {e_1 : \texttt{bool} \qquad e_2 : \tau \qquad e_3 : \tau } {\texttt{(} \ \texttt{?} \ e_1 \ e_2 \ e_3 \ \texttt{)} : \tau}

Finally, the semantics. These should also feel familiar.

\frac {n \text{ is an integer literal}} {n \Downarrow n} \qquad \frac {e_1 \Downarrow v_1 \qquad e_2 \Downarrow v_2 \qquad v_1 + v_2 = v} {\texttt{(} \ \texttt{+} \ e_1 \ e_2 \ \texttt{)} \Downarrow v} \qquad \frac {e_1 \Downarrow v_1 \qquad e_2 \Downarrow v_2 \qquad v_1 < v_2} {\texttt{(} \ \texttt{<} \ e_1 \ e_2 \ \texttt{)} \Downarrow \top}

\frac {e_1 \Downarrow v_1 \qquad e_2 \Downarrow v_2 \qquad v_1 \geq v_2} {\texttt{(} \ \texttt{<} \ e_1 \ e_2 \ \texttt{)} \Downarrow \bot} \qquad \frac {e_1 \Downarrow \top \qquad e_2 \Downarrow v_2} {\texttt{(} \ \texttt{?} \ e_1 \ e_2 \ e_3 \ \texttt{)} \Downarrow v_2} \qquad \frac {e_1 \Downarrow \bot \qquad e_3 \Downarrow v_3} {\texttt{(} \ \texttt{?} \ e_1 \ e_2 \ e_3 \ \texttt{)} \Downarrow v_3}

Problem 1: Parsing

In rough terms, lexing (or tokenizing) is the process of converting a string into a list of tokens. Tokens are the abstract units of a programming language. We convert our input program (a string) into a list of tokens so that we don't need to deal with low-level concerns like whitespace when parsing. You're given an implementation of a lexer.

type tok =

| LParen
| RParen
| TNum of int
| TAdd
| TLt
| TIf

val lex : string -> tok list option

lex s is Some tks where tks is the list of tokens representing s and None otherwise. It's a variant of tokenize from Lab 4, which you can check for more details.

We'll use an ADT to represent expressions in our language, similar to previous assignments.

type expr =

| Num of int
| Add of expr * expr
| Lt of expr * expr
| If of expr * expr * expr

val parse : string -> expr option

Implement the function parse so that parse str is

Some e if str represents a well-formed program;
None otherwise.

You've already built a parser for S-expressions in Lab 4. Repurpose that code to work for this new expr ADT (as opposed to nonempty trees) and compose it with the given function lex.

Problem 2: Type-Checking

type ty =

| TInt
| TBool

val type_of : expr -> ty option

Implement the function type_of so that type_of e is

Some t where t is the type of e if e is well-typed;
None otherwise.

An expression $e$ is well-typed if if $e : \tau$ is derivable according to the above typing rules for some type $\tau$ .

Problem 3: Evaluating

type value =

| VNum of int
| VBool of bool

val eval : expr -> value

Implement the function eval so that eval e is v is the value of e given that e is well-typed. The behavior of the implementation is undefined if e is not well-typed (i.e., if type_of e is None).

The Executable

Once you've finished the above problems, you can run your interpreter on an actual program. There are several example programs in the directory examples. If you've done everything correctly, you should be able to run the command

dune exec assign6 examples/file_name

(replace file_name with the name of one of the files in that directory) which will print out the value of the expression in the file. You can even write your own programs if you'd like.

To be clear, there are no tasks for the assignment in this section, this is just for fun. That said, please look through the file bin/main.ml. This file chains together the lexer, parser, type checker, and evaluator. In the future, we will be asking you to write this part.

Written

Ambiguity

<s> ::= A <s> B
      | <a>
<a> ::= A <a>
      | B <b>
<b> ::= B B <b>
      | B B

Determine a sentence with fewer than 8 symbols which has two distinct leftmost derivations in the above grammar. Also write down these two leftmost derivations.

Designing Grammars

Write down a grammar over the terminal symbols A and B such that every sentence recognized by the grammar is any number of As, followed by a sequence of at least 2 Bs. For example AAAABBB, BBB, and AABB should be sentences recognized by the grammar, whereas AAAB, AB, and A should not. You may introduce any nonterminal symbols that you want.