This is a rust project now, let's end the segregation

This commit is contained in:
2024-11-12 13:56:37 -06:00
parent 0fddcbd645
commit 201d7d2594
14 changed files with 1 additions and 267 deletions

View File

@@ -1,20 +0,0 @@
.PHONY: all clean
all: deelang
CC = g++
CPPFLAGS = -g
lexer.o lexer.h: parser.h lexer.l
flex -o lexer.cpp --header-file=lexer.h lexer.l
$(CC) -c lexer.cpp
rm lexer.cpp
parser.h parser.o: parser.y
bison -o parser.cpp --header=parser.h parser.y
$(CC) -c parser.cpp
rm parser.cpp
deelang: deelang.cpp lexer.o parser.o syntax.o
$(CC) $(CPPFLAGS) -o $@ $^
clean:
rm -rf *.o parser.h lexer.h deelang

View File

@@ -1,11 +0,0 @@
#include "syntax.h"
#include "lexer.h"
#include "parser.h"
using namespace std;
int main() {
yyparse();
return 0;
}

72
src/emitter.rs Normal file
View File

@@ -0,0 +1,72 @@
use std::io::Write;
use std::collections::HashSet;
use crate::parser;
pub struct LexicalContext<'a> {
parent: Option<&'a LexicalContext<'a>>,
local: HashSet<String>
}
impl <'a> LexicalContext<'a> {
pub fn toplevel() -> Self {
LexicalContext {
parent: None,
local: HashSet::new()
}
}
#[allow(dead_code)]
fn new(parent: &'a LexicalContext<'a>) -> Self {
LexicalContext {
parent: Some(parent),
local: HashSet::new(),
}
}
fn contains(&self, s: &str) -> bool {
self.local.contains(s) || self.parent.map_or(false, |c| c.contains(s))
}
fn insert(&mut self, s: String) -> bool {
self.local.insert(s)
}
}
pub fn emit_all(w: &mut dyn Write, ast: &Vec<parser::Stmt>, ctx: &mut LexicalContext) -> std::io::Result<()> {
for stmt in ast {
emit(w, stmt, ctx)?;
}
Ok(())
}
pub fn emit(w: &mut dyn Write, stmt: &parser::Stmt, ctx: &mut LexicalContext) -> std::io::Result<()> {
match &stmt {
parser::Stmt::ReplPrint(expr) => {
write!(w, "console.log((")?;
emit_expr(w, expr, ctx)?;
writeln!(w, "));")?;
}
parser::Stmt::Assignment(id, expr) => {
if !ctx.contains(id) {
ctx.insert(id.clone());
write!(w, "let ")?;
}
write!(w, "{} = ", id)?;
emit_expr(w, expr, ctx)?;
writeln!(w, ";")?;
}
_ => todo!(),
}
Ok(())
}
pub fn emit_expr(w: &mut dyn Write, expr: &parser::Expr, ctx: &mut LexicalContext) -> std::io::Result<()> {
match &expr {
parser::Expr::Id(id) => {
write!(w, "{}", id)?;
}
parser::Expr::Atom(atom) => {
write!(w, "{}", atom)?;
}
_ => todo!(),
}
Ok(())
}

137
src/evaluator.rs Normal file
View File

@@ -0,0 +1,137 @@
use std::ops;
use std::collections::HashMap;
use crate::parser::{self, GuardedBlock};
use parser::Atom;
pub struct Env<'a> {
pub parent: Option<&'a Env<'a>>,
pub values: HashMap<String, Atom>
}
impl<'a> Env<'a> {
pub fn global() -> Self {
Env {
parent: None,
values: HashMap::new(),
}
}
pub fn child(parent: &'a Env<'a>) -> Self {
Env {
parent: Some(parent),
values: HashMap::new(),
}
}
pub fn lookup(&self, id: &str) -> Atom {
if let Some(a) = self.values.get(id) {
a.clone()
} else if let Some(parent) = self.parent {
parent.lookup(id)
} else {
panic!("Variable {} not in scope.", id)
}
}
pub fn set(&mut self, id: String, atom: Atom) {
self.values.insert(id, atom);
}
}
pub fn eval(ast: &parser::Stmt, env: &mut Env) {
match &ast {
parser::Stmt::ReplPrint(expr) =>
println!("{}", eval_expr(expr, env)),
parser::Stmt::Assignment(id, expr) =>
env.set(id.clone(), eval_expr(expr, env)),
parser::Stmt::Conditional(guarded_blocks, default_block) => {
let mut matched = false;
for GuardedBlock { guard, block } in guarded_blocks {
let res = eval_expr(guard, env);
match res {
Atom::Bool(true) => {
matched = true;
for stmt in block {
eval(stmt, env);
}
break;
}
Atom::Bool(false) => continue,
_ => panic!("Conditional expression does not evaluate to a bool"),
}
}
if !matched {
if let Some(block) = default_block {
eval_expr(block, env);
}
}
}
_ => todo!(),
}
}
pub fn eval_expr(ast: &parser::Expr, env: &Env) -> Atom {
match ast {
parser::Expr::Id(a) => env.lookup(a),
parser::Expr::Atom(a) => a.clone(),
parser::Expr::UnaryMinus(a) => -eval_expr(a, env),
parser::Expr::Plus(a, b) => eval_expr(a, env) + eval_expr(b, env),
parser::Expr::Minus(a, b) => eval_expr(a, env) - eval_expr(b, env),
parser::Expr::Mult(a, b) => eval_expr(a, env) * eval_expr(b, env),
parser::Expr::Div(a, b) => eval_expr(a, env) / eval_expr(b, env),
_ => panic!("Couldn't evalute expression {{ {:?} }}", ast),
}
}
impl ops::Neg for Atom {
type Output = Self;
fn neg(self) -> Self {
match self {
Atom::Num(a) => Atom::Num(-a),
_ => panic!("Can't negate non-numeral type!"),
}
}
}
impl ops::Add for Atom {
type Output = Self;
fn add(self, other: Self) -> Self {
match (self, other) {
(Atom::Num(a), Atom::Num(b)) => Atom::Num(a + b),
_ => panic!("Can't add non-numeral types!"),
}
}
}
impl ops::Sub for Atom {
type Output = Self;
fn sub(self, other: Self) -> Self {
match (self, other) {
(Atom::Num(a), Atom::Num(b)) => Atom::Num(a - b),
_ => panic!("Can't subtract non-numeral types!"),
}
}
}
impl ops::Mul for Atom {
type Output = Self;
fn mul(self, other: Self) -> Self {
match (self, other) {
(Atom::Num(a), Atom::Num(b)) => Atom::Num(a * b),
_ => panic!("Can't multiply non-numeral types!"),
}
}
}
impl ops::Div for Atom {
type Output = Self;
fn div(self, other: Self) -> Self {
match (self, other) {
(Atom::Num(a), Atom::Num(b)) => Atom::Num(a / b),
_ => panic!("Can't divide non-numeral types!"),
}
}
}

View File

@@ -1,54 +0,0 @@
%{
#include <stack>
#include <string>
#include "syntax.h"
#include "parser.h"
using namespace std;
stack<int> s;
#define YY_USER_INIT s.push(0); BEGIN(freshline);
%}
%option noyywrap
%x freshline
digit [0-9]
letter [A-Za-z]
%%
<freshline>""/[^\t ] {
if (s.top() == 0) {
BEGIN(0);
} else {
s.pop(); yyless(0); return DEDENT;
}
}
<freshline>[ \t]+ {
if (s.top() == yyleng) {
BEGIN(0); // Same indentation, continue
} else if (s.top() > yyleng) {
s.pop(); yyless(0); return DEDENT;
// Same rule again until the stack is even
} else {
s.push(yyleng); BEGIN(0); return INDENT;
}
}
#[^\n]* // Eat comments
\"[^"]*\" yylval.sym = new string(yytext); return STRING;
if return IF;
else return ELSE;
elif return ELIF;
{digit}+|{digit}*\.{digit}+ yylval.num = atof(yytext); return NUM;
{letter}({letter}|{digit}|[?.-])* yylval.sym = new string(yytext); return ID;
"<-" return GETS;
"->" return MAPS;
".." return CAT;
[(){}.,*/+-] return yytext[0];
[\t ] // Eat whitespace not first on a line
"/"\n // Eat newlines ending in /
[\n;] BEGIN(freshline); return STOP;
. fprintf(stderr, "Scanning error!\nOffender: %s\n", yytext); exit(1);
%%

63
src/main.rs Normal file
View File

@@ -0,0 +1,63 @@
mod parser;
mod evaluator;
mod emitter;
use clap::Parser;
use std::io;
use std::io::prelude::*;
use std::fs::File;
use std::path::PathBuf;
#[derive(Parser, Debug)]
#[clap(author, version, about = "Interpreter for Deelang")]
struct Cli {
#[clap(help="Specify a file to run")]
file: Option<PathBuf>,
#[clap(short, long, help="Only parse, do not evaluate")]
parse_only: bool,
#[clap(short, long, help="Cross-compile to ECMAScript")]
ecmascript: bool,
}
fn repl(cli: &Cli) {
let mut global = evaluator::Env::global();
let mut toplevel = emitter::LexicalContext::toplevel();
let mut out = io::stdout();
loop {
let mut line = String::new();
io::stdin().read_line(&mut line).unwrap();
let tree = parser::parse_stmt(&line);
if cli.parse_only {
println!("{:#?}", tree);
} else if cli.ecmascript {
emitter::emit(&mut out, &tree, &mut toplevel).ok();
} else {
evaluator::eval(&tree, &mut global);
}
}
}
fn script(cli: &Cli) {
let mut file = File::open(cli.file.as_ref().unwrap()).expect("Could not read file");
let mut prgm = String::new();
file.read_to_string(&mut prgm).unwrap();
let tree = parser::parse(&prgm);
if cli.parse_only {
println!("{:#?}", tree);
} else if cli.ecmascript {
let mut out = io::stdout();
let mut toplevel = emitter::LexicalContext::toplevel();
emitter::emit_all(&mut out, &tree, &mut toplevel).ok();
} else {
todo!();
}
}
fn main() {
let cli = Cli::parse();
match cli.file {
None => repl(&cli),
Some(_) => script(&cli),
}
}

386
src/parser.rs Normal file
View File

@@ -0,0 +1,386 @@
use std::fmt;
#[derive(Debug,PartialEq,Clone)]
pub enum Stmt {
Assignment(String, Expr),
Funcall(Expr),
Conditional(Vec<GuardedBlock>, Option<Expr>),
ReplPrint(Expr),
}
#[derive(Debug,PartialEq,Clone)]
pub enum Expr {
Id(String),
Atom(Atom),
Funcall(String, Vec<Expr>),
Funcdef(Option<String>, Box<Expr>),
UnaryMinus(Box<Expr>),
Plus(Box<Expr>, Box<Expr>),
Minus(Box<Expr>, Box<Expr>),
Mult(Box<Expr>, Box<Expr>),
Div(Box<Expr>, Box<Expr>),
Block(Vec<Stmt>),
Object(Vec<Stmt>),
}
#[derive(Debug,PartialEq,Clone)]
pub enum Atom {
String(String),
Num(f64),
Bool(bool),
}
impl fmt::Display for Atom {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match self {
Atom::String(a) => write!(f, "\"{}\"", a),
Atom::Num(a) => write!(f, "{}", a),
Atom::Bool(a) => write!(f, "{}", a),
}
}
}
#[derive(Debug,PartialEq,Clone)]
pub struct GuardedBlock {
pub guard: Expr,
pub block: Vec<Stmt>,
}
peg::parser! {
grammar deelang_parser() for str {
pub rule program() -> Vec<Stmt>
= __* s:stmt()* { s }
pub rule stmt() -> Stmt
= a:assignment() { a } /
f:funcall() stop() { Stmt::Funcall(f) } /
c:conditional() { c } /
e:expr() stop() { Stmt::ReplPrint(e) }
rule expr() -> Expr = precedence! {
"-" _ e1:@ { Expr::UnaryMinus(Box::new(e1)) }
--
e1:(@) "+" _ e2:@ { Expr::Plus(Box::new(e1), Box::new(e2)) }
e1:(@) "-" _ e2:@ { Expr::Minus(Box::new(e1), Box::new(e2)) }
--
e1:(@) "*" _ e2:@ { Expr::Mult(Box::new(e1), Box::new(e2)) }
e1:(@) "/" _ e2:@ { Expr::Div(Box::new(e1), Box::new(e2)) }
--
"(" _ e:expr() ")" _ { e }
['"'] s:$((!['"'] [_] / r#"\""#)*) ['"'] { Expr::Atom(Atom::String(s.to_string())) }
f:funcall() { f }
f:funcdef() { f }
b:boolean() _ { Expr::Atom(Atom::Bool(b)) }
o:object() { o }
i:id() _ { Expr::Id(i) }
n:num() _ { Expr::Atom(Atom::Num(n)) }
}
rule boolean() -> bool
= b:$("true" / "false") { b.parse().unwrap() }
rule id() -> String
= i:$(letter() (letter() / digit() / ['?'|'.'|'-'])*) _ { i.to_string() }
rule assignment() -> Stmt
= i:id() "<-" _ e:expr() stop() { Stmt::Assignment(i, e) }
rule num() -> f64
= n:$(digit()+ "."? digit()* / "." digit()+) _ { n.parse().unwrap() }
rule funcall() -> Expr
= i:id() "(" _ e:(expr() ** ("," _)) ")" _ { Expr::Funcall(i, e) }
rule funcdef() -> Expr
= i:id()? "->" _ e:(expr() / block()) { Expr::Funcdef(i, Box::new(e)) }
rule conditional() -> Stmt
= i:_if() __* ei:elif()* __* e:_else()? __* { Stmt::Conditional([vec![i], ei].concat(), e) }
rule object() -> Expr
= "{" _ stop() indent() __* a:assignment()+ dedent() __* "}" _ { Expr::Object(a) }
rule _if() -> GuardedBlock
= "if" _ g:expr() b:indented_block() {
GuardedBlock {
guard: g,
block: b,
}
}
rule elif() -> GuardedBlock
= "elif" _ g:expr() b:indented_block() {
GuardedBlock {
guard: g,
block: b
}
}
rule _else() -> Expr
= "else" _ b:block() { b }
rule block() -> Expr
= i:indented_block() { Expr::Block(i) }
rule indented_block() -> Vec<Stmt>
= stop() indent() __* s:stmt()+ dedent() { s }
rule letter()
= ['A'..='Z'] / ['a'..='z']
rule digit()
= ['0'..='9']
rule stop()
= __+ / eof()
rule indent()
= ">>>"
rule dedent()
= "<<<"
rule _ // Non-meaningful whitespace
= ['\t'|' ']*
rule __ // End Of Statement (comment, newline, eof, TODO semicolon)
= comment()? newline() / comment() &eof()
rule comment()
= "#" (!newline() [_])* &(newline() / eof())
rule newline()
= "\r\n" / "\r" / "\n"
rule eof()
= ![_]
}
}
fn preprocess(input: &str) -> String {
let mut stack = vec![0];
let mut output = String::new();
for line in input.lines() {
let mut count = 0;
for c in line.chars() {
if c == ' ' || c == '\t' {
count += 1;
} else if c == '#' {
break;
} else {
use std::cmp::Ordering::*;
let curr = stack.last().unwrap();
match curr.cmp(&count) {
Less => {
stack.push(count);
output.push_str(">>>");
},
Greater => {
while stack.last().unwrap() > &count {
output.push_str("<<<");
output.push('\n');
stack.pop();
}
}
Equal => (),
}
break;
}
}
output.push_str(line.trim());
output.push('\n');
}
output
}
pub fn parse(prgm: &str) -> Vec<Stmt> {
let prgm = preprocess(prgm);
deelang_parser::program(&prgm).unwrap()
}
pub fn parse_stmt(stmt: &str) -> Stmt {
let stmt = preprocess(stmt);
deelang_parser::stmt(&stmt).unwrap()
}
#[cfg(test)]
mod test {
use super::*;
#[test]
fn test_comments() {
let prgm = r"## This is a comment
apple <- 1 ## This is too
## This comment ends the file";
let expected = vec![Stmt::Assignment("apple".to_string(), Expr::Atom(Atom::Num(1.0)))];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_funcall() {
let expected = vec![
Stmt::Funcall(Expr::Funcall("pear".to_string(), vec![])),
Stmt::Funcall(Expr::Funcall("pear".to_string(),
vec![Expr::Id("x".to_string()), Expr::Id("y".to_string())],
))
];
let prgm = r"pear()
pear(x, y)";
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_assignment() {
let prgm = r"apple <- 1
apple <- pear(x, y)";
let expected = vec![
Stmt::Assignment("apple".to_string(),
Expr::Atom(Atom::Num(1.0))),
Stmt::Assignment("apple".to_string(),
Expr::Funcall("pear".to_string(),
vec![
Expr::Id("x".to_string()),
Expr::Id("y".to_string()),
])),
];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_operators() {
let prgm = r"three <- 1 + 2
one <- 3 - 2
four <- (3 - 1) * 2";
let expected = vec![
Stmt::Assignment("three".to_string(),
Expr::Plus(
Box::new(Expr::Atom(Atom::Num(1.0))),
Box::new(Expr::Atom(Atom::Num(2.0))),
)),
Stmt::Assignment("one".to_string(),
Expr::Minus(
Box::new(Expr::Atom(Atom::Num(3.0))),
Box::new(Expr::Atom(Atom::Num(2.0))),
)),
Stmt::Assignment("four".to_string(),
Expr::Mult(
Box::new(Expr::Minus(
Box::new(Expr::Atom(Atom::Num(3.0))),
Box::new(Expr::Atom(Atom::Num(1.0))),
)),
Box::new(Expr::Atom(Atom::Num (2.0))),
))
];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_compound_expression() {
let prgm = "apple <- pear(x, y) + z";
let expected = vec![
Stmt::Assignment("apple".to_string(),
Expr::Plus(
Box::new(Expr::Funcall(
"pear".to_string(),
vec![Expr::Id("x".to_string()), Expr::Id("y".to_string())],
)),
Box::new(Expr::Id("z".to_string()))
)
),
];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_funcdef() {
let prgm = r"foo <- -> bar()
foo <- ->
>>>
bar()
baz()
<<<
foo <- x -> y -> x * y";
let expected = vec![
Stmt::Assignment(
"foo".to_string(),
Expr::Funcdef(
None,
Box::new(Expr::Funcall("bar".to_string(), vec![])),
)
),
Stmt::Assignment(
"foo".to_string(),
Expr::Funcdef(
None,
Box::new(Expr::Block(vec![
Stmt::Funcall(Expr::Funcall("bar".to_string(), vec![])),
Stmt::Funcall(Expr::Funcall("baz".to_string(), vec![])),
]))
)
),
Stmt::Assignment(
"foo".to_string(),
Expr::Funcdef(
Some("x".to_string()),
Box::new(
Expr::Funcdef(
Some("y".to_string()),
Box::new(
Expr::Mult(
Box::new(Expr::Id("x".to_string())),
Box::new(Expr::Id("y".to_string())),
)
)
)
)
)
)
];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_conditional() {
let prgm = r"if foo
>>>bar()
<<<
elif baz
>>>foobar()
<<<
else
>>>quux()
<<<";
let expected = vec![Stmt::Conditional(
vec![
GuardedBlock {
guard: Expr::Id("foo".to_string()),
block: vec![Stmt::Funcall(Expr::Funcall("bar".to_string(), vec![]))]
},
GuardedBlock {
guard: Expr::Id("baz".to_string()),
block: vec![Stmt::Funcall(Expr::Funcall("foobar".to_string(), vec![]))]
},
],
Some(Expr::Block(vec![Stmt::Funcall(Expr::Funcall("quux".to_string(), vec![]))])),
)];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_object() {
let prgm = r"fruit <- {
>>>apple <- 1
pear <- 2
<<<
}";
let expected = vec![Stmt::Assignment(
"fruit".to_string(),
Expr::Object(vec![
Stmt::Assignment("apple".to_string(), Expr::Atom(Atom::Num(1.0))),
Stmt::Assignment("pear".to_string(), Expr::Atom(Atom::Num(2.0))),
]),
)];
assert_eq!(deelang_parser::program(prgm).unwrap(), expected);
}
#[test]
fn test_preprocess() {
let prgm = r"
.
.
.
.
.
## Hello World";
let expected = r"
>>>.
>>>.
<<<
.
>>>.
<<<
<<<
.
## Hello World
";
assert_eq!(preprocess(prgm), expected);
}
}

View File

@@ -1,78 +0,0 @@
%{
#include <cstdio>
#include <string>
#include "syntax.h"
using namespace std;
int yylex();
void yyerror(const char* p) { fprintf(stderr, p); }
%}
%define parse.trace
%union {
float num;
std::string *sym;
Node *expr;
NodeList *exprlist;
}
%token <sym> ID
%token <num> NUM
%token <sym> STRING
%token STOP
%token INDENT DEDENT
%token IF ELIF ELSE
%right GETS
%left CAT
%left '+' '-'
%left '/' '*'
%left MAPS
%nterm <expr> expr
%nterm <exprlist> exprlist
%%
program: statements | statements statement;
statements: statements statement STOP
| statements STOP
| // null production ;
statement: assignment | funcall | conditional;
assignment: ID GETS expr;
assignments: assignments assignment STOP | // null production;
expr: funcdef
| funcall
| objdef
| expr CAT expr { $$ = new OpNode(CAT, $1, $3); }
| expr '+' expr { $$ = new OpNode('+', $1, $3); }
| expr '-' expr { $$ = new OpNode('-', $1, $3); }
| expr '*' expr { $$ = new OpNode('*', $1, $3); }
| expr '/' expr { $$ = new OpNode('/', $1, $3); }
| '(' expr ')' {$$ = $2;}
| ID {$$ = new IdNode($1);}
| NUM {$$ = new NumNode($1);}
| STRING {$$ = new StringNode($1);};
funcdef: param MAPS expr | param MAPS block;
param: ID | '_' | // null production;
funcall: ID '(' exprlist ')' { print_expression_list($3); }
exprlist: exprlist ',' expr { $1 -> push_back($3); $$ = $1; }
| expr { $$ = new NodeList(); $$ -> push_back($1); }
| { $$ = new NodeList(); } // null production;
conditional: IF expr block elifs
| IF expr block elifs ELSE block;
elifs: ELIF expr block elifs| // null production;
objdef: '{' block_assignments '}' | '{' '}'
block: STOP INDENT statement statements DEDENT
block_assignments: STOP INDENT assignments DEDENT
%%

View File

@@ -1,49 +0,0 @@
#include <iostream>
#include <cstring>
#include <cstdlib>
#include "syntax.h"
using namespace std;
IdNode::IdNode(string *id) {
this->id = id;
}
void IdNode::display() {
cout << *id;
}
NumNode::NumNode(float num) {
this->num = num;
}
void NumNode::display() {
cout << num;
}
StringNode::StringNode(string* stringVal) {
this->stringVal = stringVal;
}
void StringNode::display() {
cout << *stringVal;
}
OpNode::OpNode(int op, Node *first, Node *second) {
this->op = op;
this->first = first;
this->second = second;
}
void OpNode::display() {
first->display();
cout << " " << (char)op << " ";
second->display();
}
void print_expression_list(NodeList *list) {
for (Node *expr : *list) {
expr->display();
}
}

View File

@@ -1,50 +0,0 @@
#ifndef SYNTAX_H
#define SYNTAX_H
#include <vector>
#include <string>
class Node {
public:
virtual void display() = 0;
};
class IdNode : public Node {
private:
std::string *id;
public:
IdNode(std::string*);
virtual void display();
};
class NumNode : public Node {
private:
float num;
public:
NumNode(float);
virtual void display();
};
class StringNode : public Node {
private:
std::string* stringVal;
public:
StringNode(std::string*);
virtual void display();
};
class OpNode : public Node {
private:
int op;
Node *first;
Node *second;
public:
OpNode(int, Node*, Node*);
virtual void display();
};
class NodeList : public std::vector<Node*>{};
void print_expression_list(NodeList *);
#endif /* SYNTAX_H */