Divided token types

- Item = Expr | Stmt
- Added REPL (AST dump)
- Removed operator parsing because it was horrible and I need to redo it
This commit is contained in:
Erin 2021-05-01 13:44:58 +02:00 committed by ondra05
parent c26c4a14bb
commit d7cf1b013d
9 changed files with 268 additions and 75 deletions

156
Cargo.lock generated
View file

@ -9,6 +9,7 @@ dependencies = [
"clap", "clap",
"logos", "logos",
"rand", "rand",
"rustyline",
] ]
[[package]] [[package]]
@ -43,6 +44,12 @@ version = "1.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "cf1de2fe8c75bc145a2f577add951f8134889b4795d47466a54a5c846d691693" checksum = "cf1de2fe8c75bc145a2f577add951f8134889b4795d47466a54a5c846d691693"
[[package]]
name = "cc"
version = "1.0.67"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e3c69b077ad434294d3ce9f1f6143a2a4b89a8a2d54ef813d85003a4fd1137fd"
[[package]] [[package]]
name = "cfg-if" name = "cfg-if"
version = "1.0.0" version = "1.0.0"
@ -64,12 +71,49 @@ dependencies = [
"vec_map", "vec_map",
] ]
[[package]]
name = "dirs-next"
version = "2.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b98cf8ebf19c3d1b223e151f99a4f9f0690dca41414773390fc824184ac833e1"
dependencies = [
"cfg-if",
"dirs-sys-next",
]
[[package]]
name = "dirs-sys-next"
version = "0.1.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4ebda144c4fe02d1f7ea1a7d9641b6fc6b580adcfa024ae48797ecdeb6825b4d"
dependencies = [
"libc",
"redox_users",
"winapi",
]
[[package]]
name = "endian-type"
version = "0.1.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c34f04666d835ff5d62e058c3995147c06f42fe86ff053337632bca83e42702d"
[[package]] [[package]]
name = "fnv" name = "fnv"
version = "1.0.7" version = "1.0.7"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3f9eec918d3f24069decb9af1554cad7c880e2da24a9afd88aca000531ab82c1" checksum = "3f9eec918d3f24069decb9af1554cad7c880e2da24a9afd88aca000531ab82c1"
[[package]]
name = "fs2"
version = "0.4.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9564fc758e15025b46aa6643b1b77d047d1a56a1aea6e01002ac0c7026876213"
dependencies = [
"libc",
"winapi",
]
[[package]] [[package]]
name = "getrandom" name = "getrandom"
version = "0.2.2" version = "0.2.2"
@ -96,6 +140,15 @@ version = "0.2.93"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9385f66bf6105b241aa65a61cb923ef20efc665cb9f9bb50ac2f0c4b7f378d41" checksum = "9385f66bf6105b241aa65a61cb923ef20efc665cb9f9bb50ac2f0c4b7f378d41"
[[package]]
name = "log"
version = "0.4.14"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "51b9bbe6c47d51fc3e1a9b945965946b4c44142ab8792c50835a980d362c2710"
dependencies = [
"cfg-if",
]
[[package]] [[package]]
name = "logos" name = "logos"
version = "0.12.0" version = "0.12.0"
@ -120,6 +173,33 @@ dependencies = [
"utf8-ranges", "utf8-ranges",
] ]
[[package]]
name = "memchr"
version = "2.3.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0ee1c47aaa256ecabcaea351eae4a9b01ef39ed810004e298d2511ed284b1525"
[[package]]
name = "nibble_vec"
version = "0.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "77a5d83df9f36fe23f0c3648c6bbb8b0298bb5f1939c8f2704431371f4b84d43"
dependencies = [
"smallvec",
]
[[package]]
name = "nix"
version = "0.20.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "fa9b4819da1bc61c0ea48b63b7bc8604064dd43013e7cc325df098d49cd7c18a"
dependencies = [
"bitflags",
"cc",
"cfg-if",
"libc",
]
[[package]] [[package]]
name = "ppv-lite86" name = "ppv-lite86"
version = "0.2.10" version = "0.2.10"
@ -144,6 +224,16 @@ dependencies = [
"proc-macro2", "proc-macro2",
] ]
[[package]]
name = "radix_trie"
version = "0.2.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c069c179fcdc6a2fe24d8d18305cf085fdbd4f922c041943e203685d6a1c58fd"
dependencies = [
"endian-type",
"nibble_vec",
]
[[package]] [[package]]
name = "rand" name = "rand"
version = "0.8.3" version = "0.8.3"
@ -184,12 +274,66 @@ dependencies = [
"rand_core", "rand_core",
] ]
[[package]]
name = "redox_syscall"
version = "0.2.7"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "85dd92e586f7355c633911e11f77f3d12f04b1b1bd76a198bd34ae3af8341ef2"
dependencies = [
"bitflags",
]
[[package]]
name = "redox_users"
version = "0.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "528532f3d801c87aec9def2add9ca802fe569e44a544afe633765267840abe64"
dependencies = [
"getrandom",
"redox_syscall",
]
[[package]] [[package]]
name = "regex-syntax" name = "regex-syntax"
version = "0.6.23" version = "0.6.23"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "24d5f089152e60f62d28b835fbff2cd2e8dc0baf1ac13343bef92ab7eed84548" checksum = "24d5f089152e60f62d28b835fbff2cd2e8dc0baf1ac13343bef92ab7eed84548"
[[package]]
name = "rustyline"
version = "8.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b9e1b597fcd1eeb1d6b25b493538e5aa19629eb08932184b85fef931ba87e893"
dependencies = [
"bitflags",
"cfg-if",
"dirs-next",
"fs2",
"libc",
"log",
"memchr",
"nix",
"radix_trie",
"scopeguard",
"smallvec",
"unicode-segmentation",
"unicode-width",
"utf8parse",
"winapi",
]
[[package]]
name = "scopeguard"
version = "1.1.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d29ab0c6d3fc0ee92fe66e2d99f700eab17a8d57d1c1d3b748380fb20baa78cd"
[[package]]
name = "smallvec"
version = "1.6.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "fe0f37c9e8f3c5a4a66ad655a93c74daac4ad00c441533bf5c6e7990bb42604e"
[[package]] [[package]]
name = "strsim" name = "strsim"
version = "0.8.0" version = "0.8.0"
@ -216,6 +360,12 @@ dependencies = [
"unicode-width", "unicode-width",
] ]
[[package]]
name = "unicode-segmentation"
version = "1.7.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "bb0d2e7be6ae3a5fa87eed5fb451aff96f2573d2694942e40543ae0bbe19c796"
[[package]] [[package]]
name = "unicode-width" name = "unicode-width"
version = "0.1.8" version = "0.1.8"
@ -234,6 +384,12 @@ version = "1.0.4"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b4ae116fef2b7fea257ed6440d3cfcff7f190865f170cdad00bb6465bf18ecba" checksum = "b4ae116fef2b7fea257ed6440d3cfcff7f190865f170cdad00bb6465bf18ecba"
[[package]]
name = "utf8parse"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "936e4b492acfd135421d8dca4b1aa80a7bfc26e702ef3af710e0752684df5372"
[[package]] [[package]]
name = "vec_map" name = "vec_map"
version = "0.8.2" version = "0.8.2"

View file

@ -10,3 +10,4 @@ edition = "2018"
clap="*" clap="*"
logos = "0.12" logos = "0.12"
rand = "*" rand = "*"
rustyline = "8.0.0"

View file

@ -1 +0,0 @@
1 + 4

View file

@ -4,6 +4,7 @@ mod base_55;
mod error; mod error;
mod lexer; mod lexer;
mod parser; mod parser;
mod repl;
mod variables; mod variables;
use clap::{App, Arg}; use clap::{App, Arg};
@ -14,7 +15,7 @@ fn main() {
let matches = App::new("AbleScript") let matches = App::new("AbleScript")
.version(env!("CARGO_PKG_VERSION")) .version(env!("CARGO_PKG_VERSION"))
.author("Able <abl3theabove@gmail.com>") .author("Able <abl3theabove@gmail.com>")
.about("Does awesome things") .about("AbleScript interpreter")
.arg( .arg(
Arg::with_name("file") Arg::with_name("file")
.short("f") .short("f")
@ -36,8 +37,8 @@ fn main() {
println!("{:#?}", ast); println!("{:#?}", ast);
} }
None => { None => {
println!("hi"); println!("Hi [AbleScript {}] - AST Printer", env!("CARGO_PKG_VERSION"));
//start the prompt repl::repl();
} }
} }
} }

View file

@ -1,47 +1,61 @@
use std::collections::HashMap;
use crate::variables::Value; use crate::variables::Value;
#[derive(Debug, Clone, Hash, PartialEq, Eq)] #[derive(Debug, Clone, Hash, PartialEq, Eq)]
pub struct Iden(pub String); pub struct Iden(pub String);
impl From<Iden> for Expr { #[derive(Debug, Clone)]
fn from(iden: Iden) -> Self { pub enum Item {
Self::Identifier(iden) Expr(Expr),
Stmt(Stmt),
}
impl From<Expr> for Item {
fn from(e: Expr) -> Self {
Item::Expr(e)
}
}
impl From<Stmt> for Item {
fn from(s: Stmt) -> Self {
Item::Stmt(s)
} }
} }
#[derive(Debug, Clone)] #[derive(Debug, Clone)]
pub enum Expr { pub enum Expr {
Add { left: Box<Expr>, right: Box<Expr> },
Subtract { left: Box<Expr>, right: Box<Expr> },
Multiply { left: Box<Expr>, right: Box<Expr> },
Divide { left: Box<Expr>, right: Box<Expr> },
FunctionCall { iden: Iden, args: Vec<Expr> },
Literal(Value),
Identifier(Iden),
}
impl From<Iden> for Expr {
fn from(i: Iden) -> Self {
Self::Identifier(i)
}
}
#[derive(Debug, Clone)]
pub enum Stmt {
VariableDeclaration { VariableDeclaration {
iden: String, iden: Iden,
init: Option<Box<Expr>>, init: Option<Box<Item>>,
}, },
FunctionDeclaration { FunctionDeclaration {
iden: String, iden: Iden,
args: Vec<Iden>, args: Vec<Iden>,
body: Vec<Expr>, body: Vec<Item>,
}, },
BfFDeclaration { BfFDeclaration {
iden: String, iden: Iden,
body: String, body: String,
}, },
If { If {
cond: Box<Expr>, cond: Box<Item>,
body: Vec<Expr>, body: Vec<Item>,
}, },
FunctionCall {
iden: Iden,
args: Vec<Expr>,
},
Addition {
left: Box<Expr>,
right: Box<Expr>,
},
Literal(Value),
Identifier(Iden),
Melo(Iden), Melo(Iden),
} }

View file

@ -2,19 +2,22 @@ mod item;
mod ops; mod ops;
mod utils; mod utils;
use item::Expr; use item::Item;
use crate::{ use crate::{
error::{Error, ErrorKind}, error::{Error, ErrorKind},
lexer::PeekableLexer, lexer::PeekableLexer,
parser::item::{Expr, Stmt},
variables::Value, variables::Value,
}; };
use crate::{lexer::Token, parser::item::Iden}; use crate::{lexer::Token, parser::item::Iden};
pub type ParseResult = Result<Item, Error>;
/// Parser structure / state machine /// Parser structure / state machine
pub struct Parser<'a> { pub struct Parser<'a> {
lexer: PeekableLexer<'a>, lexer: PeekableLexer<'a>,
ast: Vec<Expr>, ast: Vec<Item>,
} }
impl<'a> Parser<'a> { impl<'a> Parser<'a> {
@ -26,19 +29,19 @@ impl<'a> Parser<'a> {
} }
} }
pub fn init(&mut self) -> Result<Vec<Expr>, Error> { pub fn init(&mut self) -> Result<Vec<Item>, Error> {
loop { loop {
let token = self.lexer.next(); let token = self.lexer.next();
if token.is_none() { if token.is_none() {
return Ok(self.ast.clone()); return Ok(self.ast.clone());
}; };
let expr = self.parse_expr(token)?; let expr = self.parse_item(token)?;
self.ast.push(expr); self.ast.push(expr);
} }
} }
fn parse_expr(&mut self, token: Option<Token>) -> Result<Expr, Error> { fn parse_item(&mut self, token: Option<Token>) -> ParseResult {
if matches!(token, None) { if matches!(token, None) {
return Err(Error { return Err(Error {
kind: ErrorKind::EndOfTokenStream, kind: ErrorKind::EndOfTokenStream,
@ -50,7 +53,7 @@ impl<'a> Parser<'a> {
let start = self.lexer.span().start; let start = self.lexer.span().start;
match token { match token {
Token::Identifier(_) => self.parse_ops(token), Token::Identifier(_) => self.parse_ops(token).map(|x| x.into()),
// Control flow // Control flow
Token::If => self.if_cond(), Token::If => self.if_cond(),
@ -60,17 +63,17 @@ impl<'a> Parser<'a> {
Token::BfFunction => self.bff_declaration(), Token::BfFunction => self.bff_declaration(),
// Literals // Literals
Token::String(x) => Ok(Expr::Literal(Value::Str(x))), Token::String(x) => Ok(Expr::Literal(Value::Str(x)).into()),
Token::Integer(x) => Ok(Expr::Literal(Value::Int(x))), Token::Integer(x) => Ok(Expr::Literal(Value::Int(x)).into()),
Token::Boolean(x) => Ok(Expr::Literal(Value::Bool(x))), Token::Boolean(x) => Ok(Expr::Literal(Value::Bool(x)).into()),
Token::Aboolean(x) => Ok(Expr::Literal(Value::Abool(x))), Token::Aboolean(x) => Ok(Expr::Literal(Value::Abool(x)).into()),
// Prefix keywords // Prefix keywords
// Melo - ban variable from next usage (runtime error) // Melo - ban variable from next usage (runtime error)
Token::Melo => { Token::Melo => {
let e = self.require_iden()?; let e = self.require_iden()?;
self.require(Token::Semicolon)?; self.require(Token::Semicolon)?;
Ok(Expr::Melo(Iden(e))) Ok(Stmt::Melo(e).into())
} }
_ => Err(Error { _ => Err(Error {
@ -83,14 +86,14 @@ impl<'a> Parser<'a> {
/// Parse variable declaration /// Parse variable declaration
/// ///
/// `var [iden] = [literal];` /// `var [iden] = [literal];`
fn variable_declaration(&mut self) -> Result<Expr, Error> { fn variable_declaration(&mut self) -> ParseResult {
let iden = self.require_iden()?; let iden = self.require_iden()?;
let init = match self.lexer.next() { let init = match self.lexer.next() {
Some(Token::Semicolon) => None, Some(Token::Semicolon) => None,
Some(Token::Assignment) => { Some(Token::Assignment) => {
let value = self.lexer.next(); let value = self.lexer.next();
let value = self.parse_expr(value)?; let value = self.parse_item(value)?;
self.require(Token::Semicolon)?; self.require(Token::Semicolon)?;
Some(Box::new(value)) Some(Box::new(value))
} }
@ -102,13 +105,13 @@ impl<'a> Parser<'a> {
} }
}; };
Ok(Expr::VariableDeclaration { iden, init }) Ok(Stmt::VariableDeclaration { iden, init }.into())
} }
/// Declare function /// Declare function
/// ///
/// `functio [iden] ([expr], [expr]) { ... } /// `functio [iden] ([expr], [expr]) { ... }
fn function_declaration(&mut self) -> Result<Expr, Error> { fn function_declaration(&mut self) -> ParseResult {
let iden = self.require_iden()?; let iden = self.require_iden()?;
self.require(Token::LeftParenthesis)?; self.require(Token::LeftParenthesis)?;
let args = vec![]; let args = vec![];
@ -118,13 +121,13 @@ impl<'a> Parser<'a> {
// Parse function body // Parse function body
let body = self.parse_body()?; let body = self.parse_body()?;
Ok(Expr::FunctionDeclaration { iden, args, body }) Ok(Stmt::FunctionDeclaration { iden, args, body }.into())
} }
/// Declare BF FFI Function /// Declare BF FFI Function
/// ///
/// `bff [iden] { ... }` /// `bff [iden] { ... }`
fn bff_declaration(&mut self) -> Result<Expr, Error> { fn bff_declaration(&mut self) -> ParseResult {
let iden = self.require_iden()?; let iden = self.require_iden()?;
self.require(Token::LeftBrace)?; self.require(Token::LeftBrace)?;
@ -155,23 +158,24 @@ impl<'a> Parser<'a> {
_ => return Err(self.unexpected_token(None)), _ => return Err(self.unexpected_token(None)),
}); });
} }
Ok(Expr::BfFDeclaration { iden, body }) Ok(Stmt::BfFDeclaration { iden, body }.into())
} }
/// Parse If-expression /// Parse If-expression
pub fn if_cond(&mut self) -> Result<Expr, Error> { pub fn if_cond(&mut self) -> ParseResult {
self.require(Token::LeftParenthesis)?; self.require(Token::LeftParenthesis)?;
let cond = self.lexer.next(); let cond = self.lexer.next();
let cond = self.parse_expr(cond)?; let cond = self.parse_item(cond)?;
self.require(Token::RightParenthesis)?; self.require(Token::RightParenthesis)?;
self.require(Token::LeftBrace)?; self.require(Token::LeftBrace)?;
let body = self.parse_body()?; let body = self.parse_body()?;
Ok(Expr::If { Ok(Stmt::If {
cond: Box::new(cond), cond: Box::new(cond),
body, body,
}) }
.into())
} }
} }

View file

@ -1,39 +1,27 @@
use super::*; use super::*;
type ExprResult = Result<Expr, Error>;
impl<'a> Parser<'a> { impl<'a> Parser<'a> {
/// Parse operations (got identifier/value) /// Parse operations (got identifier/value)
pub(super) fn parse_ops(&mut self, token: Token) -> Result<Expr, Error> { pub(super) fn parse_ops(&mut self, token: Token) -> ExprResult {
let iden = if let Token::Identifier(i) = token { todo!()
Iden(i)
} else {
unimplemented!()
};
let mut buf = Vec::new();
buf.push(match self.lexer.peek() {
Some(Token::LeftParenthesis) => self.fn_call(iden)?,
_ => unimplemented!(),
});
Ok(buf[0].clone())
} }
/// Parse function call /// Parse function call
fn fn_call(&mut self, iden: Iden) -> Result<Expr, Error> { fn fn_call(&mut self, iden: Iden) -> ExprResult {
self.lexer.next(); self.lexer.next();
let mut args: Vec<Expr> = Vec::new(); let mut args: Vec<Expr> = Vec::new();
while let Some(token) = self.lexer.peek() { while let Some(token) = self.lexer.peek() {
match token { match token {
Token::Identifier(id) => { Token::Identifier(id) => {
args.push(Iden(id.clone()).into()); args.push(Expr::Identifier(Iden(id.clone())));
self.lexer.next(); self.lexer.next();
} }
Token::RightParenthesis => break, Token::RightParenthesis => break,
_ => { _ => {
let next = self.lexer.next(); let next = self.lexer.next();
args.push(self.parse_expr(next)?)
} }
} }
self.require(Token::Comma)?; self.require(Token::Comma)?;

View file

@ -2,7 +2,7 @@ use crate::error::{Error, ErrorKind};
use crate::lexer::Token; use crate::lexer::Token;
use crate::variables::Abool; use crate::variables::Abool;
use super::{item::Expr, Parser}; use super::{Parser, item::{Iden, Item}};
pub fn abool2num(abool: Abool) -> i32 { pub fn abool2num(abool: Abool) -> i32 {
match abool { match abool {
@ -31,9 +31,9 @@ impl<'a> Parser<'a> {
} }
/// Require an identifier on next and return it /// Require an identifier on next and return it
pub(super) fn require_iden(&mut self) -> Result<String, Error> { pub(super) fn require_iden(&mut self) -> Result<Iden, Error> {
if let Some(Token::Identifier(id)) = self.lexer.next() { if let Some(Token::Identifier(id)) = self.lexer.next() {
Ok(id) Ok(Iden(id))
} else { } else {
Err(Error { Err(Error {
kind: ErrorKind::InvalidIdentifier, kind: ErrorKind::InvalidIdentifier,
@ -58,7 +58,7 @@ impl<'a> Parser<'a> {
} }
} }
pub(super) fn parse_body(&mut self) -> Result<Vec<Expr>, Error> { pub(super) fn parse_body(&mut self) -> Result<Vec<Item>, Error> {
let mut body = Vec::new(); let mut body = Vec::new();
loop { loop {
let token = { let token = {
@ -76,7 +76,7 @@ impl<'a> Parser<'a> {
if token == Token::RightBrace { if token == Token::RightBrace {
break; break;
} }
body.push(self.parse_expr(Some(token))?); body.push(self.parse_item(Some(token))?);
} }
Ok(body) Ok(body)
} }

30
src/repl.rs Normal file
View file

@ -0,0 +1,30 @@
use rustyline::Editor;
use crate::parser::Parser;
pub fn repl() {
let mut rl = Editor::<()>::new();
loop {
let readline = rl.readline(":: ");
match readline {
Ok(line) => {
if &line == "exit" {
println!("bye");
break;
}
let mut parser = Parser::new(&line);
let ast = parser.init();
println!("{:#?}", ast);
}
Err(rustyline::error::ReadlineError::Eof) => {
println!("bye");
break;
}
Err(rustyline::error::ReadlineError::Interrupted) => (),
Err(e) => {
println!("Error: {:?}", e);
break;
}
}
}
}