lang: basic module syntax grammar

This commit is contained in:
Schrottkatze 2024-06-03 11:22:36 +02:00
parent f6da90a354
commit 946ac879a7
No known key found for this signature in database
7 changed files with 136 additions and 8 deletions

View file

@ -111,7 +111,7 @@ impl Marker {
}
}
fn complete_node(mut self, p: &mut Parser, kind: NodeKind) -> CompletedMarker {
fn close_node(mut self, p: &mut Parser, kind: NodeKind) -> CompletedMarker {
self.bomb.defuse();
match &mut p.events[self.pos] {
Event::Start { kind: slot, .. } => *slot = kind.clone(),
@ -127,11 +127,11 @@ impl Marker {
}
pub(crate) fn complete(self, p: &mut Parser<'_, '_>, kind: SyntaxKind) -> CompletedMarker {
self.complete_node(p, NodeKind::Syntax(kind))
self.close_node(p, NodeKind::Syntax(kind))
}
pub(crate) fn error(self, p: &mut Parser, kind: SyntaxError) -> CompletedMarker {
self.complete_node(p, NodeKind::Error(kind))
self.close_node(p, NodeKind::Error(kind))
}
pub(crate) fn abandon(mut self, p: &mut Parser<'_, '_>) {

View file

@ -7,4 +7,7 @@ pub enum SyntaxError {
// if there was two space seperated items in a list
SpaceSepInList,
SemicolonInList,
CommaInMatOrVec,
UnterminatedTopLevelItem,
UnclosedModuleBody,
}

View file

@ -2,6 +2,8 @@ use std::fmt::Debug;
use crate::lst_parser::syntax_kind::SyntaxKind::*;
use self::module::{mod_body, top_level_item};
use super::{
input::Input,
output::Output,
@ -10,11 +12,13 @@ use super::{
};
mod expression;
mod module;
pub fn source_file(p: &mut Parser) {
let root = p.start("root");
expression::expression(p, false);
mod_body(p);
// expression::expression(p, false);
p.eat_succeeding_ws();
root.complete(p, ROOT);

View file

@ -25,7 +25,6 @@ pub fn vec_matrix_list(p: &mut Parser) -> CompletedMarker {
}
}
// TODO: handle semicolons, other wrong toks
fn finish_list(p: &mut Parser, list_start: Marker) -> CompletedMarker {
loop {
if p.eat(COMMA) {
@ -41,11 +40,11 @@ fn finish_list(p: &mut Parser, list_start: Marker) -> CompletedMarker {
item.precede(p, "next_item")
.complete(p, COLLECTION_ITEM)
.precede(p, "err_space_sep")
.complete_err(p, SyntaxError::SpaceSepInList);
.error(p, SyntaxError::SpaceSepInList);
} else if p.at(SEMICOLON) {
let semi_err = p.start("semicolon_err");
p.eat(SEMICOLON);
semi_err.complete_err(p, SyntaxError::SemicolonInList);
semi_err.error(p, SyntaxError::SemicolonInList);
if let Some(item) = atom(p) {
item.precede(p, "coll_item_start")
.complete(p, COLLECTION_ITEM);
@ -85,6 +84,14 @@ fn finish_mat_or_vec(p: &mut Parser, coll_start: Marker, mut row_start: Marker)
p.eat(R_BRACK);
return coll_start.complete(p, VEC);
}
} else if p.at(COMMA) {
let err_unexpected_comma = p.start("err_unexpected_comma");
p.do_bump();
err_unexpected_comma.error(p, SyntaxError::CommaInMatOrVec);
} else {
let err_unexpected = p.start("err_unexpected_tok");
p.do_bump();
err_unexpected.error(p, SyntaxError::Expected(vec![EXPR, SEMICOLON, R_BRACK]));
}
}
}

View file

@ -0,0 +1,106 @@
use enumset::enum_set;
use crate::lst_parser::{
error::SyntaxError,
grammar::expression::expression,
syntax_kind::{SyntaxKind::*, TokenSet},
CompletedMarker, Parser,
};
const TOP_LEVEL_ITEM_START: TokenSet = enum_set!(DEF_KW | MOD_KW | USE_KW);
pub fn mod_body(p: &mut Parser) {
loop {
if top_level_item(p).is_none() {
break;
}
}
}
fn mod_decl(p: &mut Parser) -> Option<CompletedMarker> {
let mod_start = p.start("module");
if !p.eat(MOD_KW) {
return None;
}
let mod_name = p.start("module_name");
if p.eat(IDENT) {
mod_name.complete(p, MODULE_NAME);
} else {
mod_name.error(p, SyntaxError::Expected(vec![IDENT]));
}
let mod_body_marker = p.start("mod_body");
if p.eat(SEMICOLON) {
mod_body_marker.abandon(p);
Some(mod_start.complete(p, MODULE))
} else if p.eat(L_BRACE) {
mod_body(p);
if !p.eat(R_BRACE) {
mod_body_marker
.complete(p, MODULE_BODY)
.precede(p, "unclosed_mod_body_err")
.error(p, SyntaxError::UnclosedModuleBody);
} else {
mod_body_marker.complete(p, MODULE_BODY);
}
Some(mod_start.complete(p, MODULE))
} else {
Some(mod_start.error(p, SyntaxError::Expected(vec![MODULE_BODY])))
}
}
pub fn top_level_item(p: &mut Parser) -> Option<CompletedMarker> {
if !TOP_LEVEL_ITEM_START.contains(p.current()) {
return None;
}
def(p).or_else(|| mod_decl(p))
}
fn def(p: &mut Parser) -> Option<CompletedMarker> {
let def_start = p.start("top_level_def");
if !p.eat(DEF_KW) {
def_start.abandon(p);
return None;
}
let def_name = p.start("def_name");
if p.eat(IDENT) {
def_name.complete(p, DEF_NAME);
} else {
def_name.error(p, SyntaxError::Expected(vec![IDENT]));
}
let maybe_expected_eq = p.start("maybe_expect_eq");
if !p.eat(EQ) {
maybe_expected_eq.error(p, SyntaxError::Expected(vec![EQ]));
} else {
maybe_expected_eq.abandon(p);
}
let body = p.start("def_body");
if expression(p, false).is_some() {
body.complete(p, DEF_BODY);
} else {
body.error(p, SyntaxError::Expected(vec![DEF_BODY]));
}
let def = def_start.complete(p, DEF);
Some(if p.eat(SEMICOLON) {
def
} else if TOP_LEVEL_ITEM_START.contains(p.current()) || p.at(EOF) {
def.precede(p, "unterminated_tl_item")
.error(p, SyntaxError::UnterminatedTopLevelItem)
} else {
def.precede(p, "err_unexpected")
.error(p, SyntaxError::Expected(vec![SEMICOLON]))
})
}
fn r#use(p: &mut Parser) -> Option<CompletedMarker> {
if !p.eat(USE_KW) {
return None;
}
todo!()
}

View file

@ -30,6 +30,7 @@ pub enum SyntaxKind {
#[token("mod")]
MOD_KW,
MODULE,
MODULE_NAME,
MODULE_BODY,
#[token("use")]
USE_KW,

View file

@ -1 +1,8 @@
meow | gay |
def hello_world = meow [ 1 2 ]
def test
mod hello {
def meow = uwu;
}