teascade/reid-llvm
1
1
Fork 0
Code Issues Pull Requests Packages Releases 6 Wiki Activity

Add formatting of the printed lines to errors

Browse Source
This commit is contained in:
Sofia 2025-07-17 18:58:05 +03:00
parent 9d1b18f083
commit df4febf1ef
3 changed files with 268 additions and 72 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

263
reid/src/error_raporting.rs
View File

@ -5,9 +5,9 @@ use std::{
use crate::{
ast,
lexer::{self, FullToken},
lexer::{self, FullToken, Position},
mir::{self, pass, Metadata, SourceModuleId},
token_stream,
token_stream::{self, TokenRange},
};
impl<T: std::error::Error + std::fmt::Display> pass::Error<T> {
@ -16,17 +16,26 @@ impl<T: std::error::Error + std::fmt::Display> pass::Error<T> {
}
}
fn label(text: &str) -> &str {
#[cfg(debug_assertions)]
{
return text;
}
#[cfg(not(debug_assertions))]
""
}
#[derive(thiserror::Error, Debug, Clone, PartialEq, Eq)]
pub enum ErrorKind {
#[error("Lexing: {}", .0.kind)]
#[error("{}{}", label("(Lexing) "), .0.kind)]
LexerError(#[from] mir::pass::Error<lexer::Error>),
#[error("Parsing: {}", .0.kind)]
#[error("{}{}", label("(Parsing) "), .0.kind)]
ParserError(#[from] mir::pass::Error<token_stream::Error>),
#[error("Typechecking: {}", .0.kind)]
#[error("{}{}", label("(TypeCheck) "), .0.kind)]
TypeCheckError(#[source] mir::pass::Error<mir::typecheck::ErrorKind>),
#[error("Type Inference: {}", .0.kind)]
#[error("{}{}", label("(TypeInference) "), .0.kind)]
TypeInferenceError(#[source] mir::pass::Error<mir::typecheck::ErrorKind>),
#[error("Linking: {}", .0.kind)]
#[error("{}{}", label("(Linker) "), .0.kind)]
LinkerError(#[from] mir::pass::Error<mir::linker::ErrorKind>),
}
@ -102,70 +111,12 @@ impl ModuleMap {
}
}
// impl TryFrom<&mir::Context> for ModuleMap {
// type Error = ();
// fn try_from(value: &mir::Context) -> Result<Self, Self::Error> {
// let mut map = HashMap::new();
// for module in &value.modules {
// if let Some(_) = map.insert(
// module.module_id,
// ErrModule {
// name: module.name.clone(),
// tokens: Some(module.clone()),
// },
// ) {
// return Err(());
// }
// }
// let module_counter = value.modules.iter().map(|m| m.module_id).max().ok_or(())?;
// Ok(ModuleMap {
// module_map: map,
// module_counter,
// })
// }
// }
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct ReidError {
map: ModuleMap,
errors: Vec<ErrorKind>,
}
impl std::error::Error for ReidError {}
impl std::fmt::Display for ReidError {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
let mut sorted_errors = self.errors.clone();
sorted_errors.sort_by(|a, b| a.cmp(&b));
sorted_errors.dedup();
let mut curr_module = None;
for error in sorted_errors {
let meta = error.get_meta();
if curr_module != Some(meta.source_module_id) {
curr_module = Some(meta.source_module_id);
writeln!(
f,
"Errors in module {}:",
color_err(format!(
"{}",
self.map
.module_map
.get(&meta.source_module_id)
.unwrap()
.name
))?
)?;
}
write!(f, " {}: ", color_err("Error")?)?;
writeln!(f, "{}", error)?;
writeln!(f, " {}: {}", color_warn("At")?, meta)?;
}
Ok(())
}
}
impl ReidError {
pub fn from_lexer<U>(
result: Result<U, lexer::Error>,
@ -214,6 +165,176 @@ impl ReidError {
}
}
impl std::error::Error for ReidError {}
impl std::fmt::Display for ReidError {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
let mut sorted_errors = self.errors.clone();
sorted_errors.sort_by(|a, b| a.cmp(&b));
sorted_errors.dedup();
let mut curr_module = None;
for error in sorted_errors {
let meta = error.get_meta();
let module = self.map.get_module(&meta.source_module_id).unwrap();
let (position_fmt, line_fmt) = if let Some(tokens) = &module.tokens {
let range_tokens = meta.range.into_tokens(&tokens);
let position = get_position(&range_tokens).unwrap();
let full_lines = get_full_lines(&tokens, position);
(
fmt_positions(get_position(&full_lines).unwrap()),
Some(fmt_tokens(&full_lines, &range_tokens)),
)
} else if let Some(position) = meta.position {
(fmt_positions((position, position)), None)
} else {
("unknown".to_owned(), None)
};
if curr_module != Some(meta.source_module_id) {
curr_module = Some(meta.source_module_id);
writeln!(
f,
"Errors in module {}:",
color_err(format!(
"{}",
self.map
.module_map
.get(&meta.source_module_id)
.unwrap()
.name
))?
)?;
}
writeln!(f)?;
write!(f, " Error: ")?;
writeln!(f, "{}", color_err(format!("{}", error))?)?;
writeln!(f, "{:>20}{}", color_warn("At: ")?, position_fmt)?;
if let Some(line_fmt) = line_fmt {
writeln!(f, "{:>20}{}", color_warn("")?, line_fmt)?;
}
}
Ok(())
}
}
impl TokenRange {
pub fn into_tokens<'v>(&self, tokens: &'v Vec<FullToken>) -> Vec<&'v FullToken> {
tokens
.iter()
.skip(self.start)
.by_ref()
.take(self.end - self.start)
.collect::<Vec<_>>()
}
}
fn get_position(tokens: &Vec<&FullToken>) -> Option<(Position, Position)> {
if let Some(first) = tokens.first() {
let last = tokens.last().unwrap();
Some((first.position, last.position.add(last.token.len() as u32)))
} else {
None
}
}
fn get_full_lines<'v>(
tokens: &'v Vec<FullToken>,
(start, end): (Position, Position),
) -> Vec<&'v FullToken> {
let (first_token_pos, _) = tokens
.iter()
.enumerate()
.find(|(_, token)| token.position.1 == start.1)
.unwrap();
tokens
.iter()
.skip(first_token_pos)
.by_ref()
.take_while(|token| token.position.1 <= end.1)
.collect::<Vec<_>>()
}
fn fmt_tokens(tokens: &Vec<&FullToken>, highlighted: &Vec<&FullToken>) -> String {
let mut text = String::new();
let mut last_likes_space = false;
for (i, token) in tokens.iter().enumerate() {
if token.token.needs_space() || (token.token.likes_space() && last_likes_space) {
text += " ";
}
last_likes_space = token.token.likes_space();
let mut token_fmt = format!("{}", token.token.to_string());
if highlighted.contains(token) {
token_fmt = color_underline(token_fmt).unwrap();
}
text += &token_fmt;
if token.token.is_newline() && i > (tokens.len() - 1) {
text += "\n"
}
}
text
}
fn fmt_positions((start, end): (Position, Position)) -> String {
if start == end {
format!("ln {}, col {}", start.1, start.0)
} else if start.1 == end.1 {
format!("ln {}, col {}-{}", start.1, start.0, end.0)
} else {
format!("{}:{} - {}:{}", start.1, start.0, end.1, end.0)
}
}
impl lexer::Token {
fn likes_space(&self) -> bool {
match self {
lexer::Token::Identifier(_) => true,
lexer::Token::DecimalValue(_) => true,
lexer::Token::StringLit(_) => true,
lexer::Token::LetKeyword => true,
lexer::Token::MutKeyword => true,
lexer::Token::ImportKeyword => true,
lexer::Token::ReturnKeyword => true,
lexer::Token::FnKeyword => true,
lexer::Token::PubKeyword => true,
lexer::Token::Arrow => true,
lexer::Token::If => true,
lexer::Token::Else => true,
lexer::Token::True => true,
lexer::Token::False => true,
lexer::Token::Extern => true,
lexer::Token::Struct => true,
lexer::Token::Equals => true,
_ => false,
}
}
fn needs_space(&self) -> bool {
match self {
lexer::Token::LetKeyword => true,
lexer::Token::MutKeyword => true,
lexer::Token::ImportKeyword => true,
lexer::Token::ReturnKeyword => true,
lexer::Token::FnKeyword => true,
lexer::Token::PubKeyword => true,
lexer::Token::Arrow => true,
lexer::Token::Equals => true,
_ => false,
}
}
fn is_newline(&self) -> bool {
match self {
lexer::Token::Semi => true,
lexer::Token::BraceOpen => true,
lexer::Token::BraceClose => true,
_ => false,
}
}
}
fn color_err(elem: impl std::fmt::Display) -> Result<String, std::fmt::Error> {
let mut text = format!("{}", elem);
@ -237,3 +358,15 @@ fn color_warn(elem: impl std::fmt::Display) -> Result<String, std::fmt::Error> {
Ok(text)
}
fn color_underline(elem: impl std::fmt::Display) -> Result<String, std::fmt::Error> {
let mut text = format!("{}", elem);
#[cfg(feature = "color")]
{
use colored::Colorize;
text = format!("{}", text.bright_yellow().underline())
}
Ok(text)
}

75
reid/src/lexer.rs
View File

@ -1,4 +1,7 @@
use std::{fmt::Debug, str::Chars};
use std::{
fmt::{Debug, Write},
str::Chars,
};
static DECIMAL_NUMERICS: &[char] = &['0', '1', '2', '3', '4', '5', '6', '7', '8', '9'];
@ -99,6 +102,54 @@ impl From<Token> for String {
}
}
impl Token {
pub fn len(&self) -> usize {
self.to_string().len()
}
}
impl ToString for Token {
fn to_string(&self) -> String {
match &self {
Token::Identifier(ident) => ident.clone(),
Token::DecimalValue(val) => val.to_string(),
Token::StringLit(lit) => format!("\"{}\"", lit),
Token::LetKeyword => String::from("let"),
Token::MutKeyword => String::from("mut"),
Token::ImportKeyword => String::from("import"),
Token::ReturnKeyword => String::from("return"),
Token::FnKeyword => String::from("fn"),
Token::PubKeyword => String::from("pub"),
Token::Arrow => String::from("=>"),
Token::If => String::from("if"),
Token::Else => String::from("else"),
Token::True => String::from("true"),
Token::False => String::from("false"),
Token::Extern => String::from("extern"),
Token::Struct => String::from("struct"),
Token::Semi => String::from(';'),
Token::Equals => String::from('='),
Token::Colon => String::from(':'),
Token::Plus => String::from('+'),
Token::Times => String::from('*'),
Token::Minus => String::from('-'),
Token::GreaterThan => String::from('>'),
Token::LessThan => String::from('<'),
Token::Et => String::from('&'),
Token::Exclamation => String::from('!'),
Token::ParenOpen => String::from('('),
Token::ParenClose => String::from(')'),
Token::BraceOpen => String::from('{'),
Token::BraceClose => String::from('}'),
Token::BracketOpen => String::from('['),
Token::BracketClose => String::from(']'),
Token::Comma => String::from(','),
Token::Dot => String::from('.'),
Token::Eof => String::new(),
}
}
}
/// A token with a position
#[derive(Clone, PartialEq, Eq, PartialOrd, Ord)]
pub struct FullToken {
@ -115,7 +166,19 @@ impl Debug for FullToken {
}
}
pub type Position = (u32, u32);
/// (Column, Line)
#[derive(Clone, Copy, Debug, Hash, PartialEq, Eq, PartialOrd, Ord)]
pub struct Position(pub u32, pub u32);
impl Position {
pub fn add(&self, num: u32) -> Position {
Position(self.0 + num, self.1)
}
pub fn sub(&self, num: u32) -> Position {
Position(self.0 - num, self.1)
}
}
struct Cursor<'a> {
pub position: Position,
@ -153,14 +216,14 @@ pub fn tokenize<T: Into<String>>(to_tokenize: T) -> Result<Vec<FullToken>, Error
let to_tokenize = to_tokenize.into();
let mut cursor = Cursor {
char_stream: to_tokenize.chars(),
position: (0, 1),
position: Position(0, 1),
};
let mut tokens = Vec::new();
while let Some(character) = &cursor.next() {
// Save "current" token first character position
let position = (cursor.position.0 - 1, cursor.position.1);
let position = cursor.position.sub(1);
let variant = match character {
// Whitespace
@ -275,9 +338,9 @@ pub fn tokenize<T: Into<String>>(to_tokenize: T) -> Result<Vec<FullToken>, Error
#[derive(thiserror::Error, Debug, Clone, PartialEq, Eq, PartialOrd, Ord)]
pub enum Error {
#[error("Invalid token '{}' at Ln {}, Col {}", .0, (.1).1, (.1).0)]
#[error("Invalid token '{}' ", .0)]
InvalidToken(char, Position),
#[error("String literal that starts at Ln {}, Col {} is never finished!", (.0).1, (.0).0)]
#[error("String literal is never finished!")]
MissingQuotation(Position),
}

2
reid/src/token_stream.rs
View File

@ -209,7 +209,7 @@ impl std::iter::Sum for TokenRange {
#[derive(thiserror::Error, Debug, Clone, PartialEq, Eq, PartialOrd, Ord)]
pub enum Error {
#[error("Expected {} at Ln {}, Col {}, got {:?}", .0, (.2).1, (.2).0, .1)]
#[error("Expected {} got {:?}", .0, .1)]
Expected(String, Token, Position),
#[error("Source file contains no tokens")]
FileEmpty,