2024-05-19 11:20:42 -05:00
|
|
|
use std::simd::cmp::SimdPartialEq;
|
|
|
|
|
2024-06-25 11:39:59 -05:00
|
|
|
const fn ascii_mask(chars: &[u8]) -> u128 {
|
|
|
|
let mut eq = 0;
|
|
|
|
let mut i = 0;
|
|
|
|
while i < chars.len() {
|
|
|
|
let b = chars[i];
|
|
|
|
eq |= 1 << b;
|
|
|
|
i += 1;
|
|
|
|
}
|
|
|
|
eq
|
|
|
|
}
|
|
|
|
|
2024-05-11 09:04:13 -05:00
|
|
|
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
|
2024-05-09 16:41:59 -05:00
|
|
|
pub struct Token {
|
|
|
|
pub kind: TokenKind,
|
|
|
|
pub start: u32,
|
|
|
|
pub end: u32,
|
|
|
|
}
|
|
|
|
|
|
|
|
impl Token {
|
|
|
|
pub fn range(&self) -> std::ops::Range<usize> {
|
|
|
|
self.start as usize..self.end as usize
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-11 15:22:08 -05:00
|
|
|
macro_rules! gen_token_kind {
|
|
|
|
($(
|
|
|
|
#[$atts:meta])*
|
|
|
|
$vis:vis enum $name:ident {
|
|
|
|
#[patterns] $(
|
|
|
|
$pattern:ident,
|
|
|
|
)*
|
|
|
|
#[keywords] $(
|
|
|
|
$keyword:ident = $keyword_lit:literal,
|
|
|
|
)*
|
|
|
|
#[punkt] $(
|
|
|
|
$punkt:ident = $punkt_lit:literal,
|
|
|
|
)*
|
|
|
|
#[ops] $(
|
2024-05-15 03:37:39 -05:00
|
|
|
#[$prec:ident] $(
|
|
|
|
$op:ident = $op_lit:literal $(=> $assign:ident)?,
|
2024-05-11 15:22:08 -05:00
|
|
|
)*
|
|
|
|
)*
|
|
|
|
}
|
|
|
|
) => {
|
|
|
|
impl std::fmt::Display for $name {
|
|
|
|
fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
|
2024-06-25 11:39:59 -05:00
|
|
|
let sf = *self as u8;
|
|
|
|
f.write_str(match *self {
|
2024-05-11 15:22:08 -05:00
|
|
|
$( Self::$pattern => concat!('<', stringify!($pattern), '>'), )*
|
|
|
|
$( Self::$keyword => stringify!($keyword_lit), )*
|
|
|
|
$( Self::$punkt => stringify!($punkt_lit), )*
|
2024-05-15 03:37:39 -05:00
|
|
|
$($( Self::$op => $op_lit,
|
|
|
|
$(Self::$assign => concat!($op_lit, "="),)?)*)*
|
2024-06-25 11:39:59 -05:00
|
|
|
_ => unsafe { std::str::from_utf8_unchecked(std::slice::from_ref(&sf)) },
|
|
|
|
})
|
2024-05-11 15:22:08 -05:00
|
|
|
}
|
|
|
|
}
|
2024-05-09 16:41:59 -05:00
|
|
|
|
2024-05-11 15:22:08 -05:00
|
|
|
impl $name {
|
|
|
|
#[inline(always)]
|
|
|
|
pub fn precedence(&self) -> Option<u8> {
|
|
|
|
Some(match self {
|
2024-05-15 03:37:39 -05:00
|
|
|
$($(Self::$op => ${ignore($prec)} ${index(1)},
|
|
|
|
$(Self::$assign => 0,)?)*)*
|
2024-05-11 15:22:08 -05:00
|
|
|
_ => return None,
|
2024-05-15 03:37:39 -05:00
|
|
|
} + 1)
|
2024-05-11 15:22:08 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
#[inline(always)]
|
|
|
|
fn from_ident(ident: &[u8]) -> Self {
|
|
|
|
match ident {
|
|
|
|
$($keyword_lit => Self::$keyword,)*
|
|
|
|
_ => Self::Ident,
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2024-06-25 11:39:59 -05:00
|
|
|
};
|
|
|
|
}
|
|
|
|
|
|
|
|
#[derive(Debug, PartialEq, Eq, Clone, Copy)]
|
|
|
|
#[repr(u8)]
|
|
|
|
pub enum TokenKind {
|
|
|
|
Not = b'!',
|
|
|
|
DQuote = b'"',
|
|
|
|
Pound = b'#',
|
|
|
|
CtIdent = b'$',
|
|
|
|
Mod = b'%',
|
|
|
|
Band = b'&',
|
|
|
|
Quote = b'\'',
|
|
|
|
LParen = b'(',
|
|
|
|
RParen = b')',
|
|
|
|
Mul = b'*',
|
|
|
|
Add = b'+',
|
|
|
|
Comma = b',',
|
|
|
|
Sub = b'-',
|
|
|
|
Dot = b'.',
|
|
|
|
Div = b'/',
|
2024-06-25 12:12:35 -05:00
|
|
|
// Unused = 2-6
|
|
|
|
Shr = b'<' - 5,
|
|
|
|
// Unused = 8
|
|
|
|
Shl = b'>' - 5,
|
2024-06-25 11:39:59 -05:00
|
|
|
Colon = b':',
|
|
|
|
Semi = b';',
|
|
|
|
Lt = b'<',
|
2024-06-25 12:12:35 -05:00
|
|
|
Assign = b'=',
|
|
|
|
Gt = b'>',
|
2024-06-25 11:39:59 -05:00
|
|
|
Que = b'?',
|
|
|
|
Directive = b'@',
|
|
|
|
|
|
|
|
Ident,
|
|
|
|
Number,
|
|
|
|
Eof,
|
|
|
|
String,
|
2024-05-11 15:22:08 -05:00
|
|
|
|
2024-06-25 11:39:59 -05:00
|
|
|
Return,
|
|
|
|
If,
|
|
|
|
Else,
|
|
|
|
Loop,
|
|
|
|
Break,
|
|
|
|
Continue,
|
|
|
|
Fn,
|
|
|
|
Struct,
|
|
|
|
True,
|
|
|
|
|
|
|
|
Ctor,
|
|
|
|
Tupl,
|
|
|
|
|
|
|
|
Or,
|
|
|
|
And,
|
|
|
|
|
|
|
|
// Unused = R-Z
|
|
|
|
LBrack = b'[',
|
|
|
|
BSlash = b'\\',
|
|
|
|
RBrack = b']',
|
|
|
|
Xor = b'^',
|
|
|
|
Tick = b'`',
|
|
|
|
// Unused = a-z
|
|
|
|
LBrace = b'{',
|
|
|
|
Bor = b'|',
|
|
|
|
RBrace = b'}',
|
|
|
|
Tilde = b'~',
|
|
|
|
|
|
|
|
Decl = b':' + 128,
|
|
|
|
Eq = b'=' + 128,
|
|
|
|
Ne = b'!' + 128,
|
|
|
|
Le = b'<' + 128,
|
|
|
|
Ge = b'>' + 128,
|
|
|
|
|
|
|
|
BorAss = b'|' + 128,
|
|
|
|
AddAss = b'+' + 128,
|
|
|
|
SubAss = b'-' + 128,
|
|
|
|
MulAss = b'*' + 128,
|
|
|
|
DivAss = b'/' + 128,
|
|
|
|
ModAss = b'%' + 128,
|
|
|
|
XorAss = b'^' + 128,
|
|
|
|
BandAss = b'&' + 128,
|
|
|
|
ShlAss = b'0' + 128,
|
|
|
|
ShrAss = b'1' + 128,
|
|
|
|
}
|
|
|
|
|
|
|
|
impl TokenKind {
|
|
|
|
pub fn assign_op(self) -> Option<Self> {
|
|
|
|
let id = (self as u8).saturating_sub(128);
|
|
|
|
if ascii_mask(b"|+-*/%^&01") & (1u128 << id) == 0 {
|
|
|
|
return None;
|
2024-05-11 15:22:08 -05:00
|
|
|
}
|
2024-06-25 11:39:59 -05:00
|
|
|
Some(unsafe { std::mem::transmute::<u8, Self>(id) })
|
|
|
|
}
|
2024-05-10 15:54:12 -05:00
|
|
|
}
|
|
|
|
|
2024-05-11 15:22:08 -05:00
|
|
|
gen_token_kind! {
|
|
|
|
pub enum TokenKind {
|
|
|
|
#[patterns]
|
2024-06-01 13:30:07 -05:00
|
|
|
CtIdent,
|
2024-05-11 15:22:08 -05:00
|
|
|
Ident,
|
|
|
|
Number,
|
|
|
|
Eof,
|
2024-06-25 11:39:59 -05:00
|
|
|
Directive,
|
2024-05-17 12:53:59 -05:00
|
|
|
String,
|
2024-05-11 15:22:08 -05:00
|
|
|
#[keywords]
|
|
|
|
Return = b"return",
|
|
|
|
If = b"if",
|
|
|
|
Else = b"else",
|
|
|
|
Loop = b"loop",
|
|
|
|
Break = b"break",
|
|
|
|
Continue = b"continue",
|
|
|
|
Fn = b"fn",
|
2024-05-12 05:16:40 -05:00
|
|
|
Struct = b"struct",
|
2024-05-12 13:10:50 -05:00
|
|
|
True = b"true",
|
2024-05-11 15:22:08 -05:00
|
|
|
#[punkt]
|
2024-05-12 05:16:40 -05:00
|
|
|
Ctor = ".{",
|
2024-05-15 03:37:39 -05:00
|
|
|
Tupl = ".(",
|
2024-05-11 15:22:08 -05:00
|
|
|
#[ops]
|
2024-05-15 03:37:39 -05:00
|
|
|
#[prec]
|
2024-05-12 05:16:40 -05:00
|
|
|
Decl = ":=",
|
2024-05-11 15:22:08 -05:00
|
|
|
Assign = "=",
|
2024-05-15 03:37:39 -05:00
|
|
|
#[prec]
|
|
|
|
Or = "||",
|
|
|
|
#[prec]
|
|
|
|
And = "&&",
|
|
|
|
#[prec]
|
|
|
|
Bor = "|" => BorAss,
|
|
|
|
#[prec]
|
|
|
|
Xor = "^" => XorAss,
|
|
|
|
#[prec]
|
|
|
|
Band = "&" => BandAss,
|
|
|
|
#[prec]
|
|
|
|
Eq = "==",
|
|
|
|
Ne = "!=",
|
|
|
|
#[prec]
|
2024-05-11 15:22:08 -05:00
|
|
|
Le = "<=",
|
2024-05-13 06:36:29 -05:00
|
|
|
Ge = ">=",
|
|
|
|
Lt = "<",
|
|
|
|
Gt = ">",
|
2024-05-15 03:37:39 -05:00
|
|
|
#[prec]
|
|
|
|
Shl = "<<" => ShlAss,
|
|
|
|
Shr = ">>" => ShrAss,
|
|
|
|
#[prec]
|
|
|
|
Add = "+" => AddAss,
|
|
|
|
Sub = "-" => SubAss,
|
|
|
|
#[prec]
|
|
|
|
Mul = "*" => MulAss,
|
|
|
|
Div = "/" => DivAss,
|
|
|
|
Mod = "%" => ModAss,
|
2024-05-10 15:54:12 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-09 16:41:59 -05:00
|
|
|
pub struct Lexer<'a> {
|
|
|
|
pos: u32,
|
|
|
|
bytes: &'a [u8],
|
|
|
|
}
|
|
|
|
|
|
|
|
impl<'a> Lexer<'a> {
|
|
|
|
pub fn new(input: &'a str) -> Self {
|
2024-06-24 10:26:00 -05:00
|
|
|
Self::restore(input, 0)
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn restore(input: &'a str, pos: u32) -> Self {
|
2024-05-09 16:41:59 -05:00
|
|
|
Self {
|
2024-06-24 10:26:00 -05:00
|
|
|
pos,
|
2024-05-09 16:41:59 -05:00
|
|
|
bytes: input.as_bytes(),
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-12 04:52:58 -05:00
|
|
|
pub fn slice(&self, tok: std::ops::Range<usize>) -> &'a str {
|
|
|
|
unsafe { std::str::from_utf8_unchecked(&self.bytes[tok]) }
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
fn peek(&self) -> Option<u8> {
|
2024-06-25 12:12:35 -05:00
|
|
|
if std::intrinsics::unlikely(self.pos >= self.bytes.len() as u32) {
|
|
|
|
None
|
|
|
|
} else {
|
|
|
|
Some(unsafe { *self.bytes.get_unchecked(self.pos as usize) })
|
|
|
|
}
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
fn advance(&mut self) -> Option<u8> {
|
|
|
|
let c = self.peek()?;
|
|
|
|
self.pos += 1;
|
|
|
|
Some(c)
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn next(&mut self) -> Token {
|
2024-05-15 03:37:39 -05:00
|
|
|
use TokenKind as T;
|
|
|
|
loop {
|
|
|
|
let mut start = self.pos;
|
|
|
|
|
|
|
|
let Some(c) = self.advance() else {
|
|
|
|
return Token {
|
|
|
|
kind: T::Eof,
|
|
|
|
start,
|
|
|
|
end: self.pos,
|
|
|
|
};
|
|
|
|
};
|
|
|
|
|
2024-06-01 13:30:07 -05:00
|
|
|
let advance_ident = |s: &mut Self| {
|
2024-06-25 11:39:59 -05:00
|
|
|
while let Some(b'a'..=b'z' | b'A'..=b'Z' | b'0'..=b'9' | b'_' | 127..) = s.peek() {
|
2024-06-01 13:30:07 -05:00
|
|
|
s.advance();
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
2024-06-25 12:12:35 -05:00
|
|
|
let identity = |s: u8| unsafe { std::mem::transmute::<u8, T>(s) };
|
|
|
|
|
|
|
|
let kind = match c {
|
2024-06-25 11:39:59 -05:00
|
|
|
..=b' ' => continue,
|
2024-06-25 12:12:35 -05:00
|
|
|
b'@' | b'$' => {
|
|
|
|
start += 1;
|
|
|
|
advance_ident(self);
|
|
|
|
identity(c)
|
|
|
|
}
|
2024-05-15 03:37:39 -05:00
|
|
|
b'0'..=b'9' => {
|
|
|
|
while let Some(b'0'..=b'9') = self.peek() {
|
|
|
|
self.advance();
|
|
|
|
}
|
|
|
|
T::Number
|
|
|
|
}
|
2024-06-25 11:39:59 -05:00
|
|
|
b'a'..=b'z' | b'A'..=b'Z' | b'_' | 127.. => {
|
2024-06-01 13:30:07 -05:00
|
|
|
advance_ident(self);
|
|
|
|
let ident = &self.bytes[start as usize..self.pos as usize];
|
|
|
|
T::from_ident(ident)
|
2024-05-15 03:37:39 -05:00
|
|
|
}
|
2024-05-17 12:53:59 -05:00
|
|
|
b'"' => {
|
|
|
|
while let Some(c) = self.advance() {
|
|
|
|
match c {
|
|
|
|
b'"' => break,
|
|
|
|
b'\\' => _ = self.advance(),
|
|
|
|
_ => {}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
T::String
|
|
|
|
}
|
2024-05-15 03:37:39 -05:00
|
|
|
b'.' if self.advance_if(b'{') => T::Ctor,
|
|
|
|
b'.' if self.advance_if(b'(') => T::Tupl,
|
2024-06-25 12:12:35 -05:00
|
|
|
b'<' | b'>' if self.advance_if(c) => {
|
|
|
|
identity(c - 5 + 128 * self.advance_if(b'=') as u8)
|
|
|
|
}
|
2024-05-15 03:37:39 -05:00
|
|
|
b'&' if self.advance_if(b'&') => T::And,
|
|
|
|
b'|' if self.advance_if(b'|') => T::Or,
|
2024-06-25 12:12:35 -05:00
|
|
|
b':' | b'=' | b'!' | b'<' | b'>' | b'|' | b'+' | b'-' | b'*' | b'/' | b'%'
|
|
|
|
| b'^' | b'&'
|
|
|
|
if self.advance_if(b'=') =>
|
|
|
|
{
|
|
|
|
identity(c + 128)
|
|
|
|
}
|
|
|
|
_ => identity(c),
|
2024-05-15 03:37:39 -05:00
|
|
|
};
|
|
|
|
|
|
|
|
return Token {
|
|
|
|
kind,
|
|
|
|
start,
|
|
|
|
end: self.pos,
|
|
|
|
};
|
|
|
|
}
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
fn advance_if(&mut self, arg: u8) -> bool {
|
|
|
|
if self.peek() == Some(arg) {
|
|
|
|
self.advance();
|
|
|
|
true
|
|
|
|
} else {
|
|
|
|
false
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-12 16:19:45 -05:00
|
|
|
pub fn line_col(&self, pos: u32) -> (usize, usize) {
|
|
|
|
line_col(self.bytes, pos)
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2024-05-19 11:20:42 -05:00
|
|
|
pub fn line_col(bytes: &[u8], pos: u32) -> (usize, usize) {
|
|
|
|
bytes[..pos as usize]
|
2024-05-12 16:19:45 -05:00
|
|
|
.split(|&b| b == b'\n')
|
2024-05-19 11:20:42 -05:00
|
|
|
.map(<[u8]>::len)
|
2024-05-12 16:19:45 -05:00
|
|
|
.enumerate()
|
2024-05-19 11:20:42 -05:00
|
|
|
.last()
|
|
|
|
.map(|(line, col)| (line + 1, col + 1))
|
2024-05-12 16:19:45 -05:00
|
|
|
.unwrap_or((1, 1))
|
|
|
|
}
|
|
|
|
|
2024-05-19 11:20:42 -05:00
|
|
|
pub struct LineMap {
|
|
|
|
lines: Box<[u8]>,
|
|
|
|
}
|
2024-05-09 16:41:59 -05:00
|
|
|
|
2024-05-19 11:20:42 -05:00
|
|
|
impl LineMap {
|
|
|
|
pub fn line_col(&self, mut pos: u32) -> (usize, usize) {
|
|
|
|
let mut line = 1;
|
2024-05-09 16:41:59 -05:00
|
|
|
|
2024-05-19 11:20:42 -05:00
|
|
|
let mut iter = self.lines.iter().copied();
|
|
|
|
|
2024-06-24 10:26:00 -05:00
|
|
|
loop {
|
2024-05-19 11:20:42 -05:00
|
|
|
let mut acc = 0;
|
2024-06-24 10:26:00 -05:00
|
|
|
let mut idx = 0;
|
|
|
|
loop {
|
|
|
|
let len = iter.next().unwrap();
|
|
|
|
acc |= ((len & 0x7F) as u32) << (7 * idx);
|
|
|
|
idx += 1;
|
|
|
|
if len & 0x80 == 0 {
|
|
|
|
break;
|
|
|
|
}
|
2024-05-19 11:20:42 -05:00
|
|
|
}
|
|
|
|
|
|
|
|
if pos < acc {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
pos = pos.saturating_sub(acc);
|
|
|
|
line += 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
(line, pos as usize + 1)
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn new(input: &str) -> Self {
|
|
|
|
let bytes = input.as_bytes();
|
|
|
|
let (start, simd_mid, end) = bytes.as_simd::<16>();
|
|
|
|
|
|
|
|
let query = std::simd::u8x16::splat(b'\n');
|
|
|
|
|
|
|
|
let nl_count = start.iter().map(|&b| (b == b'\n') as usize).sum::<usize>()
|
|
|
|
+ simd_mid
|
|
|
|
.iter()
|
|
|
|
.map(|s| s.simd_eq(query).to_bitmask().count_ones())
|
|
|
|
.sum::<u32>() as usize
|
|
|
|
+ end.iter().map(|&b| (b == b'\n') as usize).sum::<usize>();
|
|
|
|
|
|
|
|
let mut lines = Vec::with_capacity(nl_count);
|
|
|
|
let mut last_nl = 0;
|
|
|
|
|
|
|
|
let handle_rem = |offset: usize, bytes: &[u8], last_nl: &mut usize, lines: &mut Vec<u8>| {
|
|
|
|
bytes
|
|
|
|
.iter()
|
|
|
|
.copied()
|
|
|
|
.enumerate()
|
|
|
|
.filter_map(|(i, b)| (b == b'\n').then_some(i + offset))
|
|
|
|
.for_each(|i| {
|
|
|
|
lines.push((i - *last_nl + 1) as u8);
|
|
|
|
*last_nl = i + 1;
|
|
|
|
});
|
|
|
|
};
|
|
|
|
|
|
|
|
handle_rem(0, start, &mut last_nl, &mut lines);
|
|
|
|
|
|
|
|
for (i, simd) in simd_mid.iter().enumerate() {
|
|
|
|
let mask = simd.simd_eq(query);
|
|
|
|
let mut mask = mask.to_bitmask();
|
|
|
|
while mask != 0 {
|
|
|
|
let idx = mask.trailing_zeros() as usize + i * 16 + start.len();
|
|
|
|
let mut len = idx - last_nl + 1;
|
|
|
|
while len >= 0x80 {
|
2024-06-24 10:26:00 -05:00
|
|
|
lines.push(0x80 | (len & 0x7F) as u8);
|
2024-05-19 11:20:42 -05:00
|
|
|
len >>= 7;
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
2024-05-19 11:20:42 -05:00
|
|
|
lines.push(len as u8);
|
|
|
|
last_nl = idx + 1;
|
|
|
|
mask &= mask - 1;
|
|
|
|
}
|
|
|
|
}
|
2024-05-09 16:41:59 -05:00
|
|
|
|
2024-05-19 11:20:42 -05:00
|
|
|
handle_rem(bytes.len() - end.len(), end, &mut last_nl, &mut lines);
|
|
|
|
|
|
|
|
Self {
|
|
|
|
lines: Box::from(lines),
|
2024-05-09 16:41:59 -05:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2024-06-24 10:26:00 -05:00
|
|
|
|
|
|
|
#[cfg(test)]
|
|
|
|
mod test {
|
|
|
|
#[test]
|
|
|
|
fn test_smh() {
|
|
|
|
let example = include_str!("../README.md");
|
|
|
|
|
|
|
|
let nlines = super::LineMap::new(example);
|
|
|
|
|
|
|
|
fn slow_nline_search(str: &str, mut pos: usize) -> (usize, usize) {
|
|
|
|
(
|
|
|
|
str.lines()
|
|
|
|
.take_while(|l| match pos.checked_sub(l.len() + 1) {
|
|
|
|
Some(nl) => (pos = nl, true).1,
|
|
|
|
None => false,
|
|
|
|
})
|
|
|
|
.count()
|
|
|
|
+ 1,
|
|
|
|
pos + 1,
|
|
|
|
)
|
|
|
|
}
|
|
|
|
|
|
|
|
for i in 0..example.len() {
|
|
|
|
assert_eq!(slow_nline_search(example, i), nlines.line_col(i as _));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|