3 releases

0.1.2 Jan 19, 2024
0.1.1 Jan 18, 2024
0.1.0 Jan 17, 2024

#168 in Programming languages

28 downloads per month

MIT license

20KB
203 lines

Pattern Lexer

My personal implementation of a lexer.

Principle

This lexer is making use of the Pattern trait to find tokens.
The idea is to create Tokens, explain how to match them with a Pattern and build them from the matched String value.

Pattern

A string Pattern trait.

The type implementing it can be used as a pattern for &str, by default it is implemented for the following types.

Pattern type Match condition
char is contained in string
&str is substring
String is substring
&[char] any char match
&[&str] any &str match
F: Fn(&str) -> bool F returns true for substring (slow)
Regex Regex match substring

Usage

The lexer! macro match the following syntax.

lexer!(
    // Ordered by priority
    NAME(optional types, ...) {
        impl Pattern => |value: String| -> Token,
        ...,
    },
    ...,
);

It generates module gen which contains Token, LexerError, LexerResult and Lexer.

You can now call Token::tokenize to tokenize a &str, it should return a Lexer instance that implements Iterator.
Each iteration, the Lexer tries to match one of the given Pattern and returns a LexerResult<Token> built from the best match.

Example

Here is an example for a simple math lexer.

lexer!(
    // Different operators
    OPERATOR(char) {
        '+' => |_| Token::OPERATOR('+'),
        '-' => |_| Token::OPERATOR('-'),
        '*' => |_| Token::OPERATOR('*'),
        '/' => |_| Token::OPERATOR('/'),
        '=' => |_| Token::OPERATOR('='),
    },
    // Integer numbers
    NUMBER(usize) {
        |s: &str| s.chars().all(|c| c.is_digit(10))
            => |v: String| Token::NUMBER(v.parse().unwrap()),
    },
    // Variable names
    IDENTIFIER(String) {
        regex!(r"[a-zA-Z_$][a-zA-Z_$0-9]*")
            => |v: String| Token::IDENTIFIER(v),
    },
    WHITESPACE {
        [' ', '\n'] => |_| Token::WHITESPACE,
    },
);

That will expand to these enum and structs.

mod gen {
    pub enum Token {
        OPERATOR(char),
        NUMBER(usize),
        IDENTIFIER(String),
        WHITESPACE,
    }

    pub struct Lexer {...}
    pub struct LexerError {...}
    pub type LexerResult<T> = Result<T, LexerError>;
}

And you can use them afterwards.

use gen::*;

let mut lex = Token::tokenize("x_4 = 1 + 3 = 2 * 2");
assert_eq!(lex.nth(2), Some(Ok(Token::OPERATOR('='))));
assert_eq!(lex.nth(5), Some(Ok(Token::NUMBER(3))));

// Our lexer doesn't handle parenthesis...
let mut err = Token::tokenize("x_4 = (1 + 3)");
assert!(err.nth(4).is_some_and(|res| res.is_err()));

Dependencies

~2–3MB
~53K SLoC