glyph-parser 0.0.1

//! Lexical analyzer for Glyph

use logos::Logos;

#[derive(Logos, Debug, PartialEq, Clone)]
#[logos(skip r"[ \t]+")]
#[logos(skip r"#[^\n]*")]
pub enum Token {
    // Keywords
    #[token("and")]
    And,
    #[token("as")]
    As,
    #[token("assert")]
    Assert,
    #[token("async")]
    Async,
    #[token("await")]
    Await,
    #[token("break")]
    Break,
    #[token("class")]
    Class,
    #[token("continue")]
    Continue,
    #[token("def")]
    Def,
    #[token("del")]
    Del,
    #[token("elif")]
    Elif,
    #[token("else")]
    Else,
    #[token("except")]
    Except,
    #[token("False")]
    False,
    #[token("finally")]
    Finally,
    #[token("for")]
    For,
    #[token("from")]
    From,
    #[token("global")]
    Global,
    #[token("if")]
    If,
    #[token("import")]
    Import,
    #[token("in")]
    In,
    #[token("is")]
    Is,
    #[token("lambda")]
    Lambda,
    #[token("let")]
    Let,
    #[token("match")]
    Match,
    #[token("case")]
    Case,
    #[token("None")]
    None,
    #[token("nonlocal")]
    Nonlocal,
    #[token("not")]
    Not,
    #[token("or")]
    Or,
    #[token("pass")]
    Pass,
    #[token("raise")]
    Raise,
    #[token("return")]
    Return,
    #[token("True")]
    True,
    #[token("try")]
    Try,
    #[token("while")]
    While,
    #[token("with")]
    With,
    #[token("yield")]
    Yield,

    // Literals (simplified - no data attached)
    #[regex(r"-?[0-9]+")]
    Int,

    #[regex(r"-?[0-9]+\.[0-9]+")]
    Float,

    #[regex(r#""([^"\\]|\\.)*""#)]
    #[regex(r#"'([^'\\]|\\.)*'"#)]
    String,

    #[regex(r#"f"([^"\\]|\\.)*""#)]
    #[regex(r#"f'([^'\\]|\\.)*'"#)]
    FString,

    // Identifiers
    #[regex(r"[a-zA-Z_][a-zA-Z0-9_]*")]
    Identifier,

    // Operators
    #[token("+")]
    Plus,
    #[token("-")]
    Minus,
    #[token("*")]
    Star,
    #[token("**")]
    StarStar,
    #[token("/")]
    Slash,
    #[token("//")]
    SlashSlash,
    #[token("%")]
    Percent,
    #[token("@")]
    At,
    #[token("<<")]
    LeftShift,
    #[token(">>")]
    RightShift,
    #[token("&")]
    Ampersand,
    #[token("|")]
    Pipe,
    #[token("^")]
    Caret,
    #[token("~")]
    Tilde,
    #[token("<")]
    Less,
    #[token(">")]
    Greater,
    #[token("<=")]
    LessEqual,
    #[token(">=")]
    GreaterEqual,
    #[token("==")]
    EqualEqual,
    #[token("!=")]
    NotEqual,

    // Delimiters
    #[token("(")]
    LeftParen,
    #[token(")")]
    RightParen,
    #[token("[")]
    LeftBracket,
    #[token("]")]
    RightBracket,
    #[token("{")]
    LeftBrace,
    #[token("}")]
    RightBrace,
    #[token(",")]
    Comma,
    #[token(":")]
    Colon,
    #[token(".")]
    Dot,
    #[token(";")]
    Semicolon,
    #[token("=")]
    Equal,
    #[token("->")]
    Arrow,

    // Special
    #[token("\n")]
    Newline,

    // Indentation tokens (handled separately)
    Indent,
    Dedent,
}

impl Token {
    pub fn is_trivia(&self) -> bool {
        false // We're skipping whitespace and comments via logos
    }
}