This is a lexer inspired by logos. It utilizes swift macros enable easy creation.

import Kaleidoscope

let lambda: (inout LexerMachine<Tokens>) -> Substring = { $0.slice }

@kaleidoscope(skip: " |\t|\n")
enum Tokens {
    @token("not")
    case Not

    @regex("very")
    case Very

    @token("tokenizer")
    case Tokenizer

    // you could feed a closure directly to `onMatch` but swift doesn't like it for some reason
    // seems to be a compiler bug (https://github.com/apple/swift/issues/70322)
    @regex("[a-zA-Z_][a-zA-Z1-9$_]*?", onMatch: lambda) 
    case Identifier(Substring)
}


for token in Tokens.lexer(source: "not a very fast tokenizer").map({ try! $0.get() }) {
    print(token)
}

The output will be

Not
Identifier("a")
Very
Identifier("fast")
Tokenizer

The project is provides three macros: @kaleidoscope, regex, and token, and they work together to generate conformance to LexerProtocol for the decorated enums. regex takes in a regex expression for matching and token takes a string for excat matching. In addition, they can take a onMatch callback and a priority integer. The callback has access to token string slice and can futher transform it to whatever type required by the enum case. The priority are calculated by from the expression by default. However, if two exprssions have the same weight, manual specification is required to resolve the conflict.

Internally, all regex expressions and token strings are converted into a single finite automata. The finite automata consumes one character from the input at a time, until it reaches an token match or an error. This machanism is simple but works slowly. Future improvements can be established on this issue.

This package uses an internal Swift package _RegexParser included in the experimental string processing lib. Please check out the github packe for compatibility. Due to its being experimental, this library can break in the future.

• faster tokenization optomization
• cleaner code generation
• cleaner interface