macros.md

Macros

Lang supports built-in attached attribute macros with #name(...). In v1, macro expansion is token-based, runs before normal item parsing, and is limited to built-in macros only.

Syntax

• Attribute macros attach to the next struct or trait item.
• Multiple attribute lines can be stacked above one item.
• #derive(...) accepts a comma-separated list of derive names.
• #[python] ... #[endpython] is separate raw-Python interop, not part of the generic attribute macro system.

Examples

Deriving debug behavior for a struct

#derive(debug)
struct Square {
    w: int,
    h: int,
}

square = Square { w: 3, h: 4 }
print(square.debug())
print(square)
print(square.__repr__())

Expanded behavior

#derive(debug) lowers into ordinary Lang items:

trait Debug {
    fn debug(self) -> str
}

struct Square {
    w: int,
    h: int,
}

impl Debug for Square {
    fn debug(self) -> str {
        "Square { w: " + str(self.w) + ", h: " + str(self.h) + " }"
    }
}

impl Square {
    fn __str__(self) -> str {
        self.debug()
    }

    fn __repr__(self) -> str {
        self.debug()
    }
}

Notes

• #derive(debug) is the only official built-in derive in v1.
• #derive(debug) is valid on struct items only. Using it on a trait fails during macro expansion.
• Built-in prelude items such as trait Debug { ... } are injected once per module.
• Expanded items are parsed and transpiled through the normal object pipeline. No macro nodes survive in the AST.