On Nov 23, 2013, at 1:46 AM, John Clements wrote: > I'm preparing a 10-minute lightning talk on hygienic macros in rust (preview: > I'm barely going to *mention* hygiene), and in the process, I've been > surveying some of the Rust macros, and roughly categorizing them in terms of > the "three canonical categories" that Matthias described--apologies if I'm > misrepresenting him/you: > - changing evaluation order, > - implementing a data sublanguage, and > - creating new binding forms.
[That's correct. I used to say 'quoting' for the second bullet until Shriram pointed out it's really about a data sublanguage. The ideas are due to joint sessions with Eugene and Bruce in the mid 1980s.] > Some of the Rust macros seem to fall into a fourth category, which arises > from the fact that certain things are not expressions: > > - abstracting over things that are not expressions. > > For instance: > > cmp_impl!(impl Eq, eq, ne) > cmp_impl!(impl TotalEq, equals) > cmp_impl!(impl Ord, lt, gt, le, ge) > cmp_impl!(impl TotalOrd, cmp -> cmp::Ordering) > > Each of these expands into a top-level "impl" declaration, extending > implementations of, e.g., Ord, from type T to type Ratio<T>. Like someone else said, I would classify this kind of macro as a 'binding form' macro assuming you're using 'declaration' in the standard sense. > More generally, it seems to me that every time you constrain first-class-ness > by making things not-first-class (e.g. module-level stuff in Racket), you > will be required to use macros to abstract over these things. > > Thoughts? > > Back to writing my talk... Post your slides. -- Matthias ____________________ Racket Users list: http://lists.racket-lang.org/users
