为什么符号pat…表示方案宏模式中的零个或多个表达式

Question

When I'm reading the macro part of The Scheme Programming Language , it mentions that when you are trying to define a syntax extension using define-syntax , you should use pat ... to specify zero or more expression in the pattern. Why don't just use ... for zero or more expression, in which case pat ... means one or more expressions?

Also, when the author gives definition of and as follows:

(define-syntax and
  (syntax-rules ()
    [(_) #t]
    [(_ e) e]
    [(_ e1 e2 e3 ...)
     (if e1 (and e2 e3 ...) #f)]))

Why don't just write it like this:

(define-syntax and
  (syntax-rules ()
    [(_) #t]
    [(_ e) e]
    [(_ e1 e2 ...)
     (if e1 (and e2 ...) #f)]))

I have tested this definition with some cases, and I didn't find any problem with it.

Answer 1

Per R6RS , ... must always follow some identifier to get its "zero or more" meaning (it behaves like the Kleene star ), so ... and pat ... mean different things: the former is just a literal ... symbol as far as syntax-rules is concerned.

In your example, both forms seem to be the same, but that's because the pattern (_ e1 e2 ...) only fires when there are two or more conjuncts (arguments to and ) anyway, since the second pattern (_ e) already handles the case where there's only one conjunct. It doesn't always work this way, and Dybvig apparently found it cleaner to write an explicit "two or more" clause rather than "exactly one" and "one or more" clauses.