On Mon, Mar 15, 2021 at 9:23 PM messi...@gmail.com <messi.sh...@gmail.com> wrote: > > In file cycles.src there're many type declarations to demo cycle reference, > some are invalid and some are valid, I can understand why invalid ones are > invalid, but for some of the valid ones, I cannot figure out their usage > scenarios, take some as examples: > > type T *T // Does nil the only value a type T variable/const can hold? > > or with bigger cycle: > type ( // The same with above T for all T1, T2 and T3? > T1 T2 > T2 *T3 > T3 T1 > )
I'm not aware of any practical use for this. However, in Go we try to aim for simplicity and orthogonality where possible, even if that permits writing code that is not particularly useful. Rather than write a special rule like "pointer types may not be a loop," we simply permit it to work and avoid having another rule in the language. There is an amusing use of this kind of type in https://golang.org/test/peano.go, but it's not intended to be practical or useful. > or function type decl like: > > F func(F) F > > I'm wondering whether type decls like this do have practical usage, or just > for grammar learning? Types like type F func() F on the other hand, are useful for state machines. The idea is that the current state of some operation is expressed in the form of a function, and every call to the state function returns a function that implements the next state. There is an example of this at https://golang.org/src/text/template/parse/lex.go#L105. Ian -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/CAOyqgcXqKAYSLxgj%2BuZfsdgqet%3Dk7F%2BW33a4WYSTb7aXpB7UZg%40mail.gmail.com.
