Also it's probably better to use a pointer to Dog struct instead using it as value type, then it works closer to a class https://play.golang.org/p/_uS9D1i-8u . Example if you want to mutate Dog struct by adding SetAge method.
On Tuesday, January 31, 2017 at 2:51:42 PM UTC+2, jul.s...@gmail.com wrote: > > Hi, > > First, thanks for your ideas. > One downside with Henry's idea is that the attack function cannot access > potential struct attributes of the Pitbull or Dog struct (I believe, > correct me if I'm wrong). > Although there weren't in my example, our code will certainly have struct > attributes that need to be accessed during the «attack» portion of the code. > > So I'd say that goes with Henry's vision, where there isn't a one-size > fits all approach to those problems. > > I think we'll stick with the type-independent function for now as it looks > like the most versatile > > - Julien > > On Tuesday, 31 January 2017 07:34:29 UTC-5, Henry wrote: >> >> Hi, >> >> One thing that I learn about Go is that it forces you to look for a >> specific solution to your problem, and the same solution usually cannot be >> applied to other similar problems. However, the same thinking can be >> carried over from OO to Go. The two dogs share the same preparation method >> and they differ only in their attack method. Hence, you extract and >> encapsulate the attack method. Here is an example: >> https://play.golang.org/p/N1Jq7boNmZ >> >> Bear in mind that it may not work to your actual situation, but you >> should get the idea. >> >> On Monday, January 30, 2017 at 11:01:15 PM UTC+7, jul.s...@gmail.com >> wrote: >>> >>> Hi, >>> >>> Where I work, we are currently experimenting with golang in order to >>> migrate our existing Perl codebase to golang. >>> >>> Although Perl isn't a pure OO language, it allows for a lot of the OO >>> patterns to be used. One of the patterns we do use is where a superclass >>> method wraps a series of call to other methods that can be overloaded by >>> subclasses. >>> >>> Here is an example of what I mean: >>> https://goo.gl/hTnUI4 >>> >>> Now, I've tried to apply the same principle in golang but failed as the >>> method attack isn't called on the right type: >>> https://play.golang.org/p/ZHhaQhSrms >>> >>> I've scratched my head for a bit to find a solution for this, and the >>> idea I found was to wrap the prep+attack call in a function that isn't tied >>> to any of the two structs >>> https://play.golang.org/p/7nK8xEV5rj >>> >>> The example above creates the behaviour I initially expected but I don't >>> find its as clean as when using the classic OO pattern. >>> >>> Now my question is, how do people usually translate this classic OO >>> pattern into golang other than the way I found ? >>> >>> Thanks ! >>> >>> - Julien >>> >> -- 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. For more options, visit https://groups.google.com/d/optout.
