Elegant way to call a method

siberia-man has asked for the wisdom of the Perl Monks concerning the following question:

This question has been discussed by someone already: How can I elegantly call a Perl subroutine whose name is held in a variable?. But it doesn't answer the question completely. Actually I have the following code. It works but looks a bit ugly:

if ( defined &{ "some::cool::${module}::method" } ) {
    &{ \&{ "some::cool::${module}::method" } }(some arguments);
} else {
    ...
}
[download]

I can check the method availability as below but don't know how to proceed further in term of my question:

if ( "some::cool::${module}"->can("method") ) {
    ???
[download]

Comment on Elegant way to call a method Select or Download Code

Replies are listed 'Best First'.
Re: Elegant way to call a method by Corion (Patriarch) on Dec 23, 2017 at 10:44 UTC
The return value of can is a function reference: `if ( my $handler = "some::cool::${module}"->can("method") ) { $handler->( ... some arguments ... ) } else { ... }` [download]	[reply] [d/l]
Re^2: Elegant way to call a method by siberia-man (Friar) on Dec 23, 2017 at 12:10 UTC
Thank you. I was in one step behind this solution.	[reply]
Re: Elegant way to call a method by eyepopslikeamosquito (Archbishop) on Dec 23, 2017 at 10:57 UTC
See also this recent node: Rosetta Dispatch Table (Interview Question)	[reply]
Re^2: Elegant way to call a method by siberia-man (Friar) on Dec 23, 2017 at 12:09 UTC
Interesting discussion. I will try to keep this in mind on the case if something similar happen in my work.	[reply]
Re: Elegant way to call a method by Anonymous Monk on Dec 26, 2017 at 13:50 UTC
Just as you do not want a variable to contain the arbitrary name of a variable that you reference, you do not want a variable to contain the arbitrary name of a function that you call. Dispatch-tables, by any other name, are a fine solution ... so are `if..elsif` blocks. Throw an exception at a well-defined point if you are given a name that you are not somehow explicitly-prepared to recognize, because this is the only way that you will find that bug. Another technique that I like to use is to specify that the variable must be the name of an object of a specified class or parent-class, which contains a method that I will call. The invoking routine verifies that the object exists, and that it is of the proper class, and wraps the method-call in an exception handler. (This strategy provides context, contained in the object-instance.) You can do this sort of thing in many programming languages and it works everywhere.	[reply]