Re^2: Switch/case as a jump table with C-style fall-through

I also wasn't sure it would be useful, but in the case of fall-through, I can see how it might come in handy to pass the matched string. That will go into the next revision.

I'm not married to the name 'new' (and when I package it with other flavors of Case, it will certainly have a different name), but I didn't have a better name. I think it's just as legitimate to call a factory "new" as it is to call an OO constructor "new". The users are expected to read the documentation.

I left out a default identifier because I didn't want to have a string that might conflict with an alternative. But it occurred to me that a call to a subroutine that returns an empty list could be tucked in there and be absolutely transparent. That will also be in the next revision.

I don't know what a continue/finalise block should do here. Could you explain?

Thanks for your input.

Caution: Contents may have been coded under pressure.

Comment on Re^2: Switch/case as a jump table with C-style fall-through

Replies are listed 'Best First'.

Re^3: Switch/case as a jump table with C-style fall-through
by Animator (Hermit) on May 12, 2005 at 17:05 UTC

I see now that I basiclly took it from the wrong concept... (what I initially though about had more to do with something like eval then with switch)

I have no idea wheter or not it would be a good idea to implement what I was thinking about... I'll continue explaining and then you can make up your mind... I was thinking about creating a 'handler' that is run at the end if one (or more) case-statements matched...

I'm afraid I didn't explain to well and that it might be easier to explain it with an example/code, so that's what I will do now:

Assume that at the end of each case there is some default action that should be taken (update a hash with a list of matches or something). With the current implementation there are two ways to accomplish this:

a) adding the code it in each code-ref (by calling another subroutine)
b) using something like $case->("something") and some_other_code; (although in this case it would depend on the return value, but let's assume it returns a true value when the code-ref was run)

Both of these have some disadvantages, the first one has the disadvantage of the duplicate function-call and the second depends on the return value

What my suggestion was about was to be able to set up something like 'FINAL' (or 'CONTINUE') => sub { ... } which will run the code (add it to a hash or whatever) if a case returned true

[reply]
[d/l]


Perl Monk, Perl Meditation
	PerlMonks