Define method by braces

[gogotanaka]

I'm aiming to https://gist.github.com/chancancode/61f31ac0b508c8eccad7

Right now, we can write as below.

require 'java'
class MyClass
  extend JavaClass

  public int my_method(arg1, arg2) {"
    ...
  "}
end

[chancancode]

Check out https://gist.github.com/chancancode/0087d86c935db02453ae, I think you can use some of the same tricks here so you can pass real ruby code in the block instead of string.

The drawback is you'd need binding_of_caller and use method_missing to fake the local variables (turns out you can't define new dynamic local variables in Ruby)

[gogotanaka]

Wow, awesome. binding_of_caller + method_missing is really powerful..!!!
It seems we can do anything different from Ruby totally.

I'm gonna do same things. thanks.

[chancancode]

hahaha, there are many quirks :P

Just updated the gist and fixed the follow case.. previously it couldn't tell between these two:

function x() { ... }
function(x) { ... }

Turns out you can infer it based on whether the block is attached to def x or def function. If method_missing is called without a block, it's an identifier. If it's called with a block, it's a named function.

This would probably come in handy for you cause all of your method definitions will look like someIdentifier( ... ) { }, so same rules apply – no block = identifier within the ( ... ); with block = the someIdentifier call that's defining the method.

[gogotanaka]

@chancancode Hmm, it does make sense. nice idea. I'm gonna do that.

[gogotanaka]

I just try to extract the code from a block by using seattlerb/ruby2rbuy and eval. but ... I couldn't do from block made dynamic.

You'er way is most reasonable after all.

[chancancode]

I don't know about "reasonable", but it was certainly fun :P

—
Sent from Mailbox

On Thu, Dec 4, 2014 at 4:20 PM, Kazuki Tanaka [email protected]
wrote:

I just try to extract the code from a block by using seattlerb/ruby2rbuy and eval. but ... I couldn't do from block made dynamic.

You'er way is more reasonable after all.

Reply to this email directly or view it on GitHub:
#5 (comment)