Question

I have started learning Ruby, and I have read a couple of tutorials and I even bought a book (“Programming Ruby 1.9 – The Pragmatic Programmers’ Guide”), and I came across something new that I haven’t seen before in any of the other languages I know (I am working as a PHP webdeveloper).

Blocks & Procs.
I think I understand what they are, but what I don’t understand is why they are so great, and when and why I should use them.
Everywhere I look they say blocks and procs are a great feature in Ruby, but I don’t understand them.

Can anybody here give a total Ruby-newbie like me some explanations?

Answer 1

There are a lot of things that are good about blocks. The elevator pitch: Blocks let us pass around actions the same way we normally pass around data.

The most obvious level is that they let you abstract things out into functions that wouldn’t be possible otherwise. For example, let’s look at a common case where you have a list of things and you want to filter it to only include items that match some criterion:

int list[50] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50};
int evenNumbers[50] = {0};
int copyIndex = 0;
for (int i = 0; i < 50; i++) {
    if (list[i] % 2 == 0) {
        evenNumbers[copyIndex++] = list[i];
    }
}

Here’s how you write that in Ruby:

list = 1..50
listCopy = list.select {|n| n.even?}

All the common busywork is moved out of your code and into a method with a meaningful name. We don’t care about copying the array and going through indexes and all that — we just want a filtered list. And that’s what select gives us. The block allows us to pass our custom logic into this standard method.

But iterators aren’t the only place where this “hole in the middle pattern” is useful. For example, if you pass a block to File.open, it will open the file, execute the block with the file and then close the file for you.

Another thing that blocks give us is a really powerful form of callbacks. For example, without blocks, we might have to do something like this (based on how dialogs actually work in Objective-C Cocoa):

class Controller
  def delete_button_clicked(item)
    item.add_red_highlight
    context = {:item => item}
    dialog = Dialog.new("Are you sure you want to delete #{item}?")
    dialog.ok_callback = :delete_OK
    dialog.ok_receiver = self
    dialog.cancel_callback = :cancel_delete
    dialog.cancel_receiver = self
    dialog.context = context
    dialog.ask_for_confirmation
  end

  def delete_OK(sender)
    delete(sender.context[:item])
    sender.dismiss
  end

  def cancel_delete(sender)
    sender.context[:item].remove_red_highlight
    sender.dismiss
  end
end

Yowza. With blocks, we could do this instead (based on a common pattern used in many Ruby libraries):

class Controller
  def delete_button_clicked(item)
    item.add_red_highlight
    Dialog.ask_for_confirmation("Are you sure you want to delete #{item}?") do |response|
      response.ok { delete item }
      response.cancel { item.remove_red_highlight }
    end
  end
end

That’s actually two levels of blocks — the do...end block and the two {}-style blocks. But it reads pretty naturally, doesn’t it? This works because a block captures the context it’s created in, so we don’t need to pass around self and item.

As for Procs, they’re just an object wrapper for blocks. Not very much to them.