Ruby range.reduce with hash accumulator

Question

I have this method

def heights
  (60..68).reduce({}) { |h, i| h.merge!( { %(#{i/12}'#{i%12}") => i } ) }
end

it returns a hash of heights

{
  "5'0\"" => 60, "5'1\"" => 61, "5'2\"" => 62,
  "5'3\"" => 63, "5'4\"" => 64, "5'5\"" => 65,
  "5'6\"" => 66, "5'7\"" => 67, "5'8\"" => 68
}

That's what I want. However, I don't like using the merge! method. I'd much rather use the hash[key] = value syntax for assignment:

def heights
  (60..68).reduce({}) { |h, i| h[%(#{i/12}'#{i%12}")] = i }
end

But this code throws errors. I know that with reduce, in your pipes you can name your accumulator and element.

I also understand that

sum = 0
(1..5).each { |i| sum += i }

is equivalent to

(1..5).reduce(0) { |sum, i| sum + i }

So why doesn't this

hash = {}
(1..5).each { |i| hash[i.to_s] = i }

work the same as

(1..5).reduce({}) { |hash, i| hash["#{i}"] = i }

Answer 1

You could use each_with_object instead of reduce:

(60..68).each_with_object({}) { |i, h| h[%(#{i/12}'#{i%12}")] = i }

enumerable.each_with_object(obj) { ... } returns obj so you don't need the artificial-feeling ; h in the block that you'd need with reduce.

Note that the order of the arguments to the block is different than with reduce.