-
-
Save maxjacobson/7584201 to your computer and use it in GitHub Desktop.
# gem install numbers_in_words | |
require 'numbers_in_words' | |
require 'numbers_in_words/duck_punch' | |
class Array | |
def method_missing(m) | |
self[m.to_s.gsub(/\_/, ' ').in_numbers - 1] | |
end | |
end | |
nums = (1..100).to_a | |
puts nums.first #=> 1 | |
puts nums.second #=> 2 | |
puts nums.third #=> 3 | |
puts nums.fourth #=> 4 | |
puts nums.thirty_five #=> 35 | |
puts nums.thirty_fifth #=> 30 (bug) | |
# I just want to acknowledge that I'm using `#=>` incorrectly. That's supposed to indicate the return value, but I'm using it to indicate the output. |
@cjlwired I'll double check later but ... I'm pretty sure I'm not using active support. unless active support is a dependency of numbers_in_words
oh, I see now that you said that
so that means I can use .second
in my rails apps, who knew?
but I maintain that it's a bug because the interface is inconsistent. I don't mean it's a bug in that gem, I mean a bug in my implementation. if I can do [].second
then I should be able to do [].thirty_fifth
Yes, if you're tying the bug to your interface, then I agree completely. We both now know that the numbers_to_words
gem cannot handle "-nd" || "-th
processing, so that's something you'll have to account for.
Either way, this is a cool implementation and was a good excuse to pry
into someone else's code.
GET IT?! ❤️ ❤️ ❤️
@cjlwired thanks for breaking it down btw, I never would have thought it was leaning on active support if you hadn't dug into the source <3
Your implementation here is quite clever and I commend you for it.
However, it's incorrect to call
puts nums.thirty_fifth
a bug at all.I actually poked around in the source for
numbers_in_words
and found no indication that they handled the "n-th" representation of numbers. In fact, "thirty_fifth" isn't a number, but rather, it represents a relative ordering of a particular object. Interesting then... so if the code doesn't support "n-th" representations of numbers, how is your code working?I added
pry-debugger
into the project and put it intomethod_missing
like so:And this was my console output:
Whoa - so
.thirty_five
is the first method that triggersmethod_missing
.This immediately tells me that
.second
,.third
,.fourth
, etc. can't possibly be triggered by yourmethod_missing
(as I expected from the gem source code). I then threw a pry statement right before the call tonums.second
and did astep
call to see where that function ended up:So, it appears that these functions are being implemented by
ActiveSupport
and not by yourmethod_missing
. A quick grep through the source code proves this point. Check out Line 1 fromnumbers_in_words.rb
:To summarize - the code is behaving completely as expected.
The original method magic is happening through
ActiveSupport
.Therefore,
puts nums.thirty_fifth
isn't actually a bug. QED. :-)