Elixir: Enum chunk_by doesn´t work properly

[eloyucu]

Hi.

I was taking a look into the Enum info page... and I think that I had found a trouble. The chunk_by function doesn´t return the same result on my iex console that on the example:

Theese are my results:

iex(15)>  Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6], '\a\a']

iex(16)>  Enum.chunk_by([1, 2, 2, 3, 4, 4, 6], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6]]

iex(17)>  Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7, 6], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6], '\a\a', [6]]

iex(18)>  Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 8, 7, 6], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6], '\a', '\b', '\a', [6]]

iex(19)>  Enum.chunk_by([1, 7, 2, 2, 3, 4, 4, 6, 7, 8, 7, 6], &(rem(&1, 2) == 1))
[[1, 7], [2, 2], [3], [4, 4, 6], '\a', '\b', '\a', [6]]

iex(20)>  Enum.chunk_by([1, 7, 2, 2, 3, 4, 7, 4, 6, 7, 8, 7, 6], &(rem(&1, 2) == 1))
[[1, 7], [2, 2], [3], [4], '\a', [4, 6], '\a', '\b', '\a', [6]]

iex(21)>  Enum.chunk_by([1, 7, 2, 2, 3, 4, 7, 4, 6, 3, 8, 7, 6], &(rem(&1, 2) == 1))
[[1, 7], [2, 2], [3], [4], '\a', [4, 6], [3], '\b', '\a', [6]]

iex(22)>  Enum.chunk_by([1, 7, 2, 2, 3, 4, 7, 4, 6, 3, 8, 2, 7, 6], &(rem(&1, 2) == 1))
[[1, 7], [2, 2], [3], [4], '\a', [4, 6], [3], [8, 2], '\a', [6]]

The result of the example is this:

iex> Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6], [7, 7]]

My first attempt iex(15) is exactly the same that the example and the results don´t match. My quick conclusion is that thes funcition needs, at least, two consecutive even numbers to work properly.
And... about the example... shouldn´t be this one [[1], [2, 2], [3], [4, 4, 6], [7], [7]] the correct solution??

The version I am using (for windows):

Eshell V8.3 (abort with ^G)
IEx 1.4.4

[Joe-noh]

Erlang/OTP 20 [erts-9.0] [source] [64-bit] [smp:4:4] [ds:4:4:10] [async-threads:10] [hipe] [kernel-poll:false]

Interactive Elixir (1.5.1) - press Ctrl+C to exit (type h() ENTER for help)
iex(1)> Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7], &(rem(&1, 2) == 1))
[[1], [2, 2], [3], [4, 4, 6], '\a\a']

elixir 1.5.1 gives same result but I think Enum.chunk_by works correctly and this is related to char list representation. could you try following code?

iex(2)> Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7], &(rem(&1, 2) == 1)) == [[1], [2, 2], [3], [4, 4, 6], [7, 7]]
true

[jayjun]

@eloyucu Your IEx results show Enum.chunk_by works correctly. The function works by chunking _only_ when the callback’s result _changes_.

For this example,

Enum.chunk_by([1, 2, 2, 3, 4, 4, 6, 7, 7], &(rem(&1, 2) == 1))

Iterating through ...

rem(1, 2) == 1 is true. Result is [[1]].
rem(2, 2) == 1 is false (changed). Result is [[1], [2]].
rem(2, 2) == 1 is false (same). Result is [[1], [2, 2]].
rem(3, 2) == 1 is true (changed). Result is [[1], [2, 2], [3]].

... until ...

rem(7, 2) == 1 is true (changed). Result is [[1], [2, 2], [3], [4, 4, 6], [7]].
rem(7, 2) == 1 is true (same). Result is [[1], [2, 2], [3], [4, 4, 6], [7, 7]].

So [[1], [2, 2], [3], [4, 4, 6], [7, 7]] is the correct result.

[josevalim]

When all numbers in a list is a printable character, Elixir shows it using the single quoted syntax. Type i '\a\a' in IEx for more information!

[eloyucu]

Thanks guys.
You are right... the representation is the key!! :)
And sorry... but I saw this difference between my results and the example on the official web and I thought it was a mistake.