From: Artem Voroztsov <artem.voroztsov@...>
Date: 2008-03-19T00:14:50+09:00
Subject: Re: Questions on Enumerator#skip_first and Enumerable#first

2008/3/18, James Gray <james@grayproductions.net>:
> On Mar 18, 2008, at 6:20 AM, Artem Voroztsov wrote:
>
>  > I asked in ruby-talk, but did not get answer.
>
>
> I remember answering you:
>
>    http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/294422
>
>  ;)


1)

I am asking what is wrong with my code and what is the correct way to
write this method for enumerator in Ruby.


See, l we have the following
$  time ./ruby -e '(1..100000000).each.drop(2).take(5).each{|e| puts e}'
3
4
5
6
7
real     0m12.029s
user    0m11.546s
sys     0m0.585s


So it looks like 'drop' and 'take' has call 'to_a' inside.

I am asking about method 'drop' for Enumerator which just skips some
elements without creating new array.

I would like to think about Enumerator as iterator in STL.

I guest Enumerator should display conventional pipe behavior (see
'tail' and  'head' commands in Unix).


Both 'drop' and 'take' methods are slow for big collections:

$  time ./ruby -ve 'i =4; (1..100000000).each.drop(2).each{|e| puts e;
break if i==0; i-=1}'
ruby 1.9.0 (2008-03-18 revision 15793) [i686-linux]
3
4
5
6
7

real    0m11.967s
user    0m11.485s
sys     0m0.568s

$ time ./ruby -ve 'i =4; (1..100000000).each.take(100000000).each{|e|
puts e; break if i==0; i-=1}'
ruby 1.9.0 (2008-03-18 revision 15793) [i686-linux]
1
2
3
4
5

real    0m11.396s
user    0m10.929s
sys     0m0.490s


2) I am sorry. I've  found that method Enumerable#first is in trunk.


Artem