Ruby

I propose a new feature: {Enumerable,Array,Range}#first, {Array,Range}#last with block.

If no argument is passed:

• enum.first { block } works as the same as enum.find { block }
• ary.last { block } works as the same as ary.reverse_each.find { block }

If a integer is passed:

• enum.first(n) { block } works as the same as enum.lazy.select { block }.first(n)
• ary.last(n) { block } works as the same as ary.reverse_each.lazy.select { block }.first(n).reverse

Examples:

ary = [1, 2, 3, 4, 5, 6] ary.first { |i| i.even? } #=> 2 ary.first(2) { |i| i.even? } #=> [2, 4] ary.last { |i| i.odd? } #=> 5 ary.last(2) { |i| i.odd? } #=> [3, 5] 

Currently these methods just ignore given blocks.

ary.first { |i| i > 3 } #=> 1 

I sometimes write such code:

ary = ["DEBUG: a", "WARN: b", ..., "WARN: y", "DEBUG: z"] # assume a large Array ary.reverse_each.lazy.reject { |line| line.include?("DEBUG") }.first(10).reverse #=> returns last 10 non-debug logs 

But this is redundant and not intuitive. In this case, I just want the last 10 logs which is not debug message in the original order.
With the new Array#last, I can refactor it:

ary.last(10) { |line| !line.include?("DEBUG") } 

I think this represents what I want to do much more clearly.