Skip to content

Latest commit

 

History

History
99 lines (71 loc) · 2.32 KB

README.md

File metadata and controls

99 lines (71 loc) · 2.32 KB

Procer

NOTE: Experimental. Use it to experience what a default to_proc could have been. For production code, I recommend an explicit transformation, like the one provided by the gem jgomo3-func.

A reasonable good default to_proc method for all objects.

Install with:

gem install procer

When you require Procer, all objects will have a default to_proc method which will try to call one of the following methods, in the given order:

  • call
  • []
  • ===

Many methods which receive a block, can benefit greatly from this because you can now pass an object to perform the block role.

Think of the Enumerable module and all its methods.

Many objects define ===, but not to_proc. So they will be nicely usable in a case/when expression, but not in other contexts.

This is the case of classes and ranges, which you can use in case/when expressions, but they don't define to_proc.

Now they do define to_proc so they are useful in those contexts.

Examples:

require 'procer'

[1, 2, '3', '4', 5, 6].filter(&Numeric)
# => [1, 2, 5, 6]

[-10, 100, -2, 3, 20, -33].filter(&(0..50))
# => [3, 20]

Also, Hashes already implement to_proc and that is useful with Enumerator. We can use it as a transformation table with map:

table = {
  1 => 'one',
  2 => 'two',
  3 => 'three'
}

[3, 1, 2].map(&table)
# => ['three, 'one, 'two']

Sadly, Arrays, even when they have the same interface as hashes as a function of indices, don't implement to_proc and so they can't be used in the same way. Until now.

table = ['zero', 'one', 'two']
[2, 0, 1].map(&table)
# => ['two', 'zero', 'one']

Alternatively, you could have used values_at:

table.values_at([3, 1, 2]) # In the Hash example
table.values_at([2, 0, 1]) # In the Array example

But the map solution is more generic and table can be anything that implements to_proc and not something that necessarily implements values_at.

Notice that if the object implements [] that will triumph over ===. It was unexpected when I tried to use Integers as the object, as they implement [] as a way to access their binary form:

5 # b101
[5[2], 5[1], 5[0]] # [1, 0, 1]

So the proc will work like that:

[2, 4, 5].map(&5)
# Actual => [1, 0, 0]
# I was expecting => [false, false, true]