From: Gavin Sinclair <gsinclair@...>
Date: 2002-09-02T07:32:09+09:00
Subject: Re: Ruby aesthetics


----- Original Message -----
From: "Reimer Behrends" <behrends@cse.msu.edu>


> Gavin Sinclair (gsinclair@soyabean.com.au) wrote:
> [...]
>
> >  I'll conclude with one (new) feature from Python that makes me green
> >  with envy: list comprehensions.  For those who don't know what that
> >  means, here's an example:
>
> >   Ruby:  squares_less_than_10 = numbers.find_all{|i| i<10}.map{|i| i^2}
> >
> >   Python: squares_less_than_10 = [x^2 for x in numbers if x < 10]
>
> That calls for a functional abstraction (functional abstraction is
> essentially a fancy software engineering term for factoring out commonly
> occuring code sequences into a procedure, function, or method of its
> own) rather than a new language feature. This abstraction is most easily
> added to Enumerable:
>
> module Enumerable
>
>   # map all elements through a filtering block, keeping all the results
>   # that aren't nil.
>
>   def condmap
>     result = []
>     each do
>       | elem |
>       item = yield elem
>       result << item unless item.nil?
>     end
>     result
>   end
>
> end
>
> To be used as follows:
>
>     squares_less_than_10 = numbers.condmap{|x| x**2 if x < 10 }

That's very nice indeed.  Would there be any support for adding this, or
something like it, to the language standard?  Maybe not, since there's a
redundancy (collect{...}.compact).  But I'd like to see it for sure.

> In general, you do not want to expand the language when existing
> abstraction mechanisms can already capture the same thing adequately and
> with more room for customization.

True, but what about getting all pairs of number less than 10 whose combination
mulitplies to greater than 25?

result = [[a,b] for a in (1...10) for b in (1...10) if a * b > 25]

This construct hasn't had much popularity in this thread, but I think it would
be a worthwhile addition to the language.  It's simply a beautiful abstraction.
(I know a little about functional programming, but cannot be said to have a
background in it: my appreciation for list comprehensions is more
mathematical.)


> Reimer Behrends

Gavin