From: Rick DeNatale <rick.denatale@...>
Date: 2009-08-24T21:51:57+09:00
Subject: [ruby-core:25090] Re: Proposal: Simpler block format

On Mon, Aug 24, 2009 at 6:01 AM, David A. Black<dblack@rubypal.com> wrote:
> I think I can put a bit of flesh on the bones of my dislike of the
> bare {} to mean a proc.
>
> It has to do, at least in part, with left-to-right visual parsing.
> When I see this:
>
> �x = {
>
> I want to know what the { means, without having to read further to the
> right, looking for a hashrocket, and then backtracking.
>

Somehow this all reminds me of Fortran.

Ruby has a lot of optionality and flexibility in its syntax, this can
lead to problems for the parser, and sometimes the human reader.

Fortran (at least Fortran II) made whitespace optional almost
everywhere, even within a variable name.

So the compiler interpreting

     DO 10 I=1

didn't know whether it was seeing a do loop or an assignment statement
until it got past the 3 to see what's next.

    DO 10 I=1,10

is the start of a do loop whereas

    DO 10=1 + X

is (once the whitespace is eliminated is an assignment statement.

And for more complex input code (which is easier to achieve in Ruby
than Fortran) the backtracking can be even harder for a mere mortal
human reader than for the compiler.

Just an observation.


-- 
Rick DeNatale

Blog: http://talklikeaduck.denhaven2.com/
Twitter: http://twitter.com/RickDeNatale
WWR: http://www.workingwithrails.com/person/9021-rick-denatale
LinkedIn: http://www.linkedin.com/in/rickdenatale