Re: Pythonic indentation (or: beating a dead horse)

On May 20, 12:25=A0pm, Tony Arcieri <[e-mail]> wrote:
> I already responded to this. =A0I guess you didn't see it.

I responded to your response, I guess you didn't see _that_. Tell me
what's wrong with the following alternatives to your examples:

> if foo
> =A0 blah
> elsif bar
> =A0 blah2
> else
> =A0 baz

Just like with your case example before, whoops! You're missing an
end, and the parser will complain when it sees the end-of-file without
a missing end. And just as I said with your case example before, I
sure hope this missing end isn't in the middle of a long file or you
get to spend a lot of time scrolling through your code looking for the
place where it's missing.

But if we amend your example to

if foo
  blah
elsif bar
  blah2
else
  bar
end

...I really don't see what's wrong with this as an alternative:

if foo:
  blah
elsif bar:
  blah
else:
  bar

No end necessary.

> Begin statements (ala try/catch in Python)
>
> begin
> =A0 somecode
> =A0 morecode
> rescue FooError
> =A0 some_rescue_action
> rescue BarError
> =A0 another_rescue_action
> ensure
> =A0 something_gets_done

Again, you're missing an end. Applying the same pattern I applied in
the if and case examples, I do not see what the problem is. How is it
any less functional than the current Ruby?

> Each of these statements is an expression with multiple clauses. =A0How d=
o you
> tell when these expressions are complete? =A0Obviously a naive "dedent =
=3D end
> of expression" approach doesn't work in these cases.

Okay, I think I see what you're saying now. But each of those also, in
the clauses after the first one, begins with a keyword that is not
valid except as part of the previous expression. Does it solve the
problem if dedenting ends an expression unless it's immediately
followed by a keyword that makes no sense except as a continuation of
the expression?

> This doesn't work in a language where everything is an
> expression, because all expressions must be treated the same and all
> expressions must have an expression terminator (newline or semicolon)

You keep on saying this, yet I have not seen one example of code that
demonstrates that using dedent couldn't work. Basically, we're looking
for some code which, if you removed the all ends, a script following
simple deterministic rules would not be able to unambigiously decide
where to put them back. I agree that "put an end wherever you see a
dedent" is naive and wouldn't work... but "put an end wherever you see
a dedent which isn't followed by a keyword which extends the current
expression, such as elsif or rescue or ensure" might.