[#4341] DRY and embedded docs. — Hugh Sasse Staff Elec Eng <hgs@...>
If I have a here document in some ruby program:
[#4347] Re: DATA and rewind. — ts <decoux@...>
>>>>> "H" == Hugh Sasse Staff Elec Eng <hgs@dmu.ac.uk> writes:
[#4350] Re: Thirty-seven Reasons [Hal Fulton] Love[s] Ruby — "David Douthitt" <DDouthitt@...>
[#4396] Re: New Require (was: RAA development ideas (was: RE: Looking for inp ut on a 'links' page)) — Hugh Sasse Staff Elec Eng <hgs@...>
On 9 Aug 2000, Dave Thomas wrote:
[#4411] Re: RAA development ideas (was: RE: Lookin g for inp ut on a 'links' page) — Aleksi Niemel<aleksi.niemela@...>
Me:
On Thu, 10 Aug 2000, [iso-8859-1] Aleksi Niemelwrote:
[#4465] More RubyUnit questions. — Hugh Sasse Staff Elec Eng <hgs@...>
I am beginning to get a feel for this, but I still have a few more
[#4478] Re: RubyUnit. Warnings to be expected? — ts <decoux@...>
>>>>> "H" == Hugh Sasse Staff Elec Eng <hgs@dmu.ac.uk> writes:
[#4481] Invoking an extension after compilation — Dave Thomas <Dave@...>
Hi,
[#4501] What's the biggest Ruby development? — Dave Thomas <Dave@...>
[#4502] methods w/ ! giving nil — Hugh Sasse Staff Elec Eng <hgs@...>
I have got used to the idea that methods that end in '!' return nil if
[#4503] RubyUnit and encapsulation. — Hugh Sasse Staff Elec Eng <hgs@...>
My_class's instance variables are not all "attr :<name>" type variables,
[#4537] Process.wait bug + fix — Brian Fundakowski Feldman <green@...>
If your system uses the rb_waitpid() codepath of rb_f_wait(),
[#4567] Re: What's the biggest Ruby development? — Aleksi Niemel<aleksi.niemela@...>
Dave said:
Robert Feldt <feldt@ce.chalmers.se> writes:
On Sat, 26 Aug 2000, Dave Thomas wrote:
Robert Feldt <feldt@ce.chalmers.se> writes:
On Mon, 28 Aug 2000, Dave Thomas wrote:
Robert Feldt <feldt@ce.chalmers.se> writes:
[#4591] Can't get Tcl/Tk working — Stephen White <steve@...>
I can't get any of the samples in the ext/tk/sample directory working. All
I'm sure looking forwards to buying the book. :)
Stephen White <steve@deaf.org> writes:
On Sun, 27 Aug 2000, Dave Thomas wrote:
Stephen White <steve@deaf.org> writes:
[#4608] Class methods — Mark Slagell <ms@...>
Reading the thread about regexp matches made me wonder about this:
[#4611] mod_ruby 0.1.19 — shreeve@...2s.org (Steve Shreeve)
Shugo (and others),
[#4633] Printing tables — DaVinci <bombadil@...>
Hi.
[#4647] Function argument lists in parentheses? — Toby Hutton <thutton@...>
Hello,
[#4652] Andy and Dave's European Tour 2000 — Dave Thomas <Dave@...>
Hi,
[#4672] calling super from c — Robert Feldt <feldt@...>
[#4699] Double parenthesis — Klaus Spreckelsen <ks@...1.ruhr-uni-bochum.de>
Why is the first line ok, but the second line is not?
[ruby-talk:4532] Re: for and multiple assignment?
On Tue, 22 Aug 2000, Dave Thomas wrote:
> Well, I can't speak to the reason, but I can suggest why it might not
> be a Good Thing.
>
> Firstly, Ruby does allow
>
> for a,b,c in [[1,2,3],[4,5,6]]
>
> The 'for' iterates over the elements in the list, which are
> themselves arrays. The assignment of the array to the for loop
> variables is then done in parallel.
I only tried it with a flat array, so I didn't spot the case Guy Decoux
mentioned, either..
>
> So, why not allow the assignment to pull values from a linear array?
>
> Two possible reasons:
>
> 1. It would be a special case, inconsistent with the semantics of
> assignment elsewhere. Remember that the for loop actually becomes
> (effectively)
>
> for i in list => list.each do |i|
>
> So a for loop with multiple parameters is
>
> for i,j,k in list => list.each do |i,j,k|
>
> The iterator would have to change it's behavior depending on the
> block parameter count, which would be weird.
Well, it would only pull off as many as it needed. I'm not sure that
this is particularly wierd, but it is harder to implement for user
classes, so maybe people would dislike having to achieve such consistency.
>
> 2. It isn't very good OO style. This one is trickier. What you're
> coding up top is an array where every fifth element is a parameter,
> every fifth element plus one is another parameter, etc.
>
> Wouldn't it be more natural to have a class which represented an
> individual test?
>
> class TestStuff
> def initialize(p1, p2, p3, res1, res2)
> #...
> end
> end
>
>
> tests = [
> TestStuff.new(1, 2, 3, 4, 5),
> TestStuff.new(5, 6, 7, 8, 9)
> ];
>
> for t in tests
[...]
> end
Yes, this is a lot better. I keep forgetting that one can use an
object *just* as a struct!
>
>
> You could then extend it further. TestStuff could 'know' about
> My_class,
>
> class TestStuff
> def initialize(p1, p2, p3, res1, res2)
> #...
> end
> def runTest
[...]
> end
> end
>
> And the loop would be
>
> tests.each {|t| t.runTest}
That is neater still. Thank you.
>
>
> However, whether this is appropriate depends on the structure of
> your application.
>
I think it is appropriate for the for loop case I mentioned. Thank you.
>
> Regards
>
>
> Dave
>
I think you have convinced me that there is no real reason to ask for
this change, since "There's A Better Way To Do It" than I thought.
Thank you,
Hugh
hgs@dmu.ac.uk