Forum: Ruby a few thoughts for ruby...

Thought I'd pass these thoughts by the readers here before sending them
to core.  They're my current wish list :)

1) add an Object#in? method to complement the existing Array#include?

2) add some useful lists of exceptions, ex: IO.SELECT_EXCEPTIONS,
IO.READ_EXCEPTIONS, so that you can rescue the wide gamut of them
appropriately, should you desire to.

3) provide an easier way to know which platform you're on than
RUBY_PLATFORM =~ /mswin32|dgjpp|mingw/

4) (this one's controversial) remove the extra # for code in strings
(i.e. "string#{code}") -> "string{code}" less typing.

5) add a BigDecimal(float) method.
 -> BigDecimal.new("%f" % float)

6) add Dir.directory?
 -> File.directory?

Feedback?
Thanks!
-=r

Roger Pack wrote:
> Thought I'd pass these thoughts by the readers here before sending them
> to core.  They're my current wish list :)
> 
> 1) add an Object#in? method to complement the existing Array#include?

It's not really symmetrical. Array has an #include? method because it is
a collection. Object has #in? because it is--what? A potential member of
a collection?

This would conflict with lots of potential other uses of #in? that have
nothing to do with collections:

Clutch#in?

Doctor#in?

Fix#in?

(to start just with the 0-ary ones).

> 2) add some useful lists of exceptions, ex: IO.SELECT_EXCEPTIONS,
> IO.READ_EXCEPTIONS, so that you can rescue the wide gamut of them
> appropriately, should you desire to.

I've wanted that, and ended up with:

   ALL_NETWORK_ERRORS = [Errno::ECONNRESET, Errno::ECONNABORTED,
     Errno::ECONNREFUSED, Errno::EPIPE, IOError, Errno::ETIMEDOUT]

   ALL_NETWORK_ERRORS << Errno::EPROTO if defined?(Errno::EPROTO)
   ALL_NETWORK_ERRORS << Errno::ENETUNREACH if 
defined?(Errno::ENETUNREACH)

Maybe a better way would be to include a module, AnyNetworkError, in
each of these. Then you could

  rescue AnyNetworkError

Like this:

module E; end

class E1 < StandardError; include E; end
class E2 < StandardError; include E; end

begin
   raise E1, "foo"
rescue E => ex
   p ex
end

> 4) (this one's controversial) remove the extra # for code in strings
> (i.e. "string#{code}") -> "string{code}" less typing.

Would not like having to escape { in strings...

Thanks for the reply Joel.

>> 1) add an Object#in? method to complement the existing Array#include?
> 
> It's not really symmetrical. Array has an #include? method because it is
> a collection. Object has #in? because it is--what? A potential member of
> a collection?

Yeah.  I'm not saying it's symmetric as much as useful and 
intuitive--some of the big reasons I like Ruby :)
Would #within? be better?

> I've wanted that, and ended up with:
> 
>    ALL_NETWORK_ERRORS = [Errno::ECONNRESET, Errno::ECONNABORTED,
>      Errno::ECONNREFUSED, Errno::EPIPE, IOError, Errno::ETIMEDOUT]
> 
>    ALL_NETWORK_ERRORS << Errno::EPROTO if defined?(Errno::EPROTO)
>    ALL_NETWORK_ERRORS << Errno::ENETUNREACH if defined?(Errno::ENETUNREACH)

Looks like it would indeed be useful--especially for cross platform 
where you 'wish you didnt have to know them all for every platform' 
(there really are a lot of them -- see 
http://betterlogic.com/roger/?p=1223 as an example).

> Maybe a better way would be to include a module, AnyNetworkError, in
> each of these. Then you could...

Yeah, or have them all descend from a single ancestor.  I would say just 
descend from SocketError except that SocketError is actually raised as 
an exception itself, every so often, so it has its own meaning 
currently...though I guess that doesn't stop it from being an ancestor.

> module E; end
...
> begin
>    raise E1, "foo"
> rescue E => ex
>    p ex
> end

That's pretty elegant.  With an array you can [non intuitively] also get 
by:

> begin
>    raise SocketError "foo"
> rescue *ALL_NETWORK_ERRORS  => ex
>    p ex
> end
(I'm sure you knew that).


>> 4) (this one's controversial) remove the extra # for code in strings
>> (i.e. "string#{code}") -> "string{code}" less typing.
> 
> Would not like having to escape { in strings...

Ok. This next isn't meant as an attacking question but...do you use { in 
normal strings often?  Granted probably more than #, but...?

Thanks!
-=r

Hi,

At Sat, 11 Apr 2009 15:30:54 +0900,
Roger Pack wrote in [ruby-talk:333612]:
> Ok. This next isn't meant as an attacking question but...do you use { in 
> normal strings often?  Granted probably more than #, but...?

gettext, and printf in 1.9.

  printf "%{foo}\n", foo:1  #=> 1

On 11.04.2009 06:52, Roger Pack wrote:
> Thought I'd pass these thoughts by the readers here before sending them
> to core.  They're my current wish list :)
> 
> 1) add an Object#in? method to complement the existing Array#include?

-0

I don't see the benefit but I'm also not strongly against.  I do see
Joel's point about the reversion.  Basically it is strange that every
object should be able to answer a question that only the collection can
answer.  Plus, you can easily add it yourself if you need it.

> 2) add some useful lists of exceptions, ex: IO.SELECT_EXCEPTIONS,
> IO.READ_EXCEPTIONS, so that you can rescue the wide gamut of them
> appropriately, should you desire to.

A good idea.  But I do not know how feasible this is.  Not all OS have
the same error reporting mechanisms and the mapping from OS errors to
exception types would have to be maintained of all platforms.

> 3) provide an easier way to know which platform you're on than
> RUBY_PLATFORM =~ /mswin32|dgjpp|mingw/

-1

This would make a Ruby specific unification of operating systems
necessary.  This means that not only maintainers of automake need to
keep track of operating systems but also maintainers of Ruby.  Other
difficulties are: how much level of detail do you provide?  For one
application it may be enough to know it's running on Linux, the other
one needs to know the kernel version and a third one does not bother
about versions but must know the distro.  I see too much effort for too
little benefit.

> 4) (this one's controversial) remove the extra # for code in strings
> (i.e. "string#{code}") -> "string{code}" less typing.

-1

Definitively a don't as the overhead of typing # isn't too big (plus, it
is more easily spotted) and the potential for damage caused by this is
large.

> 5) add a BigDecimal(float) method.
>  -> BigDecimal.new("%f" % float)

0

Seems reasonable at first sight but the absence might have a reason.
For example, by making the conversion to String explicit it is more
obvious that float and BigDecimal are not really compatible.

> 6) add Dir.directory?
>  -> File.directory?

0

Btw, there is also this nice idiom for those tests

if test ?d, "some dir"

Kind regards

  robert

Hi,

At Sat, 11 Apr 2009 13:52:49 +0900,
Roger Pack wrote in [ruby-talk:333607]:
> 6) add Dir.directory?
>  -> File.directory?

Dir.exist? exists in 1.9.

>> 1) add an Object#in? method to complement the existing Array#include?
> 
> -0
> 
> I don't see the benefit but I'm also not strongly against.  I do see
> Joel's point about the reversion.  Basically it is strange that every
> object should be able to answer a question that only the collection can
> answer.  Plus, you can easily add it yourself if you need it.

I kind of agree with Joel on this one, too.

A few other observations:

re: in?
Currently with #select you've got one in Kernel [which is IO.select] but 
Arrays seem to have their own #select.  So it is "conceivably possible" 
to have a "default #in?" and have it overridden by Clutch#in? or 
House#in? if desired.

Another option would be included? -- might be more ruby-y :)

>> 2) add some useful lists of exceptions, ex: IO.SELECT_EXCEPTIONS,
>> IO.READ_EXCEPTIONS, so that you can rescue the wide gamut of them
>> appropriately, should you desire to.
> 
> A good idea.  But I do not know how feasible this is.  Not all OS have
> the same error reporting mechanisms and the mapping from OS errors to
> exception types would have to be maintained of all platforms.

True, mapping exceptions directly from OS to OS would be problematic. 
And knowing which ones are on each OS is also annoying.
I am proposing more of a (platform dependent) container of all possible 
exceptions, regardless of what they may mean.  Or have them all include 
a common ancestor--same result.

>> 3) provide an easier way to know which platform you're on than
>> RUBY_PLATFORM =~ /mswin32|dgjpp|mingw/
> 
> -1
> 
> This would make a Ruby specific unification of operating systems
> necessary.  This means that not only maintainers of automake need to
> keep track of operating systems but also maintainers of Ruby.  Other
> difficulties are: how much level of detail do you provide?  For one
> application it may be enough to know it's running on Linux, the other
> one needs to know the kernel version and a third one does not bother
> about versions but must know the distro.  I see too much effort for too
> little benefit.

True maintaining this is annoying, but I'd also propose that it's 
useful.  Currently in 1.9 we have:
>> RUBY_VERSION
=> "1.9.2"
>> RUBY_PLATFORM
=> "x86_64-linux"
>> RUBY_ENGINE
=> "ruby"

Typically "enough" OS information is given in RUBY_PLATFORM to determine 
the platform--it's just "hard" to use that for such.  My example being 
that knowing if you're on windows is something like RUBY_PLATFORM =~ 
/dgjpp|mingw|mswin/
which seems overly complex for me.  And very hard to get right the first 
time (ex: RUBY_PLATFORM =~ /win/ doesn't work--that includes darwin).

>> 4) (this one's controversial) remove the extra # for code in strings
>> (i.e. "string#{code}") -> "string{code}" less typing.
> 
> -1
> 
> Definitively a don't as the overhead of typing # isn't too big (plus, it
> is more easily spotted) and the potential for damage caused by this is
> large.

True it's not too hard to type--I just think its absence would be less 
typing, and that the # is "too" easily spotted, but again, that's just 
my take on it.  Actually you may have a reasonable point (easy to spot 
is good).

The kicker is also that # is ingrained so much in existing ruby 
code...it would be a pretty dramatic change.


>> 5) add a BigDecimal(float) method.
>>  -> BigDecimal.new("%f" % float)
> 
> 0
> 
> Seems reasonable at first sight but the absence might have a reason.
> For example, by making the conversion to String explicit it is more
> obvious that float and BigDecimal are not really compatible.

Yeah I wonder that myself. I was just hoping to make it easier to use 
BigDecimal, since Floats are so imprecise to use for decimal numbers :)

> if test ?d, "some dir"

Could you explain that again?  Not sure I do understand the idiom. 
Looks like bash?
Much thanks.
-=r

>> Roger Pack
>>
>> Thought I'd pass these thoughts by the readers here before sending them
>> to core.  They're my current wish list :)
>>
>> 1) add an Object#in? method to complement the existing Array#include?

> Joel VanderWerf
>
> It's not really symmetrical. Array has an #include? method because it is
> a collection. Object has #in? because it is--what? A potential member of
> a collection?

> Robert Klemme
>
> I don't see the benefit but I'm also not strongly against.  I do see
> Joel's point about the reversion.  Basically it is strange that every
> object should be able to answer a question that only the collection can
> answer.  Plus, you can easily add it yourself if you need it.

Sometimes the best (or better) usage depends on the context.
For example, on some occasions we use:  Integer === object
and on other occasions we use:          object.kind_of?( Integer )
So I'm not against this on principle, and for me the question is
what is the balance between not adding things unless we need to
and does this make things easier in a significant number of cases.
I don't know the answer!


>> Roger Pack
>> 4) (this one's controversial) remove the extra # for code in strings
>> (i.e. "string#{code}") -> "string{code}" less typing.

I'm not in favour, for Robert Klemme's reasons.


>>> Roger Pack
>>>
>>> 5) add a BigDecimal(float) method.
>>>  -> BigDecimal.new("%f" % float)

>> Robert Klemme
>>
>> Seems reasonable at first sight but the absence might have a reason.
>> For example, by making the conversion to String explicit it is more
>> obvious that float and BigDecimal are not really compatible.

> Roger Pack
>
> Yeah I wonder that myself. I was just hoping to make it easier to use
> BigDecimal, since Floats are so imprecise to use for decimal numbers :)

Then probably best to start with BigDecimal and keep using it?
We can always use  BD = BigDecimal ; BD.new( "123.456" )
and I seem to recall that there's a way to assign a method to a variable
which could reduce this to BD( "123.456" ). I think it's important
to normally use strings to get BigDecimal values to avoid precision 
errors.

Robert Klemme makes a good point: Float has less precision than 
BigDecimal
and in general I think that you should be *very* wary about converting
from lower precision to higher precision - at least not without being
fully aware of what you are doing - because later on you may be mislead
into thinking that your calculation is more accurate than it actually 
is.
(See "Here's one I made earlier" below.)

You can't gain precision by such a conversion, which is why my start
position is not to do it. That said, converting Float to BigDecimal
may sometimes be useful when it can prevent further loss of precision.
For example, ( float_a - float_b ) can reduce the accuracy drastically,
and conversion to BigDecimal might be useful there.
 ( Caveat: Numerical Analysis - study of ... - can be extremely tricky,
   and I only know enough about it to know that:
   1. I don't understand it sufficiently to make definitive statements.
   2. In my own stuff I probably ignore it more than I should. )

So I am against easy conversion by using something like BigDecimal( 
float ).
I think any standard conversion method from Float to BigDecimal
should be relatively ugly and messy. I did tbink of suggesting a rather 
long
method name, maybe 
BigDecimal.do_you_really_want_to_convert_this_float(fnum)
but someone can always alias that to a much shorter name.

If you  require "bigdecimal/util"  this does, amongst other things:
# BigDecimal utility library. ... The following methods are provided
# to convert other types to BigDecimals: ...
class Float < Numeric
  def to_d
    BigDecimal(self.to_s)
  end
end

So you can require that library, or just use  BigDecimal( float.to_s ),
I think  BigDecimal( float.to_s )  satisfies my general thought here
that whatever the conversion method used is, it shouldn't be so simple
that you can use it without being in some way being reminded
that what you are doing might not be a good idea.
So making it sort of ugly fits in with that.
(In fact, I'm not sure that having a Float#to_d method is a good idea.)

Doing a quick bit of testing suggested that using  "%f" % float
might be better than using float.to_s because in some cases
the former preserves information that is actually in the float
which the latter uses. In fact, I was going to suggest you propose
to use  "%f" % self  in Float#to_d instead of  self.to_s.
But after some more thought and a bit of testing, I found that
for me in IRB  "%f" % float  only gives results to 6 decimal places
but  float.to_s  seems to give results to 14 or 15 significant figures
so overall using  float.to_s  for the conversion seems better.

If we want a BigDecimal method (and in view of  BigDecimal( float.to_s )
I don't think we do - but  BigDecimal( float.to_s )  should perhaps
be mentioned prominently in BigDecimal documentation)
how about something like (not tested):
  def BigDecimal.from_f( f )
    if f.kind_of?( Float ) then BigDecimal( f.to_s )
    else raise "BigDecimal.from_f: argument must be a Float"
    end
  end

An argument for this is that by providing it you give a reasonably
simple standard way to do it which might reduce potentially
problematic "build your own" solutions.
(I'm not against "build your own" in principle: at the least
 it can be a useful way to find out how things work,
 and it may result in an improvement, and as long as it's
 your own time and your own decision to do it, why not?
 But where there are hidden pitfalls, a reliable and
 reasonably easy to use solution is a good idea.)


***** For those who want to know a bit more *****

*** Here's one I made earlier. ***

(Really! Two days ago, in fact. For those not in the UK the reference
 is to a long-running BBC television programme for children:
   http://en.wikipedia.org/wiki/Blue_peter
   ... The show is also famous for its "makes", which are demonstrations
   of how to construct a useful object or prepare food. These have given 
rise
   to the oft-used phrase "Here's one I made earlier", as presenters 
bring out
   a perfect and completed version of the object they are making. ...
 Well, one thing we try to do in Ruby is make (almost) perfect objects! 
)

I'm looking (for my own purposes) at the various Numeric classes,
and I think it's useful to consider them in terms of precision
potentially being lost on conversion and how we should cope with that.

(I'm using Microsoft Windows Vista (*1): if the behaviour shown doesn't 
work,
 just increase n until it does! Admittedly this uses Integer not 
BigDecimal,
 but it shows the potential problems if precision is lost.)

irb

n = 53
i = 2 ** n           #=>  9007199254740992
ii = i + 1           #=>  9007199254740993
f = Float( i )       #=> 9.00719925474099e+015
ff = Float( ii )     #=> 9.00719925474099e+015
iii = Integer( ff )  #=>  9007199254740992
i  == ii    #=> false
i  == f     #=> true
ii == f     #=> true
f  == ff    #=> true
i  <=> ii   #=> -1
i  <=> f    #=> 0
ii <=> f    #=> 0

Colin Bartlett

(*1) Question: Why do you like Linux and BSD Unix so much?
               You've never used either of them.
     Answer:   No, but I have used Microsoft Windows.
with apologies to Karl Kraus http://en.wikipedia.org/wiki/Karl_Kraus

Roger Pack wrote:
> Thought I'd pass these thoughts by the readers here before sending them
> to core.  They're my current wish list :)
> 
> 1) add an Object#in? method to complement the existing Array#include?
> 
> 2) add some useful lists of exceptions, ex: IO.SELECT_EXCEPTIONS,
> IO.READ_EXCEPTIONS, so that you can rescue the wide gamut of them
> appropriately, should you desire to.
> 
> 3) provide an easier way to know which platform you're on than
> RUBY_PLATFORM =~ /mswin32|dgjpp|mingw/
> 
> 4) (this one's controversial) remove the extra # for code in strings
> (i.e. "string#{code}") -> "string{code}" less typing.
> 
> 5) add a BigDecimal(float) method.
>  -> BigDecimal.new("%f" % float)
> 
> 6) add Dir.directory?
>  -> File.directory?
> 
> Feedback?
> Thanks!
> -=r

Object#in? can be very useful in keeping code legible; it more likely to 
be problematic in some circumstances than others. I'm using it in the 
game I'm writing right now and it's working just fine, but I don't 
really think it should be part of Ruby Core.

That said, it's very very easy to add yourself if you need it:

class Object
  def in?(object)
    if object.respond_to?(:include?) then
      object.include? self
    else
      false
    end
  end #def in?
end #class Object

Roger Pack wrote:
>>> 4) (this one's controversial) remove the extra # for code in strings
>>> (i.e. "string#{code}") -> "string{code}" less typing.
>> Would not like having to escape { in strings...
> 
> Ok. This next isn't meant as an attacking question but...do you use { in 
> normal strings often?  Granted probably more than #, but...?

Yes, quite extensively when generating C code. Would hate to have to use
\{...} instead of {...} for C blocks. Also, but less often, inside of

   instance_eval "...",

or

   eval "...", etc.

On 11.04.2009 20:43, Roger Pack wrote:
> A few other observations:
> 
> re: in?
> Currently with #select you've got one in Kernel [which is IO.select] but 
> Arrays seem to have their own #select.  So it is "conceivably possible" 
> to have a "default #in?" and have it overridden by Clutch#in? or 
> House#in? if desired.

That's not exactly an override situation because the "global" select is
used without an instance or with a class instance (IO) while
Enumerable#select is always used with an instance.

> And knowing which ones are on each OS is also annoying.
> I am proposing more of a (platform dependent) container of all possible 
> exceptions, regardless of what they may mean.  Or have them all include 
> a common ancestor--same result.

There is SystemCallError already:

ObjectSpace.each_object(Class) do |cl|
   p cl if cl.ancestors.include? SystemCallError
end

>> about versions but must know the distro.  I see too much effort for too
> 
> Typically "enough" OS information is given in RUBY_PLATFORM to determine 
> the platform--it's just "hard" to use that for such.  My example being 
> that knowing if you're on windows is something like RUBY_PLATFORM =~ 
> /dgjpp|mingw|mswin/
> which seems overly complex for me.  And very hard to get right the first 
> time (ex: RUBY_PLATFORM =~ /win/ doesn't work--that includes darwin).

And how do you want to resolve the issues I have raised in my posting?
The question really is "what is a platform"?  When using cygwin, are you
"on Windows" or not? etc.

>>> 5) add a BigDecimal(float) method.
>>>  -> BigDecimal.new("%f" % float)
>> 0
>>
>> Seems reasonable at first sight but the absence might have a reason.
>> For example, by making the conversion to String explicit it is more
>> obvious that float and BigDecimal are not really compatible.
> 
> Yeah I wonder that myself. I was just hoping to make it easier to use 
> BigDecimal, since Floats are so imprecise to use for decimal numbers :)

Which is exactly the reason why conversion to a BigDecimal should not be
too easy.  String with decimal encoded number as input format for a
BigDecimal is really the proper type as you can see from the name
"BigDecimal". :-)

>> if test ?d, "some dir"
> 
> Could you explain that again?  Not sure I do understand the idiom. 
> Looks like bash?

[robert@ora01 ~]$ ruby19 -e 'puts "yeah!" if test ?d, "."'
yeah!
[robert@ora01 ~]$

Happy Easter!

  robert

On Sat, Apr 11, 2009 at 1:35 PM, Colin Bartlett 
<colinb2r@googlemail.com> wrote:
>> a collection. Object has #in? because it is--what? A potential member of
> For example, on some occasions we use:  Integer === object
>
>>> Seems reasonable at first sight but the absence might have a reason.
> and I seem to recall that there's a way to assign a method to a variable
> which could reduce this to BD( "123.456" ). I think it's important
> to normally use strings to get BigDecimal values to avoid precision errors.

True, so I'd agree that encouraging Float -> BigDecimal conversions
isn't a great idea. OTOH, what I would like to see is BigDecimal (and
maybe even Rational) literal syntax in Ruby in the future, e.g.

1.23  # Float
1.23d # BigDecimal
a =1 # Fixnum
a/2 # => 0
b = 1r # Rational
b/2 # == Rational("1/2")

Essentially, provide a way to make working with exact numbers
convenient without necessarily adopting all the behavioral changes of
the mathn library (which you may want to avoid in order not to break
older code.)

On Apr 11, 2:27 pm, Adam Gardner <adam.oddfel...@gmail.com> wrote:
>
> end #class Object
Before reading your email I was writing my own Object#in? and came up
with a very similar method without testing :include? support:

class Object
  def in?(collection)
    collection.include? self
  end
end

This way you can detect you're calling Object#in? with an "incorrect"
argument. For example,

3.include? [1,2,3] => true
3.include? 5 => NoMethodError: undefined method `include?' for
5:Fixnum

My $0.02.

Best regards,
Edgardo

Why aren't they? What do you mean by this out of interest?

Blog: http://random8.zenunit.com/
Learn rails: http://sensei.zenunit.com/

On 11/04/2009, at 6:45 PM, Robert Klemme <shortcutter@googlemail.com>

Please do not top post.

2009/4/14 Julian Leviston <julian@coretech.net.au>:
> Why aren't they? What do you mean by this out of interest?

09:41:50 ~$ ruby19 x.rb
Float              0.00000000010000000827
BigDecimal         0.00000000010000000000
Float              0.00000000000000000827
BigDecimal         0.00000000000000000000
false
true
09:41:52 ~$ cat x.rb

require 'bigdecimal'

f = (1.0 + 1.0e-10) - 1.0
bd = (BigDecimal.new("1.0") + BigDecimal.new("1.0e-10")) - 
BigDecimal.new("1.0")

printf "%-10s %30.20f\n", f.class, f
printf "%-10s %30.20f\n", bd.class, bd

f -= 1.0e-10
bd -= BigDecimal.new "1.0e-10"

printf "%-10s %30.20f\n", f.class, f
printf "%-10s %30.20f\n", bd.class, bd

puts f == 0.0, bd == 0.0
09:41:57 ~$

> On 11/04/2009, at 6:45 PM, Robert Klemme <shortcutter@googlemail.com> wrote:
>
>> Seems reasonable at first sight but the absence might have a reason. For
>> example, by making the conversion to String explicit it is more obvious that
>> float and BigDecimal are not really compatible.

Cheers

robert

> 09:41:50 ~$ ruby19 x.rb
> Float              0.00000000010000000827
> BigDecimal         0.00000000010000000000
> Float              0.00000000000000000827
> BigDecimal         0.00000000000000000000
> false
> true
> 09:41:52 ~$ cat x.rb

Good examples.
I should explain better my original thought perhaps--I don't actually 
want to convert "rounded" data but yes make it "cleaner" and "simpler" 
to make BigDecimals.
I think I can achieve this by allowing for BigDecimal(float) but, 
similar to how String#to_s works currently in trunk:
"only accept default values of floats"
i.e.

BigDecimal(0.9)
=>  #<BigDecimal:b7db9e14,'0.9E0',4(8)>

>> BigDecimal((2.0-1.1).to_s) # 2.0-1.1 != 0.9
# BOOM not allowed, since that's a float that's been rounded away from 
its pure form (0.9's default form)

You can ascertain this a la
("%g" % (2.0-1.1)).to_f != (2.0-1.1) # blow up in this case--those are 
rounded floats we don't want to convert them.

Somewhat sane?

-=r

> Thought I'd pass these thoughts by the readers here before sending them
> to core.  They're my current wish list :)


and here's a few more, for any feedbacks:

1) have load not "necessarily" require the .rb suffix, (i.e. make it 
behave more like require).  Rationale--both load and require are 
ultimately used to import ruby scripts, Therefore having one be suffix 
insensitive and the other not is surprising.

2) having "string" + something default to
"string" + something.to_s

History: when I mentioned this proposal once, Matz said something like 
"it used to work that way, but it was changed because it hid bugs"
I think the reason that it hides bugs is because if the "something" is 
nil, the concatenation silently concatenated with nothing--you would 
have hoped it would have raised an exception, but instead you just have 
strings with mysteriously shorter lengths.

(i.e. nil should *not* default to .to_s--this hides bugs)

therefore I would propose that string#+ default to
string + something.to_s
unless that something is nil -- then raise an exception.
I just really miss this from java, and dislike having to type in .to_s 
quite frequently (and I do know you can leverage .to_str for this same 
effect, I just wish it were default--I'm quite lazy and want to share 
this "nicety" from java).

Thoughts?
Thanks!
-=r

some_var = '/tmp'


print 'Is a directory.' if Dir.directory? some_var

vs

print 'Is a directory.' if File.directory? some_var

The difference is small, and it is not really important but from a 
logical point of view, since we don't treat both as Inodes anyway in 
ruby, using Dir.directory? seems syntactically more logical than using 
File.directory?.

But it is not anything which really bothers me.

Roger Pack wrote:
> 2) having "string" + something default to
> "string" + something.to_s

What about getting in the habit of

   "#{s}#{t}"

instead of

   s + t

to take care of the #to_s for you? Or course, that doesn't help with

   s += t

Hi --

Roger Pack wrote:
> Thought I'd pass these thoughts by the readers here before sending them
> to core.  They're my current wish list :)

Could you possibly start a different thread for each topic? Bunching
them together makes it very labor-intensive to try to follow the thread
if one is interested in a specific point, and impossible to tell who has
responded to what without going through six topics' worth of posts.

> 4) (this one's controversial) remove the extra # for code in strings
> (i.e. "string#{code}") ->  "string{code}" less typing.

It's not extra, though. One way or another you have to flag the
difference between interpolation and literal {. Having to escape literal
{ because they have special meaning seems to me to be the long way 
'round.


David

> Could you possibly start a different thread for each topic? Bunching
> them together makes it very labor-intensive to try to follow the thread
> if one is interested in a specific point, and impossible to tell who has
> responded to what without going through six topics' worth of posts.

Great idea.
-=r

> Doing a quick bit of testing suggested that using  "%f" % float
> might be better than using float.to_s because in some cases
> the former preserves information that is actually in the float
> which the latter uses. In fact, I was going to suggest you propose
> to use  "%f" % self  in Float#to_d instead of  self.to_s.
> But after some more thought and a bit of testing, I found that
> for me in IRB  "%f" % float  only gives results to 6 decimal places
> but  float.to_s  seems to give results to 14 or 15 significant figures
> so overall using  float.to_s  for the conversion seems better.

The good news is that with current ruby trunk I believe that Float#to_s 
is as accurate as doing a ("%f" % self) [it no longer loses accuracy].


> If we want a BigDecimal method (and in view of  BigDecimal( float.to_s )
> I don't think we do - but  BigDecimal( float.to_s )  should perhaps
> be mentioned prominently in BigDecimal documentation)
> how about something like (not tested):
>   def BigDecimal.from_f( f )
>     if f.kind_of?( Float ) then BigDecimal( f.to_s )
>     else raise "BigDecimal.from_f: argument must be a Float"
>     end
>   end

That would be nice, though the way you suggested earlier (require 
'bigdecimal/util'; 3.3.to_d) works for me :)
-=

Hello,

>> Ok. This next isn't meant as an attacking question but...do you use { in
>> normal strings often?  Granted probably more than #, but...?
>
> Yes, quite extensively when generating C code. Would hate to have to use
> \{...} instead of {...} for C blocks.

So do I when generating LaTeX code.

Cheers,