From noreply at rubyforge.org  Thu Nov  1 22:44:22 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Thu,  1 Nov 2007 22:44:22 -0400 (EDT)
Subject: [Facets] [ facets-Patches-15277 ] Enhanced Module::attr_accessor
	and Module::attr_reader supporting default values
Message-ID: <20071102024422.26E6E18585D7@rubyforge.org>

Patches item #15277, was opened at 2007-11-01 21:44
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3171&aid=15277&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Nils Jonsson (njonsson)
Assigned to: Nobody (None)
Summary: Enhanced Module::attr_accessor and Module::attr_reader supporting default values

Initial Comment:
This is a backwards-compatible enhancement to Module::attr_accessor and Module::attr_reader. They accept a hash of attribute names with default values of those attributes. That is to say, you can specify a value for an attribute that it will have until you set it to something else. If the default value is a Proc then the Proc is called in order to obtain a default value.

This kind of thing is useful in dependency-injection scenarios. The default value of an attribute represents a runtime dependency value. At test-time, the value can be set to a mock value.

Note that the normal behavior of attr_accessor and attr_reader is still supported. The hash attributes are optional.

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3171&aid=15277&group_id=804

From noreply at rubyforge.org  Fri Nov 23 08:04:37 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Fri, 23 Nov 2007 08:04:37 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15864 ] Error instaling facets-2.1.2.gem
Message-ID: <20071123130437.80BE4185864F@rubyforge.org>

Bugs item #15864, was opened at 2007-11-23 08:04
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15864&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Nobody (None)
Assigned to: Nobody (None)
Summary: Error instaling facets-2.1.2.gem

Initial Comment:


U:...> gem install facets-2.1.2.gem
ERROR:  While executing gem ... (RuntimeError)
    Error instaling facets-2.1.2.gem:
        invalid gem format for facets-2.1.2.gem

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15864&group_id=804

From transfire at gmail.com  Fri Nov 23 08:58:01 2007
From: transfire at gmail.com (Trans)
Date: Fri, 23 Nov 2007 05:58:01 -0800 (PST)
Subject: [Facets] [ facets-Bugs-15864 ] Error instaling facets-2.1.2.gem
In-Reply-To: <20071123130437.80BE4185864F@rubyforge.org>
References: <20071123130437.80BE4185864F@rubyforge.org>
Message-ID: <de2129a8-41d6-4a82-ad56-db8b5d53283a@p69g2000hsa.googlegroups.com>



On Nov 23, 8:04 am, <nore... at rubyforge.org> wrote:
> Bugs item #15864, was opened at 2007-11-23 08:04
> You can respond by visiting:http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15864&group_i...
>
> Category: None
> Group: None
> Status: Open
> Resolution: None
> Priority: 3
> Submitted By: Nobody (None)
> Assigned to: Nobody (None)
> Summary: Error instaling facets-2.1.2.gem
>
> Initial Comment:
>
> U:...> gem install facets-2.1.2.gem
> ERROR:  While executing gem ... (RuntimeError)
>     Error instaling facets-2.1.2.gem:
>         invalid gem format for facets-2.1.2.gem

Thanks for the report. I'm not having problems with it, so my first
thought, is whether this to do with the recent RubyGems update. What
version of RubyGems are you running? Also what OS?

Thanks,
T.

From noreply at rubyforge.org  Fri Nov 23 22:41:46 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Fri, 23 Nov 2007 22:41:46 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15864 ] Error instaling facets-2.1.2.gem
Message-ID: <20071124034146.1F5031858646@rubyforge.org>

Bugs item #15864, was opened at 2007-11-23 08:04
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15864&group_id=804

>Category: Insallation
Group: None
Status: Open
Resolution: None
>Priority: 5
Submitted By: Nobody (None)
Assigned to: Nobody (None)
Summary: Error instaling facets-2.1.2.gem

Initial Comment:


U:...> gem install facets-2.1.2.gem
ERROR:  While executing gem ... (RuntimeError)
    Error instaling facets-2.1.2.gem:
        invalid gem format for facets-2.1.2.gem

----------------------------------------------------------------------

>Comment By: 7rans  (transami)
Date: 2007-11-23 22:41

Message:
I'm not having issues with it. I wonder if this has something to do with the latest release of RubyGems, which I have not upgraded to yet. What version of RubyGems are you running? And what OS are you on. Thanks.

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15864&group_id=804

From noreply at rubyforge.org  Tue Nov 27 19:18:01 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Tue, 27 Nov 2007 19:18:01 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15981 ] Enumerable#collate has unexpected
	results
Message-ID: <20071128001801.A7F00185865C@rubyforge.org>

Bugs item #15981, was opened at 2007-11-27 18:18
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Andrew Dudzik (adudzik)
Assigned to: Nobody (None)
Summary: Enumerable#collate has unexpected results

Initial Comment:
* It seems to me that [[],[]].collate should return {[]=>[]}, but it returns {}.  This seems to be a consequence of the use of flatten!.

* Also, [[]].collate{ |x| [x,x] } should return {[]=>[]}, but returns {nil=>nil}. (this seems to be a similar problem, but is not fixed by removing flatten)

* Here is my proposed fix:

module Enumerable
  def collate(&yld)
    if yld
      inject({}) do |h, *kv|      # Used to be inject({}) do |h,kv|
        nk, nv = *yld[*kv].to_a   # Used to be nk, nv = *yld[*kv].to_a.flatten
        h[nk] = nv
        h
      end
    else
      Hash[*self.to_a]            # Used to be Hash[*self.to_a.flatten]
    end
  end
end


* Caveats:

Note that this will change the behavior of e.g. [[1,2]].collate; instead of returning {1=>2}, it will throw an ArgumentError.  But in my opinion, if collate is being used in this way, the user should instead call flatten.collate.

This will also break the behavior of collate on hashes.  If we *really* want to maintain that, we can do it explicitly:

class Hash
  def collate(&yld)
    to_a.flatten.collate(&yld)
  end
end

Is there a method that we can use besides flatten in these situations, that is not recursive--this is to say, which only goes one level deep?

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

From noreply at rubyforge.org  Tue Nov 27 19:32:49 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Tue, 27 Nov 2007 19:32:49 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15981 ] Enumerable#collate has unexpected
	results
Message-ID: <20071128003249.2FA3B1858669@rubyforge.org>

Bugs item #15981, was opened at 2007-11-27 18:18
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Andrew Dudzik (adudzik)
Assigned to: Nobody (None)
Summary: Enumerable#collate has unexpected results

Initial Comment:
* It seems to me that [[],[]].collate should return {[]=>[]}, but it returns {}.  This seems to be a consequence of the use of flatten!.

* Also, [[]].collate{ |x| [x,x] } should return {[]=>[]}, but returns {nil=>nil}. (this seems to be a similar problem, but is not fixed by removing flatten)

* Here is my proposed fix:

module Enumerable
  def collate(&yld)
    if yld
      inject({}) do |h, *kv|      # Used to be inject({}) do |h,kv|
        nk, nv = *yld[*kv].to_a   # Used to be nk, nv = *yld[*kv].to_a.flatten
        h[nk] = nv
        h
      end
    else
      Hash[*self.to_a]            # Used to be Hash[*self.to_a.flatten]
    end
  end
end


* Caveats:

Note that this will change the behavior of e.g. [[1,2]].collate; instead of returning {1=>2}, it will throw an ArgumentError.  But in my opinion, if collate is being used in this way, the user should instead call flatten.collate.

This will also break the behavior of collate on hashes.  If we *really* want to maintain that, we can do it explicitly:

class Hash
  def collate(&yld)
    to_a.flatten.collate(&yld)
  end
end

Is there a method that we can use besides flatten in these situations, that is not recursive--this is to say, which only goes one level deep?

----------------------------------------------------------------------

>Comment By: Andrew Dudzik (adudzik)
Date: 2007-11-27 18:32

Message:
I suppose that inject won't be taking more than 2 block
parameters any time soon, so both splats are unnecessary:

      inject({}) do |h, kv|      # Used to be inject({}) do
|h,kv|
        nk, nv = *yld[kv].to_a   # Used to be nk, nv =
*yld[*kv].to_a.flatten

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

From noreply at rubyforge.org  Tue Nov 27 22:50:08 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Tue, 27 Nov 2007 22:50:08 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15981 ] Enumerable#collate has unexpected
	results
Message-ID: <20071128035008.F415418585FC@rubyforge.org>

Bugs item #15981, was opened at 2007-11-27 19:18
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Andrew Dudzik (adudzik)
Assigned to: Nobody (None)
Summary: Enumerable#collate has unexpected results

Initial Comment:
* It seems to me that [[],[]].collate should return {[]=>[]}, but it returns {}.  This seems to be a consequence of the use of flatten!.

* Also, [[]].collate{ |x| [x,x] } should return {[]=>[]}, but returns {nil=>nil}. (this seems to be a similar problem, but is not fixed by removing flatten)

* Here is my proposed fix:

module Enumerable
  def collate(&yld)
    if yld
      inject({}) do |h, *kv|      # Used to be inject({}) do |h,kv|
        nk, nv = *yld[*kv].to_a   # Used to be nk, nv = *yld[*kv].to_a.flatten
        h[nk] = nv
        h
      end
    else
      Hash[*self.to_a]            # Used to be Hash[*self.to_a.flatten]
    end
  end
end


* Caveats:

Note that this will change the behavior of e.g. [[1,2]].collate; instead of returning {1=>2}, it will throw an ArgumentError.  But in my opinion, if collate is being used in this way, the user should instead call flatten.collate.

This will also break the behavior of collate on hashes.  If we *really* want to maintain that, we can do it explicitly:

class Hash
  def collate(&yld)
    to_a.flatten.collate(&yld)
  end
end

Is there a method that we can use besides flatten in these situations, that is not recursive--this is to say, which only goes one level deep?

----------------------------------------------------------------------

>Comment By: 7rans  (transami)
Date: 2007-11-27 22:50

Message:
Thanks for catching this. Using Hash[*enum.to_a.flatten] is a pattern I've mistakenly used a few times. Hopefully we've gotten them all now, but I'll have to do a grep and make sure. I think we can use your version of collate as is. The blockless Hash case was a null-op anyway, so it's better to have it do something different rather than nothing at all.

----------------------------------------------------------------------

Comment By: Andrew Dudzik (adudzik)
Date: 2007-11-27 19:32

Message:
I suppose that inject won't be taking more than 2 block
parameters any time soon, so both splats are unnecessary:

      inject({}) do |h, kv|      # Used to be inject({}) do
|h,kv|
        nk, nv = *yld[kv].to_a   # Used to be nk, nv =
*yld[*kv].to_a.flatten

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

From noreply at rubyforge.org  Tue Nov 27 23:33:19 2007
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Tue, 27 Nov 2007 23:33:19 -0500 (EST)
Subject: [Facets] [ facets-Bugs-15981 ] Enumerable#collate has unexpected
	results
Message-ID: <20071128043319.B30C718585C5@rubyforge.org>

Bugs item #15981, was opened at 2007-11-27 19:18
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

Category: None
Group: None
Status: Open
Resolution: None
Priority: 3
Submitted By: Andrew Dudzik (adudzik)
Assigned to: Nobody (None)
Summary: Enumerable#collate has unexpected results

Initial Comment:
* It seems to me that [[],[]].collate should return {[]=>[]}, but it returns {}.  This seems to be a consequence of the use of flatten!.

* Also, [[]].collate{ |x| [x,x] } should return {[]=>[]}, but returns {nil=>nil}. (this seems to be a similar problem, but is not fixed by removing flatten)

* Here is my proposed fix:

module Enumerable
  def collate(&yld)
    if yld
      inject({}) do |h, *kv|      # Used to be inject({}) do |h,kv|
        nk, nv = *yld[*kv].to_a   # Used to be nk, nv = *yld[*kv].to_a.flatten
        h[nk] = nv
        h
      end
    else
      Hash[*self.to_a]            # Used to be Hash[*self.to_a.flatten]
    end
  end
end


* Caveats:

Note that this will change the behavior of e.g. [[1,2]].collate; instead of returning {1=>2}, it will throw an ArgumentError.  But in my opinion, if collate is being used in this way, the user should instead call flatten.collate.

This will also break the behavior of collate on hashes.  If we *really* want to maintain that, we can do it explicitly:

class Hash
  def collate(&yld)
    to_a.flatten.collate(&yld)
  end
end

Is there a method that we can use besides flatten in these situations, that is not recursive--this is to say, which only goes one level deep?

----------------------------------------------------------------------

>Comment By: 7rans  (transami)
Date: 2007-11-27 23:33

Message:
I'm thinking this is a little more appropriate now:

  def collate(&yld)
    if yld
      inject({}) do |h, *kv|
        nk, nv = *yld[*kv].to_a
        h[nk] = nv
        h
      end
    else
      Enumerator.new(self,:collate)
    end
  end


----------------------------------------------------------------------

Comment By: 7rans  (transami)
Date: 2007-11-27 22:50

Message:
Thanks for catching this. Using Hash[*enum.to_a.flatten] is a pattern I've mistakenly used a few times. Hopefully we've gotten them all now, but I'll have to do a grep and make sure. I think we can use your version of collate as is. The blockless Hash case was a null-op anyway, so it's better to have it do something different rather than nothing at all.

----------------------------------------------------------------------

Comment By: Andrew Dudzik (adudzik)
Date: 2007-11-27 19:32

Message:
I suppose that inject won't be taking more than 2 block
parameters any time soon, so both splats are unnecessary:

      inject({}) do |h, kv|      # Used to be inject({}) do
|h,kv|
        nk, nv = *yld[kv].to_a   # Used to be nk, nv =
*yld[*kv].to_a.flatten

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=3169&aid=15981&group_id=804

From transfire at gmail.com  Wed Nov 28 18:05:42 2007
From: transfire at gmail.com (Trans)
Date: Wed, 28 Nov 2007 15:05:42 -0800 (PST)
Subject: [Facets] [ facets-Bugs-15864 ] Error instaling facets-2.1.2.gem
In-Reply-To: <20071124034146.1F5031858646@rubyforge.org>
References: <20071124034146.1F5031858646@rubyforge.org>
Message-ID: <7424993d-5029-45e5-9829-aa055622f414@n20g2000hsh.googlegroups.com>

Lets see if 2.1.3 does the trick.

T.