From ertai at lrde.epita.fr  Sun Jan  1 11:22:30 2006
From: ertai at lrde.epita.fr (Nicolas Pouillard)
Date: Sun,  1 Jan 2006 17:22:30 +0100 (CET)
Subject: [Uttk-Patches] [core_ex] 446: Extend Inflector::Inflections.
Message-ID: <20060101162230.F22966E4C2@smtp1-g19.free.fr>

svn://svn.feydakins.org/ruby_ex/trunk/core_ex


Index: ChangeLog
from  Nicolas Pouillard  <ertai at lrde.epita.fr>

	Extend Inflector::Inflections.

	* lib/core_ex.rb: Extend Inflector::Inflections to deal with custom
	inflections (e.g. URI -> uri not u_r_i).
	* lib/core_ex/string.rb: Rename import to import! to match with the
	others.

 core_ex.rb        |   72 +++++++++++++++++++++++++++++++++++++++++++++++++++++-
 core_ex/string.rb |    4 +--
 2 files changed, 73 insertions(+), 3 deletions(-)

Index: lib/core_ex/string.rb
--- lib/core_ex/string.rb	(revision 445)
+++ lib/core_ex/string.rb	(working copy)
@@ -26,8 +26,8 @@
       self
     end
 
-    def import
-      to_path.import
+    def import!
+      to_path.import!
     end
 
     def to_path
Index: lib/core_ex.rb
--- lib/core_ex.rb	(revision 445)
+++ lib/core_ex.rb	(working copy)
@@ -83,8 +83,78 @@
   # <<< little active_support patch
   silence_warnings do
     module Inflector
+      class Inflections
+        # <<<
+        attr_reader :underscore_rules, :camelize_rules
+        # >>>
+        def initialize
+          # <<<
+          clear_all
+          # >>>
+        end
+        # <<<
+        def clear_all
+          @plurals, @singulars, @uncountables, @underscore_rules, @camelize_rules = [], [], [], [], []
+        end
+        # >>>
+        def clear ( scope=:all )
+          case scope
+            when :all
+              # <<<
+              clear_all
+              # >>>
+            else
+              instance_variable_set "@#{scope}", []
+          end
+        end
+        def underscore(rule, replacement)
+          @underscore_rules.unshift [rule, replacement]
+        end
+        def camelize(rule, replacement)
+          @camelize_rules.unshift [rule, replacement]
+        end
+        def fixed_case ( camel_cased_word,
+                         lower_case_and_underscored_word=camel_cased_word.downcase )
+          camelize(lower_case_and_underscored_word, camel_cased_word)
+          underscore(camel_cased_word, lower_case_and_underscored_word)
+        end
+        def fixed_cases ( *words )
+          for word in words
+            if word.is_a? Array
+              fixed_case(*word)
+            else
+              fixed_case word
+            end
+          end
+        end
+      end
+      # Since Inflections is already loaded and is a Singleton we need to initialize
+      # our variables by hand:
+      Inflections.instance.instance_eval { @underscore_rules, @camelize_rules = [], [] }
+
       def underscore(camel_cased_word)
-        camel_cased_word.to_s.gsub(/::/, '/').gsub(/([A-Z])/,'_\1').gsub(/(^|\/)_/, '\1').downcase
+        # <<<
+        result = camel_cased_word.to_s.split '::'
+        result.map! do |word|
+          inflections.underscore_rules.each { |(rule, replacement)| break if word.gsub!(rule, replacement) }
+          word.gsub!(/([A-Z])/, '_\1')
+          word.gsub!(/^_/, '')
+          word.downcase!
+          word
+        end
+        result.join '/'
+        # >>>
+      end
+      def camelize(lower_case_and_underscored_word)
+        # <<<
+        result = lower_case_and_underscored_word.to_s.split '/'
+        result.map! do |word|
+          inflections.camelize_rules.each { |(rule, replacement)| break if word.gsub!(rule, replacement) }
+          word.gsub!(/(^|_)(.)/) { $2.upcase }
+          word
+        end
+        result.join '::'
+        # >>>
       end
     end # module Inflector
   end
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 186 bytes
Desc: OpenPGP digital signature
Url : http://rubyforge.org/pipermail/ttk-patches/attachments/20060101/9af04e82/signature.bin
-------------- next part --------------
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches

From ertai at lrde.epita.fr  Sun Jan  1 11:29:59 2006
From: ertai at lrde.epita.fr (Nicolas Pouillard)
Date: Sun,  1 Jan 2006 17:29:59 +0100 (CET)
Subject: [Uttk-Patches] [ruby_ex] 447: Fix the URI::* lazy loading.
Message-ID: <20060101162959.E62ED4409D@smtp3-g19.free.fr>

svn://svn.feydakins.org/ruby_ex/trunk


Index: ChangeLog
from  Nicolas Pouillard  <ertai at lrde.epita.fr>

	Fix the URI::* lazy loading.

	* lib/uri_ex.rb: Setup some inflection exceptions and add some test.
	* lib/commands/command.rb: Extend the sleep time to 2 seconds.

 commands/command.rb |    5 +++--
 uri_ex.rb           |   38 +++++++++++++++++++++++++++++++++++---
 2 files changed, 38 insertions(+), 5 deletions(-)

Index: lib/uri_ex.rb
--- lib/uri_ex.rb	(revision 446)
+++ lib/uri_ex.rb	(working copy)
@@ -6,6 +6,12 @@
 require 'uri'
 
 module UriEx
+
+  # Setup exceptions about the camel_case of URI and its classes.
+  Inflector.inflections do |i|
+    i.fixed_cases 'URI', 'DRuby', 'FTP', 'HTTP', 'MySQL', 'PgSQL'
+  end
+
 end # module UriEx
 
 module URI
@@ -102,16 +108,42 @@
         end
       end
 
+      def teardown
+        URI::Generic.remove_subclasses
+      end
+
       def test_mk_custom_opts
         assert_equal([], @foo.mk_custom_opts)
         assert_equal(%w[-a b -c -d e -f --ghi --jkl mno], @cmplx.mk_custom_opts)
       end
 
-    end # class GenericTest
+      def test_lazy_loading_one
+        assert_nothing_raised do
+          assert_equal 'URI::DRuby', URI::DRuby.name
+          assert_equal URI::Generic, URI::DRuby.superclass
+        end
+      end
+
+      def test_lazy_loading_one2
+        assert_equal 'File', ::File.name
+        assert_nothing_raised do
+          assert_equal 'URI::File', 'URI::File'.constantize.name
+          assert_equal URI::Generic, 'URI::File'.constantize.superclass
+        end
   end
 
-  PathList[__FILE__.to_path.dirname + 'uri/**/*.rb'].each do |path|
-    require path.to_s
+      def test_lazy_loading_full
+        PathList[__FILE__.to_path.dirname + '(uri/**/*).rb'].each do |full, path|
+          assert_nothing_raised do
+            class_name = path.to_s.camelize.gsub(/Ex$/, '')
+            class_object = class_name.constantize
+            assert_equal class_name, class_object.name
+            assert_equal URI::Generic, class_object.superclass
+          end
+        end
+      end
+
+    end # class GenericTest
   end
 
 end # module URI
Index: lib/commands/command.rb
--- lib/commands/command.rb	(revision 446)
+++ lib/commands/command.rb	(working copy)
@@ -416,12 +416,13 @@
       end
 
       def test_fork
-        cmd = (@touch + @tmp) + (@sleep + 1)
+        sleep_arg = 2
+        cmd = (@touch + @tmp) + (@sleep + sleep_arg)
         diff = DTime.diff { @data = cmd.fork }
         assert(@tmp.exist?)
         @data.waitpid
         assert(@tmp.exist?)
-        assert(diff.to_f < 1)
+        assert(diff.to_f < sleep_arg, "#{diff} should be < #{sleep_arg}")
       end
 
       def test_ruby
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 186 bytes
Desc: OpenPGP digital signature
Url : http://rubyforge.org/pipermail/ttk-patches/attachments/20060101/ae6123da/signature-0001.bin
-------------- next part --------------
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches

From nicolas.despres at gmail.com  Mon Jan  2 15:14:10 2006
From: nicolas.despres at gmail.com (=?ISO-8859-1?Q?Nicolas_Despr=E8s?=)
Date: Mon, 2 Jan 2006 21:14:10 +0100
Subject: [Uttk-Patches] (backend-refactoring) 771: Introduce
	Logger::Backend.
In-Reply-To: <20051220145702.DEC35B0C05@smtp.feydakins.org>
References: <20051220145702.DEC35B0C05@smtp.feydakins.org>
Message-ID: <840b875c0601021214w62ec67fncea47fc577212ad5@mail.gmail.com>

On 12/20/05, Nicolas Pouillard <ertai at lrde.epita.fr> wrote:
> svn://svn.feydakins.org/uttk/branches/without_the_up_notif
>
> You can also view this changeset here:
>
> http://http://dev.uttk.org/repository/changesets/771
>
> Index: ChangeLog
> from  Nicolas Pouillard  <ertai at lrde.epita.fr>
>
>         Introduce Logger::Backend.
>
>         * lib/uttk/logger/backend.rb: New. A common base class for D::Dumper
>         and F::Filter.

[...]

> +    # - or even better to avoid the leaf encoding (by ourself),
> +    #   we send the path of the leaf and the leaf:

So new_node is useless?
I don't understand well this last option.

[...]

Ok.

--
Nicolas Despr?s
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches


From nicolas.despres at gmail.com  Tue Jan  3 08:07:27 2006
From: nicolas.despres at gmail.com (=?ISO-8859-1?Q?Nicolas_Despr=E8s?=)
Date: Tue, 3 Jan 2006 14:07:27 +0100
Subject: [Uttk-Patches] (backend-refactoring) 771: Introduce
	Logger::Backend.
In-Reply-To: <cd67f63a0601021334o5fcc28c3ge1b72760744c59d8@mail.gmail.com>
References: <20051220145702.DEC35B0C05@smtp.feydakins.org>
	<840b875c0601021214w62ec67fncea47fc577212ad5@mail.gmail.com>
	<cd67f63a0601021334o5fcc28c3ge1b72760744c59d8@mail.gmail.com>
Message-ID: <840b875c0601030507y3cdc998awe002d2c91e84aba8@mail.gmail.com>

On 1/2/06, Nicolas Pouillard <nicolas.pouillard at gmail.com> wrote:
> On 1/2/06, Nicolas Despr?s <nicolas.despres at gmail.com> wrote:
> > On 12/20/05, Nicolas Pouillard <ertai at lrde.epita.fr> wrote:
> > > svn://svn.feydakins.org/uttk/branches/without_the_up_notif
> > >
> > > You can also view this changeset here:
> > >
> > > http://http://dev.uttk.org/repository/changesets/771
> > >
> > > Index: ChangeLog
> > > from  Nicolas Pouillard  <ertai at lrde.epita.fr>
> > >
> > >         Introduce Logger::Backend.
> > >
> > >         * lib/uttk/logger/backend.rb: New. A common base class for D::Dumper
> > >         and F::Filter.
> >
> > [...]
> >
> > > +    # - or even better to avoid the leaf encoding (by ourself),
> > > +    #   we send the path of the leaf and the leaf:
> >
> > So new_node is useless?
> > I don't understand well this last option.
>
> Yes it's useless if the path is send completly:

That's what I thought. So new_node is useless in the implementation.
--
Nicolas Despr?s
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches


From nicolas.despres at gmail.com  Tue Jan  3 18:38:38 2006
From: nicolas.despres at gmail.com (=?ISO-8859-1?Q?Nicolas_Despr=E8s?=)
Date: Wed, 4 Jan 2006 00:38:38 +0100
Subject: [Uttk-Patches] (backend-refactoring) 771: Introduce
	Logger::Backend.
In-Reply-To: <cd67f63a0601030538r42e0c86cycb278357cba03b0@mail.gmail.com>
References: <20051220145702.DEC35B0C05@smtp.feydakins.org>
	<840b875c0601021214w62ec67fncea47fc577212ad5@mail.gmail.com>
	<cd67f63a0601021334o5fcc28c3ge1b72760744c59d8@mail.gmail.com>
	<840b875c0601030507y3cdc998awe002d2c91e84aba8@mail.gmail.com>
	<cd67f63a0601030538r42e0c86cycb278357cba03b0@mail.gmail.com>
Message-ID: <840b875c0601031538l11621b1ds8fb6c40af517bea1@mail.gmail.com>

On 1/3/06, Nicolas Pouillard <nicolas.pouillard at gmail.com> wrote:
> On 1/3/06, Nicolas Despr?s <nicolas.despres at gmail.com> wrote:
> > On 1/2/06, Nicolas Pouillard <nicolas.pouillard at gmail.com> wrote:
> > > On 1/2/06, Nicolas Despr?s <nicolas.despres at gmail.com> wrote:
> > > > On 12/20/05, Nicolas Pouillard <ertai at lrde.epita.fr> wrote:
> > > > > svn://svn.feydakins.org/uttk/branches/without_the_up_notif
> > > > >
> > > > > You can also view this changeset here:
> > > > >
> > > > > http://http://dev.uttk.org/repository/changesets/771
> > > > >
> > > > > Index: ChangeLog
> > > > > from  Nicolas Pouillard  <ertai at lrde.epita.fr>
> > > > >
> > > > >         Introduce Logger::Backend.
> > > > >
> > > > >         * lib/uttk/logger/backend.rb: New. A common base class for D::Dumper
> > > > >         and F::Filter.
> > > >
> > > > [...]
> > > >
> > > > > +    # - or even better to avoid the leaf encoding (by ourself),
> > > > > +    #   we send the path of the leaf and the leaf:
> > > >
> > > > So new_node is useless?
> > > > I don't understand well this last option.
> > >
> > > Yes it's useless if the path is send completly:
> >
> > That's what I thought. So new_node is useless in the implementation.
>
> That depending on the point of view:
>
> 1/ It's useless to transmit it through the filters, dumpers, network....
>
> 2/ In some filter/dumper implementation it's easier to do some task at
> the node creation
>     so this service is still needed.
>

Which entity is supposed to call new_node? Logger? abstract filters?

--
Nicolas Despr?s
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches


From nicolas.despres at gmail.com  Tue Jan  3 19:47:04 2006
From: nicolas.despres at gmail.com (=?ISO-8859-1?Q?Nicolas_Despr=E8s?=)
Date: Wed, 4 Jan 2006 01:47:04 +0100
Subject: [Uttk-Patches] [ruby_ex] 441: Export Concrete from Abstract.
In-Reply-To: <20051231155530.BD5278FFB1@smtp.feydakins.org>
References: <20051231155530.BD5278FFB1@smtp.feydakins.org>
Message-ID: <840b875c0601031647qac7786fo19ad6595665d5ecb@mail.gmail.com>

On 12/31/05, Nicolas Pouillard <ertai at lrde.epita.fr> wrote:
> svn://svn.feydakins.org/ruby_ex/trunk
>
>
> Index: ChangeLog
> from  Nicolas Pouillard  <ertai at lrde.epita.fr>
>
>         Export Concrete from Abstract.
>
>         * lib/abstract.rb: Export Concrete to make it lazy loadable without
>         saying Abstract before.
>         * lib/concrete.rb: New.
>

Ok, why not.

--
Nicolas Despr?s
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches


From ertai at lrde.epita.fr  Mon Jan  9 03:55:42 2006
From: ertai at lrde.epita.fr (Nicolas Pouillard)
Date: Mon,  9 Jan 2006 09:55:42 +0100 (CET)
Subject: [Uttk-Patches] [core_ex] 448: Suppress some warnings.
Message-ID: <20060109085542.6529D91526@smtp.feydakins.org>

svn://svn.feydakins.org/ruby_ex/trunk/core_ex


Index: ChangeLog
from  Nicolas Pouillard  <ertai at lrde.epita.fr>

	Suppress some warnings.

	* lib/core_ex/lazy_loading.rb: Add an alias.
	* lib/core_ex.rb: Use $VERBOSE directly since silence_warnings is not
	yet available.

 core_ex.rb              |    4 ++++
 core_ex/lazy_loading.rb |    1 +
 2 files changed, 5 insertions(+)

Index: lib/core_ex/lazy_loading.rb
--- lib/core_ex/lazy_loading.rb	(revision 447)
+++ lib/core_ex/lazy_loading.rb	(working copy)
@@ -140,6 +140,7 @@
   def name
     the_standard_name.sub(/^Controllers::/, '')
   end
+  alias_method :core_ex_const_missing, :const_missing
   def const_missing ( aConst )
     Binding.of_caller do |binding|
       nesting = eval "::Module.nesting", binding
Index: lib/core_ex.rb
--- lib/core_ex.rb	(revision 447)
+++ lib/core_ex.rb	(working copy)
@@ -77,8 +77,12 @@
     require name
   end
 
+  verbose = $VERBOSE
+  $VERBOSE = false
+
   core_ex_require 'active_support', 'activesupport', 'activesupport', '~> 1.2.5'
 
+  $VERBOSE = verbose
 
   # <<< little active_support patch
   silence_warnings do
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 186 bytes
Desc: OpenPGP digital signature
Url : http://rubyforge.org/pipermail/ttk-patches/attachments/20060109/8cb54070/signature.bin
-------------- next part --------------
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches

From nicolas.despres at gmail.com  Mon Jan  9 20:04:47 2006
From: nicolas.despres at gmail.com (=?ISO-8859-1?Q?Nicolas_Despr=E8s?=)
Date: Tue, 10 Jan 2006 02:04:47 +0100
Subject: [Uttk-Patches] [ruby_ex] 439: Enforce the double load protection.
In-Reply-To: <20051231155047.E79498FFB0@smtp.feydakins.org>
References: <20051231155047.E79498FFB0@smtp.feydakins.org>
Message-ID: <840b875c0601091704y577187d7g1ebdf2822d6c4985@mail.gmail.com>

On 12/31/05, Nicolas Pouillard <ertai at lrde.epita.fr> wrote:
> svn://svn.feydakins.org/ruby_ex/trunk
>
>
> Index: ChangeLog
> from  Nicolas Pouillard  <ertai at lrde.epita.fr>
>
>         Enforce the double load protection.
>
>         * lib/ruby_ex.rb: Enforce the double load protection.
>         * test/unit-suite.yml: Likewise but in a different way.
>

[...]

>
>    test:
>      <<it_name>>: !S::RUnit
> -      load_path: <<pwd>>/../lib
> -      requires: ruby_ex
> -      input: !path <<it_file>>
> +      load_path: [<<pwd>>/../lib, <<pwd>>/algorithms]
> +      requires: <<pwd>>/../lib/ruby_ex # The full path is here to ensure we
> +                                       # load the good one.
> +      input: !path <<it_name>>
>        uttk_unit_options: --uttk no
>        verbose: true
>

I don't kown that much about RubyOnRails import system but it seems
strange to me to be oblige to "requires" a path which have been
already loaded by the "load_path" tag. Am I wrong?

--
Nicolas Despr?s
_______________________________________________
Uttk-Patches mailing list
Uttk-Patches at lists.feydakins.org
http://mailman.feydakins.org/mailman/listinfo/uttk-patches