upmark 0.1.0 → 0.1.1

data/.gitignore

@@ -1,5 +1,4 @@
 *.gem
 .bundle
-Gemfile.lock
-pkg/*
+coverage
 .rvmrc

data/Gemfile.lock

@@ -0,0 +1,34 @@
+PATH
+  remote: .
+  specs:
+    upmark (0.1.1)
+      parslet
+
+GEM
+  remote: http://rubygems.org/
+  specs:
+    blankslate (2.1.2.4)
+    diff-lcs (1.1.3)
+    multi_json (1.0.3)
+    parslet (1.2.1)
+      blankslate (~> 2.0)
+    rspec (2.6.0)
+      rspec-core (~> 2.6.0)
+      rspec-expectations (~> 2.6.0)
+      rspec-mocks (~> 2.6.0)
+    rspec-core (2.6.4)
+    rspec-expectations (2.6.0)
+      diff-lcs (~> 1.1.2)
+    rspec-mocks (2.6.0)
+    simplecov (0.5.3)
+      multi_json (~> 1.0.3)
+      simplecov-html (~> 0.5.3)
+    simplecov-html (0.5.3)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  rspec
+  simplecov
+  upmark!

data/lib/upmark.rb

@@ -5,20 +5,20 @@ require "core_ext/array"
 require "upmark/parser/xml"
 require 'upmark/transform_helpers'
 require "upmark/transform/markdown"
-require "upmark/transform/pass_through"
+require "upmark/transform/normalise"
 require "upmark/transform/preprocess"
 require "upmark/version"
 
 module Upmark
   def self.convert(html)
     xml          = Parser::XML.new
+    normalise    = Transform::Normalise.new
     preprocess   = Transform::Preprocess.new
-    pass_through = Transform::PassThrough.new
     markdown     = Transform::Markdown.new
 
     ast = xml.parse(html.strip)
+    ast = normalise.apply(ast)
     ast = preprocess.apply(ast)
-    ast = pass_through.apply(ast)
     ast = markdown.apply(ast)
 
     # The result is either a String or an Array.

data/lib/upmark/transform/markdown.rb

@@ -1,9 +1,7 @@
 module Upmark
   module Transform
-    # The Markdown transform class.
-    #
-    # Transforms an abstract syntax tree (AST) into a Markdown document.
-    #
+    # A transform class which converts an abstract syntax tree (AST) into
+    # a Markdown document.
     class Markdown < Parslet::Transform
       include TransformHelpers
 

data/lib/upmark/transform/normalise.rb

@@ -0,0 +1,39 @@
+module Upmark
+  module Transform
+    # A transform class withich normalises start/end/empty tags into the
+    # same structure.
+    class Normalise < Parslet::Transform
+      rule(
+        element: {
+          start_tag: {name: simple(:name), attributes: subtree(:attributes)},
+          end_tag:   {name: simple(:name)},
+          children:  subtree(:children)
+        }
+      ) do
+        {
+          element: {
+            name:       name,
+            attributes: attributes,
+            children:   children,
+            ignore:     false
+          }
+        }
+      end
+
+      rule(
+        element: {
+          empty_tag: {name: simple(:name), attributes: subtree(:attributes)}
+        }
+      ) do
+        {
+          element: {
+            name:       name,
+            attributes: attributes,
+            children:   [],
+            ignore:     false
+          }
+        }
+      end
+    end
+  end
+end

data/lib/upmark/transform/preprocess.rb

@@ -1,35 +1,19 @@
+require "upmark/transform/ignore"
+
 module Upmark
   module Transform
-    # A preprocess transform which normalises start/end/empty tags into the same structure.
+    # A transform class which marks block-level elements as ignored.
+    # i.e. These elements should not be converted to Markdown.
     class Preprocess < Parslet::Transform
-      rule(
-        element: {
-          start_tag: {name: simple(:name), attributes: subtree(:attributes)},
-          end_tag:   {name: simple(:name)},
-          children:  subtree(:children)
-        }
-      ) do
-        {
-          element: {
-            name:       name,
-            attributes: attributes,
-            children:   children,
-            ignore:     false
-          }
-        }
-      end
+      include TransformHelpers
 
-      rule(
-        element: {
-          empty_tag: {name: simple(:name), attributes: subtree(:attributes)}
-        }
-      ) do
+      element(:div, :table, :pre) do |element|
         {
           element: {
-            name:       name,
-            attributes: attributes,
-            children:   [],
-            ignore:     false
+            name:       element[:name],
+            attributes: element[:attributes],
+            children:   Ignore.new.apply(element[:children]),
+            ignore:     true
           }
         }
       end

data/lib/upmark/version.rb

@@ -1,3 +1,3 @@
 module Upmark
-  VERSION = "0.1.0"
+  VERSION = "0.1.1"
 end

data/spec/acceptance/upmark_spec.rb

@@ -109,20 +109,18 @@ organic
     MD
   end
 
-  context "<div>" do
-    let(:html) { <<-HTML.strip }
+  context "block-level elements" do
+    context "<div>" do
+      let(:html) { <<-HTML.strip }
 <div>messenger <strong>bag</strong> skateboard</div>
 <div id="tofu" class="art party">messenger <strong>bag</strong> skateboard</div>
-    HTML
+      HTML
 
-    it { should == <<-MD.strip }
-<div>messenger <strong>bag</strong> skateboard</div>
-<div id="tofu" class="art party">messenger <strong>bag</strong> skateboard</div>
-    MD
-  end
+      it { should == html }
+    end
 
-  context "<table>" do
-    let(:html) { <<-HTML.strip }
+    context "<table>" do
+      let(:html) { <<-HTML.strip }
 <table>
   <tr>
     <td>messenger</td>
@@ -134,48 +132,33 @@ organic
     <td>skateboard</td>
   </tr>
 </table>
-    HTML
+      HTML
 
-    it { should == <<-MD.strip }
-<table>
-  <tr>
-    <td>messenger</td>
-  </tr>
-  <tr>
-    <td><strong>bag</strong></td>
-  </tr>
-  <tr>
-    <td>skateboard</td>
-  </tr>
-</table>
-    MD
-  end
+      it { should == html }
+    end
 
-  context "<pre>" do
-    let(:html) { <<-HTML.strip }
+    context "<pre>" do
+      let(:html) { <<-HTML.strip }
 <pre>
   <code>
     messenger bag skateboard
   </code>
 </pre>
-    HTML
+      HTML
 
-    it { should == <<-MD.strip }
-<pre>
-  <code>
-    messenger bag skateboard
-  </code>
-</pre>
-    MD
+      it { should == html }
+    end
   end
 
-  context "<span>" do
-    let(:html) { <<-HTML.strip }
+  context "span-level elements" do
+    context "<span>" do
+      let(:html) { <<-HTML.strip }
 <span>messenger <strong>bag</strong> skateboard</span>
-    HTML
+      HTML
 
-    it { should == <<-MD.strip }
+      it { should == <<-MD.strip }
 <span>messenger **bag** skateboard</span>
-    MD
+      MD
+    end
   end
 end

data/spec/spec_helper.rb

@@ -1,3 +1,7 @@
 require "parslet/rig/rspec"
 require "rspec"
+require "simplecov"
+
+SimpleCov.start
+
 require "upmark"

data/upmark.gemspec

@@ -9,18 +9,18 @@ Gem::Specification.new do |s|
   s.version     = Upmark::VERSION
   s.authors     = ["Josh Bassett", "Gus Gollings"]
   s.email       = "dev@theconversation.edu.au"
-  s.homepage    = "https://github.com/conversation/upmark"
+  s.homepage    = "http://github.com/conversation/upmark"
   s.summary     = %q{A HTML to Markdown converter.}
   s.description = %q{Upmark has the skills to convert your HTML to Markdown.}
 
   s.rubyforge_project = "upmark"
 
-  s.files         = `git ls-files`.split("\n")
-  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
-  s.executables   = `git ls-files -- bin/*`.split("\n").map {|f| File.basename(f) }
-  s.require_paths = ["lib"]
+  s.files       = `git ls-files`.split("\n")
+  s.test_files  = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.executables = `git ls-files -- bin/*`.split("\n").map {|f| File.basename(f) }
 
   s.add_development_dependency "rspec"
+  s.add_development_dependency "simplecov"
 
   s.add_runtime_dependency "parslet"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: upmark
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -10,11 +10,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-09-25 00:00:00.000000000 Z
+date: 2011-09-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &70218294771840 !ruby/object:Gem::Requirement
+  requirement: &70363456228720 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -22,10 +22,21 @@ dependencies:
         version: '0'
   type: :development
   prerelease: false
-  version_requirements: *70218294771840
+  version_requirements: *70363456228720
+- !ruby/object:Gem::Dependency
+  name: simplecov
+  requirement: &70363456228280 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: *70363456228280
 - !ruby/object:Gem::Dependency
   name: parslet
-  requirement: &70218294771320 !ruby/object:Gem::Requirement
+  requirement: &70363456227680 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -33,7 +44,7 @@ dependencies:
         version: '0'
   type: :runtime
   prerelease: false
-  version_requirements: *70218294771320
+  version_requirements: *70363456227680
 description: Upmark has the skills to convert your HTML to Markdown.
 email: dev@theconversation.edu.au
 executables:
@@ -44,16 +55,18 @@ files:
 - .gitignore
 - .rspec
 - Gemfile
-- LICENSE.md
+- Gemfile.lock
+- LICENSE
 - README.md
 - Rakefile
 - bin/upmark
+- coverage/.gitkeep
 - lib/core_ext/array.rb
 - lib/upmark.rb
 - lib/upmark/parser/xml.rb
 - lib/upmark/transform/ignore.rb
 - lib/upmark/transform/markdown.rb
-- lib/upmark/transform/pass_through.rb
+- lib/upmark/transform/normalise.rb
 - lib/upmark/transform/preprocess.rb
 - lib/upmark/transform_helpers.rb
 - lib/upmark/version.rb
@@ -62,7 +75,7 @@ files:
 - spec/unit/lib/upmark/parser/xml_spec.rb
 - spec/unit/lib/upmark/transform/markdown_spec.rb
 - upmark.gemspec
-homepage: https://github.com/conversation/upmark
+homepage: http://github.com/conversation/upmark
 licenses: []
 post_install_message: 
 rdoc_options: []

data/lib/upmark/transform/pass_through.rb

@@ -1,20 +0,0 @@
-require "upmark/transform/ignore"
-
-module Upmark
-  module Transform
-    class PassThrough < Parslet::Transform
-      include TransformHelpers
-
-      element(:div, :table, :pre) do |element|
-        {
-          element: {
-            name:       element[:name],
-            attributes: element[:attributes],
-            children:   Ignore.new.apply(element[:children]),
-            ignore:     true
-          }
-        }
-      end
-    end
-  end
-end