gatchaman 0.2.0 → 0.2.1

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 83669db5ec638c57ccfd8f56068d4db631a061d4
+  data.tar.gz: 4f143d83b8e109c79b6d463dac40806433e991ad
+SHA512:
+  metadata.gz: e3c44f552269b19291486c5aec3d00e4035f4f66bcee93e42f516fe8b55a96219fe395ec709d707175c8de9c76aed7dd8b92961b1d91bfb525870a32421058e0
+  data.tar.gz: f4471f7a121ae95344d9f155e6fdb9c1752ce8a8853b575442657bdfd7119333aaf93043219c4aeb12a1d2b91b5394d4b252ff5460891b748b87ee9619a05fcb

data/ChangeLog

@@ -10,3 +10,14 @@
 
 	* RADME.md add usage of Gatchaman class
 	* relaese version 0.1.1
+
+2013-02-20 maruyama shinepi <shinpeim@gmail.com>
+
+	* release version 0.2.0
+	* now gatchan expand js and css(thanks to wneko)
+
+2013-03-18 maruyama shinpei <shinpeim@gmail.com>
+
+	* release version 0.2.1
+	* some bug fixes
+	* refactor (thanks to wneko)

data/README.md

@@ -25,6 +25,6 @@ or
 4. Push to the branch (`git push origin my-new-feature`)
 5. Create new Pull Request
 
-## Contributers
+## Contributors
 
 wneko [js and css expand options]

data/Rakefile

@@ -1 +1,10 @@
 require "bundler/gem_tasks"
+
+begin
+  require 'rspec/core/rake_task'
+  RSpec::Core::RakeTask.new(:spec) do |spec|
+    spec.pattern = 'spec/**/*_spec.rb'
+    spec.rspec_opts = ['-cfs']
+  end
+rescue LoadError => e
+end

data/bin/gatchan

@@ -24,8 +24,8 @@ options = parse_options
 input_file = ARGV.shift or help
 
 begin
-  print Gatchaman.new(options[:document_root], options[:current_directory] || File.dirname(input_file)).
-    data_uri_schemize(IO.read(input_file))
+  options[:current_directory] ||= File.dirname(input_file)
+  print Gatchaman.new(options).data_uri_schemize(IO.read(input_file))
 rescue => e
   warn e.message
   exit 2

data/gatchaman.gemspec

@@ -20,4 +20,5 @@ Gem::Specification.new do |gem|
   gem.add_runtime_dependency "nokogiri", "~> 1.5.6"
   gem.add_runtime_dependency "mime-types", "~> 1.19"
   gem.add_development_dependency "rspec", "~> 2.12.0"
+  gem.add_development_dependency "rake", "~> 10.0.3"
 end

data/lib/gatchaman.rb

@@ -1,91 +1,41 @@
 require "gatchaman/version"
-require "nokogiri"
-require "mime/types"
-require "open-uri"
-require "base64"
+require "gatchaman/embedder"
+
 class Gatchaman
-  SCHEMIZE_TARGET_ELEMENTS = [
-    :img,
-    :audio,
-    :video
+  EMBED_OPTION_KEYS = [
+    :document_root,
+    :current_dir,
   ].freeze
-  
+
   def initialize(options = {})
-    @document_root = chomp_last_slash(options[:document_root])
-    @current_dir   = chomp_last_slash(options[:current_dir])
+    @options = options
+    @embed_options = {}
+    EMBED_OPTION_KEYS.each do |key|
+      @embed_options[key] = @options[key] if @options[key]
+    end
+  end
 
-    @expand_js  = options[:expand_js]  || false
-    @expand_css = options[:expand_css] || false
+  def embedders
+    if @embedders.nil?
+      @embedders = []
+      @embedders << Gatchaman::Embedder::Media.new(@embed_options)
+      @embedders << Gatchaman::Embedder::JS.new(@embed_options)  if @options[:expand_js]
+      @embedders << Gatchaman::Embedder::CSS.new(@embed_options) if @options[:expand_css]
+    end
+    @embedders
   end
 
   def data_uri_schemize(html_string)
-    doc = Nokogiri::HTML::DocumentFragment.parse(html_string)
+    doc = if Nokogiri.XML(html_string).root.name == "html"
+            Nokogiri::HTML::Document.parse(html_string)
+          else
+            Nokogiri::HTML::DocumentFragment.parse(html_string)
+          end
 
-    doc.css(SCHEMIZE_TARGET_ELEMENTS.join(',')).each do |element|
-      element[:src] = to_data_scheme(extract_path(element[:src]))
+    embedders.each do |embedder|
+      embedder.embed(doc)
     end
 
-    expand_js(doc)  if @expand_js
-    expand_css(doc) if @expand_css
-
     doc.to_s
   end
-
-  private
-  def chomp_last_slash(str)
-    str.gsub(/\/$/,"")
-  end
-  def extract_path(path)
-    if path =~ /^http/
-      return path
-    elsif path =~ /^\//
-      return @document_root + path
-    else
-      return @current_dir + "/" + path
-    end
-  end
-  def to_data_scheme(path)
-    mime = MIME::Types.type_for(path).first
-    image_binary = open(path, "r:ASCII-8BIT"){|f|
-      f.read(nil)
-    }
-    base64 = Base64.encode64(image_binary).gsub(/\s/,'')
-    "data:#{mime.to_s};base64,#{base64}"
-  end
-  def inner_html_from_file(filename)
-    ["", open(extract_path(filename)).read.chomp, ""].join("\n")
-  end
-  def expand_js(doc)
-    doc.css('script').each do |element|
-      expand_js_element(element)
-    end
-    doc
-  end
-  def expand_css(doc)
-    doc.css('link[rel=stylesheet]').each do |element|
-      expand_css_element(element)
-    end
-    doc
-  end
-  def expand_js_element(element)
-    return if element[:src].nil? or element[:src].size == 0
-    element.inner_html += inner_html_from_file(element[:src])
-    [:src, :charset].each {|attr| element.delete(attr.to_s)}
-  end
-  def expand_css_element(element)
-    element.name = "style"
-    content = inner_html_from_file(element[:href])
-    content = expand_css_url(content)
-    element.inner_html = content
-    element.attributes.keys.each do |attr_name|
-      next if [:media, :type].include? attr_name.to_sym
-      element.delete(attr_name)
-    end
-  end
-  def expand_css_url(css_content)
-    css_content.gsub(/\burl\b\(([^)]+)\)/) do
-      data_scheme = to_data_scheme(extract_path($1))
-      "url(#{data_scheme})"
-    end
-  end
 end

data/lib/gatchaman/data_scheme.rb

@@ -0,0 +1,28 @@
+require "mime/types"
+require "open-uri"
+require "base64"
+
+class Gatchaman
+  class DataScheme < String
+    def initialize(path)
+      @path = path
+      super(to_s)
+    end
+
+    def mime_type
+      @mime_type ||= MIME::Types.type_for(@path).first.to_s
+    end
+
+    def binary
+      @binary ||= open(@path, "r:ASCII-8BIT") {|f| f.read(nil) }
+    end
+
+    def base64
+      @base64 ||= Base64.encode64(binary).gsub(/\s/, '')
+    end
+
+    def to_s
+      "data:#{mime_type};base64,#{base64}"
+    end
+  end
+end

data/lib/gatchaman/embedder.rb

@@ -0,0 +1,12 @@
+require 'open-uri'
+require "nokogiri"
+require "gatchaman/data_scheme"
+require "gatchaman/path"
+require "gatchaman/embedder/media"
+require "gatchaman/embedder/css"
+require "gatchaman/embedder/js"
+
+class Gatchaman
+  module Embedder
+  end
+end

data/lib/gatchaman/embedder/css.rb

@@ -0,0 +1,31 @@
+require 'gatchaman/embedder/generic'
+
+class Gatchaman::Embedder::CSS < Gatchaman::Embedder::Generic
+  def embed(doc)
+    doc.css('link[rel=stylesheet]').each do |element|
+      embed_element(element)
+    end
+    doc.css('style').each do |element|
+      element.content = embed_all_url_in_css(element.content)
+    end
+    doc
+  end
+
+  private
+  def embed_element(element)
+    element.name = "style"
+    content = inner_html_from_file(element[:href])
+    element.inner_html = content
+    element.attributes.keys.each do |attr_name|
+      next if [:media, :type].include? attr_name.to_sym
+      element.delete(attr_name)
+    end
+  end
+
+  def embed_all_url_in_css(css_content)
+    css_content.gsub(/\burl\b\(([^)]+)\)/) do
+      data_scheme = Gatchaman::DataScheme.new(path($1))
+      "url(#{data_scheme})"
+    end
+  end
+end

data/lib/gatchaman/embedder/generic.rb

@@ -0,0 +1,22 @@
+class Gatchaman
+  module Embedder
+    class Generic
+      def initialize(options)
+        @options = options
+      end
+
+      def expand(doc)
+        doc
+      end
+
+      private
+      def path(path)
+        Path.new(path, @options)
+      end
+
+      def inner_html_from_file(filename)
+        ["", open(path(filename)).read.chomp, ""].join("\n")
+      end
+    end
+  end
+end

data/lib/gatchaman/embedder/js.rb

@@ -0,0 +1,21 @@
+require 'gatchaman/embedder/generic'
+
+class Gatchaman::Embedder::JS < Gatchaman::Embedder::Generic
+  def embed(doc)
+    doc.css('script').each do |element|
+      embed_element(element)
+    end
+    doc
+  end
+
+  private
+  def embed_element(element)
+    return if element[:src].nil? or element[:src].size == 0
+    element.inner_html += html_comment_out(inner_html_from_file(element[:src]))
+    [:src, :charset].each {|attr| element.delete(attr.to_s)}
+  end
+
+  def html_comment_out(content)
+    ["<!--", content, "-->"].join("\n")
+  end
+end

data/lib/gatchaman/embedder/media.rb

@@ -0,0 +1,15 @@
+require 'gatchaman/embedder/generic'
+
+class Gatchaman::Embedder::Media < Gatchaman::Embedder::Generic
+  TARGET_ELEMENTS = [
+    :img,
+    :audio,
+    :video
+  ].freeze
+
+  def embed(doc)
+    doc.css(TARGET_ELEMENTS.join(',')).each do |element|
+      element[:src] = Gatchaman::DataScheme.new(path(element[:src]))
+    end
+  end
+end

data/lib/gatchaman/path.rb

@@ -0,0 +1,33 @@
+require 'uri'
+require 'pathname'
+
+class Gatchaman
+  class Path < String
+    HTTP_REGEX = /^(?:shttp|http|https):/
+
+    def initialize(path, options = {})
+      @path = path
+      @document_root = options[:document_root]
+      @current_dir   = options[:current_dir]
+      super(normalized)
+    end
+
+    def normalized
+      if @path =~ HTTP_REGEX
+        @path
+      else
+        path = Pathname.new(@path)
+        if path.relative?
+          Pathname.new(@current_dir).join(@path).to_s
+        else
+          Pathname.new(@document_root).join(path.relative_path_from(root)).to_s
+        end
+      end
+    end
+
+    private
+    def root
+      Pathname.new("/")
+    end
+  end
+end

data/lib/gatchaman/version.rb

@@ -1,3 +1,3 @@
 class Gatchaman
-  VERSION = "0.2.0"
+  VERSION = "0.2.1"
 end

data/spec/gatchaman_spec.rb

@@ -7,7 +7,7 @@ describe Gatchaman do
     let(:document_root) {File.dirname(File.dirname(__FILE__))}
     let(:current_dir) {File.dirname(__FILE__)}
     let(:gatchaman){Gatchaman.new(document_root: document_root, current_dir: current_dir, expand_js: true, expand_css: true)}
-    let(:base64_encoded_resouce){
+    let(:base64_encoded_resource){
       "iVBORw0KGgoAAAANSUhEUgAAAAQAAAAGCAIAAABrW6giAAAACXBIWXMAA
        BYlAAAWJQFJUiTwAAABy2lUWHRYTUw6Y29tLmFkb2JlLnhtcAAAAAAAPH
        g6eG1wbWV0YSB4bWxuczp4PSJhZG9iZTpuczptZXRhLyIgeDp4bXB0az0
@@ -23,39 +23,54 @@ describe Gatchaman do
        bXBtZXRhPgrjpclbAAAAFUlEQVQIHWOUlJRkgAEmGANEk8MBABhmAFcTc
        YlpAAAAAElFTkSuQmCC".gsub(/[\s\n]/,'')
     }
-    let(:test_css_content) { open("#{current_dir}/resouces/test.css").read.chomp }
-    let(:test_js_content)  { open("#{current_dir}/resouces/test.js").read.chomp }
+    let(:test_css_content) { open("#{current_dir}/resources/test.css").read.chomp }
+    let(:test_js_content)  { open("#{current_dir}/resources/test.js").read.chomp }
 
     it "絶対urlのsrcをdata schemeで置き換えてくれること" do
-      gatchaman.should_receive(:open).with("http://example.com/img/test.png", "r:ASCII-8BIT").
-        and_return(Base64.decode64(base64_encoded_resouce))
+      Gatchaman::DataScheme.any_instance.should_receive(:open).with("http://example.com/img/test.png", "r:ASCII-8BIT").
+        and_return(Base64.decode64(base64_encoded_resource))
       gatchaman.data_uri_schemize("<img src='http://example.com/img/test.png'>").
-        should == "<img src=\"data:image/png;base64,#{base64_encoded_resouce}\">"
+        should == "<img src=\"data:image/png;base64,#{base64_encoded_resource}\">"
     end
 
     it "相対urlの相対パスsrcをdata schemeで置き換えてくれること" do
-      gatchaman.data_uri_schemize('<img src="resouces/test.png">').
-        should == "<img src=\"data:image/png;base64,#{base64_encoded_resouce}\">"
+      gatchaman.data_uri_schemize('<img src="resources/test.png">').
+        should == "<img src=\"data:image/png;base64,#{base64_encoded_resource}\">"
     end
 
     it "相対urlの絶対パスsrcをdata schemeで置き換えてくれること" do
-      gatchaman.data_uri_schemize('<img src="/spec/resouces/test.png">').
-        should == "<img src=\"data:image/png;base64,#{base64_encoded_resouce}\">"
+      gatchaman.data_uri_schemize('<img src="/spec/resources/test.png">').
+        should == "<img src=\"data:image/png;base64,#{base64_encoded_resource}\">"
+    end
+
+    it "ルートにhtmlタグを含む場合、htmlページとして出力すること" do
+      gatchaman.data_uri_schemize('<html><body><h1>nyan!</h1></body></html>').
+        should match /^<html>/
+    end
+
+    it "ルートにhtmlタグを含まない場合、htmlの断片として出力すること" do
+      gatchaman.data_uri_schemize('<body><h1>nyan!</h1></body>').
+        should_not match /^<html>/
     end
 
     it "cssを展開してくれること" do
-      gatchaman.data_uri_schemize('<link rel="stylesheet" type="text/css" media="screen" href="resouces/test.css">').
+      gatchaman.data_uri_schemize('<link rel="stylesheet" type="text/css" media="screen" href="resources/test.css">').
         should == %[<style type="text/css" media="screen">\n#{test_css_content}\n</style>]
     end
 
     it "css内部のurl参照を展開してくれること" do
-      gatchaman.data_uri_schemize('<link rel="stylesheet" type="text/css" media="screen" href="resouces/test_2.css">').
-        include?(base64_encoded_resouce).should be_true
+      gatchaman.data_uri_schemize('<link rel="stylesheet" type="text/css" media="screen" href="resources/test_2.css">').
+        include?(base64_encoded_resource).should be_true
+    end
+
+    it "styleタグ内部のURLを展開してくれること" do
+      gatchaman.data_uri_schemize('<style> * { background: url(/spec/resources/test.png); } </style>').
+        include?(base64_encoded_resource).should be_true
     end
 
     it "jsを展開してくれること" do
-      gatchaman.data_uri_schemize('<script src="resouces/test.js" type="text/javascript" charset="utf-8"></script>').
-        should == %[<script type="text/javascript">\n#{test_js_content}\n</script>]
+      gatchaman.data_uri_schemize('<script src="resources/test.js" type="text/javascript" charset="utf-8"></script>').
+        should == %[<script type="text/javascript"><!--\n\n#{test_js_content}\n\n--></script>]
     end
   end
 end

data/spec/resouces/test_2.css

@@ -3,3 +3,15 @@
 body {
   background: url(resouces/test.png);
 }
+
+.a {
+  background: url(resouces/test.png);
+}
+
+.b {
+  background: url(resouces/test.png);
+}
+
+.c {
+  background: url(resouces/test.png);
+}

data/spec/resources/test_2.css

@@ -0,0 +1,5 @@
+/* Here is Comment */
+
+body {
+  background: url(resources/test.png);
+}

metadata

@@ -1,20 +1,18 @@
 --- !ruby/object:Gem::Specification
 name: gatchaman
 version: !ruby/object:Gem::Version
-  version: 0.2.0
-  prerelease: 
+  version: 0.2.1
 platform: ruby
 authors:
 - Shinpei Maruyama
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-02-20 00:00:00.000000000 Z
+date: 2013-03-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -22,7 +20,6 @@ dependencies:
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -30,7 +27,6 @@ dependencies:
 - !ruby/object:Gem::Dependency
   name: mime-types
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -38,7 +34,6 @@ dependencies:
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -46,7 +41,6 @@ dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -54,11 +48,24 @@ dependencies:
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
         version: 2.12.0
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 10.0.3
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 10.0.3
 description: Gatchaman is a gem to replace src values in HTML documents with data
   URI scheme
 email:
@@ -77,42 +84,50 @@ files:
 - bin/gatchan
 - gatchaman.gemspec
 - lib/gatchaman.rb
+- lib/gatchaman/data_scheme.rb
+- lib/gatchaman/embedder.rb
+- lib/gatchaman/embedder/css.rb
+- lib/gatchaman/embedder/generic.rb
+- lib/gatchaman/embedder/js.rb
+- lib/gatchaman/embedder/media.rb
+- lib/gatchaman/path.rb
 - lib/gatchaman/version.rb
 - spec/gatchaman_spec.rb
-- spec/resouces/test.css
-- spec/resouces/test.js
-- spec/resouces/test.png
 - spec/resouces/test_2.css
+- spec/resources/test.css
+- spec/resources/test.js
+- spec/resources/test.png
+- spec/resources/test_2.css
 - spec/spec_helper.rb
 homepage: https://github.com/Shinpeim/Gatchaman
 licenses: []
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.24
+rubygems_version: 2.0.0
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Gatchaman is a gem to replace src values in HTML documents with data URI
   scheme
 test_files:
 - spec/gatchaman_spec.rb
-- spec/resouces/test.css
-- spec/resouces/test.js
-- spec/resouces/test.png
 - spec/resouces/test_2.css
+- spec/resources/test.css
+- spec/resources/test.js
+- spec/resources/test.png
+- spec/resources/test_2.css
 - spec/spec_helper.rb