mislav-remark 0.1.0

data/README.markdown

@@ -0,0 +1,17 @@
+Remark
+======
+
+A Ruby tool that parses HTML and delivers proper Markup.
+
+Usage
+-----
+
+From command-line:
+
+    ruby -Ilib -rubygems bin/remark spec/sample.html
+
+(You can also give input to STDIN instead as file argument.)
+
+From Ruby code:
+
+    Remark.new('<h1>My document</h1><p>Some content</p>').to_markdown

data/Rakefile

@@ -0,0 +1,25 @@
+desc "generates .gemspec file"
+task :gemspec do
+  spec = Gem::Specification.new do |gem|
+    gem.name = "remark"
+    gem.summary = "HTML to Markdown converter"
+    gem.email = "mislav.marohnic@gmail.com"
+    gem.homepage = "http://github.com/mislav/remark"
+    gem.authors = ["Mislav Marohnić"]
+    gem.has_rdoc = false
+    
+    gem.version = '0.1.0'
+    gem.files = FileList['Rakefile', '{bin,lib,rails,spec}/**/*', 'README*', 'LICENSE*']
+    gem.executables = Dir['bin/*'].map { |f| File.basename(f) }
+  end
+  
+  spec_string = spec.to_ruby
+  
+  begin
+    Thread.new { eval("$SAFE = 3\n#{spec_string}", binding) }.join 
+  rescue
+    abort "unsafe gemspec: #{$!}"
+  else
+    File.open("#{spec.name}.gemspec", 'w') { |file| file.write spec_string }
+  end
+end

data/bin/remark

@@ -0,0 +1,4 @@
+#!/usr/bin/env ruby
+require 'remark'
+
+puts Remark.new(ARGF.read).to_markdown

data/lib/remark.rb

@@ -0,0 +1,99 @@
+require 'hpricot'
+
+class Remark
+  def initialize(source)
+    @doc = Hpricot(source)
+  end
+  
+  def to_markdown
+    remark_children(@doc).join("\n\n")
+  end
+  
+  IGNORE = %w(script head style)
+  
+  private
+  
+  def valid_attributes?(elem)
+    case elem.name
+    when 'a'
+      (elem.attributes.keys - %w(title)) == %w(href)
+    when 'img'
+      (elem.attributes.keys - %w(title)).sort == %w(alt src)
+    else
+      elem.attributes.empty?
+    end
+  end
+  
+  def remark_children(node)
+    remarked = []
+    node.children.each do |item|
+      result = remark_item(item)
+      remarked << result if result
+    end
+    remarked
+  end
+  
+  def remark_item(item)
+    if item.text?
+      item.to_s.gsub(/\n+/, ' ') unless item.to_s =~ /^\s*$/
+    elsif item.elem?
+      if IGNORE.include?(item.name)
+        nil
+      elsif valid_attributes?(item)
+        remark_element(item)
+      else
+        item
+      end
+    end
+  end
+  
+  def remark_element(elem)
+    case elem.name
+    when 'p'
+      remark_inline(elem)
+    when /^h([1-6])$/
+      ('#' * $1.to_i) + ' ' + remark_inline(elem)
+    when 'ul', 'ol'
+      remark_list(elem)
+    when 'li'
+      remark_inline(elem)
+    when 'pre'
+      elem.inner_text.gsub(/^/, ' '*4)
+    when 'em'
+      "_#{elem.inner_text}_"
+    when 'strong'
+      "**#{elem.inner_text}**"
+    when 'code'
+      "`#{elem.inner_text}`"
+    when 'a'
+      remark_link(elem.inner_html, elem.attributes['href'], elem.attributes['title'])
+    when 'img'
+      '!' + remark_link(elem.attributes['alt'], elem.attributes['src'], elem.attributes['title'])
+    when 'blockquote'
+      remark_children(elem).join("\n\n").gsub(/^/, '> ')
+    else
+      elem
+    end
+  end
+  
+  def remark_link(text, href, title = nil)
+    title_markup = title ? %( "#{title}") : ''
+    "[#{text}](#{href}#{title_markup})"
+  end
+  
+  def remark_inline(elem)
+    remark_children(elem).join('')
+  end
+  
+  def remark_list(list)
+    unordered = list.name == 'ul'
+    marker = unordered ? '*' : 0
+    remark_children(list).map do |item|
+      if unordered
+        marker + ' ' + item
+      else
+        (marker += 1).to_s + '. ' + item
+      end
+    end.join("\n")
+  end
+end

data/spec/remark_spec.rb

@@ -0,0 +1,96 @@
+require 'remark'
+
+describe Remark do
+  def remark(source)
+    described_class.new(source).to_markdown
+  end
+  
+  it "should let through text content" do
+    remark("Foo bar").should == 'Foo bar'
+    remark("Foo bar\nbaz").should == 'Foo bar baz'
+  end
+  
+  it "should split paragraphs with an empty line" do
+    remark("<p>Foo bar</p>").should == 'Foo bar'
+    remark("<p>Foo bar</p><p>baz").should == "Foo bar\n\nbaz"
+    remark("<p>Foo bar</p>baz").should == "Foo bar\n\nbaz"
+  end
+  
+  it "should output title syntax" do
+    remark("<h1>Foo bar</h1>").should == '# Foo bar'
+    remark("<h2>Foo bar</h2>").should == '## Foo bar'
+  end
+  
+  it "should preserve elements in remarked blocks" do
+    remark("<p>Foo <ins>bar</ins></p>").should == 'Foo <ins>bar</ins>'
+    remark("<h2>Foo <ins>bar</ins></h2>").should == '## Foo <ins>bar</ins>'
+  end
+  
+  it "should unescape HTML entities" do
+    remark("Foo&amp;bar").should == 'Foo&bar'
+    remark("<p>If you&#8217;re doing all your development on the &#8220;master&#8221; branch, you&#8217;re not using git").should == "If you’re doing all your development on the “master” branch, you’re not using git"
+  end
+  
+  it "should ignore tags without user-facing content" do
+    remark("<script>foo</script>").should == ''
+    remark("<head>foo</head>").should == ''
+  end
+  
+  it "should leave known elements with attributes intact" do
+    remark("<p class='notice'>Kittens attack!</p>").should == '<p class="notice">Kittens attack!</p>'
+  end
+  
+  it "should leave unknown elements intact" do
+    remark(<<-HTML).should == "Foo\n\n<table>data</table>\n\nBar"
+      <p>Foo</p>
+      <table>data</table>
+      <p>Bar</p>
+    HTML
+  end
+  
+  it "should support lists" do
+    remark(<<-HTML).should == "* foo\n* bar"
+      <ul>
+        <li>foo</li>
+        <li>bar</li>
+      </ul>
+    HTML
+    
+    remark(<<-HTML).should == "1. foo\n2. bar"
+      <ol>
+        <li>foo</li>
+        <li>bar</li>
+      </ol>
+    HTML
+  end
+  
+  it "should support preformatted blocks" do
+    remark("<pre>def foo\n  bar\nend</pre>").should == "    def foo\n      bar\n    end"
+    remark("<pre><code>def foo\n  &lt;bar&gt;\nend</code></pre>").should == "    def foo\n      <bar>\n    end"
+  end
+  
+  it "should remark inline elements" do
+    remark("<p>I'm so <strong>strong</strong></p>").should == "I'm so **strong**"
+    remark("<p>I'm so <em>emo</em></p>").should == "I'm so _emo_"
+    remark("<p>Write more <code>code</code></p>").should == "Write more `code`"
+    remark("<ul><li><em>Inline</em> stuff in <strong>lists</strong></li></ul>").should == "* _Inline_ stuff in **lists**"
+    remark("<h1>Headings <em>too</em></h1>").should == '# Headings _too_'
+  end
+  
+  it "should support hyperlinks" do
+    remark("<p>Click <a href='http://mislav.uniqpath.com'>here</a></p>").should ==
+      "Click [here](http://mislav.uniqpath.com)"
+    remark("<a href='/foo' title='bar'>baz</a>").should == '[baz](/foo "bar")'
+  end
+  
+  it "should support blockquotes" do
+    remark("<blockquote>Cogito, ergo sum</blockquote>").should == '> Cogito, ergo sum'
+    remark("<blockquote><p>I think</p><p>therefore I am</p></blockquote>").should == "> I think\n> \n> therefore I am"
+  end
+  
+  it "should support image tags" do
+    remark("<img src='moo.jpg' alt='cow'>").should == '![cow](moo.jpg)'
+    remark("<img src='moo.jpg' alt='cow' width='16'>").should == '<img src="moo.jpg" alt="cow" width="16" />'
+  end
+end
+

data/spec/sample.html

@@ -0,0 +1,40 @@
+<h1>Remark &mdash; HTML to Markdown converter</h1>
+
+<p>This is a sample document which will get updated as Remark understands more HTML.
+It reflects what's currently supported.</p>
+
+<p class="nice">Known block elements are left intact if they have attributes.
+Markdown doesn't have a syntax for them.</p>
+
+<table>
+  <tr>
+    <td>Elements that can't be represented in Markdown are left intact.</td>
+  </tr>
+</table>
+
+<p>SCRIPT and HEAD tags are swallowed, as browsers don't render them as content.</p>
+
+<script type="text/javascript">
+  alert("I will not survive")
+</script>
+
+<p>Remark supports Markdown syntax for <em>inline</em> markup.
+<a href="http://github.com/mislav">Hyperlinks</a> and <code>code spans</code> are a must.</p>
+
+<ul>
+  <li>List items too;</li>
+  <li>ordered or unordered.</li>
+</ul>
+
+<pre><code>And who would forget
+Preformatted code blocks :)</code></pre>
+
+<h2>TODO</h2>
+
+<p>Remark should probably support BR elements in paragraphs,<br>
+although people tend to abuse them.</p>
+
+<div class="content">
+  <p>What to do with pieces of content inside wrapper elements,
+  like DIV, is still undecided.</p>
+</div>

metadata

@@ -0,0 +1,58 @@
+--- !ruby/object:Gem::Specification 
+name: mislav-remark
+version: !ruby/object:Gem::Version 
+  version: 0.1.0
+platform: ruby
+authors: 
+- "Mislav Marohni\xC4\x87"
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-06-23 00:00:00 -07:00
+default_executable: remark
+dependencies: []
+
+description: 
+email: mislav.marohnic@gmail.com
+executables: 
+- remark
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- Rakefile
+- bin/remark
+- lib/remark.rb
+- spec/remark_spec.rb
+- spec/sample.html
+- README.markdown
+has_rdoc: false
+homepage: http://github.com/mislav/remark
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.2.0
+signing_key: 
+specification_version: 3
+summary: HTML to Markdown converter
+test_files: []
+