mislav-remark 0.2.1 → 0.3.0

data/Rakefile

@@ -1,14 +1,20 @@
+desc "renders the spec/sample.html to Markdown"
+task :sample do
+  system %(ruby -Ilib -rubygems bin/remark spec/sample.html)
+end
+
 desc "generates .gemspec file"
 task :gemspec do
   spec = Gem::Specification.new do |gem|
     gem.name = "remark"
+    gem.version = '0.3.0'
+    
     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 = '0.2.1'
     gem.files = FileList['Rakefile', '{bin,lib,rails,spec}/**/*', 'README*', 'LICENSE*'] & `git ls-files`.split("\n")
     gem.executables = Dir['bin/*'].map { |f| File.basename(f) }
   end

data/bin/remark

@@ -1,4 +1,22 @@
 #!/usr/bin/env ruby
+require 'optparse'
 require 'remark'
 
-puts Remark.new(ARGF.read).to_markdown
+options = {}
+OptionParser.new do |opts|
+  opts.banner = "Usage: remark [options] [FILE]"
+
+  opts.on("-n", "--inline-links", "Render link URLs inline (instead of reference-style)") do |inline|
+    options[:reference_links] = !inline
+  end
+  
+  opts.on("-s", "--scope EXPR", "Scope to a spefic CSS/XPath expression in the HTML document") do |scope|
+    options[:scope] = scope
+  end
+  
+  opts.on("-i", "--ignore EXPR", "Ignore elements that match CSS/XPath expression") do |expr|
+    (options[:ignores] ||= []) << expr
+  end
+end.parse!
+
+puts Remark.new(ARGF.read, options).to_markdown

data/lib/remark.rb

@@ -1,16 +1,27 @@
-require 'hpricot'
+require 'remark/hpricot_ext'
 
 class Remark
-  def initialize(source)
+  DEFAULT_OPTIONS = { :reference_links => true }
+  
+  def initialize(source, options = {})
     @doc = Hpricot(source)
+    @options = DEFAULT_OPTIONS.merge options
+    @links = []
+    @ignored_elements = nil
   end
   
   def to_markdown
-    remark_block(scope)
+    parent = scope
+    collect_ignored_elements(parent)
+    links = @options[:links] = [] unless inline_links?
+    result = parent.to_markdown(@options)
+    result + (inline_links? || links.empty?? '' : "\n\n\n" + output_reference_links(links))
   end
   
   def scope
-    if body = @doc.at('/html/body')
+    if scope = @options[:scope]
+      @doc.at(scope)
+    elsif body = @doc.at('/html/body')
       candidates = (body / 'p').inject(Hash.new(0)) do |memo, para|
         memo[para.parent] += 1
         memo
@@ -22,135 +33,29 @@ class Remark
     end
   end
   
-  IGNORE = %w(script head style)
-  BLOCK = %w(p blockquote h1 h2 h3 h4 h5 h6 pre)
-  
-  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_block(elem)
-    remark_children(elem).
-      reject { |item| item.blank? }.
-      join("\n\n")
+  def inline_links?
+    !@options[:reference_links]
   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.blank?
-    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'
-      elem.children.any? { |e| e.elem? and BLOCK.include?(e.name) } ?
-        remark_block(elem).indent : remark_inline(elem)
-    when 'pre'
-      elem.inner_text.indent
-    when 'em'
-      "_#{elem.inner_text}_"
-    when 'strong'
-      "**#{elem.inner_text}**"
-    when 'code'
-      code = elem.inner_text
-      code.index('`') ? "`` #{code} ``" : "`#{code}`"
-    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").indent('> ')
-    when 'br'
-      "  \n" + elem.inner_html
-    else
-      elem
+  def output_reference_links(links)
+    references = []
+    links.each_with_index do |(href, title), i|
+      references << "[#{i + 1}]: #{href}#{title ? '  ' + title.inspect : ''}"
     end
+    references.join("\n")
   end
   
-  def remark_link(text, href, title = nil)
-    title_markup = title ? %( "#{title}") : ''
-    "[#{text}](#{href}#{title_markup})"
-  end
+  private
   
-  def remark_inline(elem)
-    remark_children(elem).join('').strip.gsub(/ {2,}(?!\n)/, ' ').gsub(/(\n) +/, '\1')
+  def ignore_element?(elem)
+    IGNORE.include?(elem.name) or (@ignored_elements and @ignored_elements.include?(elem))
   end
   
-  def remark_list(list)
-    unordered = list.name == 'ul'
-    marker = unordered ? '*' : 0
-    nested = false
-    
-    items = remark_children(list).map do |item|
-      current = unordered ? marker : "#{marker += 1}."
-      if item =~ /\A\s/
-        nested = true
-        item[0, current.length] = current
-        item
-      else
-        current + ' ' + item
-      end
+  def collect_ignored_elements(scope)
+    if @options[:ignores]
+      @options[:ignored_elements] = @options[:ignores].map do |expr|
+        scope.search(expr).to_a
+      end.flatten.uniq
     end
-    
-    items.join("\n" * (nested ? 2 : 1))
   end
 end
-
-Object.class_eval do
-  def blank?() false end
-end
-
-NilClass.class_eval do
-  def blank?() true end
-end
-
-String.class_eval do
-  def blank?
-    self.empty? or !!(self =~ /\A\s+\Z/)
-  end
-  
-  def squish
-    self.strip.gsub!(/\s+/, ' ')
-  end
-  
-  def indent(with = ' ' * 4)
-    self.gsub(/^/, with)
-  end
-end
-
-Hpricot::Text.class_eval do
-  def blank?() to_s.blank? end
-end

data/lib/remark/core_ext.rb

@@ -0,0 +1,25 @@
+Object.class_eval do
+  def blank?() false end
+end
+
+NilClass.class_eval do
+  def blank?() true end
+end
+
+String.class_eval do
+  def blank?
+    self.empty? or !!(self =~ /\A\s+\Z/)
+  end
+  
+  def squeeze_whitespace
+    self.tr("\n\t", ' ').squeeze(' ')
+  end
+  
+  def indent(with = ' ' * 4)
+    self.gsub(/^/, with)
+  end
+end
+
+Hpricot::Text.module_eval do
+  def blank?() to_s.blank? end
+end

data/lib/remark/hpricot_ext.rb

@@ -0,0 +1,196 @@
+require 'hpricot'
+require 'remark/core_ext'
+
+# this applies the default behavior to virtually all Hpricot classes
+Hpricot::Node.module_eval do
+  def to_markdown(options = {}) nil end
+  def markdown_block?() false end
+end
+
+# nothing special to process on Text or CData
+Hpricot::Text.module_eval do
+  def to_markdown(options = {}) to_s.squeeze_whitespace end
+end
+
+Hpricot::CData.module_eval do
+  def to_markdown(options = {}) to_s.squeeze_whitespace end
+end
+
+# elements that have children
+Hpricot::Container.module_eval do
+  def to_markdown(options = {})
+    return '' unless self.children
+    previous_was_block = false
+    parent_is_block = self.markdown_block?
+    
+    # recurse over this element's children
+    content = self.children.inject([]) do |all, child|
+      current_is_block = child.markdown_block?
+      child_content = child.to_markdown(options)
+      
+      # skip this node if its markdown is nil, empty or, in case
+      # that the previous element was a block, all-whitespace
+      unless child_content.nil? or child_content.empty? or (previous_was_block and child_content.blank?)
+        # handle separating of adjacent markdown blocks with an empty line
+        if not all.empty? and current_is_block or previous_was_block
+          # strip trailing whitespace if we're opening a new block
+          all.last.blank?? all.pop : all.last.rstrip!
+          # guard against adding a newline at the beginning
+          all << "\n\n" if all.any?
+        end
+        
+        unless 'pre' == child.name
+          # strip whitespace from the left if ...
+          child_content.lstrip! if previous_was_block or # we're adjacent to a block
+            (parent_is_block and child == self.children.first) or # this is the first child
+            (not all.empty? and all.last =~ / ( \n)?$/) # we're following a space or a forced line break token
+          
+          
+          # strip whitespace from the right if this is the last node in a block
+          child_content.rstrip! if parent_is_block and self.children.last == child
+        end
+        
+        all << child_content
+      end
+      
+      previous_was_block = current_is_block
+      all
+    end
+    
+    result = content.join('')
+    return result
+  end
+end
+
+# elements without children
+Hpricot::Leaf.module_eval do
+  def to_markdown(options = {})
+    inner_text.squeeze_whitespace if elem?
+  end
+end
+
+Hpricot::Elem.module_eval do
+  IGNORE = %w(script head style)
+  ALLOWED_EMPTY = %w(img br hr )
+  MARKDOWN_BLOCK = %w(p blockquote h1 h2 h3 h4 h5 h6 pre hr)
+  MARKDOWN_INLINE = %w(em strong code a img br)
+  MARKDOWN_RECOGNIZED = MARKDOWN_BLOCK + MARKDOWN_INLINE + %w(div)
+  HTML_BLOCK = MARKDOWN_BLOCK + %w(ul ol dl div noscript form table address fieldset)
+  
+  def to_markdown(options = {})
+    return nil if markdown_ignored?(options)
+    return '' if markdown_empty?
+    return to_s unless markdown_supported_attributes?
+
+    case name
+    when 'div', 'noscript'
+      super
+    when 'p'
+      super
+    when /^h([1-6])$/
+      ('#' * $1.to_i) + ' ' + super
+    when 'ul', 'ol'
+      remark_list(options)
+    when 'li'
+      content = super
+      content = content.indent if children.any? { |e| e.markdown_block? }
+      content
+    when 'pre'
+      inner_text.rstrip.indent
+    when 'em'
+      "_#{super}_"
+    when 'strong'
+      "**#{super}**"
+    when 'code'
+      code = inner_text
+      code.index('`') ? "`` #{code} ``" : "`#{code}`"
+    when 'a'
+      remark_link(super, attributes['href'], attributes['title'], options)
+    when 'img'
+      '!' + remark_link(attributes['alt'], attributes['src'], attributes['title'], :reference_links => false)
+    when 'blockquote'
+      super.indent('> ')
+    when 'br'
+      "  \n" + inner_html
+    else
+      to_s
+    end
+  end
+  
+  def remark_list(options = {})
+    unordered = self.name == 'ul'
+    marker = unordered ? '*' : 0
+    nested = false
+    
+    items = self.children_of_type('li').map do |item|
+      item = item.to_markdown(options)
+      current = unordered ? marker : "#{marker += 1}."
+      if item =~ /\A\s/
+        nested = true
+        item[0, current.length] = current
+        item
+      else
+        current + ' ' + item
+      end
+    end
+    
+    items.join("\n" * (nested ? 2 : 1))
+  end
+  
+  def markdown_block?
+    HTML_BLOCK.include?(name)
+  end
+  
+  def markdown_recognized?
+    MARKDOWN_RECOGNIZED.include?(name)
+  end
+  
+  protected
+  
+  def markdown_ignored?(options)
+    IGNORE.include?(name) or
+      (options[:ignored_elements] and options[:ignored_elements].include?(self))
+  end
+  
+  def markdown_empty?
+    empty? and markdown_recognized? and not ALLOWED_EMPTY.include?(name)
+  end
+  
+  def markdown_supported_attributes?
+    case name
+    when 'div'
+      true
+    when 'a'
+      attribute_names_match?('href', 'title')
+    when 'img'
+      attribute_names_match?(%w(alt src), 'title')
+    when 'ol', 'ul'
+      attributes.empty? and children.all? do |item|
+        not item.elem? or (item.name == 'li' and item.attributes.empty?)
+      end
+    else
+      attributes.empty?
+    end
+  end
+  
+  def attribute_names_match?(only, optional = nil)
+    names = attributes.keys.sort
+    names -= Array(optional) if optional
+    names == Array(only)
+  end
+  
+  def remark_link(text, href, title = nil, options = {})
+    if options[:reference_links]
+      if existing = options[:links].find { |h, t| href == h }
+        num = options[:links].index(existing) + 1
+      else
+        options[:links] << [href, title]
+        num = options[:links].length
+      end
+      "[#{text}][#{num}]"
+    else
+      title_markup = title ? %( "#{title}") : ''
+      "[#{text}](#{href}#{title_markup})"
+    end
+  end
+end

data/spec/hpricot_ext_spec.rb

@@ -0,0 +1,138 @@
+require 'remark/hpricot_ext'
+
+describe Hpricot, "remark extensions" do
+  before(:all) do
+    @doc = Hpricot(<<-HTML.strip)
+      <?xml version="moo" ?>
+      <!DOCTYPE html>
+      <html>
+        <head>
+          <title>Sample document</title>
+        </head>
+        <body>
+          <h1>Sample <strong>Remark</strong> document</h1>
+          <p>
+            A paragraph with <em>nested</em> <strong>content</strong>
+            and <i>Remark</i>-supported elements.
+          </p>
+          
+          <a name="content"> </a>
+          <h2>The content</h2>
+          <div id="content">
+            <p>First</p>
+            <p>Second</p>
+            Some content
+            <em>in-between</em>
+            <p>Third</p>
+          </div>
+          <p class="foo">I has classname</p>
+          
+          <div id="empty"></div>
+          <blockquote>
+            Some famous quote
+            <blockquote>Nested famous quote</blockquote>
+          </blockquote>
+          <div class="code">
+            <p>Sample code:</p>
+            <pre>def preformatted
+  text
+end
+            </pre>
+          </div>
+          <img src='moo.jpg' alt='cow'>
+          <img src='moo.jpg' alt='cow' width='16'>
+          
+          <code>simple</code> <code>comp ` lex</code> <code>&lt;tag&gt;</code>
+          
+          <div id="br">
+            <p>Foo<br>bar</p>
+            <p>Foo<br>
+            bar <code>baz</code></p>
+            <p>Foo</p><br><br><p>Bar</p><br>
+          </div>
+          
+          <ul>
+            <li>First</li>
+            <li>Second</li>
+          </ul>
+          <ol>
+            <li>First</li>
+            <li>Second</li>
+          </ol>
+        </body>
+      </html>
+    HTML
+  end
+  
+  def remark(elem, options = {})
+    (String === elem ? @doc.at(elem) : elem).to_markdown(options)
+  end
+  
+  it "should return empty string for empty document" do
+    remark(Hpricot('')).should == ''
+  end
+  
+  it "should ignore DOCTYPE, HEAD and XML processing instructions" do
+    remark('head').should be_nil
+    remark(@doc.children[0]).should be_nil # doctype
+    remark(@doc.children[2]).should be_nil # xmldecl
+  end
+  
+  it "should have whitespace nodes respond to blank" do
+    @doc.at('a[@name]').children.first.blank?
+  end
+  
+  it "should support headings" do
+    remark('h1').should == "# Sample **Remark** document"
+    remark('h2').should == "## The content"
+  end
+  
+  it "should support paragraphs" do
+    remark('p').should == "A paragraph with _nested_ **content** and <i>Remark</i>-supported elements."
+  end
+  
+  it "should split paragraphs with an empty line" do
+    remark('#content').should == "First\n\nSecond\n\nSome content _in-between_\n\nThird"
+  end
+  
+  it "should keep full HTML for paragraphs if they have attributes" do
+    remark('p.foo').should == '<p class="foo">I has classname</p>'
+  end
+  
+  it "should not break on empty DIV" do
+    remark('#empty').should == ""
+  end
+  
+  it "should support blockquotes" do
+    remark('blockquote > blockquote').should == "> Nested famous quote"
+    remark('blockquote').should == "> Some famous quote\n> \n> > Nested famous quote"
+  end
+  
+  it "should support preformatted text" do
+    remark('div.code').should == "Sample code:\n\n    def preformatted\n      text\n    end"
+  end
+  
+  it "should support image tags" do
+    remark('img[@alt]').should == '![cow](moo.jpg)'
+    remark('img[@width]').should == '<img src="moo.jpg" alt="cow" width="16" />'
+  end
+  
+  it "should support code spans" do
+    remark('code').should == "`simple`"
+    remark('code ~ code').should == "`` comp ` lex ``"
+    remark('code ~ code ~ code').should == "`<tag>`"
+  end
+  
+  it "should support BR" do
+    remark('#br').should == "Foo  \nbar\n\nFoo  \nbar `baz`\n\nFoo\n\nBar"
+  end
+  
+  it "should support unordered list" do
+    remark('ul').should == "* First\n* Second"
+  end
+  
+  it "should support ordered list" do
+    remark('ol').should == "1. First\n2. Second"
+  end
+end
+

data/spec/remark_spec.rb

@@ -1,8 +1,9 @@
 require 'remark'
 
 describe Remark do
-  def remark(source)
-    described_class.new(source).to_markdown
+  def remark(source, options = {})
+    options = {:reference_links => false}.merge(options)
+    described_class.new(source, options).to_markdown
   end
   
   it "should let through text content" do
@@ -10,17 +11,6 @@ describe Remark do
     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>'
@@ -31,15 +21,6 @@ describe Remark do
     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>
@@ -48,109 +29,129 @@ describe Remark do
     HTML
   end
   
-  it "should strip excess whitespace" do
-    remark(<<-HTML).should == "Foo bar"
-      <p>
-        Foo
-        bar
-      </p>
-    HTML
-  end
+  describe "whitespace" do
+    it "should strip excess whitespace" do
+      remark(<<-HTML).should == "Foo bar"
+        <p>
+          Foo
+          bar
+        </p>
+      HTML
+    end
   
-  it "should strip whitespace in text nodes between main content" do
-    pending
-    remark(<<-HTML).should == "Foo\n\nbar\n\nBaz"
-      <p>Foo</p>
+    it "should strip whitespace in text nodes between processed nodes" do
+      remark(<<-HTML).should == "Foo\n\nbar\n\nBaz"
+        <p>Foo</p>
         
-           bar
-      <p>Baz</p>
-    HTML
-  end
-  
-  it "should support lists" do
-    remark(<<-HTML).should == "* foo\n* bar"
-      <ul>
-        <li>foo</li>
-        <li>bar</li>
-      </ul>
-    HTML
+             bar
+        <p>Baz</p>
+      HTML
+    end
+  end
+  
+  describe "lists" do
+    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 lists with nested content" do
-    remark(<<-HTML).should == "*   foo\n    \n    bar\n\n*   baz"
-      <ul>
-        <li><p>foo</p><p>bar</p></li>
-        <li><p>baz</p></li>
-      </ul>
-    HTML
+      remark(<<-HTML).should == "1. foo\n2. bar"
+        <ol>
+          <li>foo</li>
+          <li>bar</li>
+        </ol>
+      HTML
+    end
+  
+    it "should support lists with nested content" do
+      remark(<<-HTML).should == "*   foo\n    \n    bar\n\n*   baz"
+        <ul>
+          <li><p>foo</p><p>bar</p></li>
+          <li><p>baz</p></li>
+        </ul>
+      HTML
+    end
+  
+    it "should output malformed lists as HTML" do
+      remark(<<-HTML).should == "<ul>\n          <span>bar</span>\n        </ul>"
+        <ul>
+          <span>bar</span>
+        </ul>
+      HTML
+    end
   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("<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 remark inline code" do
-    remark("<p>Write more <code>code</code></p>").should == "Write more `code`"
-    remark("<p>Even with <code>`backticks`</code></p>").should == "Even with `` `backticks` ``"
-    remark("<p>Or HTML <code>&lt;tags&gt;</code></p>").should == "Or HTML `<tags>`"
-  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
-  
-  it "should not have BR ruin all the fun" do
-    remark("<p>Foo<br>bar</p>").should == "Foo  \nbar"
-    remark("<p>Foo<br>\nbar <code>baz</code></p>").should == "Foo  \nbar `baz`"
-    remark("<p>Foo</p><br><p>Bar</p>").should == "Foo\n\nBar"
-  end
-  
-  it "should scope to the most likely element that holds content" do
-    remark(<<-HTML).should == "Wow, 3 paragraphs\n\nThis must be where the content is\n\nI'm sure"
-      <html>
-        <body>
-          <div id="div1">
-            <p>Only 1 paragraph</p>
-          </div>
-          <div id="div3">
-            <p>Wow, 3 paragraphs</p>
-            <p>This must be where the content is</p>
-            <p>I'm sure</p>
-          </div>
-          <div id="div2">
-            <p>Only 2 paragraphs</p>
-            <p>How disappointing</p>
-          </div>
-        </body>
-      </html>
-    HTML
+    remark("<pre>def foo\n</pre>").should == "    def foo"
+  end
+  
+  describe "inline" do
+    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("<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 handle nested inline elements" do
+      remark("<p>I <strong>love <code>code</code></strong></p>").should == "I **love `code`**"
+      remark("<p>I <a href='#'>am <em>fine</em></a></p>").should == "I [am _fine_](#)"
+    end
+  end
+  
+  describe "hyperlinks" do
+    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 have reference-style hyperlinks" do
+      remark("<p>Click <a href='foo' title='mooslav'>here</a> and <a href='bar'>there</a></p>", :reference_links => true).should ==
+        "Click [here][1] and [there][2]\n\n\n[1]: foo  \"mooslav\"\n[2]: bar"
+      remark("<p>Click <a href='foo'>here</a> and <a href='foo'>there</a></p>", :reference_links => true).should ==
+        "Click [here][1] and [there][1]\n\n\n[1]: foo"
+      remark("", :reference_links => true).should == ""
+    end
+  end
+  
+  it "should support ignores" do
+    remark("<p>Foo <span>bar</span> baz</p>", :ignores => ['span']).should == "Foo baz"
+  end
+  
+  describe "scoping" do
+    before do
+      @html = <<-HTML
+        <html>
+          <body>
+            <div id="div1">
+              <p>Only 1 paragraph</p>
+            </div>
+            <div id="div3">
+              <p>Wow, 3 paragraphs</p>
+              <p>This must be where the content is</p>
+              <p>I'm sure</p>
+            </div>
+            <div id="div2">
+              <p>Only 2 paragraphs</p>
+              <p>How disappointing</p>
+            </div>
+          </body>
+        </html>
+      HTML
+    end
+    
+    it "should scope to the most likely element that holds content" do
+      remark(@html).should == "Wow, 3 paragraphs\n\nThis must be where the content is\n\nI'm sure"
+    end
+    
+    it "should scope to the explicit scope" do
+      remark(@html, :scope => '#div2').should == "Only 2 paragraphs\n\nHow disappointing"
+    end
   end
 end
-

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: mislav-remark
 version: !ruby/object:Gem::Version 
-  version: 0.2.1
+  version: 0.3.0
 platform: ruby
 authors: 
 - "Mislav Marohni\xC4\x87"
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-06-24 00:00:00 -07:00
+date: 2009-07-30 00:00:00 -07:00
 default_executable: remark
 dependencies: []
 
@@ -24,12 +24,16 @@ extra_rdoc_files: []
 files: 
 - Rakefile
 - bin/remark
+- lib/remark/core_ext.rb
+- lib/remark/hpricot_ext.rb
 - lib/remark.rb
+- spec/hpricot_ext_spec.rb
 - spec/remark_spec.rb
 - spec/sample.html
 - README.markdown
 has_rdoc: false
 homepage: http://github.com/mislav/remark
+licenses: 
 post_install_message: 
 rdoc_options: []
 
@@ -50,7 +54,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.2.0
+rubygems_version: 1.3.5
 signing_key: 
 specification_version: 3
 summary: HTML to Markdown converter