reverse-markdown 0.1.1 → 0.1.2

data/lib/reverse-markdown.rb

@@ -1,15 +1,22 @@
-#!/usr/bin/env ruby
-
-# file:  reverse-markdown.rb
-
-# it is forked from https://github.com/xijo/reverse-markdown/blob/master/reverse_markdown.rb
 require 'rexml/document'
 require 'benchmark'
 include REXML
 include Benchmark
 
+# reverse markdown for ruby
+# author: JO
+# e-mail: xijo@gmx.de
+# date: 14.7.2009
+# version: 0.1
+# license: GPL
+# taken from https://github.com/xijo/reverse-markdown/raw/master/reverse_markdown.rb
+
+# TODO
+# - ol numbering is buggy, in fact doesn't matter for markdown code
+# -
+
 class ReverseMarkdown
-  
+
   # set basic variables:
   # - @li_counter: numbering list item (li) tags in an ordered list (ol)
   # - @links:      hold the links for adding them to the bottom of the @output
@@ -24,20 +31,17 @@ class ReverseMarkdown
     @indent = 0
     @errors = []
   end
-  
+
   # Invokes the HTML parsing by using a string. Returns the markdown code in @output.
   # To garantuee well-formed xml for REXML a <root> element will be added, but has no effect.
   # After parsing all elements, the 'reference style'-links will be inserted.
   def parse_string(string)
     doc = Document.new("<root>\n"+string+"\n</root>")
-    root = doc.root
-    root.elements.each do |element|
-      parse_element(element, :root)
-    end
+    parse_element(doc.root, :none)
     insert_links()
     @output
   end
-  
+
   # Parsing an element and its children (recursive) and writing its markdown code to @output
   # 1. do indent for nested list items
   # 2. add the markdown opening tag for this element
@@ -50,33 +54,33 @@ class ReverseMarkdown
     @output << indent() if name.eql?(:li)
     # 2.
     @output << opening(element, parent)
-    
+
     # 3a.
     if (element.has_text? and element.children.size < 2)
       @output << text_node(element, parent)
     end
-    
+
     # 3b.
     if element.has_elements?
       element.children.each do |child|
         # increase indent if nested list
         @indent += 1 if element.name=~/(ul|ol)/ and parent.eql?(:li)
-        
+
         if child.node_type.eql?(:element)
           parse_element(child, element.name.to_sym)
         else
           if parent.eql?(:blockquote)
-            @output << child.to_s.gsub("\n", "\n>")
+            @output << child.to_s.gsub("\n ", "\n>")
           else
             @output << child.to_s
           end
         end
-        
+
         # decrease indent if end of nested list
         @indent -= 1 if element.name=~/(ul|ol)/ and parent.eql?(:li)
       end
     end
-    
+
     # 4.
     @output << ending(element, parent)
   end
@@ -95,16 +99,22 @@ class ReverseMarkdown
         ""
       when :h2
         "## "
+      when :h3
+        "### "
+      when :h4
+        "#### "
+      when :h5
+        "##### "
+      when :h6
+        "###### "
       when :em
         "*"
-      when :br
-        "\n"
-        "\n>" if parent.eql?(:blockquote)
       when :strong
         "**"
       when :blockquote
-        type.delete(type.elements[1]) if type.first.name.to_sym.eql?(:br) 
-        ">"
+        # remove leading newline
+        type.children.first.value = ""
+        "> "
       when :code
         parent.eql?(:pre) ? "    " : "`"
       when :a
@@ -113,12 +123,14 @@ class ReverseMarkdown
         "!["
       when :hr
         "----------\n\n"
+      when :root
+        ""
       else
         @errors << "unknown start tag: "+type.name.to_s
         ""
     end
   end
-  
+
   # Returns the closing markdown tag, like opening()
   def ending(type, parent)
     case type.name.to_sym
@@ -126,6 +138,14 @@ class ReverseMarkdown
         " #\n\n"
       when :h2
         " ##\n\n"
+      when :h3
+        " ###\n\n"
+      when :h4
+        " ####\n\n"
+      when :h5
+        " #####\n\n"
+      when :h6
+        " ######\n\n"
       when :p
         parent.eql?(:root) ? "\n\n" : "\n"
       when :ol
@@ -149,12 +169,14 @@ class ReverseMarkdown
         @links << type.attribute('src').to_s
         "" + type.attribute('alt').to_s + "][" + @links.size.to_s + "] "
         "#{type.attribute('alt')}][#{@links.size}] "
+      when :root
+        ""
       else
         @errors << "  unknown end tag: "+type.name.to_s
         ""
     end
   end
-  
+
   # Perform indent: two space, @indent times - quite simple! :)
   def indent
     str = ""
@@ -163,7 +185,7 @@ class ReverseMarkdown
     end
     str
   end
-  
+
   # Return the content of element, which should be just text.
   # If its a code block to indent of 4 spaces.
   # For block quotation add a leading '>'
@@ -171,12 +193,12 @@ class ReverseMarkdown
     if element.name.to_sym.eql?(:code) and parent.eql?(:pre)
       element.text.gsub("\n","\n    ") << "\n"
     elsif parent.eql?(:blockquote)
-      element.text.gsub("\n","\n>")
+      element.text.gsub!("\n ","\n>")
     else
       element.text
     end
   end
-  
+
   # Insert the mentioned reference style links.
   def insert_links
     @output << "\n"
@@ -184,14 +206,14 @@ class ReverseMarkdown
       @output << "  [#{index+1}]: #{@links[index]}\n"
     end
   end
-  
+
   # Print out all errors, that occured and have been written to @errors.
   def print_errors
     @errors.each do |error|
       puts error
     end
   end
-  
+
   # Perform a benchmark on a given string n-times.
   def speed_benchmark(string, n)
     initialize()
@@ -199,5 +221,77 @@ class ReverseMarkdown
       test.report("reverse markdown:")    { n.times do; parse_string(string); initialize(); end; }
     end
   end
-  
+
 end
+
+if __FILE__ == $0
+
+  # Example HTML Code for parsing
+  example = <<-EOF
+This text, though not within an element, should also be shown.
+
+<h2>heading 1.1</h2>
+
+<p>text *italic* and **bold**.</p>
+
+<pre><code>text *italic* and **bold**.
+sdfsdff
+sdfsd
+sdf sdfsdf
+</code></pre>
+
+<blockquote>
+  <p>text <em>italic</em> and <strong>bold</strong>. sdfsdff
+  sdfsd sdf sdfsdf</p>
+</blockquote>
+
+<p>asdasd <code>sdfsdfsdf</code> asdad <a href="http://www.bla.de">link text</a></p>
+
+<p><a href="http://www.bla.de">link <strong>text</strong></a></p>
+
+<ol>
+<li>List item</li>
+<li>List <em>item</em>
+<ol><li>List item</li>
+<li>dsfdsf
+<ul><li>dfwe</li>
+<li>dsfsdfsdf</li></ul></li>
+<li>lidsf <img src="http://www.dfgdfg.de/dsf.jpe" alt="item" title="" /></li></ol></li>
+<li>sdfsdfsdf
+<ul><li>sdfsdfsdf</li>
+<li>sdfsdfsdf <strong>sdfsdf</strong></li></ul></li>
+</ol>
+
+<blockquote>
+  <p>Lorem ipsum dolor sit amet, consetetur
+  voluptua. At vero eos et accusam et
+  justo duo dolores et ea rebum. Stet
+  clita kasd gubergren, no sea takimata
+  sanctus est Lorem ipsum dolor sit
+  amet. <em>italic</em></p>
+</blockquote>
+
+<hr />
+
+<blockquote>
+  <p>Lorem ipsum dolor sit amet, consetetur
+  sadipscing elitr, sed diam nonumy
+  eirmod tempor invidunt ut labore et
+  dolore magna aliquyam erat, sed</p>
+</blockquote>
+
+This should also be shown, even if it's not wrapped in an element.
+
+<p>nur ein text! nur eine maschine!</p>
+
+This text should not be invisible!
+EOF
+
+  r = ReverseMarkdown.new
+
+  puts r.parse_string(example)
+
+  #r.print_errors
+
+  #r.speed_benchmark(example, 100)
+end

metadata

@@ -2,16 +2,16 @@
 name: reverse-markdown
 version: !ruby/object:Gem::Version 
   prerelease: 
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors: 
 - James Robertson
+- JO
 autorequire: 
 bindir: bin
 cert_chain: []
 
-date: 2011-10-09 00:00:00 +01:00
-default_executable: 
+date: 2012-05-26 00:00:00 Z
 dependencies: []
 
 description: 
@@ -24,7 +24,6 @@ extra_rdoc_files: []
 
 files: 
 - lib/reverse-markdown.rb
-has_rdoc: true
 homepage: 
 licenses: []
 
@@ -48,7 +47,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.5.2
+rubygems_version: 1.8.23
 signing_key: 
 specification_version: 3
 summary: reverse-markdown