Bayeux 0.1.0

data/lib/bayeux/html_gen.rb

@@ -0,0 +1,506 @@
+require 'tidy_ffi'
+require 'uv'
+
+require 'log4r'
+include Log4r
+
+class BayeuxHTMLGen
+  
+  # Default Constructor
+  def initialize(syntax_tree)
+    @syntax_tree = syntax_tree
+    @html_string = String.new
+  end
+  
+  # Return the AST as an HTML string (stored in @html_string)
+  def generate
+    
+    # Set-up the logger
+    @html_gen_log = Logger.new('bayex_html_gen')
+    @html_gen_log.outputters = Outputter.stdout
+    
+    # Create a walker for the trees in the forest
+    walker = TreeWalker.new
+    
+    walker.on_before_down = self.method(:generate_node_start)
+    walker.on_after_up = self.method(:generate_node_end)
+    
+    walker.on_no_siblings = self.method(:generate_full_node)
+    walker.on_no_children = self.method(:generate_full_node)
+        
+    # Clear the internal HTML representation
+    @html_string.clear
+    
+    # Walk the forest
+    @syntax_tree.block_forest.each{|tree|
+      walker.walk_tree(tree)
+    }
+
+    #@html_gen_log.debug {@syntax_tree.to_s}
+    #@html_gen_log.info {@html_string}
+    
+    # Remove extra spaces between entities and tags
+    @html_string.gsub!(/;[\s]+</,';<')
+    
+    # Return the generated string to the caller
+    tidy = TidyFFI::Tidy.new(@html_string)
+  
+    tidy.options.fix_uri = 1
+    tidy.options.indent = 1
+    tidy.options.join_classes = 1
+    tidy.options.markup = 1
+    tidy.options.output_html = 1
+    tidy.options.punctuation_wrap = 1
+    tidy.options.show_body_only = 1  
+        
+    return tidy.clean  
+  end
+  
+  # Return the table of contents, extract from the AST
+  def doc_toc
+    
+    # ToC to return to the caller
+    toc = String.new
+    
+    # The current heading level
+    current_level = 0
+
+    # The last heading level
+    # Note: The very top level is provided
+    # by the enclosing template. Effectivly
+    # we are skipping the enclosures for the L1
+    # headers
+    last_level = 2
+    
+    # Set containing valid header tags (ignore level 1 tags)
+    header_tags = Set.new [:h2, :h3, :h4, :h5, :h6]
+    
+    # Walk the forest
+    @syntax_tree.block_forest.each{|tree|
+      
+      # Ignore if not a header
+      if header_tags.include?(tree.content.type) then
+      
+        # Set the current level based on the header
+        case tree.content.type
+          when :h1
+            current_level = 1
+          when :h2
+            current_level = 2
+          when :h3
+            current_level = 3
+          when :h4
+            current_level = 4
+          when :h5
+            current_level = 5
+          when :h6
+            current_level = 6        
+        end
+        
+        # Write out the toc
+        if current_level < last_level then
+          
+          # Moving up: so close off the last level,
+          # create a new level, and record the current
+          # item
+          
+          unless last_level == 0 then
+            toc << "</ul>"
+          end
+          
+          toc << "#{indent(last_level)}<ul>\n"
+          toc << "#{indent(current_level)}<li><a href=\"##{header_id(tree.content.content)}\">#{tree.content.content}</a></li>\n"
+          
+        elsif current_level > last_level then
+          
+          # Moving down: so add a new list as a sub-list of
+          # the current list, and record the current item
+          
+          toc << "#{indent(last_level)}<ul>\n"
+          toc << "#{indent(current_level)}<li><a href=\"##{header_id(tree.content.content)}\">#{tree.content.content}</a></li>\n"
+          
+        else
+          # This header is at the same level, so just output the
+          # list item
+          toc << "#{indent(current_level)}<li><a href=\"##{header_id(tree.content.content)}\">#{tree.content.content}</a></li>\n"
+        
+        end
+        
+        # Record the current level
+        last_level = current_level
+      end
+    }
+    
+    # Close off the lists
+    current_level.downto(3){|index|
+      toc << "#{indent(index)}</ul>\n"
+    }
+        
+    return toc
+  end
+  
+  ###
+  ### Walker Hooks. These are called by the tree walker on specific
+  ### events
+  ###
+  
+  # Output the full contents of the node, properly bracketed as an
+  # HTML expression. This is only called if we have no sub-nodes to
+  # deal with
+  def generate_full_node(block)
+    
+    # Check if we need to add a space
+    # Check if we need to add a space
+    if not block.content[0] =~ /\s|[:;.,?!]/
+      @html_string << " "
+    end
+                
+    case block.type
+      
+      # Headers
+      when :h1
+        @html_string << "<h1 id=\"#{header_id(block.content)}\">#{block.content}</h1>"
+      when :h2
+        @html_string << "<h2 id=\"#{header_id(block.content)}\">#{block.content}</h2>"
+      when :h3
+        @html_string << "<h3 id=\"#{header_id(block.content)}\">#{block.content}</h3>"
+      when :h4
+        @html_string << "<h4 id=\"#{header_id(block.content)}\">#{block.content}</h4>"
+      when :h5
+        @html_string << "<h5 id=\"#{header_id(block.content)}\">#{block.content}</h5>"
+      when :h5
+        @html_string << "<h6 id=\"#{header_id(block.content)}\">#{block.content}</h6>"
+      
+      # Ordinary paragraphs
+      when :paragraph
+        @html_string << "<p>#{typeset(block.content)}</p>"
+
+      when :none
+        @html_string << "#{typeset(block.content)}"
+        
+      # Special paragraphs
+      when :block_quote
+        @html_string << "<blockquote>#{typeset(block.content)}</blockquote>"  
+      when :single_quote
+        @html_string << "&lsquo;&#8202;#{typeset(block.content)}&#8202;&rsquo;"
+      when :double_quote
+        @html_string << "&ldquo;&#8202;#{typeset(block.content)}&#8202;&rdquo;"
+        
+      when :note
+        @html_string << "<p class=\"note\"><span class=\"note_header\">Note:</span>#{typeset(block.content)}</span></p>"
+        
+      when :command
+        @html_string << "<p class=\"command\">#{block.content}</span>"
+        
+      when :code_language
+        @code_language = block.content
+      when :code_start_number
+        @code_start_number = block.content  
+        
+      when :file
+        @html_string << "<pre class=\"file\">#{block.content}</pre>"
+      when :output
+        @html_string << "<pre class=\"output\">#{block.content}</pre>"
+        
+      # Special Characters
+      when :em_dash
+        @html_string << "&mdash;"
+      when :en_dash
+        @html_string << "&ndash;"
+      when :elipses
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;"
+      when :elipses_stop
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;."
+        
+      # Links
+      when :link_target
+        @link_target = block.content
+      when :link_text
+        @link_text = block.content
+        
+      # Lists
+      when :item
+        @html_string << "<li>#{typeset(block.content)}</li>"
+      
+      when :dl_header
+        @html_string << "<dt>#{block.content}</dt>"
+      when :dl_text
+        @html_string << "<dd>#{block.content}</dd>"
+              
+      # Tags
+      when :ac
+        # Do we know anything about this acronymn?
+        if $reference.acronym_list.include?(block.content) then
+          
+          # Assemble from the reference list        
+          if $reference.acronym_list[block.content].include?('text_html') then
+            acronym = $reference.acronym_list[block.content]['text_html']
+          else
+            acronym = $reference.acronym_list[block.content]['text']
+          end
+          
+          definition = $reference.acronym_list[block.content]['def']
+          
+          @html_string << "<acronym title=\"#{definition}\">#{acronym}</acronym>"
+        else
+          # Do what we can
+          @html_string << "<acronym>#{block.content}</acronym>"
+        end
+            
+      when :emph 
+        @html_string << "<em>#{block.content}</em>"
+      
+      when :tt
+        @html_string << "<tt>#{block.content}</tt>"
+        
+    end
+  end
+  
+  # Output only the start of a node
+  def generate_node_start(block)
+    
+    # Check if we need to add a space
+    if not block.content[0] =~ /\s|[:;.,?!]/
+      @html_string << " "
+    end
+    
+    case block.type
+      
+      # Headers
+      when :h1
+        @html_string << "<h1 id=\"#{header_id(block.content)}\">"
+      when :h2
+        @html_string << "<h2 id=\"#{header_id(block.content)}\">"
+      when :h3
+        @html_string << "<h3 id=\"#{header_id(block.content)}\">"
+      when :h4
+        @html_string << "<h4 id=\"#{header_id(block.content)}\">"
+      when :h5
+        @html_string << "<h5 id=\"#{header_id(block.content)}\">"
+      when :h5
+        @html_string << "<h6 id=\"#{header_id(block.content)}\">"
+      
+      # Ordinary paragraphs
+      when :paragraph
+        @html_string << "<p>#{typeset(block.content)}"
+        
+      when :none
+        @html_string << "#{block.content}"
+        
+      # Special paragraphs
+      when :block_quote
+        @html_string << "<blockquote>#{typeset(block.content)}"
+      when :single_quote
+        @html_string << "&lsquo;&#8202;#{typeset(block.content)}"
+      when :double_quote
+        @html_string << "&ldquo;&#8202;#{typeset(block.content)}"
+        
+      when :note
+        @html_string << "<p class=\"note\"><span class=\"note_header\">Note:</span>#{typeset(block.content)}"
+        
+      when :command
+        @html_string << "<p class=\"command\">#{typeset(block.content)}"
+        
+      when :code_language
+        @code_language = block.content
+      when :code_start_number
+        @code_start_number = block.content
+        
+      # Special Characters
+      when :em_dash
+        @html_string << "&mdash;"
+      when :en_dash
+        @html_string << "&ndash;"
+      when :elipses
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;"
+      when :elipses_stop
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;."
+        
+      # Links
+      when :link_target
+        @link_target = block.content
+      when :link_text
+        @link_text = block.content
+        
+      # Lists
+      when :ol
+        @html_string << "<ol>" 
+      when :ul
+        @html_string << "<ul>" 
+      when :item
+        @html_string << "<li>#{block.content}"
+      
+      when :dl
+        @html_string << "<dl>" 
+      when :dl_header
+        @html_string << "<dt>#{block.content}"
+      when :dl_text
+        @html_string << "<dd>#{block.content}"
+      
+      # Tags        
+      when :ac
+        # Do we know anything about this acronymn?
+        if $reference.acronym_list.include?(block.content) then
+        
+          # Assemble from the reference list        
+          if $reference.acronym_list[block.content].include?('text_html') then
+            acronym = $reference.acronym_list[block.content]['text_html']
+          else
+            acronym = $reference.acronym_list[block.content]['text']
+          end
+        
+          definition = $reference.acronym_list[block.content]['def']
+        
+          @html_string << "<acronym title=\"#{definition}\">#{acronym}"
+        else
+          # Do what we can
+          @html_string << "<acronym>#{block.content}"
+        end
+                
+      when :emph 
+        @html_string << "<em>#{block.content}"
+        
+      when :tt
+        @html_string << "<tt>#{block.content}"
+        
+    end
+  end
+  
+  def generate_node_end(block)
+        
+    case block.type
+      
+      # Headers
+      when :h1
+        @html_string << "</h1>"
+      when :h2
+        @html_string << "</h2>"
+      when :h3
+        @html_string << "</h3>"
+      when :h4
+        @html_string << "</h4>"
+      when :h5
+        @html_string << "</h5>"
+      when :h6
+        @html_string << "</h6>"
+      
+      # Ordinary paragraphs
+      when :paragraph
+        @html_string << "</p>"
+      
+      # Special paragraphs
+      when :block_quote
+        @html_string << "</blockquote>"
+      when :single_quote
+        @html_string << "&#8202;&rsquo;"
+      when :double_quote
+        @html_string << "&#8202;&rdquo;"
+        
+      when :note, :command
+        @html_string << "</p>"
+        
+      when :code
+        # Assemble the code block from the sub-tree nodes
+        begin
+          pretty_code = Uv.parse(block.content, "xhtml", @code_language, @code_start_number.to_i, "dawn")
+          @html_string << pretty_code
+        rescue
+          pretty_code = Uv.parse(block.content, "xhtml", "plain_text", @code_start_number.to_i, "dawn")
+          @html_string << pretty_code          
+        end
+                
+      # Special Characters
+      when :em_dash
+        @html_string << "&mdash;"
+      when :en_dash
+        @html_string << "&ndash;"
+      when :elipses
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;"
+      when :elipses_stop
+        @html_string << "&#8202;.&thinsp;&#8202;.&thinsp;&#8202;.&thinsp;."
+        
+      # Links
+      when :link
+        # Assemble the link from the sub-tree nodes
+        @html_string << "<a href=\"#{@link_target}\">#{@link_text}</a>"
+      when :link_target
+        @link_target = block.content
+      when :link_text
+        @link_text = block.content
+          
+      # Lists
+      when :ol
+        @html_string << "</ol>" 
+      when :ul
+        @html_string << "</ul>" 
+      when :item
+        @html_string << "</li>"
+      
+      when :dl
+        @html_string << "</dl>" 
+      when :dl_header
+        @html_string << "</dt>"
+      when :dl_text
+        @html_string << "</dd>"
+      
+      # Tags
+      when :ac
+        @html_string << "</acronym>"
+            
+      when :emph 
+        @html_string << "</em>"
+      
+      when :tt
+        @html_string << "</tt>"
+    end
+  end
+
+  ###
+  ### Helper functions
+  ###
+
+  # Return the header id, given a string of text
+  def self.header_id(header_text)
+    return header_text.gsub(" ", "_").downcase
+  end
+  
+  # Fix Typography, according to HTML standards
+  def typeset(string)
+    return_str = string
+        
+    return_str.gsub!(';','&#8202;:')
+    return_str.gsub!(':','&#8202;:')
+    
+    return_str.gsub!("'", "&rsquo;")
+
+    return_str.gsub!('.','&#8202;.&thinsp;')
+    return_str.gsub!('?','&#8202;?&thinsp;')
+    return_str.gsub!('!','&#8202;!&thinsp;')
+
+    return_str.gsub!('e&#8202;.&thinsp;g&#8202;.&thinsp;','e.g. ')
+    return_str.gsub!('i&#8202;.&thinsp;e&#8202;.&thinsp;','i.e. ')
+    
+    return return_str
+  end
+  
+  # Return the correct indent for the given level
+  def indent(level)
+    case level
+      when 1
+        return ""
+      when 2
+        return "  "
+      when 3
+        return "    "
+      when 4
+        return "      "
+      when 5
+        return "        "
+      when 6
+        return "          "
+    else
+      return ""
+    end
+  end
+
+end