swordfish 0.0.0 → 0.0.1

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    NzVjN2FlM2EwNzY2ZWQxOTJiZjdiZmVlZTZiZDdhMTE2MjYyNWQ1Ng==
+    MGMzODc2Y2FlNzUzMzdiMGNmMjlmODA1MjUwYjg3MWJhMzYwZWViOQ==
   data.tar.gz: !binary |-
-    MTU0ZjQzMTRiOTkzMGU5NDdkMDk0MzAyZTc5NTkyNTBiNzQwOGNiMA==
+    ODgzMmMyNDBkNTEzZDg1M2NkNWFiY2RlN2VlZDBjYmE2M2I3NTNlYw==
 SHA512:
   metadata.gz: !binary |-
-    ZWM2M2IzZjJmMmExMGYxNGU3MTllNmE2ZjQ2YTdlZDhiOTE0YWU3YjYwYzBl
-    MjBiY2ZiYTE0MWQ2OWRlYTkyYzE3ZTg1Y2I4NDIwOWYxNDY3MDk1ZWM0NjYw
-    YzcxYmVjZjEzNWEyNjI5NmU5OWM1Y2IyZTg3YzBhNWFhNjliNjg=
+    ZjY4M2JiYmZlZGM1MTMxY2FlYWI2MjVlOTg1YmU0OTMxZDZmOTY1MWRmZjc3
+    ZTFiYTJiZTAyODJiZjA0OWM1MjAzNDE0YWQyODVjYTY4ZjYwYzlmNmFmM2Y1
+    MzE5NThjZDQwMGRmNDJjNjc4YzI3NmZkNWM3Y2I3Nzc4ZjQ2MWQ=
   data.tar.gz: !binary |-
-    NmI2NGQ4ZjIxOWI2YzEzNWI5OTEwMzE4NzEyYjRmZjg2MmJjOTcwMDQ1OWYx
-    NzZjZjBjZmEyMzhhMTcwNjEyMzE1M2RkNzM2ZWQwZTE1YzhiMjVhNWRmOTQx
-    NWJlZTdkMTY1MDg0YTY5NDcwMGUzODdkM2I4ODFmOWEzMWM1MjE=
+    OTRhZmZkYjQzOWQ5NGIxOGQ1NmNiNGE1MWUwNTBjYTIyMWFmOWUxNDdkN2U1
+    MzBmYWZhM2UwOGYxNmFkN2M0ZjY1ZWI1Y2MzZjczMzgxMGU5MmY2M2MwZWVh
+    N2QxYjIzMTkzYTdkODkzMDg2Mjg3OWMwYTcwZGQ0ZmI4Yzk1MjM=

data/lib/swordfish/formats/docx.rb

@@ -50,12 +50,16 @@ module Swordfish
       @xml.xpath('//w:body').children.each do |node|
         case node.name
           when 'p'
-            if node.xpath('.//w:numPr').length == 0
+            if node.xpath('.//w:numPr').length == 0 && (@buffer.is_a?(Swordfish::Node::List) ? node.xpath('.//w:ind').length.zero? : true)
               # Regular paragraph
+              # (The buffer check makes sure that this isn't an indented paragraph immediately after a list item,
+              # which means we're most likely dealing with a multi-paragraph list item)
               flush
               @swordfish_doc.append _node_parse_paragraph(node)
-            else
+            elsif node.xpath('.//w:numPr/ancestor::w:pPrChange').length.zero?
               # List paragraph
+              # (must have a numPr node, but cannot have a pPrChange ancestor, since that means
+              # we are just looking at historical changes)
               # (Don't flush because we need to first ensure the list is fully parsed)
               _node_parse_list(node)
             end
@@ -86,6 +90,12 @@ module Swordfish
             elsif style_node['w:val'] == 'subscript'
               swordfish_node.stylize :subscript
             end
+          when 'rStyle'
+            if style_node['w:val'] == 'Strong'
+              swordfish_node.stylize :strong
+            elsif style_node['w:val'] == 'Emphasis'
+              swordfish_node.stylize :emphasis
+            end
         end
       end
     end
@@ -142,9 +152,13 @@ module Swordfish
           when 'r'
             # A true run node
             text = Swordfish::Node::Text.new
-            text.content = run_xml.xpath('./w:t')[0].content
-            get_styles_for_node(text, run_xml.xpath('./w:rPr')[0])
-            texts << text
+            if run_xml.xpath('./w:t').length > 0
+              # Only examine the run if it includes text codes. The run may also include
+              # things like comment nodes, which should be ignored.
+              text.content = run_xml.xpath('./w:t')[0].content
+              get_styles_for_node(text, run_xml.xpath('./w:rPr')[0])
+              texts << text
+            end
           when 'hyperlink'
             # Hyperlink nodes are placed amongst other run nodes, but
             # they themselves also contain runs. Hyperlinks include
@@ -174,6 +188,15 @@ module Swordfish
       # since the only way to tell the list has been fully parsed is to encounter
       # a non-list element.
 
+      # Handle paragraphs with no level, which represent multi-paragraph list items
+      if node.xpath(".//w:numPr/w:ilvl").length.zero?
+        para = Swordfish::Node::Paragraph.new
+        _node_parse_runs(node).each {|r| para.append(r)}
+        @buffer.last_list_item(:recurse => true).wrap_children(Swordfish::Node::Text, Swordfish::Node::Paragraph)
+        @buffer.last_list_item(:recurse => true).append para
+        return
+      end
+
       # Get the list item's abstract numbering and level
       list_item = Swordfish::Node::ListItem.new
       _node_parse_runs(node).each {|r| list_item.append(r)}
@@ -184,22 +207,23 @@ module Swordfish
       unless @buffer
         @buffer = Swordfish::Node::List.new
         @buffer.stylize @numbering[numbering_scheme][level].to_sym
+        @buffer_initial_value = level # Lists may have an arbitrary initial level
       end
 
       # Compare the level of this list item to the bottommost node in
       # the build buffer to determine where in the hierarchy to add
       # this node (i.e., are we dealing with list nesting or not?)
-      if @buffer.depth_of_final_node >= level
+      if @buffer.depth_of_final_node >= level || @buffer.children.empty?
         # Add sibling to existing list
         target = @buffer
-        level.times do
+        (level - @buffer_initial_value).times do
           target = target.last_list_item.nested_list
         end
         target.append list_item
       elsif @buffer.depth_of_final_node < level
         # Add new nested list
         target = @buffer
-        (level - 1).times do
+        (level - @buffer_initial_value- 1).times do
           target = target.last_list_item.nested_list
         end
         list = Swordfish::Node::List.new

data/lib/swordfish/nodes/base.rb

@@ -5,7 +5,7 @@ module Swordfish
     class Base
 
       attr_accessor :content
-      attr_reader :children
+      attr_accessor :children
       attr_reader :style
 
       # Initialize with a blank stylesheet and no children
@@ -21,6 +21,11 @@ module Swordfish
         @children.flatten!
       end
 
+      # Replace a child node at a given index
+      def replace(node, idx)
+        @children[idx] = node
+      end
+
       # Take a style or styles and add them to this node's stylesheet
       def stylize(styles)
         @style.merge styles
@@ -39,6 +44,22 @@ module Swordfish
         end
       end
 
+      # Delete all child nodes
+      def clear_children
+        @children = []
+      end
+
+      # Wrap all children of type child_class with a new node of type wrapper_class
+      def wrap_children(child_class, wrapper_class)
+        new_node = wrapper_class.new
+        new_node.append @children.select{|n| n.is_a? child_class}
+        unless new_node.children.empty?
+          idx = @children.find_index(new_node.children[0])
+          @children = @children - new_node.children
+          @children.insert idx, new_node
+        end
+      end
+
     end
 
     class BadContentError < Exception

data/lib/swordfish/nodes/list.rb

@@ -23,9 +23,27 @@ module Swordfish
         depth
       end
 
-      # Return the final child list item (no nesting)
-      def last_list_item
-        @children.last
+      # Return the final child list
+      def last_list
+        node = self
+        while node.children && node.last_list_item.nested_list
+          node = node.last_list_item.nested_list
+        end
+        node
+      end
+
+      # Return the final child list item
+      def last_list_item(opts = {})
+        if opts[:recurse]
+          node = self
+          li = @children.last
+          while node.children && node = node.last_list_item.nested_list
+            li = node.children.last
+          end
+          li
+        else
+          @children.last
+        end
       end
 
     end

data/lib/swordfish/nodes/paragraph.rb

@@ -8,7 +8,8 @@ module Swordfish
         if @content
           "<p>#{@content}</p>"
         else
-          "<p>#{@children.map(&:to_html).join}</p>"
+          text = @children.map(&:to_html).join
+          "<p>#{text}</p>" unless text.length.zero?
         end
       end
 

data/lib/swordfish/nodes/text.rb

@@ -17,6 +17,8 @@ module Swordfish
         html = "<strike>#{html}</strike>" if @style.strikethrough?
         html = "<sup>#{html}</sup>" if @style.superscript?
         html = "<sub>#{html}</sub>" if @style.subscript?
+        html = "<strong>#{html}</strong>" if @style.strong?
+        html = "<em>#{html}</em>" if @style.emphasis?
         html
       end
 

data/lib/swordfish/stylesheet.rb

@@ -6,7 +6,7 @@ module Swordfish
     # Define all supported values here
     SUPPORTED_STYLES = [
       # Inline styles
-      :bold, :italic, :underline, :superscript, :subscript, :strikethrough,
+      :bold, :italic, :underline, :superscript, :subscript, :strikethrough, :strong, :emphasis,
       # List enumeration styles
       :bullet, :decimal, :lowerLetter, :lowerRoman
     ]

data/lib/swordfish.rb

@@ -4,10 +4,10 @@ require 'swordfish/formats/docx'
 module Swordfish
   
   # Main entry point into the parser. Pass in a filepath and return a parsed document.
-  def self.open(filepath)
-    extension = filepath.split('.').last.downcase
+  def self.open(filepath, opts={})
+    extension = (opts[:extension] || filepath.split('.').last).downcase.to_sym
     case extension
-      when 'docx'
+      when :docx
         Swordfish::DOCX.open(filepath)
       else
         raise UnsupportedFormatError, "'#{extension}' is not a recognized file format"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: swordfish
 version: !ruby/object:Gem::Version
-  version: 0.0.0
+  version: 0.0.1
 platform: ruby
 authors:
 - Martin Posthumus