swordfish 0.0.1 → 0.0.2

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    MGMzODc2Y2FlNzUzMzdiMGNmMjlmODA1MjUwYjg3MWJhMzYwZWViOQ==
+    NDQyYjhkZjNiNTRjNTE3ZWRlYmFmNWMzOGI5Nzg0OTU3MjNhMWRkOQ==
   data.tar.gz: !binary |-
-    ODgzMmMyNDBkNTEzZDg1M2NkNWFiY2RlN2VlZDBjYmE2M2I3NTNlYw==
+    YjA5YWZmZDMyMWY4NzgxMmQ1MWVhNTU4YmQ0MDA4YmU4NDUyOTRhNg==
 SHA512:
   metadata.gz: !binary |-
-    ZjY4M2JiYmZlZGM1MTMxY2FlYWI2MjVlOTg1YmU0OTMxZDZmOTY1MWRmZjc3
-    ZTFiYTJiZTAyODJiZjA0OWM1MjAzNDE0YWQyODVjYTY4ZjYwYzlmNmFmM2Y1
-    MzE5NThjZDQwMGRmNDJjNjc4YzI3NmZkNWM3Y2I3Nzc4ZjQ2MWQ=
+    OTg0NzAwNTFjMzhhNDYzNWY3MjgyMzQ3MDdkMDVhNjE1NWYzMTc0ZjFhZWJj
+    YzM1M2YzNzQ4YjEyZDg2YTQwOWY3OGJjY2JhZmE4MGUyZGQ1NmEwNzkwMDk1
+    N2FjMzQ2MjRjNDcwYWY3YmVmODhkMGNhMjEwMTc2OTRmYzFmODY=
   data.tar.gz: !binary |-
-    OTRhZmZkYjQzOWQ5NGIxOGQ1NmNiNGE1MWUwNTBjYTIyMWFmOWUxNDdkN2U1
-    MzBmYWZhM2UwOGYxNmFkN2M0ZjY1ZWI1Y2MzZjczMzgxMGU5MmY2M2MwZWVh
-    N2QxYjIzMTkzYTdkODkzMDg2Mjg3OWMwYTcwZGQ0ZmI4Yzk1MjM=
+    NzRhNWRiOWFjMzAwNGNlZjQ0OTZiZWM3MDYwM2NkYzNlMDQ0YWNjYTk0MGJk
+    Zjk3NmZhYzA0NmVhYWJjYjFkZTg0MDY0OGY1MzFmMjM5MTE3ZjgwNGU5NTVk
+    NmEwZWZiOGU3MGFkMTk3MTZjOGJmMGZlZGQ4YjMzMjBjNjkyM2M=

data/lib/swordfish/document.rb

@@ -8,17 +8,20 @@ require 'swordfish/nodes/hyperlink'
 require 'swordfish/nodes/table'
 require 'swordfish/nodes/table_row'
 require 'swordfish/nodes/table_cell'
+require 'swordfish/nodes/image'
 
 # Swordfish::Document is the internal representation of a parsed document.
 
 module Swordfish
   class Document
     
-    attr_reader :nodes   # An array of all top-level elements in the document
+    attr_reader :nodes    # An array of all top-level elements in the document
+    attr_accessor :images # Stored image assets
 
     # On initialization, set the nodes list to an empty array
     def initialize
       @nodes = []
+      @images = {}
     end
 
     # Pass in a node and append it to the nodes array
@@ -30,8 +33,36 @@ module Swordfish
       end
     end
 
+    # Retrieve an image by name
+    def get_image(name)
+      @images[name]
+    end
+
+    # Save an image to a specified directory
+    def save_image(image, dest)
+      @images[image].open
+      File.open(dest, 'w') { |f| f.write(@images[image].read) }
+      @images[image].close
+    end
+
+    # Change the value that an image should report its source to be
+    def update_image_path(original_name, new_path)
+      find_nodes_by_type(Swordfish::Node::Image).each do |image_node|
+        if image_node.original_name == original_name
+          image_node.path = new_path
+        end
+      end
+    end
+
     def to_html
       @nodes.map(&:to_html).join
     end
+
+    private
+
+    # Return all nodes of a given type
+    def find_nodes_by_type(klass)
+      @nodes.collect{|n| n.find_nodes_by_type(klass)}.flatten
+    end
   end
 end

data/lib/swordfish/formats/docx.rb

@@ -8,6 +8,8 @@ module Swordfish
   class DOCX
 
     attr_reader :swordfish_doc   # The Swordfish::Document corresponding to the parsed document
+    attr_reader :docx_archive    # The source archive
+    attr_reader :namespaces      # A hash of XML namespaces used in this doc
     
     # Parse a document and return a Swordfish::Document object
     def self.open(filepath)
@@ -20,11 +22,12 @@ module Swordfish
       relationships = docx_archive.read 'word/_rels/document.xml.rels'
 
       # Parse the XML files and generate the Swordfish::Document
-      swordfish_docx = new document, styles, numbering, relationships
+      swordfish_docx = new docx_archive, document, styles, numbering, relationships
       swordfish_docx.swordfish_doc
     end
 
-    def initialize(document_xml, styles_xml, numbering_xml, relationships_xml)
+    def initialize(archive, document_xml, styles_xml, numbering_xml, relationships_xml)
+      @docx_archive = archive
       @swordfish_doc = Swordfish::Document.new
       parse_styles styles_xml
       parse_numbering numbering_xml
@@ -45,6 +48,7 @@ module Swordfish
     # Parse the document structure XML
     def parse(document_xml)
       @xml = Nokogiri::XML(document_xml)
+      @namespaces = @xml.collect_namespaces
 
       # Iterate over each element node and dispatch it to the appropriate parser
       @xml.xpath('//w:body').children.each do |node|
@@ -138,6 +142,14 @@ module Swordfish
       end
     end
 
+    # Extract an image resource as a tempfile
+    def read_image(image_name)
+      tempfile = Tempfile.new(image_name)
+      tempfile.write @docx_archive.get_input_stream("word/media/#{image_name}").read
+      tempfile.close
+      tempfile
+    end
+
     # NODE PARSERS
     # Each of the methods below (beginning with '_node') are specialized parsers for handling
     # a particular type of XML element.
@@ -151,13 +163,22 @@ module Swordfish
         case run_xml.name
           when 'r'
             # A true run node
-            text = Swordfish::Node::Text.new
             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 = 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
+            elsif run_xml.xpath('.//pic:pic', :pic => @namespaces['xmlns:pic']).length > 0
+              # An image run
+              image = Swordfish::Node::Image.new
+              relationship_id = run_xml.xpath('.//pic:pic/pic:blipFill/a:blip', :pic => @namespaces['xmlns:pic'], :a => @namespaces['xmlns:a'])[0]['r:embed'] rescue nil
+              if relationship_id
+                image.original_name = @relationships[relationship_id].split('/').last
+                @swordfish_doc.images[image.original_name] = read_image(image.original_name)
+                texts << image
+              end
             end
           when 'hyperlink'
             # Hyperlink nodes are placed amongst other run nodes, but

data/lib/swordfish/nodes/base.rb

@@ -60,6 +60,13 @@ module Swordfish
         end
       end
 
+      # Find all descendant nodes of a given type
+      def find_nodes_by_type(klass)
+        nodes = @children.collect{|n| n.find_nodes_by_type(klass)}.flatten
+        nodes << self if self.is_a?(klass)
+        nodes.compact
+      end
+
     end
 
     class BadContentError < Exception

data/lib/swordfish/nodes/image.rb

@@ -0,0 +1,26 @@
+# An image node
+# Actual image data is stored at the document level, and can be
+# retrieved by calling get_image(image_image) on the document
+# object.
+
+module Swordfish
+  module Node
+    class Image < Base
+
+      # @original_name holds the name of the file as it is reported by the source document
+      attr_accessor :original_name
+      # @path holds a new name for the image that must be assigned explicitly
+      attr_accessor :path
+
+      # Override Base append because an image node should never have children
+      def append(node)
+        raise BadContentError
+      end
+
+      def to_html
+        "<img src='#{@path ? @path : @original_name}'>"
+      end
+
+    end
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: swordfish
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Martin Posthumus
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-05-27 00:00:00.000000000 Z
+date: 2014-06-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -65,6 +65,7 @@ files:
 - lib/swordfish/formats/docx.rb
 - lib/swordfish/nodes/base.rb
 - lib/swordfish/nodes/hyperlink.rb
+- lib/swordfish/nodes/image.rb
 - lib/swordfish/nodes/list.rb
 - lib/swordfish/nodes/list_item.rb
 - lib/swordfish/nodes/paragraph.rb