htmltoword 0.5.1 → 0.7.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a0deb2026e95be6bde1c5af01b676e231dea63c0
-  data.tar.gz: 94971919c3ad4bfae9dadf00a1d810828f33140c
+  metadata.gz: 7034d51bc5f9b9251aec22334f9ee4427bc3d520
+  data.tar.gz: b5925e2169eca6f9de534788a2f62850da43ac4d
 SHA512:
-  metadata.gz: 19f266330f3e4c953f108b01e960517e8b00c3937de7a1ad9f443f6a7bbca2abc94678a0c2926bb8905fe04708ec8665014edb73605e3b68826b098443176d7e
-  data.tar.gz: 93d0cf96fc81fa058ba6eda905152352d6b3c9dfbef4473c03cbf900c512af00ab9d8244dfa271df77a2a2b6d649b38e76ad292e793c582e44a61b03a260fc8c
+  metadata.gz: c5ac9f6e7c6f26acbb06ee4c85a042cce12318802aabb08a836df3db136b408f6622b34366e0993ec4b8de7b2183b7f4e83c639c858342b5aa2586b264c2ed89
+  data.tar.gz: 6747b7abea40e7b43d52322ebe4283e4edc5f97c95487cb3de57e0f66bcfd209f379012572a88a96b4f95fe95651984eaeaf2b4cfb7b9410e648a80a0b7a3426

data/lib/htmltoword.rb

@@ -1,8 +1,7 @@
 # encoding: UTF-8
-# require 'action_controller'
-# require 'action_view'
 require 'nokogiri'
 require 'zip'
+require 'open-uri'
 require_relative 'htmltoword/configuration'
 
 module Htmltoword
@@ -15,12 +14,16 @@ module Htmltoword
       @configuration ||= Configuration.new
     end
 
-    alias :config :configuration
+    alias_method :config, :configuration
   end
 end
 
-
 require_relative 'htmltoword/version'
-require_relative 'htmltoword/htmltoword_helper'
+require_relative 'htmltoword/helpers/templates_helper'
+require_relative 'htmltoword/helpers/xslt_helper'
 require_relative 'htmltoword/document'
-require_relative 'htmltoword/action_controller'
+
+if defined?(Rails)
+  require_relative 'htmltoword/renderer'
+  require_relative 'htmltoword/railtie'
+end

data/lib/htmltoword/document.rb

@@ -1,26 +1,26 @@
 module Htmltoword
   class Document
-    class << self
-      include HtmltowordHelper
+    include XSLTHelper
 
+    class << self
+      include TemplatesHelper
       def create(content, template_name = nil, extras = false)
         template_name += extension if template_name && !template_name.end_with?(extension)
         document = new(template_file(template_name))
-        document.replace_file(content, Document.doc_xml_file, extras)
+        document.replace_files(content, extras)
         document.generate
       end
 
       def create_and_save(content, file_path, template_name = nil, extras = false)
-        File.open(file_path, "wb") do |out|
+        File.open(file_path, 'wb') do |out|
           out << create(content, template_name, extras)
         end
       end
 
-      def create_with_content(template, content, set=nil, extras = false)
+      def create_with_content(template, content, extras = false)
         template += extension unless template.end_with?(extension)
-        content = replace_values(content, set) if set
         document = new(template_file(template))
-        document.replace_file(content, Document.doc_xml_file, extras)
+        document.replace_files(content, extras)
         document.generate
       end
 
@@ -40,22 +40,15 @@ module Htmltoword
         'word/_rels/document.xml.rels'
       end
 
-      def numbering_xslt
-        File.join(Htmltoword.config.default_xslt_path, 'numbering.xslt')
-      end
-
-      def relations_xslt
-        File.join(Htmltoword.config.default_xslt_path, 'relations.xslt')
-      end
-
-      def xslt_template(extras = false)
-        File.join(Htmltoword.config.default_xslt_path, (extras ? 'htmltoword.xslt' : 'base.xslt'))
+      def content_types_xml_file
+        '[Content_Types].xml'
       end
     end
 
     def initialize(template_path)
       @replaceable_files = {}
       @template_path = template_path
+      @image_files = []
     end
 
     #
@@ -67,35 +60,95 @@ module Htmltoword
           template_zip.each do |entry|
             out.put_next_entry entry.name
             if @replaceable_files[entry.name] && entry.name == Document.doc_xml_file
-              source = entry.get_input_stream.read.sub(/(<w:body>)((.|\n)*?)(<w:sectPr)/, "\\1#{@replaceable_files[entry.name]}\\4")
+              source = entry.get_input_stream.read
+              # Change only the body of document. TODO: Improve this...
+              source = source.sub(/(<w:body>)((.|\n)*?)(<w:sectPr)/, "\\1#{@replaceable_files[entry.name]}\\4")
               out.write(source)
             elsif @replaceable_files[entry.name]
               out.write(@replaceable_files[entry.name])
+            elsif entry.name == Document.content_types_xml_file
+              raw_file = entry.get_input_stream.read
+              content_types = @image_files.empty? ? raw_file : inject_image_content_types(raw_file)
+
+              out.write(content_types)
             else
               out.write(template_zip.read(entry.name))
             end
           end
+          unless @image_files.empty?
+          #stream the image files into the media folder using open-uri
+            @image_files.each do |hash|
+              out.put_next_entry("word/media/#{hash[:filename]}")
+              open(hash[:url], 'rb') do |f|
+                out.write(f.read)
+              end
+            end
+          end
         end
         buffer.string
       end
     end
 
-    def replace_file(html, file_name = Document.doc_xml_file, extras = false)
-      html = html.presence || '<body></body>'
+    def replace_files(html, extras = false)
+      html = '<body></body>' if html.nil? || html.empty?
       source = Nokogiri::HTML(html.gsub(/>\s+</, '><'))
-      transform_and_replace(source, Document.numbering_xslt, Document.numbering_xml_file)
-      transform_and_replace(source, Document.relations_xslt, Document.relations_xml_file)
-      cleaned_source = Nokogiri::XSLT(File.open(File.join(Htmltoword.config.default_xslt_path, 'inline_elements.xslt'))).transform(source)
-      transform_and_replace(cleaned_source, Document.xslt_template(extras), file_name, extras)
+      transform_and_replace(source, xslt_path('numbering'), Document.numbering_xml_file)
+      transform_and_replace(source, xslt_path('relations'), Document.relations_xml_file)
+      transform_doc_xml(source, extras)
+      local_images(source)
+    end
+
+    def transform_doc_xml(source, extras = false)
+      transformed_source = xslt(stylesheet_name: 'cleanup').transform(source)
+      transformed_source = xslt(stylesheet_name: 'inline_elements').transform(transformed_source)
+      transform_and_replace(transformed_source, document_xslt(extras), Document.doc_xml_file, extras)
     end
 
     private
 
-    def transform_and_replace source, xslt, file, remove_ns = false
-      xslt = Nokogiri::XSLT(File.open(xslt))
-      content = xslt.apply_to(source)
-      content.gsub!(/\s*xmlns:(\w+)="(.*?)\s*"/,'') if remove_ns
-      @replaceable_files[file] =  content
+    def transform_and_replace(source, stylesheet_path, file, remove_ns = false)
+      stylesheet = xslt(stylesheet_path: stylesheet_path)
+      content = stylesheet.apply_to(source)
+      content.gsub!(/\s*xmlns:(\w+)="(.*?)\s*"/, '') if remove_ns
+      @replaceable_files[file] = content
+    end
+
+    #generates an array of hashes with filename and full url
+    #for all images to be embeded in the word document
+    def local_images(source)
+      source.css('img').each_with_index do |image,i|
+        filename = image['data-filename'] ? image['data-filename'] : image['src'].split("/").last
+        ext = File.extname(filename).delete(".").downcase
+
+        @image_files << { filename: "image#{i+1}.#{ext}", url: image['src'], ext: ext }
+      end
+    end
+
+    #get extension from filename and clean to match content_types
+    def content_type_from_extension(ext)
+      ext == "jpg" ? "jpeg" : ext
+    end
+
+    #inject the required content_types into the [content_types].xml file...
+    def inject_image_content_types(source)
+      doc = Nokogiri::XML(source)
+
+      #get a list of all extensions currently in content_types file
+      existing_exts = doc.css("Default").map { |node| node.attribute("Extension").value }.compact
+
+      #get a list of extensions we need for our images
+      required_exts = @image_files.map{ |i| i[:ext] }
+
+      #workout which required extensions are missing from the content_types file
+      missing_exts = (required_exts - existing_exts).uniq
+
+      #inject missing extensions into document
+      missing_exts.each do |ext|
+        doc.at_css("Types").add_child( "<Default Extension='#{ext}' ContentType='image/#{content_type_from_extension(ext)}'/>")
+      end
+
+      #return the amended source to be saved into the zip
+      doc.to_s
     end
   end
 end

data/lib/htmltoword/helpers/templates_helper.rb

@@ -0,0 +1,9 @@
+module Htmltoword
+  module TemplatesHelper
+    def template_file(template_file_name = nil)
+      default_path = File.join(::Htmltoword.config.default_templates_path, 'default.docx')
+      template_path = template_file_name.nil? ? '' : File.join(::Htmltoword.config.custom_templates_path, template_file_name)
+      File.exist?(template_path) ? template_path : default_path
+    end
+  end
+end

data/lib/htmltoword/helpers/xslt_helper.rb

@@ -0,0 +1,17 @@
+module Htmltoword
+  module XSLTHelper
+    def document_xslt(extras = false)
+      file_name = extras ? 'htmltoword' : 'base'
+      xslt_path(file_name)
+    end
+
+    def xslt_path(template_name)
+      File.join(Htmltoword.config.default_xslt_path, "#{template_name}.xslt")
+    end
+
+    def xslt(stylesheet_name: nil, stylesheet_path: nil)
+      return Nokogiri::XSLT(File.open(stylesheet_path)) if stylesheet_path
+      Nokogiri::XSLT(File.open(xslt_path(stylesheet_name)))
+    end
+  end
+end

data/lib/htmltoword/railtie.rb

@@ -0,0 +1,25 @@
+module Htmltoword
+  class Railtie < ::Rails::Railtie
+    initializer 'htmltoword.setup' do
+      if defined?(Mime) and Mime[:docx].nil?
+        Mime::Type.register 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', :docx
+      end
+
+      ActionController::Renderers.add :docx do |file_name, options|
+        Htmltoword::Renderer.send_file(self, file_name, options)
+      end
+
+      if defined? ActionController::Responder
+        ActionController::Responder.class_eval do
+          def to_docx
+            if @default_response
+              @default_response.call(options)
+            else
+              controller.render({ docx: controller.action_name }.merge(options))
+            end
+          end
+        end
+      end
+    end
+  end
+end

data/lib/htmltoword/renderer.rb

@@ -0,0 +1,43 @@
+module Htmltoword
+  class Renderer
+    class << self
+      def send_file(context, filename, options = {})
+        new(context, filename, options).send_file
+      end
+    end
+
+    def initialize(context, filename, options)
+      @word_template = options[:word_template].presence
+      @disposition = options.fetch(:disposition, 'attachment')
+      @use_extras = options.fetch(:extras, false)
+      @file_name = file_name(filename, options)
+      @context = context
+      define_template(filename, options)
+      @content = options[:content] || @context.render_to_string(options)
+    end
+
+    def send_file
+      document = Htmltoword::Document.create(@content, @word_template, @use_extras)
+      @context.send_data(document, filename: @file_name, type: Mime[:docx], disposition: @disposition)
+    end
+
+    private
+
+    def define_template(filename, options)
+      if options[:template] == @context.action_name
+        if filename =~ %r{^([^\/]+)/(.+)$}
+          options[:prefixes] ||= []
+          options[:prefixes].unshift $1
+          options[:template] = $2
+        else
+          options[:template] = filename
+        end
+      end
+    end
+
+    def file_name(filename, options)
+      name = options[:filename].presence || filename
+      name =~ /\.docx$/ ? name : "#{name}.docx"
+    end
+  end
+end

data/lib/htmltoword/version.rb

@@ -1,3 +1,3 @@
 module Htmltoword
-  VERSION = "0.5.1"
+  VERSION = '0.7.0'
 end

data/lib/htmltoword/xslt/base.xslt

@@ -19,7 +19,7 @@
   <xsl:include href="./functions.xslt"/>
   <xsl:include href="./tables.xslt"/>
   <xsl:include href="./links.xslt"/>
-
+  <xsl:include href="./images.xslt"/>
   <xsl:template match="/">
     <xsl:apply-templates />
   </xsl:template>
@@ -204,6 +204,12 @@
         <xsl:when test="self::a[starts-with(@href, 'http://') or starts-with(@href, 'https://')]">
           <xsl:call-template name="link" />
         </xsl:when>
+        <xsl:when test="self::img">
+          <xsl:comment>
+            This template adds images.
+          </xsl:comment>
+          <xsl:call-template name="image"/>
+        </xsl:when>
         <xsl:otherwise>
           <xsl:apply-templates />
         </xsl:otherwise>
@@ -277,12 +283,12 @@
   <xsl:template match="text()">
     <xsl:if test="string-length(.) > 0">
       <w:r>
-        <xsl:if test="ancestor::i or ancestor::em">
+        <xsl:if test="ancestor::i">
           <w:rPr>
             <w:i />
           </w:rPr>
         </xsl:if>
-        <xsl:if test="ancestor::b or ancestor::strong">
+        <xsl:if test="ancestor::b">
           <w:rPr>
             <w:b />
           </w:rPr>
@@ -292,6 +298,21 @@
             <w:u w:val="single"/>
           </w:rPr>
         </xsl:if>
+        <xsl:if test="ancestor::s">
+          <w:rPr>
+            <w:strike w:val="true"/>
+          </w:rPr>
+        </xsl:if>
+        <xsl:if test="ancestor::sub">
+          <w:rPr>
+            <w:vertAlign w:val="subscript"/>
+          </w:rPr>
+        </xsl:if>
+        <xsl:if test="ancestor::sup">
+          <w:rPr>
+            <w:vertAlign w:val="superscript"/>
+          </w:rPr>
+        </xsl:if>
         <w:t xml:space="preserve"><xsl:value-of select="."/></w:t>
       </w:r>
     </xsl:if>

data/lib/htmltoword/xslt/cleanup.xslt

@@ -0,0 +1,70 @@
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:str="http://exslt.org/strings"
+                xmlns:func="http://exslt.org/functions"
+                xmlns:fn="http://www.w3.org/2005/xpath-functions"
+                version="1.0"
+                exclude-result-prefixes="java msxsl ext w o v WX aml w10"
+                extension-element-prefixes="func">
+  <xsl:output method="html" encoding="utf-8" omit-xml-declaration="yes" indent="yes"/>
+
+  <xsl:template match="/">
+    <xsl:apply-templates/>
+  </xsl:template>
+
+  <xsl:template match="head"/>
+
+  <!-- Elements not supported -->
+  <xsl:template match="applet"/>
+  <xsl:template match="area"/>
+  <xsl:template match="audio"/>
+  <xsl:template match="base"/>
+  <xsl:template match="basefont"/>
+  <xsl:template match="canvas"/>
+  <xsl:template match="command"/>
+  <xsl:template match="font"/>
+  <xsl:template match="iframe"/>
+  <xsl:template match="isindex"/>
+  <xsl:template match="map"/>
+  <xsl:template match="noframes"/>
+  <xsl:template match="noscript"/>
+  <xsl:template match="object"/>
+  <xsl:template match="param"/>
+  <xsl:template match="script"/>
+  <xsl:template match="source"/>
+  <xsl:template match="style"/>
+  <xsl:template match="video"/>
+
+  <!-- Elements currently being handled as normal text. Remove tags only -->
+  <xsl:template match="abbr"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="acronym"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="bdi"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="bdo"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="big"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="code"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="kbd"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="samp"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="small"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="tt"><xsl:apply-templates/></xsl:template>
+  <xsl:template match="var"><xsl:apply-templates/></xsl:template>
+
+  <!-- Inline elements transformations -->
+  <xsl:template match="cite"><i><xsl:apply-templates/></i></xsl:template>
+  <xsl:template match="del"><s><xsl:apply-templates/></s></xsl:template>
+  <xsl:template match="dfn"><i><xsl:apply-templates/></i></xsl:template>
+  <xsl:template match="em"><i><xsl:apply-templates/></i></xsl:template>
+  <xsl:template match="ins"><u><xsl:apply-templates/></u></xsl:template>
+  <xsl:template match="mark"><span class="h" data-style="yellow"><xsl:apply-templates/></span></xsl:template>
+  <xsl:template match="q">"<xsl:apply-templates/>"</xsl:template>
+  <xsl:template match="strike"><s><xsl:apply-templates/></s></xsl:template>
+  <xsl:template match="strong"><b><xsl:apply-templates/></b></xsl:template>
+
+  <!-- Block elements transformations -->
+  <xsl:template match="section"><div class="{@class}" style="{@style}"><xsl:apply-templates/></div></xsl:template>
+  <xsl:template match="article"><div  class="{@class}" style="{@style}"><xsl:apply-templates/></div></xsl:template>
+
+  <xsl:template match="@*|node()">
+    <xsl:copy>
+      <xsl:apply-templates select="@*|node()"/>
+    </xsl:copy>
+  </xsl:template>
+</xsl:stylesheet>

data/lib/htmltoword/xslt/functions.xslt

@@ -31,4 +31,7 @@
       </xsl:choose>
     </func:result>
   </func:function>
+
+  <!-- template as function used to return the relationship id of the element (currently links or images) -->
+  <xsl:template name="relationship-id">rId<xsl:value-of select="count(preceding::a[starts-with(@href, 'http://') or starts-with(@href, 'https://')])+count(preceding::img)+8"/></xsl:template>
 </xsl:stylesheet>

data/lib/htmltoword/xslt/image_functions.xslt

@@ -0,0 +1,148 @@
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main"
+                xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"
+                xmlns:o="urn:schemas-microsoft-com:office:office"
+                xmlns:v="urn:schemas-microsoft-com:vml"
+                xmlns:WX="http://schemas.microsoft.com/office/word/2003/auxHint"
+                xmlns:aml="http://schemas.microsoft.com/aml/2001/core"
+                xmlns:w10="urn:schemas-microsoft-com:office:word"
+                xmlns:pkg="http://schemas.microsoft.com/office/2006/xmlPackage"
+                xmlns:msxsl="urn:schemas-microsoft-com:xslt"
+                xmlns:ext="http://www.xmllab.net/wordml2html/ext"
+                xmlns:java="http://xml.apache.org/xalan/java"
+                xmlns:str="http://exslt.org/strings"
+                xmlns:func="http://exslt.org/functions"
+                xmlns:fn="http://www.w3.org/2005/xpath-functions"
+                version="1.0"
+                exclude-result-prefixes="java msxsl ext w o v WX aml w10"
+                extension-element-prefixes="func">
+
+  <!-- template as function used to return the file extension of an image. -->
+  <xsl:template name="image-extension">
+    <xsl:param name="data-filename" select="." />
+    <xsl:param name="source" select="." />
+    <xsl:variable name="filename">
+      <xsl:call-template name="image-name">
+        <xsl:with-param name="source" select="$source"/>
+        <xsl:with-param name="data-filename" select="$data-filename"/>
+      </xsl:call-template>
+    </xsl:variable>
+    <xsl:value-of select="substring-after($filename,'.')"/>
+  </xsl:template>
+
+  <!-- template as function used to return the name of an image. -->
+  <xsl:template name="image-name">
+    <xsl:param name="data-filename" select="." />
+    <xsl:param name="source" select="." />
+    <xsl:choose>
+      <xsl:when test="string-length($data-filename) > 0">
+        <xsl:value-of select="$data-filename"/>
+      </xsl:when>
+      <xsl:otherwise>
+        <xsl:call-template name="extract-filename-from-path">
+          <xsl:with-param name="path" select="$source"/>
+        </xsl:call-template>
+      </xsl:otherwise>
+    </xsl:choose>
+  </xsl:template>
+
+  <!-- template as function used to extract the filename from the image source. Can't use tokenize or other functions which return a fragment tree as xpath can't process them. -->
+  <xsl:template name="extract-filename-from-path">
+      <xsl:param name="path" select="." />
+      <xsl:choose>
+        <xsl:when test="not(contains($path, '/'))">
+            <xsl:value-of select="$path" />
+        </xsl:when>
+        <xsl:otherwise>
+          <xsl:call-template name="extract-filename-from-path">
+            <xsl:with-param name="path" select="substring-after($path, '/')" />
+          </xsl:call-template>
+        </xsl:otherwise>
+      </xsl:choose>
+  </xsl:template>
+
+  <!--
+    Returns the unqualified dimension from a length specification copied from:
+  http://docbook.sourceforge.net/release/xsl/1.76.1/doc/lib/length-magnitude.html
+   -->
+
+  <xsl:template name="length-magnitude">
+    <xsl:param name="length" select="'0pt'"></xsl:param>
+
+    <xsl:choose>
+      <xsl:when test="string-length($length) = 0"></xsl:when>
+      <xsl:when test="substring($length,1,1) = '0' or substring($length,1,1) = '1'                                      or substring($length,1,1) = '2' or substring($length,1,1) = '3'                                      or substring($length,1,1) = '4' or substring($length,1,1) = '5'                                      or substring($length,1,1) = '6' or substring($length,1,1) = '7'                                      or substring($length,1,1) = '8' or substring($length,1,1) = '9'                                      or substring($length,1,1) = '.'">
+        <xsl:value-of select="substring($length,1,1)"></xsl:value-of>
+        <xsl:call-template name="length-magnitude">
+          <xsl:with-param name="length" select="substring($length,2)"></xsl:with-param>
+        </xsl:call-template>
+      </xsl:when>
+    </xsl:choose>
+  </xsl:template>
+
+
+
+
+  <!-- Convert em and pixel sizes to inches. Inspired by from:
+  http://docbook.sourceforge.net/release/xsl/1.76.1/doc/lib/length-in-points.html
+  EMU info from:  https://startbigthinksmall.wordpress.com/2010/01/04/points-inches-and-emus-measuring-units-in-office-open-xml/
+   -->
+  <xsl:template name="length-in-emus">
+    <xsl:param name="length" select="'0px'"/>
+    <xsl:param name="em.size" select="6.02250006023"/>
+    <xsl:param name="pixels.per.inch" select="90"/>
+    <xsl:variable name="emus.per.inch" select="914400"/>
+
+    <xsl:variable name="magnitude">
+      <xsl:call-template name="length-magnitude">
+        <xsl:with-param name="length" select="$length"/>
+      </xsl:call-template>
+    </xsl:variable>
+
+    <xsl:variable name="units">
+      <xsl:value-of select="substring($length, string-length($magnitude)+1)"/>
+    </xsl:variable>
+
+    <xsl:variable name="inches">
+      <xsl:choose>
+        <xsl:when test="$units = 'px'">
+          <xsl:value-of select="$magnitude div $pixels.per.inch"/>
+        </xsl:when>
+        <xsl:when test="$units = 'em'">
+          <xsl:value-of select="$magnitude div $em.size"/>
+        </xsl:when>
+        <xsl:otherwise>
+          <xsl:message>
+            <xsl:text>Unrecognized unit of measure: </xsl:text>
+            <xsl:value-of select="$units"></xsl:value-of>
+            <xsl:text>.</xsl:text>
+          </xsl:message>
+        </xsl:otherwise>
+      </xsl:choose>
+    </xsl:variable>
+
+    <xsl:value-of select="$inches * $emus.per.inch"/>
+  </xsl:template>
+
+  <!-- template as function used to get the width or height of an image in points. -->
+  <xsl:template name="image-dimention">
+    <xsl:param name="style" />
+    <xsl:param name="data-value" />
+    <xsl:param name="type" />
+
+    <xsl:variable name="size">
+      <xsl:choose>
+        <xsl:when test="string-length($data-value) > 0">
+          <xsl:value-of select="$data-value" />
+        </xsl:when>
+        <xsl:when test="contains($style, concat($type,':'))">
+          <xsl:value-of select="translate(str:tokenize(substring-after($style, concat($type,':')), ';')[1],'&#x20;&#x9;&#xD;&#xA;','')" />
+        </xsl:when>
+        <xsl:otherwise>1</xsl:otherwise>
+      </xsl:choose>
+    </xsl:variable>
+    <xsl:call-template name="length-in-emus">
+      <xsl:with-param name="length" select="$size"/>
+    </xsl:call-template>
+  </xsl:template>
+</xsl:stylesheet>

data/lib/htmltoword/xslt/images.xslt

@@ -0,0 +1,129 @@
+<xsl:stylesheet xmlns:wpc="http://schemas.microsoft.com/office/word/2010/wordprocessingCanvas"
+                xmlns:mo="http://schemas.microsoft.com/office/mac/office/2008/main"
+                xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006"
+                xmlns:mv="urn:schemas-microsoft-com:mac:vml"
+                xmlns:o="urn:schemas-microsoft-com:office:office"
+                xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"
+                xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math"
+                xmlns:v="urn:schemas-microsoft-com:vml"
+                xmlns:wp14="http://schemas.microsoft.com/office/word/2010/wordprocessingDrawing"
+                xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing"
+                xmlns:w10="urn:schemas-microsoft-com:office:word"
+                xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main"
+                xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml"
+                xmlns:w15="http://schemas.microsoft.com/office/word/2012/wordml"
+                xmlns:wpg="http://schemas.microsoft.com/office/word/2010/wordprocessingGroup"
+                xmlns:wpi="http://schemas.microsoft.com/office/word/2010/wordprocessingInk"
+                xmlns:wne="http://schemas.microsoft.com/office/word/2006/wordml"
+                xmlns:wps="http://schemas.microsoft.com/office/word/2010/wordprocessingShape"
+                xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main"
+                xmlns:a14="http://schemas.microsoft.com/office/drawing/2010/main"
+                xmlns:pic="http://schemas.openxmlformats.org/drawingml/2006/picture"
+                xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
+                xmlns:WX="http://schemas.microsoft.com/office/word/2003/auxHint"
+                xmlns:aml="http://schemas.microsoft.com/aml/2001/core"
+                xmlns:pkg="http://schemas.microsoft.com/office/2006/xmlPackage"
+                xmlns:msxsl="urn:schemas-microsoft-com:xslt"
+                xmlns:ext="http://www.xmllab.net/wordml2html/ext"
+                xmlns:java="http://xml.apache.org/xalan/java"
+                xmlns:str="http://exslt.org/strings"
+                xmlns:func="http://exslt.org/functions"
+                xmlns:fn="http://www.w3.org/2005/xpath-functions"
+                mc:Ignorable="w14 w15 wp14 a14"
+                version="1.0"
+                exclude-result-prefixes="java msxsl ext w o v WX aml w10"
+                extension-element-prefixes="func">
+
+  <xsl:include href="./image_functions.xslt"/>
+
+  <xsl:template match="img|body/img" name="image">
+    <w:drawing>
+      <wp:inline distT="0" distB="0" distL="0" distR="0">
+        <wp:extent>
+          <xsl:call-template name="image-dimention-attributes"/>
+        </wp:extent>
+        <wp:effectExtent l="0" t="0" r="0" b="0"/>
+        <wp:docPr>
+          <xsl:attribute name="id"><xsl:value-of select="count(preceding::img)+1" /></xsl:attribute>
+          <xsl:attribute name="name">Picture <xsl:value-of select="count(preceding::img)+1" /></xsl:attribute>
+        </wp:docPr>
+        <wp:cNvGraphicFramePr>
+          <a:graphicFrameLocks noChangeAspect="1"/>
+        </wp:cNvGraphicFramePr>
+        <a:graphic>
+          <a:graphicData uri="http://schemas.openxmlformats.org/drawingml/2006/picture">
+            <pic:pic>
+              <pic:nvPicPr>
+                <pic:cNvPr>
+                  <xsl:attribute name="id"><xsl:value-of select="count(preceding::img)+1" /></xsl:attribute>
+                  <xsl:attribute name="title"><xsl:value-of select="@alt" /></xsl:attribute>
+                  <xsl:attribute name="name"><xsl:call-template name="image-name">
+                      <xsl:with-param name="source" select="@src"/>
+                      <xsl:with-param name="data-filename" select="@data-filename"/>
+                    </xsl:call-template></xsl:attribute>
+                </pic:cNvPr>
+                <pic:cNvPicPr/>
+              </pic:nvPicPr>
+              <pic:blipFill>
+                <a:blip>
+                  <xsl:attribute name="r:embed"><xsl:call-template name="relationship-id"/></xsl:attribute>
+                  <a:extLst>
+                    <a:ext uri="{{28A0092B-C50C-407E-A947-70E740481C1C}}">
+                      <a14:useLocalDpi val="0"/>
+                    </a:ext>
+                  </a:extLst>
+                </a:blip>
+                <a:stretch>
+                  <a:fillRect/>
+                </a:stretch>
+              </pic:blipFill>
+              <pic:spPr>
+                <a:xfrm>
+                  <a:off x="0" y="0"/>
+                  <a:ext>
+                    <xsl:call-template name="image-dimention-attributes"/>
+                  </a:ext>
+                </a:xfrm>
+                <a:prstGeom prst="rect">
+                  <a:avLst/>
+                </a:prstGeom>
+              </pic:spPr>
+            </pic:pic>
+          </a:graphicData>
+        </a:graphic>
+      </wp:inline>
+    </w:drawing>
+  </xsl:template>
+
+  <!--
+    A style specifiying the width and height is required to render an image correctly in word.
+    Styles can be provided either via data attributes:
+    img src="pathtosomeimage" data-width="200px" data-height="250px"
+
+    or via the html style attribute
+    img src="pathtosomeimage" style="width:150px;height:200px"
+
+    If both a style and data attributes are provided then data attributes take president.
+    e.g.
+    img src="pathtosomeimage" data-width="200em" style="width:150px;height:200px"
+    becomes: width:200em;height:200px
+
+    Note: All sizes must be in pixles or em
+  -->
+  <xsl:template name="image-dimention-attributes">
+    <xsl:attribute name="cx">
+      <xsl:call-template name="image-dimention">
+        <xsl:with-param name="style" select="@style" />
+        <xsl:with-param name="data-value" select="@data-width" />
+        <xsl:with-param name="type" select="'width'" />
+      </xsl:call-template>
+    </xsl:attribute>
+    <xsl:attribute name="cy">
+      <xsl:call-template name="image-dimention">
+        <xsl:with-param name="style" select="@style" />
+        <xsl:with-param name="data-value" select="@data-height" />
+        <xsl:with-param name="type" select="'height'" />
+      </xsl:call-template>
+    </xsl:attribute>
+  </xsl:template>
+</xsl:stylesheet>

data/lib/htmltoword/xslt/inline_elements.xslt

@@ -11,29 +11,30 @@
   </xsl:template>
 
   <!-- get first inline element of a sequence or text having block element siblings... -->
-  <xsl:template match="node()[self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()][parent::div|parent::li|parent::td]">
+  <xsl:template match="node()[self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()][parent::div|parent::li|parent::td]">
     <div>
       <xsl:attribute name="class"><xsl:value-of select="../@class"/></xsl:attribute>
       <xsl:attribute name="style"><xsl:value-of select="../@style"/></xsl:attribute>
       <xsl:call-template name="inlineElement"/>
     </div>
-    <xsl:apply-templates select="following-sibling::node()[not((self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text())[parent::div|parent::li|parent::td])][1]"/>
+    <xsl:apply-templates select="following-sibling::node()[not((self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text())[parent::div|parent::li|parent::td])][1]"/>
   </xsl:template>
 
   <!-- get following inline elements... -->
-  <xsl:template match="a[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |b[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |em[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |i[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |small[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |span[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |strong[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |u[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]
-    |text()[preceding-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]]"
+  <xsl:template match="
+     a[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |b[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |i[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |s[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |span[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |sub[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |sup[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |u[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]
+    |text()[preceding-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]]"
     name="inlineElement">
     <xsl:copy>
       <xsl:apply-templates select="@*|node()[1]"/>
     </xsl:copy>
-    <xsl:apply-templates select="following-sibling::node()[1][self::a|self::b|self::em|self::i|self::small|self::span|self::strong|self::u|self::text()]"/>
+    <xsl:apply-templates select="following-sibling::node()[1][self::a|self::b|self::i|self::s|self::span|self::sub|self::sup|self::u|self::text()]"/>
   </xsl:template>
 </xsl:stylesheet>

data/lib/htmltoword/xslt/links.xslt

@@ -17,9 +17,10 @@
                 exclude-result-prefixes="java msxsl ext w o v WX aml w10"
                 extension-element-prefixes="func">
 
+
   <xsl:template match="a[starts-with(@href, 'http://') or starts-with(@href, 'https://')]" name="link">
     <w:hyperlink>
-      <xsl:attribute name="r:id">rId<xsl:value-of select="count(preceding::a[starts-with(@href, 'http://') or starts-with(@href, 'https://')]) + 8" /></xsl:attribute>
+      <xsl:attribute name="r:id"><xsl:call-template name="relationship-id"/></xsl:attribute>
       <w:r>
         <w:rPr>
           <w:rStyle w:val="Hyperlink"/>

data/lib/htmltoword/xslt/relations.xslt

@@ -3,13 +3,45 @@
                 version="1.0">
   <xsl:output method="xml" encoding="utf-8" omit-xml-declaration="yes" indent="yes" />
 
+  <xsl:include href="./image_functions.xslt"/>
+  <xsl:include href="./functions.xslt"/>
+
   <xsl:template match="a[starts-with(@href, 'http://') or starts-with(@href, 'https://')]" priority="1">
     <Relationship Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/hyperlink" Target="{@href}" TargetMode="External">
-      <xsl:attribute name="Id">rId<xsl:value-of select="count(preceding::a[starts-with(@href, 'http://') or starts-with(@href, 'https://')]) + 8"/></xsl:attribute>
+      <xsl:attribute name="Id"><xsl:call-template name="relationship-id"/></xsl:attribute>
+    </Relationship>
+  </xsl:template>
+
+  <!--
+    Images can either be embedded in the document or referenced via the external url.
+    By default images are embedded and a copy will be stored in the final zip folder.
+    If you want your images to be referenced externally (don't store the image files) then add  data-external="true" to your image tag(s).
+    If you are embedding them images need a filename. You can either define one by passing the name and extension to data-filename or rely on the defaul which is the last part of the source url.
+    **Word files can be corrupted if an extension isn't present so if you are relying on the source url ensure that is has an extension.**
+  -->
+  <xsl:template match="img" priority="0">
+    <Relationship Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/image">
+      <xsl:choose>
+        <xsl:when test="@data-external = 'true'">
+          <xsl:attribute name="Target"><xsl:value-of select="@src"/></xsl:attribute>
+          <xsl:attribute name="TargetMode">External</xsl:attribute>
+        </xsl:when>
+
+        <xsl:otherwise>
+          <xsl:attribute name="Target">media/image<xsl:value-of select="count(preceding::img)+1"/>.<xsl:call-template name="image-extension">
+              <xsl:with-param name="source" select="@src"/>
+              <xsl:with-param name="data-filename" select="@data-filename"/>
+            </xsl:call-template>
+          </xsl:attribute>
+        </xsl:otherwise>
+      </xsl:choose>
+
+      <xsl:attribute name="Id"><xsl:call-template name="relationship-id"/></xsl:attribute>
     </Relationship>
   </xsl:template>
 
   <xsl:template match="/">
+    <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
     <Relationships>
       <Relationship Id="rId3" Type="http://schemas.microsoft.com/office/2007/relationships/stylesWithEffects" Target="stylesWithEffects.xml"/>
       <Relationship Id="rId4" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/settings" Target="settings.xml"/>

data/lib/htmltoword/xslt/tables.xslt

@@ -152,6 +152,9 @@
         <xsl:variable name="cell-bg" select="str:tokenize(substring-after(@class, 'ms-fill-'), ' ')[1]"/>
         <w:shd w:val="clear" w:color="auto" w:fill="{$cell-bg}" />
       </xsl:if>
+      <xsl:if test="@colspan &gt; 1">
+        <w:gridSpan w:val="{@colspan}"/>
+      </xsl:if>
     </w:tcPr>
   </xsl:template>
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: htmltoword
 version: !ruby/object:Gem::Version
-  version: 0.5.1
+  version: 0.7.0
 platform: ruby
 authors:
 - Nicholas Frandsen, Cristina Matonte
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-11-24 00:00:00.000000000 Z
+date: 2017-11-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: actionpack
@@ -134,16 +134,21 @@ files:
 - Rakefile
 - bin/htmltoword
 - lib/htmltoword.rb
-- lib/htmltoword/action_controller.rb
 - lib/htmltoword/configuration.rb
 - lib/htmltoword/document.rb
-- lib/htmltoword/htmltoword_helper.rb
+- lib/htmltoword/helpers/templates_helper.rb
+- lib/htmltoword/helpers/xslt_helper.rb
+- lib/htmltoword/railtie.rb
+- lib/htmltoword/renderer.rb
 - lib/htmltoword/templates/default.docx
 - lib/htmltoword/version.rb
 - lib/htmltoword/xslt/base.xslt
+- lib/htmltoword/xslt/cleanup.xslt
 - lib/htmltoword/xslt/extras.xslt
 - lib/htmltoword/xslt/functions.xslt
 - lib/htmltoword/xslt/htmltoword.xslt
+- lib/htmltoword/xslt/image_functions.xslt
+- lib/htmltoword/xslt/images.xslt
 - lib/htmltoword/xslt/inline_elements.xslt
 - lib/htmltoword/xslt/links.xslt
 - lib/htmltoword/xslt/numbering.xslt
@@ -170,7 +175,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.8
+rubygems_version: 2.6.14
 signing_key: 
 specification_version: 4
 summary: This simple gem allows you to create MS Word docx documents from simple html

data/lib/htmltoword/action_controller.rb

@@ -1,54 +0,0 @@
-require 'rack/test'
-require 'action_controller'
-
-unless defined? Mime::DOCX
-  Mime::Type.register 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', :docx
-end
-
-ActionController::Renderers.add :docx do |filename, options|
-  formats << :docx unless formats.include?(:docx) || Rails.version < '3.2'
-
-  # This is ugly and should be solved with regular file utils
-  if options[:template] == action_name
-    if filename =~ %r{^([^\/]+)/(.+)$}
-      options[:prefixes] ||= []
-      options[:prefixes].unshift $1
-      options[:template] = $2
-    else
-      options[:template] = filename
-    end
-  end
-
-  # disposition / filename
-  disposition = options.delete(:disposition) || 'attachment'
-  if file_name = options.delete(:filename)
-    file_name += '.docx' unless file_name =~ /\.docx$/
-  else
-    file_name = "#{filename.gsub(/^.*\//, '')}.docx"
-  end
-
-  # other properties
-  word_template = options.delete(:word_template) || nil
-  extras = options.delete(:extras) || false
-  # content will come from property content unless not specified
-  # then it will look for a template.
-  content = options.delete(:content) || render_to_string(options)
-
-  document = Htmltoword::Document.create(content, word_template, extras)
-  send_data document, filename: file_name, type: Mime::DOCX, disposition: disposition
-end
-
-if defined? ActionController::Responder
-  module ActionController
-    # For respond_with default
-    class Responder
-      def to_docx
-        if @default_response
-          @default_response.call(options)
-        else
-          controller.render({ docx: controller.action_name }.merge(options))
-        end
-      end
-    end
-  end
-end

data/lib/htmltoword/htmltoword_helper.rb

@@ -1,35 +0,0 @@
-module Htmltoword
-  module HtmltowordHelper
-
-    def template_file template_file_name=nil
-      default_path = File.join(::Htmltoword.config.default_templates_path, 'default.docx')
-      template_path = template_file_name.present? ? File.join(::Htmltoword.config.custom_templates_path, template_file_name) : ''
-      File.exist?(template_path) ? template_path : default_path
-    end
-
-    def replace_values content, set
-      doc = Nokogiri::HTML(content)
-      set.each_pair do |key, value|
-        fields = (doc/"//span[@data-id='#{key}']")
-        fields.each do |f|
-          date_format = f.attr("date-format") || "long"
-          data_transform = f.attr("data-transform")
-          if value.is_a? Hash
-            view = ActionView::Base.new(ActionController::Base.view_paths, {})
-            final_value = view.render 'partials/answer_table', answer: value
-            fragment = doc.root.parse(final_value).first
-            new_node = doc.root.add_child(fragment)
-            f.parent.replace new_node
-          elsif value.is_a? Time
-            f.content = I18n.l(value.to_date, format: date_format.to_sym)
-          elsif data_transform == 'capitalized'
-            f.content = value.mb_chars.capitalize rescue value
-          else
-            f.content = value
-          end
-        end
-      end
-      doc.to_s
-    end
-  end
-end