reverse_adoc 0.2.3 → 0.2.8

data/lib/reverse_asciidoctor/config.rb

@@ -3,10 +3,11 @@ require 'tmpdir'
 module ReverseAsciidoctor
   class Config
     attr_accessor :unknown_tags, :tag_border, :mathml2asciimath, :external_images,
-      :destination, :sourcedir, :image_counter, :image_counter_pattern
+      :destination, :sourcedir, :image_counter, :image_counter_pattern, :input_format
 
     def initialize
       @unknown_tags     = :pass_through
+      @input_format     = :html
       @mathml2asciimath = false
       @external_images  = false
 

data/lib/reverse_asciidoctor/converters/a.rb

@@ -9,10 +9,14 @@ module ReverseAsciidoctor
         title = extract_title(node)
         id = node['id'] || node['name']
 
-        if !id.nil? && !id.empty?
+        id = id&.gsub(/\s/, "")&.gsub(/__+/, "_")
+
+        if /^_Toc\d+$|^_GoBack$/.match id
+          ""
+        elsif !id.nil? && !id.empty?
           "[[#{id}]]"
         elsif href.to_s.start_with?('#')
-            href = href.sub(/^#/, "")
+          href = href.sub(/^#/, "").gsub(/\s/, "").gsub(/__+/, "_")
           if name.empty?
             "<<#{href}>>"
           else

data/lib/reverse_asciidoctor/converters/example.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+module ReverseAsciidoctor
+  module Converters
+    class Example < Base
+      def convert(node, state = {})
+        <<~TEMPLATE
+          [example]
+          ====
+          #{treat_children(node, state)}
+          ====
+        TEMPLATE
+      end
+    end
+    register :example, Example.new
+  end
+end

data/lib/reverse_asciidoctor/converters/express_ref.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module ReverseAsciidoctor
+  module Converters
+    class ExpressRef < Base
+      def convert(node, _state = {})
+        "express_ref:[#{node['linkend']}]"
+      end
+    end
+    register :express_ref, ExpressRef.new
+  end
+end

data/lib/reverse_asciidoctor/converters/ext_description.rb

@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module ReverseAsciidoctor
+  module Converters
+    class ExtDescription < Base
+      def convert(node, state = {})
+        <<~TEMPLATE
+          (*"#{node['linkend']}"
+          #{treat_children(node, state)}
+          *)
+        TEMPLATE
+      end
+    end
+    register :ext_description, ExtDescription.new
+  end
+end

data/lib/reverse_asciidoctor/converters/ext_descriptions.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module ReverseAsciidoctor
+  module Converters
+    class ExtDescriptions < Base
+      def convert(node, state = {})
+        treat_children(node, state)
+      end
+    end
+    register :ext_descriptions, ExtDescriptions.new
+  end
+end

data/lib/reverse_asciidoctor/converters/h.rb

@@ -3,9 +3,24 @@ module ReverseAsciidoctor
     class H < Base
       def convert(node, state = {})
         id = node['id']
-        anchor = id ? "[[#{id}]]\n" : ""
+        anchor = id ? "[[#{id}]]" : ""
+        internal_anchor = treat_children_anchors(node, state) || ""
+        anchor.empty? and anchor = internal_anchor
+        anchor.empty? or anchor += "\n"
         prefix = '=' * (node.name[/\d/].to_i + 1)
-        ["\n", anchor, prefix, ' ', treat_children(node, state), "\n"].join
+        ["\n", anchor, prefix, ' ', treat_children_no_anchors(node, state), "\n"].join
+      end
+
+      def treat_children_no_anchors(node, state)
+        node.children.reject { |a| a.name == "a" }.inject('') do |memo, child|
+          memo << treat(child, state)
+        end
+      end
+
+      def treat_children_anchors(node, state)
+        node.children.select { |a| a.name == "a" }.inject('') do |memo, child|
+          memo << treat(child, state)
+        end
       end
     end
 

data/lib/reverse_asciidoctor/converters/img.rb

@@ -1,4 +1,8 @@
-require 'fileutils'
+require "fileutils"
+require "pathname"
+require "tempfile"
+require "base64"
+require "mimemagic"
 
 module ReverseAsciidoctor
   module Converters
@@ -22,25 +26,7 @@ module ReverseAsciidoctor
         images_dir = dest_dir + 'images'
         FileUtils.mkdir_p(images_dir)
 
-        ext = ""
-
-        if imgdata
-          file = Tempfile.new(["radoc", ".jpg"]) do |f|
-            f.binmode
-            f.write(Base64.strict_decode64(imgdata))
-            f.rewind
-            ext = MimeMagic.by_magic(f)
-          end
-
-          image_src_path = file.path
-          # puts "tempfile: #{file}"
-
-        else
-          ext = File.extname(src).strip.downcase[1..-1]
-          image_src_path = Pathname.new(ReverseAsciidoctor.config.sourcedir) + src
-
-        end
-
+        ext, image_src_path = determine_image_src_path(src, imgdata)
         image_dest_path = images_dir + "#{image_number}.#{ext}"
 
         # puts "image_dest_path: #{image_dest_path.to_s}"
@@ -52,6 +38,23 @@ module ReverseAsciidoctor
         image_dest_path.relative_path_from(dest_dir)
       end
 
+      def determine_image_src_path(src, imgdata)
+        return copy_temp_file(imgdata) if imgdata
+
+        ext = File.extname(src).strip.downcase[1..-1]
+        [ext, Pathname.new(ReverseAsciidoctor.config.sourcedir) + src]
+      end
+
+      def copy_temp_file(imgdata)
+        Tempfile.open(['radoc', '.jpg']) do |f|
+          f.binmode
+          f.write(Base64.strict_decode64(imgdata))
+          f.rewind
+          ext = MimeMagic.by_magic(f).subtype
+          [ext, f.path]
+        end
+      end
+
       def convert(node, state = {})
         alt   = node['alt']
         src   = node['src']

data/lib/reverse_asciidoctor/converters/note.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+module ReverseAsciidoctor
+  module Converters
+    class Note < Base
+      def convert(node, state = {})
+        <<~TEMPLATE
+          [NOTE]
+          --
+          #{treat_children(node, state)}
+          --
+        TEMPLATE
+      end
+    end
+    register :note, Note.new
+  end
+end

data/lib/reverse_asciidoctor/html_converter.rb

@@ -0,0 +1,56 @@
+# frozen_string_literal: true
+
+require 'reverse_asciidoctor/converters/a'
+require 'reverse_asciidoctor/converters/aside'
+require 'reverse_asciidoctor/converters/audio'
+require 'reverse_asciidoctor/converters/blockquote'
+require 'reverse_asciidoctor/converters/br'
+require 'reverse_asciidoctor/converters/bypass'
+require 'reverse_asciidoctor/converters/code'
+require 'reverse_asciidoctor/converters/div'
+require 'reverse_asciidoctor/converters/drop'
+require 'reverse_asciidoctor/converters/em'
+require 'reverse_asciidoctor/converters/figure'
+require 'reverse_asciidoctor/converters/h'
+require 'reverse_asciidoctor/converters/head'
+require 'reverse_asciidoctor/converters/hr'
+require 'reverse_asciidoctor/converters/ignore'
+require 'reverse_asciidoctor/converters/img'
+require 'reverse_asciidoctor/converters/mark'
+require 'reverse_asciidoctor/converters/li'
+require 'reverse_asciidoctor/converters/ol'
+require 'reverse_asciidoctor/converters/p'
+require 'reverse_asciidoctor/converters/pass_through'
+require 'reverse_asciidoctor/converters/pre'
+require 'reverse_asciidoctor/converters/q'
+require 'reverse_asciidoctor/converters/strong'
+require 'reverse_asciidoctor/converters/sup'
+require 'reverse_asciidoctor/converters/sub'
+require 'reverse_asciidoctor/converters/table'
+require 'reverse_asciidoctor/converters/td'
+require 'reverse_asciidoctor/converters/th'
+require 'reverse_asciidoctor/converters/text'
+require 'reverse_asciidoctor/converters/tr'
+require 'reverse_asciidoctor/converters/video'
+require 'reverse_asciidoctor/converters/math'
+
+module ReverseAsciidoctor
+  class HtmlConverter
+    def self.convert(input, options = {})
+      root = if input.is_a?(String)
+              then Nokogiri::HTML(input).root
+            elsif input.is_a?(Nokogiri::XML::Document)
+              then input.root
+            elsif input.is_a?(Nokogiri::XML::Node)
+              then input
+            end
+
+      root || (return '')
+
+      ReverseAsciidoctor.config.with(options) do
+        result = ReverseAsciidoctor::Converters.lookup(root.name).convert(root)
+        ReverseAsciidoctor.cleaner.tidy(result)
+      end
+    end
+  end
+end

data/lib/reverse_asciidoctor/smrl_description_converter.rb

@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+require 'reverse_asciidoctor/converters/a'
+require 'reverse_asciidoctor/converters/blockquote'
+require 'reverse_asciidoctor/converters/bypass'
+require 'reverse_asciidoctor/converters/br'
+require 'reverse_asciidoctor/converters/code'
+require 'reverse_asciidoctor/converters/drop'
+require 'reverse_asciidoctor/converters/em'
+require 'reverse_asciidoctor/converters/example'
+require 'reverse_asciidoctor/converters/ext_descriptions'
+require 'reverse_asciidoctor/converters/ext_description'
+require 'reverse_asciidoctor/converters/express_ref'
+require 'reverse_asciidoctor/converters/head'
+require 'reverse_asciidoctor/converters/hr'
+require 'reverse_asciidoctor/converters/ignore'
+require 'reverse_asciidoctor/converters/note'
+require 'reverse_asciidoctor/converters/p'
+require 'reverse_asciidoctor/converters/pass_through'
+require 'reverse_asciidoctor/converters/q'
+require 'reverse_asciidoctor/converters/strong'
+require 'reverse_asciidoctor/converters/sup'
+require 'reverse_asciidoctor/converters/sub'
+require 'reverse_asciidoctor/converters/text'
+
+module ReverseAsciidoctor
+  class SmrlDescriptionConverter
+    def self.convert(input, options = {})
+      root = if input.is_a?(String)
+                then Nokogiri::XML(input).root
+              elsif input.is_a?(Nokogiri::XML::Document)
+                then input.root
+              elsif input.is_a?(Nokogiri::XML::Node)
+                then input
+              end
+
+      root || (return '')
+
+      ReverseAsciidoctor.config.with(options) do
+        result = ReverseAsciidoctor::Converters.lookup(root.name).convert(root)
+        ReverseAsciidoctor.cleaner.tidy(result)
+      end
+    end
+  end
+end

data/lib/reverse_asciidoctor/version.rb

@@ -1,3 +1,3 @@
 module ReverseAsciidoctor
-  VERSION = '0.2.3'
+  VERSION = '0.2.8'
 end

data/reverse_adoc.gemspec

@@ -22,6 +22,7 @@ Gem::Specification.new do |s|
   # specify any dependencies here; for example:
   s.add_dependency 'nokogiri', ">= 1.10.4"
   s.add_dependency 'mathml2asciimath'
+  s.add_dependency 'mimemagic'
   s.add_development_dependency 'rspec'
   s.add_development_dependency 'simplecov'
   s.add_development_dependency 'rake'

data/spec/assets/anchors.html

@@ -14,9 +14,17 @@
 
     some text...
 
+    <a href="#a__b spaced">Double __ anchor with space</a>
+
+    Double __ anchor with space <a id="a__b spaced"/>
+
+    <a id="_Toc12345"/>
+
     <img src="http://foobar.com/logo.png">
     <img alt="foobar image" src="http://foobar.com/foobar.png">
     <img alt="foobar image 2" title="this is the foobar image 2" src="http://foobar.com/foobar2.png">
     some text...
+
+    <h1><a id="a__Foreword"></a>Text</h1>
   </body>
 </html>

data/spec/assets/external_images.html

@@ -0,0 +1,35 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1 plus MathML 2.0//EN" "http://www.w3.org/Math/DTD/mathml2/xhtml-math11-f.dtd">
+<?xml version="1.0" encoding="UTF-8"??><html xmlns="http://www.w3.org/1999/xhtml" style="margin: 0;">
+<!--This file was converted to xhtml by LibreOffice - see http://cgit.freedesktop.org/libreoffice/core/tree/filter/source/xslt for the code.--><head profile="http://dublincore.org/documents/dcmi-terms/">
+<meta http-equiv="Content-Type" content="application/xhtml+xml; charset=utf-8">
+<meta name="DCTERMS.title" content="" xml:lang="en-US">
+<meta name="DCTERMS.language" content="en-US" scheme="DCTERMS.RFC4646">
+<meta name="DCTERMS.source" content="http://xml.openoffice.org/odf2xhtml">
+<meta name="DCTERMS.creator" content="Nick Nicholas">
+<meta name="DCTERMS.issued" content="2019-11-21T08:48:00" scheme="DCTERMS.W3CDTF">
+<meta name="DCTERMS.contributor" content="Nick Nicholas">
+<meta name="DCTERMS.modified" content="2019-11-21T10:01:00" scheme="DCTERMS.W3CDTF">
+<meta name="DCTERMS.provenance" content="" xml:lang="en-US">
+<meta name="DCTERMS.subject" content="," xml:lang="en-US">
+<link rel="schema.DC" href="http://purl.org/dc/elements/1.1/" hreflang="en">
+<link rel="schema.DCTERMS" href="http://purl.org/dc/terms/" hreflang="en">
+<link rel="schema.DCTYPE" href="http://purl.org/dc/dcmitype/" hreflang="en">
+<link rel="schema.DCAM" href="http://purl.org/dc/dcam/" hreflang="en">
+</head>
+<body dir="ltr" style="max-width: 21.001cm; margin: 2.54cm 3.175cm;">
+<h1 class="P3" style="clear: both; color: #2f5496; font-size: 16pt; font-family: Calibri Light; writing-mode: lr-tb; margin: 0.423cm 0 0cm;" align="left ! important">
+<a id="a__Hello" style="margin: 0;"><span style="margin: 0;"></span></a><span class="T1" style="margin: 0;">Hello</span>
+</h1>
+<p class="P1" style="font-size: 12pt; font-family: Calibri; writing-mode: lr-tb; margin: 0;" align="left ! important"> </p>
+<p class="Standard" style="font-size: 12pt; font-family: Calibri; writing-mode: lr-tb; margin: 0;" align="left ! important"><span class="T1" style="margin: 0;">H</span><h1 class="T3" style="vertical-align: super; font-size: 58%; margin: 0;">2</h1><span class="T1" style="margin: 0;">0</span></p>
+<p class="P1" style="font-size: 12pt; font-family: Calibri; writing-mode: lr-tb; margin: 0;" align="left ! important"> </p>
+<!--Next 'div' was a 'text:p'.--><div class="Standard" style="font-size: 12pt; font-family: Calibri; writing-mode: lr-tb; margin: 0;" align="left ! important">
+<!--Next ' 			span' is a draw:frame. 		--><span style="margin: 0;"><math xmlns="http://www.w3.org/1998/Math/MathML" display="block" style="margin: 0;"><mrow style="margin: 0;"><mrow style="margin: 0;"><mrow style="margin: 0;"></mrow><mrow style="margin: 0;"><mi style="margin: 0;">i</mi><mo stretchy="false" style="margin: 0;">=</mo><mn style="margin: 0;">1</mn></mrow></mrow><mrow style="margin: 0;"><mrow style="margin: 0;"></mrow><mi style="margin: 0;">n</mi><msubsup style="margin: 0;"><mi style="margin: 0;">β</mi><mn style="margin: 0;">2</mn><mi style="margin: 0;">i</mi></msubsup></mrow></mrow></math></span>
+</div>
+<div style="clear: both; line-height: 0; width: 0; height: 0; margin: 0; padding: 0;"> </div>
+<!--Next 'div' was a 'text:p'.--><div class="Standard" style="font-size: 12pt; font-family: Calibri; writing-mode: lr-tb; margin: 0;" align="left ! important">
+<a id="_GoBack" style="margin: 0;"></a><!--Next ' 			span' is a draw:frame. 		--><span style="height: 1.764cm; margin: 0cm; padding: 0; border: none; width: 1.764cm; font-size: 12pt; font-family: Calibri; text-align: center; vertical-align: top; background-color: transparent;" class="fr1" id="Εικόνα_2"><img style="height: 1.764cm; width: 1.764cm; margin: 0;" alt="" src="data:image/gif;base64,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"></span>
+</div>
+<div style="clear: both; line-height: 0; width: 0; height: 0; margin: 0; padding: 0;"> </div>
+</body>
+</html>

data/spec/bin/reverse_adoc_spec.rb

@@ -0,0 +1,32 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+describe 'bin/revrse_adoc' do
+  subject(:convert) do
+    ShellUtils.execute!("./bin/reverse_adoc -e -o test1 #{input_file_path}")
+  end
+  unless Gem.win_platform?
+    context 'when external images present' do
+      let(:input_file_path) { 'spec/assets/external_images.html' }
+      let(:images_folder) { 'images' }
+
+      after do
+        FileUtils.rm_rf(images_folder) if File.directory?(images_folder)
+      end
+
+      it 'Does not raise error' do
+        expect { convert }.to_not raise_error
+      end
+
+      it 'exatracts images from source html' do
+        expect { convert }
+          .to(change do
+            Dir["#{images_folder}/*gif"]
+              .map { |entry| File.basename(entry) }
+              .sort
+          end.from([]).to(['001.gif']))
+      end
+    end
+  end
+end

data/spec/bin/w2a_spec.rb

@@ -0,0 +1,35 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+describe 'bin/w2a' do
+  subject(:convert) do
+    ShellUtils.execute!("./bin/w2a -e -o test1 #{input_file_path}")
+  end
+  # TODO: fix github actions integration with libreoffice, currently it hangs
+  # when trying to use soffice binary
+  unless Gem.win_platform? ||
+         (Gem::Platform.local.os == 'darwin' && !ENV['GITHUB_ACTION'].nil?)
+    context 'when external images present' do
+      let(:input_file_path) { 'spec/assets/external_images.docx' }
+      let(:images_folder) { 'images' }
+
+      after do
+        FileUtils.rm_rf(images_folder) if File.directory?(images_folder)
+      end
+
+      it 'Does not raise error' do
+        expect { convert }.to_not raise_error
+      end
+
+      it 'exatracts images from source html' do
+        expect { convert }
+          .to(change do
+            Dir["#{images_folder}/*gif"]
+              .map { |entry| File.basename(entry) }
+              .sort
+          end.from([]).to(['001.gif', '002.gif']))
+      end
+    end
+  end
+end