reverse_adoc 0.2.4 → 0.2.9

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/li.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Li < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/mark.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Mark < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/math.rb

@@ -1,15 +1,15 @@
-# Unless run with ReverseAsciidoctor.config.mathml2asciimath,
+# Unless run with ReverseAdoc.config.mathml2asciimath,
 # this is cheating: we're injecting MathML into Asciidoctor, but
 # Asciidoctor only understands AsciiMath or LaTeX
 
 require "mathml2asciimath"
 
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Math < Base
       def convert(node, state = {})
         stem = node.to_s.gsub(/\n/, " ")
-        stem = MathML2AsciiMath.m2a(stem) if ReverseAsciidoctor.config.mathml2asciimath
+        stem = MathML2AsciiMath.m2a(stem) if ReverseAdoc.config.mathml2asciimath
         stem = stem.gsub(/\[/, "\\[").gsub(/\]/, "\\]").gsub(/\(\(([^\)]+)\)\)/, "(\\1)") unless stem.nil?
         " stem:[" << stem << "] "
       end

data/lib/reverse_adoc/converters/note.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+module ReverseAdoc
+  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 → reverse_adoc}/converters/ol.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Ol < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/p.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class P < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/pass_through.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class PassThrough < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/pre.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Pre < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/q.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Q < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/strong.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Strong < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/sub.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Sub < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/sup.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Sup < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/table.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Table < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/td.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Td < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/text.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Text < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/th.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Th < Td
       def cellstyle(node)

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/tr.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Tr < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/converters/video.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     class Video < Base
       def convert(node, state = {})

data/lib/{reverse_asciidoctor → reverse_adoc}/errors.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   class Error < StandardError
   end
 

data/lib/reverse_adoc/html_converter.rb

@@ -0,0 +1,56 @@
+# frozen_string_literal: true
+
+require 'reverse_adoc/converters/a'
+require 'reverse_adoc/converters/aside'
+require 'reverse_adoc/converters/audio'
+require 'reverse_adoc/converters/blockquote'
+require 'reverse_adoc/converters/br'
+require 'reverse_adoc/converters/bypass'
+require 'reverse_adoc/converters/code'
+require 'reverse_adoc/converters/div'
+require 'reverse_adoc/converters/drop'
+require 'reverse_adoc/converters/em'
+require 'reverse_adoc/converters/figure'
+require 'reverse_adoc/converters/h'
+require 'reverse_adoc/converters/head'
+require 'reverse_adoc/converters/hr'
+require 'reverse_adoc/converters/ignore'
+require 'reverse_adoc/converters/img'
+require 'reverse_adoc/converters/mark'
+require 'reverse_adoc/converters/li'
+require 'reverse_adoc/converters/ol'
+require 'reverse_adoc/converters/p'
+require 'reverse_adoc/converters/pass_through'
+require 'reverse_adoc/converters/pre'
+require 'reverse_adoc/converters/q'
+require 'reverse_adoc/converters/strong'
+require 'reverse_adoc/converters/sup'
+require 'reverse_adoc/converters/sub'
+require 'reverse_adoc/converters/table'
+require 'reverse_adoc/converters/td'
+require 'reverse_adoc/converters/th'
+require 'reverse_adoc/converters/text'
+require 'reverse_adoc/converters/tr'
+require 'reverse_adoc/converters/video'
+require 'reverse_adoc/converters/math'
+
+module ReverseAdoc
+  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 '')
+
+      ReverseAdoc.config.with(options) do
+        result = ReverseAdoc::Converters.lookup(root.name).convert(root)
+        ReverseAdoc.cleaner.tidy(result)
+      end
+    end
+  end
+end

data/lib/reverse_adoc/smrl_description_converter.rb

@@ -0,0 +1,45 @@
+# frozen_string_literal: true
+
+require 'reverse_adoc/converters/a'
+require 'reverse_adoc/converters/blockquote'
+require 'reverse_adoc/converters/bypass'
+require 'reverse_adoc/converters/br'
+require 'reverse_adoc/converters/code'
+require 'reverse_adoc/converters/drop'
+require 'reverse_adoc/converters/em'
+require 'reverse_adoc/converters/example'
+require 'reverse_adoc/converters/ext_descriptions'
+require 'reverse_adoc/converters/ext_description'
+require 'reverse_adoc/converters/express_ref'
+require 'reverse_adoc/converters/head'
+require 'reverse_adoc/converters/hr'
+require 'reverse_adoc/converters/ignore'
+require 'reverse_adoc/converters/note'
+require 'reverse_adoc/converters/p'
+require 'reverse_adoc/converters/pass_through'
+require 'reverse_adoc/converters/q'
+require 'reverse_adoc/converters/strong'
+require 'reverse_adoc/converters/sup'
+require 'reverse_adoc/converters/sub'
+require 'reverse_adoc/converters/text'
+
+module ReverseAdoc
+  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 '')
+
+      ReverseAdoc.config.with(options) do
+        result = ReverseAdoc::Converters.lookup(root.name).convert(root)
+        ReverseAdoc.cleaner.tidy(result)
+      end
+    end
+  end
+end

data/lib/reverse_adoc/version.rb

@@ -0,0 +1,3 @@
+module ReverseAdoc
+  VERSION = '0.2.9'
+end

data/reverse_adoc.gemspec

@@ -1,10 +1,10 @@
 # -*- encoding: utf-8 -*-
 $:.push File.expand_path("../lib", __FILE__)
-require "reverse_asciidoctor/version"
+require "reverse_adoc/version"
 
 Gem::Specification.new do |s|
   s.name        = "reverse_adoc"
-  s.version     = ReverseAsciidoctor::VERSION
+  s.version     = ReverseAdoc::VERSION
   s.authors       = ["Ribose Inc."]
   s.email         = ["open.source@ribose.com"]
 
@@ -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

data/spec/components/anchors_spec.rb

@@ -1,10 +1,10 @@
 require 'spec_helper'
 
-describe ReverseAsciidoctor do
+describe ReverseAdoc do
 
   let(:input)    { File.read('spec/assets/anchors.html') }
   let(:document) { Nokogiri::HTML(input) }
-  subject { ReverseAsciidoctor.convert(input) }
+  subject { ReverseAdoc.convert(input) }
 
   it { is_expected.to include 'http://foobar.com[Foobar]' }
   it { is_expected.to include 'http://foobar.com[Fubar]' }
@@ -18,4 +18,9 @@ describe ReverseAsciidoctor do
   it { is_expected.to include ' link <<content2>>internal jumplinks without anchors ' }
   it { is_expected.to include ' treat [[content]] as bookmarks ' }
 
+  it { is_expected.to include "<<a_bspaced,Double \\_\\_ anchor with space>>" }
+  it { is_expected.to include "[[a_bspaced]]" }
+  it { is_expected.to include "[[a_Foreword]]\n== Text" }
+  it { is_expected.not_to include "[[_Toc12345]]" }
+
 end

data/spec/components/basic_spec.rb

@@ -1,10 +1,10 @@
 require 'spec_helper'
 
-describe ReverseAsciidoctor do
+describe ReverseAdoc do
 
   let(:input)    { File.read('spec/assets/basic.html') }
   let(:document) { Nokogiri::HTML(input) }
-  subject { ReverseAsciidoctor.convert(input) }
+  subject { ReverseAdoc.convert(input) }
 
   it { is_expected.to match /plain text ?\n/ }
   it { is_expected.to match /\n== h1\n/ }