reverse_adoc 0.2.5 → 0.2.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: de7821c8ca445117a045f3c2c28ae78f20a2185f61af5de3fff7ed064dd98c3e
-  data.tar.gz: 53151bcdd028e693da6b496df31c7a7da5ffa1dea8b0ee22696011a91fae074a
+  metadata.gz: abd0bf0dce3d437c176a1c3486d565fedd235d4b4e58cfdb4259188a1a94e090
+  data.tar.gz: '08926fcde9721aecc4672ac81c1dce18bf750763fe60621f57eb4cc32605eb70'
 SHA512:
-  metadata.gz: 4ae2e50d201e868dc7caa44a599695c6d03f155dd72013888d3cdee5505b074aa37cfc93cfc96435e43fa9bc0096ed46742dd2ff77280957c99cde1ca6454cee
-  data.tar.gz: 8a628ecfe850737831d5f1e10ef30ee65390f9599d7d75f7323de4ca39445c2d873fbaca49db7178b913d696a8fce85c50ce780a7e1715a37a12dce7c4e63cc4
+  metadata.gz: 39bf9ee3868bbb02440e2d853839a267f4c385e84204aedaeabe1000a7fa3fdc3f9e6baf381025548633b02b5c9e8f7d87925044c3ec992ca1eca2b83e7722b6
+  data.tar.gz: 5c34f9d71fa904a05c2e726b627d944f61a09f08e50ff587a102cc6e42c107600e08ad4b583fb5c9a07d678c09f06bc86fe25c15b94ae2dbefa0f29d046796bf

data/.github/workflows/macos.yml

@@ -1,27 +1,41 @@
-# Auto-generated !!! Do not edit it manually
-# use ci-master https://github.com/metanorma/metanorma-build-scripts
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
 name: macos
 
-on: [push]
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    paths-ignore:
+      - .github/workflows/ubuntu.yml
+      - .github/workflows/windows.yml
 
 jobs:
   test-macos:
     name: Test on Ruby ${{ matrix.ruby }} macOS
     runs-on: macos-latest
+    continue-on-error: ${{ matrix.experimental }}
     strategy:
+      fail-fast: false
       matrix:
         ruby: [ '2.6', '2.5', '2.4' ]
+        experimental: [false]
+        include:
+          - ruby: '2.7'
+            experimental: true
     steps:
       - uses: actions/checkout@master
       - name: Use Ruby
         uses: actions/setup-ruby@v1
         with:
           ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
       - name: Update gems
         run: |
-          sudo gem install bundler -v "~> 2" --force
+          sudo gem install bundler --force
           bundle install --jobs 4 --retry 3
+      - name: Install LibreOffice
+        run: |
+          brew cask install libreoffice
       - name: Run specs
         run: |
           bundle exec rake

data/.github/workflows/ubuntu.yml

@@ -1,26 +1,39 @@
-# Auto-generated !!! Do not edit it manually
-# use ci-master https://github.com/metanorma/metanorma-build-scripts
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
 name: ubuntu
 
-on: [push]
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    paths-ignore:
+      - .github/workflows/macos.yml
+      - .github/workflows/windows.yml
 
 jobs:
   test-linux:
     name: Test on Ruby ${{ matrix.ruby }} Ubuntu
     runs-on: ubuntu-latest
+    continue-on-error: ${{ matrix.experimental }}
     strategy:
+      fail-fast: false
       matrix:
         ruby: [ '2.6', '2.5', '2.4' ]
+        experimental: [false]
+        include:
+          - ruby: '2.7'
+            experimental: true
     steps:
       - uses: actions/checkout@master
       - name: Use Ruby
         uses: actions/setup-ruby@v1
         with:
           ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
+      - name: Install LibreOffice
+        run: sudo apt-get -y --no-install-recommends install libreoffice
       - name: Update gems
         run: |
-          gem install bundler -v "~> 2"
+          gem install bundler
           bundle install --jobs 4 --retry 3
       - name: Run specs
         run: |

data/.github/workflows/windows.yml

@@ -1,29 +1,43 @@
-# Auto-generated !!! Do not edit it manually
-# use ci-master https://github.com/metanorma/metanorma-build-scripts
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
 name: windows
 
-on: [push]
+on:
+  push:
+    branches: [ master ]
+  pull_request:
+    paths-ignore:
+      - .github/workflows/macos.yml
+      - .github/workflows/ubuntu.yml
 
 jobs:
   test-windows:
     name: Test on Ruby ${{ matrix.ruby }} Windows
     runs-on: windows-latest
+    continue-on-error: ${{ matrix.experimental }}
     strategy:
+      fail-fast: false
       matrix:
         ruby: [ '2.6', '2.5', '2.4' ]
+        experimental: [false]
+        include:
+          - ruby: '2.7'
+            experimental: true
     steps:
       - uses: actions/checkout@master
       - name: Use Ruby
         uses: actions/setup-ruby@v1
         with:
           ruby-version: ${{ matrix.ruby }}
-          architecture: 'x64'
+      - name: Install LibreOffice
+        run: |
+          choco install libreoffice-still
+          echo "::add-path::C:\Program Files\LibreOffice\program"
       - name: Update gems
         shell: pwsh
         run: |
-          gem install bundler -v "~> 2"
+          gem install bundler
           bundle config --local path vendor/bundle
-          bundle update
           bundle install --jobs 4 --retry 3
       - name: Run specs
         run: |

data/Gemfile

@@ -1,8 +1,8 @@
 source "https://rubygems.org"
-  
+
 git_source(:github) {|repo_name| "https://github.com/#{repo_name}" }
 
-# Specify your gem's dependencies in reverse_asciidoctor.gemspec
+# Specify your gem's dependencies in reverse_adoc.gemspec
 gemspec
 
 gem "byebug"

data/Gemfile.lock

@@ -0,0 +1,85 @@
+PATH
+  remote: .
+  specs:
+    reverse_adoc (0.2.10)
+      mathml2asciimath
+      mimemagic
+      nokogiri (>= 1.10.4)
+      premailer (~> 1.11.0)
+      word-to-markdown
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    addressable (2.7.0)
+      public_suffix (>= 2.0.2, < 5.0)
+    byebug (11.1.1)
+    cliver (0.3.2)
+    codeclimate-test-reporter (1.0.7)
+      simplecov
+    css_parser (1.7.1)
+      addressable
+    descriptive_statistics (2.5.1)
+    diff-lcs (1.3)
+    docile (1.3.2)
+    ffi (1.13.1)
+    htmlentities (4.3.4)
+    mathml2asciimath (0.0.10)
+      htmlentities (~> 4.3.4)
+      nokogiri (>= 1.10.4)
+    mimemagic (0.3.5)
+    mini_portile2 (2.4.0)
+    nokogiri (1.10.10)
+      mini_portile2 (~> 2.4.0)
+    nokogiri-styles (0.1.2)
+      nokogiri
+    premailer (1.11.1)
+      addressable
+      css_parser (>= 1.6.0)
+      htmlentities (>= 4.0.0)
+    public_suffix (4.0.6)
+    rake (13.0.1)
+    redcarpet (3.5.0)
+    reverse_markdown (1.4.0)
+      nokogiri
+    rspec (3.9.0)
+      rspec-core (~> 3.9.0)
+      rspec-expectations (~> 3.9.0)
+      rspec-mocks (~> 3.9.0)
+    rspec-core (3.9.1)
+      rspec-support (~> 3.9.1)
+    rspec-expectations (3.9.0)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.9.0)
+    rspec-mocks (3.9.1)
+      diff-lcs (>= 1.2.0, < 2.0)
+      rspec-support (~> 3.9.0)
+    rspec-support (3.9.2)
+    simplecov (0.18.5)
+      docile (~> 1.1)
+      simplecov-html (~> 0.11)
+    simplecov-html (0.12.1)
+    sys-proctable (1.2.6)
+      ffi
+    word-to-markdown (1.1.8)
+      cliver (~> 0.3)
+      descriptive_statistics (~> 2.5)
+      nokogiri-styles (~> 0.1)
+      premailer (~> 1.8)
+      reverse_markdown (~> 1.0)
+      sys-proctable (~> 1.0)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  byebug
+  codeclimate-test-reporter
+  rake
+  redcarpet
+  reverse_adoc!
+  rspec
+  simplecov
+
+BUNDLED WITH
+   2.0.2

data/README.adoc

@@ -1,11 +1,12 @@
 = AsciiDoc from HTML and Microsoft Word: reverse_adoc
 
-image:https://img.shields.io/gem/v/reverse_asciidoctor.svg["Gem Version", link="https://rubygems.org/gems/reverse_asciidoctor"]
-image:https://travis-ci.com/metanorma/reverse_asciidoctor.svg["Build Status", link="https://travis-ci.com/metanorma/reverse_asciidoctor"]
-image:https://ci.appveyor.com/api/projects/status/9dui2fs4pc590f4k?svg=true["Appveyor Build Status", link="https://ci.appveyor.com/project/metanorma/reverse-asciidoctor"]
-image:https://codeclimate.com/github/metanorma/reverse_asciidoctor/badges/gpa.svg["Code Climate", link="https://codeclimate.com/github/metanorma/reverse_asciidoctor"]
-image:https://img.shields.io/github/issues-pr-raw/metanorma/reverse_asciidoctor.svg["Pull Requests", link="https://github.com/metanorma/reverse_asciidoctor/pulls"]
-image:https://img.shields.io/github/commits-since/metanorma/reverse_asciidoctor/latest.svg["Commits since latest",link="https://github.com/metanorma/reverse_asciidoctor/releases"]
+https://github.com/metanorma/reverse_adoc[reverse_adoc] image:https://img.shields.io/gem/v/reverse_adoc.svg["Gem Version", link="https://rubygems.org/gems/reverse_adoc"]::
+image:https://github.com/metanorma/reverse_adoc/workflows/macos/badge.svg["Build Status", link="https://github.com/metanorma/reverse_adoc/actions?workflow=macos"]
+image:https://github.com/metanorma/reverse_adoc/workflows/windows/badge.svg["Build Status", link="https://github.com/metanorma/reverse_adoc/actions?workflow=windows"]
+image:https://github.com/metanorma/reverse_adoc/workflows/ubuntu/badge.svg["Build Status", link="https://github.com/metanorma/reverse_adoc/actions?workflow=ubuntu"]
+image:https://codeclimate.com/github/metanorma/reverse_adoc/badges/gpa.svg["Code Climate", link="https://codeclimate.com/github/metanorma/reverse_adoc"]
+image:https://img.shields.io/github/issues-pr-raw/metanorma/reverse_adoc.svg["Pull Requests", link="https://github.com/metanorma/reverse_adoc/pulls"]
+image:https://img.shields.io/github/commits-since/metanorma/reverse_adoc/latest.svg["Commits since latest",link="https://github.com/metanorma/reverse_adoc/releases"]
 
 == Purpose
 
@@ -43,20 +44,6 @@ $ reverse_adoc file.html > file.adoc
 $ cat file.html | reverse_adoc > file.adoc
 ----
 
-[source,console]
-----
-$ reverse_adoc -h
-Usage: reverse_adoc [options] <file>
-    -m, --mathml2asciimath           Convert MathML to AsciiMath
-    -o, --output=FILENAME            Output file to write to
-    -e, --external-images            Export images if data URI
-    -u [pass_through, drop, bypass, raise],
-        --unknown_tags               Unknown tag handling (default: pass_through)
-    -v, --version                    Version information
-    -h, --help                       Prints this help
-----
-
-
 
 === Microsoft Word to AsciiDoc: `w2a`
 
@@ -235,7 +222,7 @@ The gem does not support:
 === MathML support
 
 If you are using this gem in the context of https://www.metanorma.com[Metanorma],
-Metanorma Asciidoctor accepts MathML as a native mathematical format. So you do not need
+Metanorma AsciiDoc accepts MathML as a native mathematical format. So you do not need
 to convert the MathML to AsciiMath.
 
 The gem will optionally invoke the https://github.com/metanorma/mathml2asciimath
@@ -253,7 +240,7 @@ uses the output) as such; so this is still much better than nothing for stem pro
 
 === Word cleanup
 
-This gem is routinely used in the Metanorma project to export Word documents to Asciidoctor.
+This gem is routinely used in the Metanorma project to export Word documents to AsciiDoc.
 The HTML export from Word that the gem uses, from LibreOffice, is much cleaner than the
 native HTML 4 export from Word; but it has some infelicities which this gem cleans up:
 
@@ -272,7 +259,7 @@ Simple to use.
 
 [source,ruby]
 ----
-result = ReverseAsciidoctor.convert input
+result = ReverseAdoc.convert input
 result.inspect # " *feelings* "
 ----
 
@@ -282,7 +269,7 @@ Just pass your chosen configuration options in after the input. The given option
 
 [source,ruby]
 ----
-ReverseAsciidoctor.convert(input, unknown_tags: :raise, mathml2asciimath: true)
+ReverseAdoc.convert(input, unknown_tags: :raise, mathml2asciimath: true)
 ----
 
 
@@ -292,7 +279,7 @@ Or configure it block style on a initializer level. These configurations will la
 
 [source,ruby]
 ----
-ReverseAsciidoctor.config do |config|
+ReverseAdoc.config do |config|
   config.unknown_tags      = :bypass
   config.mathml2asciimath  = true
   config.tag_border  = ''

data/Rakefile

@@ -10,5 +10,5 @@ task :default => :spec
 
 desc 'Open an irb session preloaded with this library'
 task :console do
-  sh 'irb -rubygems -I lib -r reverse_asciidoctor.rb'
+  sh 'irb -rubygems -I lib -r reverse_adoc.rb'
 end

data/bin/reverse_adoc

@@ -1,31 +1,34 @@
 #!/usr/bin/env ruby
-# Usage: reverse_asciidoctor [FILE]...
-# Usage: cat FILE | reverse_asciidoctor
-require 'reverse_asciidoctor'
+# Usage: reverse_adoc [FILE]...
+# Usage: cat FILE | reverse_adoc
+require 'rubygems'
+require 'bundler/setup'
+
+require 'reverse_adoc'
 require 'optparse'
 require 'fileutils'
 
 OptionParser.new do |opts|
   opts.banner = "Usage: reverse_adoc [options] <file>"
   opts.on('-m', '--mathml2asciimath', 'Convert MathML to AsciiMath') do |v|
-    ReverseAsciidoctor.config.mathml2asciimath = true
+    ReverseAdoc.config.mathml2asciimath = true
   end
 
   opts.on('-oFILENAME', '--output=FILENAME', 'Output file to write to') do |v|
-    ReverseAsciidoctor.config.destination = File.expand_path(v)
-    # puts "output goes to #{ReverseAsciidoctor.config.destination}"
+    ReverseAdoc.config.destination = File.expand_path(v)
+    # puts "output goes to #{ReverseAdoc.config.destination}"
   end
 
   opts.on('-e', '--external-images', 'Export images if data URI') do |v|
-    ReverseAsciidoctor.config.external_images = true
+    ReverseAdoc.config.external_images = true
   end
 
   opts.on('-u', '--unknown_tags [pass_through, drop, bypass, raise]', 'Unknown tag handling (default: pass_through)') do |v|
-    ReverseAsciidoctor.config.unknown_tags = v
+    ReverseAdoc.config.unknown_tags = v
   end
 
   opts.on('-v', '--version', 'Version information') do |v|
-    puts "reverse_adoc: v#{ReverseAsciidoctor::VERSION}"
+    puts "reverse_adoc: v#{ReverseAdoc::VERSION}"
     exit
   end
 
@@ -38,30 +41,30 @@ end.parse!
 
 if filename = ARGV.pop
   input_content = IO.read(filename)
-  ReverseAsciidoctor.config.sourcedir = File.dirname(File.expand_path(filename))
+  ReverseAdoc.config.sourcedir = File.dirname(File.expand_path(filename))
 else
-  if ReverseAsciidoctor.config.external_images
+  if ReverseAdoc.config.external_images
     raise "The -e | --external-images feature cannot be used with STDIN input. Exiting."
   end
 
   input_content = ARGF.read
 end
 
-if ReverseAsciidoctor.config.external_images && ReverseAsciidoctor.config.destination.nil?
+if ReverseAdoc.config.external_images && ReverseAdoc.config.destination.nil?
   raise "The -e | --external-images feature must be used with -o | --output. Exiting."
 end
 
 # Read from STDIN
-adoc_content = ReverseAsciidoctor.convert(input_content)
+adoc_content = ReverseAdoc.convert(input_content)
 
 # Print to STDOUT
-unless ReverseAsciidoctor.config.destination
+unless ReverseAdoc.config.destination
   puts adoc_content
   exit
 end
 
-# Write output to ReverseAsciidoctor.config.destination
-FileUtils.mkdir_p(File.dirname(ReverseAsciidoctor.config.destination))
-File.open(ReverseAsciidoctor.config.destination, "w") do |file|
+# Write output to ReverseAdoc.config.destination
+FileUtils.mkdir_p(File.dirname(ReverseAdoc.config.destination))
+File.open(ReverseAdoc.config.destination, "w") do |file|
   file.write(adoc_content)
 end

data/bin/w2a

@@ -1,44 +1,32 @@
 #!/usr/bin/env ruby
 # frozen_string_literal: true
 
+require 'rubygems'
+require 'bundler/setup'
+
 require 'word-to-markdown'
 require 'optparse'
-require 'reverse_asciidoctor'
-
-def scrub_whitespace(string)
-  string = string.dup
-  string.gsub!(/ |\ |\u00a0/i, ' ')       # HTML encoded spaces
-  string.sub!(/^\A[[:space:]]+/m, '') # document leading whitespace
-  string.sub!(/[[:space:]]+\z$/m, '') # document trailing whitespace
-  string.gsub!(/([ ]+)$/, ' ')       # line trailing whitespace
-  string.gsub!(/\n\n\n\n/, "\n\n")  # Quadruple line breaks
-  #string.delete!(' ')               # Unicode non-breaking spaces, injected as tabs
-  # following added by me
-  string.gsub!(%r{<h[1-9][^>]*></h1>}, " ") # I don't know why Libre Office is inserting them, but they need to go
-  string.gsub!(%r{<h1[^>]* style="vertical-align: super;[^>]*>([^<]+)</h1>},
-              "<sup>\\1</sup>")         # I absolutely don't know why Libre Office is rendering superscripts as h1
-  string
-end
+require 'reverse_adoc'
 
 ARGV.push('-h') if ARGV.empty?
 
 OptionParser.new do |opts|
   opts.banner = "Usage: w2a [options] <file>"
   opts.on('-m', '--mathml2asciimath', 'Convert MathML to AsciiMath') do |v|
-    ReverseAsciidoctor.config.mathml2asciimath = true
+    ReverseAdoc.config.mathml2asciimath = true
   end
 
   opts.on('-oFILENAME', '--output=FILENAME', 'Output file to write to') do |v|
-    ReverseAsciidoctor.config.destination = File.expand_path(v)
-    # puts "output goes to #{ReverseAsciidoctor.config.destination}"
+    ReverseAdoc.config.destination = File.expand_path(v)
+    # puts "output goes to #{ReverseAdoc.config.destination}"
   end
 
   opts.on('-e', '--external-images', 'Export images if data URI') do |v|
-    ReverseAsciidoctor.config.external_images = true
+    ReverseAdoc.config.external_images = true
   end
 
   opts.on('-v', '--version', 'Version information') do |v|
-    puts "reverse_adoc: v#{ReverseAsciidoctor::VERSION}"
+    puts "reverse_adoc: v#{ReverseAdoc::VERSION}"
     puts "[dependency] WordToMarkdown: v#{WordToMarkdown::VERSION}"
     unless Gem.win_platform?
       puts "[dependency] LibreOffice: v#{WordToMarkdown.soffice.version}"
@@ -58,29 +46,28 @@ end.parse!
 filename = ARGV.pop
 raise "Please provide an input file to process. Exiting." unless filename
 
-if ReverseAsciidoctor.config.external_images && ReverseAsciidoctor.config.destination.nil?
+if ReverseAdoc.config.external_images && ReverseAdoc.config.destination.nil?
   raise "The -e | --external-images feature must be used with -o | --output. Exiting."
 end
 
-ReverseAsciidoctor.config.sourcedir = Dir.mktmpdir
-# puts "ReverseAsciidoctor.config.sourcedir #{ReverseAsciidoctor.config.sourcedir}"
+ReverseAdoc.config.sourcedir = Dir.mktmpdir
 
-doc = WordToMarkdown.new(filename, ReverseAsciidoctor.config.sourcedir)
+doc = WordToMarkdown.new(filename, ReverseAdoc.config.sourcedir)
 #File.open("test.html", "w:UTF-8") { |f| f.write doc.document.html }
-adoc_content = ReverseAsciidoctor.convert(
-  scrub_whitespace(doc.document.html),
+adoc_content = ReverseAdoc.convert(
+  ReverseAdoc.cleaner.preprocess_word_html(doc.document.html),
   WordToMarkdown::REVERSE_MARKDOWN_OPTIONS
 )
 # puts scrub_whitespace(doc.document.html)
 
 # Print to STDOUT
-unless ReverseAsciidoctor.config.destination
+unless ReverseAdoc.config.destination
   puts adoc_content
   exit
 end
 
-# Write output to ReverseAsciidoctor.config.destination
-FileUtils.mkdir_p(File.dirname(ReverseAsciidoctor.config.destination))
-File.open(ReverseAsciidoctor.config.destination, "w") do |file|
+# Write output to ReverseAdoc.config.destination
+FileUtils.mkdir_p(File.dirname(ReverseAdoc.config.destination))
+File.open(ReverseAdoc.config.destination, "w") do |file|
   file.write(adoc_content)
 end