reverse_adoc 0.2.7 → 0.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8f60fbebeee21569fdf9af0bb7153a06ac7f3a8d3431d143750e96185c701c7e
-  data.tar.gz: 72f47514bce560d201ac2b22c6b6034a02550f2e50f76200e96fa003628756f5
+  metadata.gz: 22407f5a4ca3839adba7a35be1c31492f2c051a2a9862b84ecb44f7992b1ece1
+  data.tar.gz: 55bbbb5169f22a48ed4cac0d5653a30ddf92163827d8e2cc7ba0699098ac0d70
 SHA512:
-  metadata.gz: 82babe7f90eb4445d2d2bebb16dcf5a1af869da69ac5fc4da0d14e24d456a00d80f390097b10b597dbc8ec69bf782dd04d1464ef0a73cd513b4f164267ca67e1
-  data.tar.gz: 38b3887266e6c666b117d85d8df810b8e3d168b1a03af71c06fcd1218a3b99af0228686896248074767a207b32bdc6e7d16a8826ba1b11680b4371648f956576
+  metadata.gz: e3860543e71ec3c4c372cdf9b96a51128bf1579e53e0d110879e39ccab90b4b08754f13fa28edfa96a24b4cbb2974a7aa65e23c8d6326e42e85fa7ff55b05063
+  data.tar.gz: 0e043787955792419c2bd521a1c2a96cf2c3a0f022ad00c2ff8e74107d5e89d3c745774c55857f5a857f394ef05c6c82123402e1d18320670f4c4cc38e6fe376

data/.editorconfig

@@ -0,0 +1,15 @@
+# EditorConfig is awesome: http://EditorConfig.org
+
+# top-most EditorConfig file
+root = true
+
+# Unix-style newlines with a newline ending every file
+[*]
+charset = utf-8
+end_of_line = lf
+
+[{*.adoc,*.html,*.js,*.json,*.rake,*.rb,*.rf,*.yaml,*.yml,Rakefile,rakefile}]
+indent_style = space
+indent_size = 2
+insert_final_newline = true
+trim_trailing_whitespace = true

data/.github/workflows/rake.yml

@@ -0,0 +1,42 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+name: rake
+
+on:
+  push:
+    branches: [ master, main ]
+    tags: [ v* ]
+  pull_request:
+
+jobs:
+  rake:
+    name: Test on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    continue-on-error: ${{ matrix.experimental }}
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: [ '2.7', '2.6', '2.5', '2.4' ]
+        os: [ ubuntu-latest, windows-latest, macos-latest ]
+        experimental: [ false ]
+        include:
+          - ruby: '3.0'
+            os: 'ubuntu-latest'
+            experimental: true
+          - ruby: '3.0'
+            os: 'windows-latest'
+            experimental: true
+          - ruby: '3.0'
+            os: 'macos-latest'
+            experimental: true
+    steps:
+      - uses: actions/checkout@master
+
+      - uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+
+      - uses: metanorma/metanorma-build-scripts/libreoffice-setup-action@master
+
+      - run: bundle exec rake

data/.gitignore

@@ -0,0 +1,10 @@
+/coverage
+/Gemfile.lock
+
+.DS_Store
+
+# RuboCop's remote style books
+.rubocop-https---*
+
+# Some odd files created by specs
+/test1

data/.rubocop.yml

@@ -5,6 +5,16 @@
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/master/ci/rubocop.yml
 AllCops:
-  TargetRubyVersion: 2.3
+  TargetRubyVersion: 2.4
 Rails:
-  Enabled: true
+  Enabled: false
+
+# Dependencies are ordered for purpose.
+Gemspec/OrderedDependencies:
+  Enabled: false
+
+# TODO Rather than disabling, change severity to "info" in Rubocop 1.9+.
+Layout:
+  Enabled: false
+Style/StringLiterals:
+  Enabled: false

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/README.adoc

@@ -1,9 +1,7 @@
 = AsciiDoc from HTML and Microsoft Word: reverse_adoc
 
 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://github.com/metanorma/reverse_adoc/workflows/rake/badge.svg["Build Status", link="https://github.com/metanorma/reverse_adoc/actions?workflow=rake"]
 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"]
@@ -44,20 +42,6 @@ $ 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`
 
@@ -236,7 +220,7 @@ 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
@@ -254,7 +238,7 @@ 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:
 
@@ -273,7 +257,7 @@ Simple to use.
 
 [source,ruby]
 ----
-result = ReverseAsciidoctor.convert input
+result = ReverseAdoc.convert input
 result.inspect # " *feelings* "
 ----
 
@@ -283,7 +267,7 @@ 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)
 ----
 
 
@@ -293,7 +277,7 @@ 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 → exe}/reverse_adoc

@@ -1,6 +1,6 @@
 #!/usr/bin/env ruby
-# Usage: reverse_asciidoctor [FILE]...
-# Usage: cat FILE | reverse_asciidoctor
+# Usage: reverse_adoc [FILE]...
+# Usage: cat FILE | reverse_adoc
 require 'rubygems'
 require 'bundler/setup'
 
@@ -11,24 +11,24 @@ 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
 
@@ -41,30 +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 → exe}/w2a

@@ -1,31 +1,32 @@
 #!/usr/bin/env ruby
 # frozen_string_literal: true
+
 require 'rubygems'
 require 'bundler/setup'
 
 require 'word-to-markdown'
 require 'optparse'
-require 'reverse_asciidoctor'
+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}"
@@ -45,29 +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)
-File.open("test.html", "w:UTF-8") { |f| f.write doc.document.html }
-adoc_content = ReverseAsciidoctor.convert(
-  ReverseAsciidoctor.cleaner.preprocess_word_html(doc.document.html),
+doc = WordToMarkdown.new(filename, ReverseAdoc.config.sourcedir)
+#File.open("test.html", "w:UTF-8") { |f| f.write 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

data/lib/reverse_adoc.rb

@@ -0,0 +1,27 @@
+# frozen_string_literal: true
+
+require 'digest'
+require 'nokogiri'
+require 'reverse_adoc/version'
+require 'reverse_adoc/errors'
+require 'reverse_adoc/cleaner'
+require 'reverse_adoc/config'
+require 'reverse_adoc/converters'
+require 'reverse_adoc/converters/base'
+require "reverse_adoc/html_converter"
+
+module ReverseAdoc
+  def self.convert(input, options = {})
+    ReverseAdoc::HtmlConverter.convert(input, options)
+  end
+
+  def self.config
+    @config ||= Config.new
+    yield @config if block_given?
+    @config
+  end
+
+  def self.cleaner
+    @cleaner ||= Cleaner.new
+  end
+end

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

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   class Cleaner
 
     def tidy(string)
@@ -36,19 +36,19 @@ module ReverseAsciidoctor
     # Same for underscores and brackets.
     def clean_tag_borders(string)
       result = string.gsub(/\s?\*{2,}.*?\*{2,}\s?/) do |match|
-        preserve_border_whitespaces(match, default_border: ReverseAsciidoctor.config.tag_border) do
+        preserve_border_whitespaces(match, default_border: ReverseAdoc.config.tag_border) do
           match.strip.sub('** ', '**').sub(' **', '**')
         end
       end
 
       result = result.gsub(/\s?\_{2,}.*?\_{2,}\s?/) do |match|
-        preserve_border_whitespaces(match, default_border: ReverseAsciidoctor.config.tag_border) do
+        preserve_border_whitespaces(match, default_border: ReverseAdoc.config.tag_border) do
           match.strip.sub('__ ', '__').sub(' __', '__')
         end
       end
 
       result = result.gsub(/\s?~{2,}.*?~{2,}\s?/) do |match|
-        preserve_border_whitespaces(match, default_border: ReverseAsciidoctor.config.tag_border) do
+        preserve_border_whitespaces(match, default_border: ReverseAdoc.config.tag_border) do
           match.strip.sub('~~ ', '~~').sub(' ~~', '~~')
         end
       end

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

@@ -1,12 +1,13 @@
 require 'tmpdir'
 
-module ReverseAsciidoctor
+module ReverseAdoc
   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 → reverse_adoc}/converters.rb

@@ -1,4 +1,4 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   module Converters
     def self.register(tag_name, converter)
       @@converters ||= {}
@@ -16,17 +16,17 @@ module ReverseAsciidoctor
     private
 
     def self.default_converter(tag_name)
-      case ReverseAsciidoctor.config.unknown_tags.to_sym
+      case ReverseAdoc.config.unknown_tags.to_sym
       when :pass_through
-        ReverseAsciidoctor::Converters::PassThrough.new
+        ReverseAdoc::Converters::PassThrough.new
       when :drop
-        ReverseAsciidoctor::Converters::Drop.new
+        ReverseAdoc::Converters::Drop.new
       when :bypass
-        ReverseAsciidoctor::Converters::Bypass.new
+        ReverseAdoc::Converters::Bypass.new
       when :raise
         raise UnknownTagError, "unknown tag: #{tag_name}"
       else
-        raise InvalidConfigurationError, "unknown value #{ReverseAsciidoctor.config.unknown_tags.inspect} for ReverseAsciidoctor.config.unknown_tags"
+        raise InvalidConfigurationError, "unknown value #{ReverseAdoc.config.unknown_tags.inspect} for ReverseAdoc.config.unknown_tags"
       end
     end
   end