reverse_adoc 0.2.8 → 0.3.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: eb34effa1c1af2e72ef8df7b6b1553c7c2574bcca9608e02cfaf561ed8a2bfec
-  data.tar.gz: e18b3b9cf2d936a48fdf2e6d5f265aa5c945c9493bd232a4d8fb311df780e81f
+  metadata.gz: c47e4d9eb1123a3b243cbb8f525eb3a4e5292d3cb721d09b1a0dc5a308521c24
+  data.tar.gz: 15602c27effe49fd669aa0114c655b30879d71fefe1019a5333962860e88375d
 SHA512:
-  metadata.gz: ca8454b96ad3064c16018f09198f8450f9a15e6525c1df6e63b7a0e3dd3254f66f2bee56e69e8c7088087962911480513f3608c759f955489b2e602dc2251890
-  data.tar.gz: 49c2198645f22b9ac26ea78aa36e1073a27807dca8364ad076250dccfc6b307055775894962d70d6400d7c8ec6b3dfa02cb31c84c922c60d8c18cbbf5ee55171
+  metadata.gz: ce264bc1decfc19683b4fc5c5f72d96a71be82bfb5ffbbe7f6c1e2428ca78eae48f6df66d473e712c925f04afa8fca887c858549dddf2fe0568dd953b2fe6a01
+  data.tar.gz: 460f744b26878bf13783393f8a0776dd74c15450072cc9b7654221da3b122c28457353aec3560a8f3bb5c88db70cf8568e84cafae63b29154539ba789f0d1045

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 +1,32 @@
+# 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: [ '3.0', '2.7', '2.6', '2.5', '2.4' ]
+        os: [ ubuntu-latest, windows-latest, macos-latest ]
+        experimental: [ false ]
+    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/.github/workflows/release.yml

@@ -0,0 +1,45 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
+name: release
+
+on:
+  workflow_dispatch:
+    inputs:
+      next_version:
+        description: |
+          Next release version. Possible values: x.y.z, major, minor, patch or pre|rc|etc
+        required: true
+        default: 'skip'
+  push:
+    tags: [ v* ]
+
+jobs:
+  release:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+
+      - uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: '2.6'
+          bundler-cache: true
+
+      - run: gem install gem-release
+
+      - if: ${{ github.event_name == 'workflow_dispatch' }} && ${{ github.event.inputs.next_version != 'skip' }}
+        run: |
+          git config user.name github-actions
+          git config user.email github-actions@github.com
+          gem bump --version ${{ github.event.inputs.next_version }} --tag --push
+
+      - name: publish to rubygems.org
+        env:
+          RUBYGEMS_API_KEY: ${{secrets.METANORMA_CI_RUBYGEMS_API_KEY}}
+        run: |
+          gem install gem-release
+          envsubst << 'EOF' > ~/.gem/credentials
+          ---
+          :rubygems_api_key: ${RUBYGEMS_API_KEY}
+          EOF
+          chmod 0600 ~/.gem/credentials
+          gem release

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/.hound.yml

@@ -1,3 +1,5 @@
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
 ruby:
-  Enabled: true
+  enabled: true
   config_file: .rubocop.yml

data/.rubocop.yml

@@ -1,10 +1,10 @@
-# This project follows the Ribose OSS style guide.
-# https://github.com/riboseinc/oss-guides
-# All project-specific additions and overrides should be specified in this file.
-
+# Auto-generated by Cimas: Do not edit it manually!
+# See https://github.com/metanorma/cimas
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/master/ci/rubocop.yml
+
+# local repo-specific modifications
+# ...
+
 AllCops:
-  TargetRubyVersion: 2.3
-Rails:
-  Enabled: true
+  TargetRubyVersion: 2.4

data/Gemfile

@@ -2,7 +2,7 @@ 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"]
@@ -44,31 +42,6 @@ $ reverse_adoc file.html > file.adoc
 $ cat file.html | reverse_adoc > file.adoc
 ----
 
-=== XML smrl description to AsciiDoc: `reverse_adoc`
-
-Convert XML smrl description files to AsciiDoc:
-
-[source,console]
-----
-$ reverse_adoc -f smrl_description descriptions.xml > file.adoc
-$ cat descriptions.xml | reverse_adoc -f smrl_description > file.adoc
-----
-
-[source,console]
-----
-$ reverse_adoc -h
-Usage: reverse_adoc [options] <file>
-    -f, --input_format               Supply input format to choose converter(html or smrl_description)
-    -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`
 
@@ -247,7 +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
@@ -265,7 +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:
 
@@ -284,7 +257,7 @@ Simple to use.
 
 [source,ruby]
 ----
-result = ReverseAsciidoctor.convert input
+result = ReverseAdoc.convert input
 result.inspect # " *feelings* "
 ----
 
@@ -294,7 +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)
 ----
 
 
@@ -304,7 +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/exe/reverse_adoc

@@ -0,0 +1,70 @@
+#!/usr/bin/env ruby
+# 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|
+    ReverseAdoc.config.mathml2asciimath = true
+  end
+
+  opts.on('-oFILENAME', '--output=FILENAME', 'Output file to write to') do |v|
+    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|
+    ReverseAdoc.config.external_images = true
+  end
+
+  opts.on('-u', '--unknown_tags [pass_through, drop, bypass, raise]', 'Unknown tag handling (default: pass_through)') do |v|
+    ReverseAdoc.config.unknown_tags = v
+  end
+
+  opts.on('-v', '--version', 'Version information') do |v|
+    puts "reverse_adoc: v#{ReverseAdoc::VERSION}"
+    exit
+  end
+
+  opts.on("-h", "--help", "Prints this help") do
+    puts opts
+    exit
+  end
+
+end.parse!
+
+if filename = ARGV.pop
+  input_content = IO.read(filename)
+  ReverseAdoc.config.sourcedir = File.dirname(File.expand_path(filename))
+else
+  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 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 = ReverseAdoc.convert(input_content)
+
+# Print to STDOUT
+unless ReverseAdoc.config.destination
+  puts adoc_content
+  exit
+end
+
+# 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

@@ -6,27 +6,27 @@ 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}"
@@ -46,28 +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
+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(
-  ReverseAsciidoctor.cleaner.preprocess_word_html(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

@@ -1,6 +1,5 @@
-module ReverseAsciidoctor
+module ReverseAdoc
   class Cleaner
-
     def tidy(string)
       result = remove_inner_whitespaces(string)
       result = remove_newlines(result)
@@ -14,7 +13,7 @@ module ReverseAsciidoctor
     end
 
     def remove_leading_newlines(string)
-      string.gsub(/\A\n+/, '')
+      string.gsub(/\A\n+/, "")
     end
 
     def remove_inner_whitespaces(string)
@@ -25,7 +24,7 @@ module ReverseAsciidoctor
       end
       string.each_line.inject("") do |memo, line|
         memo + preserve_border_whitespaces(line) do
-          line.strip.gsub(/[ \t]{2,}/, ' ')
+          line.strip.gsub(/[ \t]{2,}/, " ")
         end
       end
     end
@@ -36,65 +35,68 @@ 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
-          match.strip.sub('** ', '**').sub(' **', '**')
+        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
-          match.strip.sub('__ ', '__').sub(' __', '__')
+      result = result.gsub(/\s?_{2,}.*?_{2,}\s?/) do |match|
+        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
-          match.strip.sub('~~ ', '~~').sub(' ~~', '~~')
+        preserve_border_whitespaces(match, default_border: ReverseAdoc.config.tag_border) do
+          match.strip.sub("~~ ", "~~").sub(" ~~", "~~")
         end
       end
 
       result.gsub(/\s?\[.*?\]\s?/) do |match|
         preserve_border_whitespaces(match) do
-          match.strip.sub('[ ', '[').sub(' ]', ']')
+          match.strip.sub("[ ", "[").sub(" ]", "]")
         end
       end
     end
 
     def clean_punctuation_characters(string)
-      string.gsub(/(\*\*|~~|__)\s([\.!\?'"])/, "\\1".strip + "\\2")
+      string.gsub(/(\*\*|~~|__)\s([.!?'"])/, "\\1".strip + "\\2")
     end
 
-     # preprocesses HTML, rather than postprocessing it
+    # preprocesses HTML, rather than postprocessing it
     def preprocess_word_html(string)
       clean_headings(scrub_whitespace(string.dup))
     end
 
     def scrub_whitespace(string)
-      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
+      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
       string
     end
 
     # following added by me
     def clean_headings(string)
-      string.gsub!(%r{<h([1-9])[^>]*></h\1>}, " ") # I don't know why Libre Office is inserting them, but they need to go
+      string.gsub!(%r{<h([1-9])[^>]*></h\1>}, " ")
+      # I don't know why Libre Office is inserting them, but they need to go
       string.gsub!(%r{<h([1-9])[^>]* style="vertical-align: super;[^>]*>(.+?)</h\1>},
-                   "<sup>\\2</sup>")         # I absolutely don't know why Libre Office is rendering superscripts as h1
+                   "<sup>\\2</sup>")
+      # I absolutely don't know why Libre Office is rendering superscripts as h1
       string
     end
 
     private
 
-    def preserve_border_whitespaces(string, options = {}, &block)
-      return string if string =~ /\A\s*\Z/
-      default_border = options.fetch(:default_border, '')
+    def preserve_border_whitespaces(string, options = {})
+      return string if /\A\s*\Z/.match?(string)
+
+      default_border = options.fetch(:default_border, "")
       # If the string contains part of a link so the characters [,],(,)
       # then don't add any extra spaces
-      default_border = '' if string =~ /[\[\(\]\)]/
+      default_border = "" if /[\[(\])]/.match?(string)
       string_start   = present_or_default(string[/\A\s*/], default_border)
       string_end     = present_or_default(string[/\s*\Z/], default_border)
       result         = yield
@@ -108,6 +110,5 @@ module ReverseAsciidoctor
         string
       end
     end
-
   end
 end