From b37e67f664e55daec65cc08b7061c1e3abed4e3b Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 11:12:15 +0100
Subject: [PATCH 1/6] whitelist options accepted by pandoc

---
 Gemfile.lock             | 10 ++++------
 lib/pandoc-ruby.rb       | 26 ++++++++++++++++++++++----
 test/test_pandoc_ruby.rb | 17 -----------------
 3 files changed, 26 insertions(+), 27 deletions(-)

diff --git a/Gemfile.lock b/Gemfile.lock
index 5ed44a7..1c32749 100644
--- a/Gemfile.lock
+++ b/Gemfile.lock
@@ -1,14 +1,12 @@
 GEM
   remote: http://rubygems.org/
   specs:
-    json (1.8.3)
     metaclass (0.0.4)
-    minitest (5.8.4)
+    minitest (5.8.5)
     mocha (1.1.0)
       metaclass (~> 0.0.1)
-    rake (11.1.2)
-    rdoc (4.2.2)
-      json (~> 1.4)
+    rake (12.0.0)
+    rdoc (5.0.0)
 
 PLATFORMS
   ruby
@@ -20,4 +18,4 @@ DEPENDENCIES
   rdoc
 
 BUNDLED WITH
-   1.12.2
+   1.12.5
diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index c7c47f5..e8ee615 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -57,6 +57,26 @@ class PandocRuby
   # All of the available Writers.
   WRITERS = STRING_WRITERS.merge(BINARY_WRITERS)
 
+  # Options understood by pandoc, taken from http://pandoc.org/MANUAL.html.
+  # Ignore all other options passed to pandoc
+  AVAILABLE_OPTIONS = Set.new %w(from read to write output data-dir strict
+    parse-raw smart old-dashes base-header-level indented-code-classes filter
+    normalize preserve-tabs tab-stop track-changes extract-media standalone
+    template metadata variable print-default-template print-default-data-file
+    no-wrap columns toc table-of-contents toc-depth no-highlight
+    highlight-style include-in-header include-before-body include-after-body
+    self-contained offline html5 html-q-tags ascii reference-links atx-headers
+    chapters number-sections number-offsetS no-tex-ligatures listings
+    incremental slide-level section-divs default-image-extension
+    email-obfuscation id-prefix title-prefix css reference-odt reference-docx
+    epub-stylesheet epub-cover-image epub-metadata epub-embed-font
+    epub-chapter-level latex-engine latex-engine-opt bibliography csl
+    citation-abbreviations natbib biblatex latexmathml asciimathml mathml
+    mimetex webtex jsmath mathjax katex katex-stylesheet gladtex trace
+    dump-args ignore-args verbose bash-completion)
+  ALIAS_OPTIONS = Set.new %w(f r t w o R S F p s M V D H B A 5 N i T c m)
+  ALLOWED_OPTIONS = AVAILABLE_OPTIONS + ALIAS_OPTIONS
+
   # To use run the pandoc command with a custom executable path, the path
   # to the pandoc executable can be set here.
   def self.pandoc_path=(path)
@@ -243,11 +263,11 @@ def prepare_options(opts = [])
     # Takes a flag and optional argument, uses it to set any relevant options
     # used by the library, and returns string with the option formatted as a
     # command line options. If the option has an argument, it is also included.
+    # Only whitelisted options are sent to pandoc.
     def create_option(flag, argument = nil)
-      return '' unless flag
+      return '' unless flag && ALLOWED_OPTIONS.include?(flag.to_s.gsub('_', '-'))
       flag = flag.to_s
       set_pandoc_ruby_options(flag, argument)
-      return '' if flag == 'timeout' # pandoc doesn't accept timeouts yet
       if !argument.nil?
         "#{format_flag(flag)} #{argument}"
       else
@@ -272,8 +292,6 @@ def set_pandoc_ruby_options(flag, argument = nil)
       when 't', 'to'
         self.writer = argument.to_s
         self.binary_output = true if BINARY_WRITERS.keys.include?(self.writer)
-      when 'timeout'
-        @timeout = argument
       end
     end
 
diff --git a/test/test_pandoc_ruby.rb b/test/test_pandoc_ruby.rb
index 11f6e52..6d3c5e4 100644
--- a/test/test_pandoc_ruby.rb
+++ b/test/test_pandoc_ruby.rb
@@ -92,12 +92,6 @@
     assert converter.convert
   end
 
-  it 'raises RuntimeError from pandoc executable error' do
-    assert_raises(RuntimeError) do
-      PandocRuby.new('# hello', 'badopt').to_html5
-    end
-  end
-
   PandocRuby::READERS.each_key do |r|
     it "converts from #{r} with PandocRuby.#{r}" do
       converter = PandocRuby.send(r, @string)
@@ -152,17 +146,6 @@
     end
   end
 
-  it 'gracefully times out when pandoc hangs due to malformed input' do
-    file = File.join(File.dirname(__FILE__), 'files', 'bomb.tex')
-    contents = File.read(file)
-
-    assert_raises(RuntimeError) do
-      PandocRuby.convert(
-        contents, :from => :latex, :to => :html, :timeout => 1
-      )
-    end
-  end
-
   it 'has reader and writer constants' do
     assert_equal PandocRuby::READERS,
                  'html'      =>  'HTML',

From a2754007cf29a37d1f09045992bde66c7d794fb5 Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 12:03:13 +0100
Subject: [PATCH 2/6] added ignore_whitelist option

---
 lib/pandoc-ruby.rb       | 12 +++++++++++-
 test/test_pandoc_ruby.rb | 23 +++++++++++++++++++++++
 2 files changed, 34 insertions(+), 1 deletion(-)

diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index e8ee615..b5b513c 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -101,6 +101,11 @@ def options
     @options ||= []
   end
 
+  attr_writer :ignore_whitelist
+  def ignore_whitelist
+    @ignore_whitelist
+  end
+
   attr_writer :option_string
   def option_string
     @option_string ||= ''
@@ -130,6 +135,7 @@ def initialize(*args)
       self.input_files = args.shift.join(' ')
     end
     self.options = args
+    self.ignore_whitelist = self.options.find { |o| o == :ignore_whitelist }
   end
 
   # Run the conversion. The convert method can take any number of arguments,
@@ -265,9 +271,11 @@ def prepare_options(opts = [])
     # command line options. If the option has an argument, it is also included.
     # Only whitelisted options are sent to pandoc.
     def create_option(flag, argument = nil)
-      return '' unless flag && ALLOWED_OPTIONS.include?(flag.to_s.gsub('_', '-'))
+      return '' unless flag
       flag = flag.to_s
       set_pandoc_ruby_options(flag, argument)
+      return '' unless @ignore_whitelist ||
+                       ALLOWED_OPTIONS.include?(flag.gsub('_', '-'))
       if !argument.nil?
         "#{format_flag(flag)} #{argument}"
       else
@@ -292,6 +300,8 @@ def set_pandoc_ruby_options(flag, argument = nil)
       when 't', 'to'
         self.writer = argument.to_s
         self.binary_output = true if BINARY_WRITERS.keys.include?(self.writer)
+      when 'timeout'
+        @timeout = argument
       end
     end
 
diff --git a/test/test_pandoc_ruby.rb b/test/test_pandoc_ruby.rb
index 6d3c5e4..b5f41f8 100644
--- a/test/test_pandoc_ruby.rb
+++ b/test/test_pandoc_ruby.rb
@@ -92,6 +92,18 @@
     assert converter.convert
   end
 
+  it 'ignores options not whitelisted' do
+    converter = PandocRuby.new('# hello', 'badopt')
+    converter.expects(:execute).with('pandoc').returns(true)
+    assert converter.convert
+  end
+
+  it 'raises RuntimeError from pandoc executable error' do
+    assert_raises(RuntimeError) do
+      PandocRuby.new('# hello', 'badopt', :ignore_whitelist).to_html5
+    end
+  end
+
   PandocRuby::READERS.each_key do |r|
     it "converts from #{r} with PandocRuby.#{r}" do
       converter = PandocRuby.send(r, @string)
@@ -146,6 +158,17 @@
     end
   end
 
+  it 'gracefully times out when pandoc hangs due to malformed input' do
+    file = File.join(File.dirname(__FILE__), 'files', 'bomb.tex')
+    contents = File.read(file)
+
+    assert_raises(RuntimeError) do
+      PandocRuby.convert(
+        contents, :from => :latex, :to => :html, :timeout => 1
+      )
+    end
+  end
+
   it 'has reader and writer constants' do
     assert_equal PandocRuby::READERS,
                  'html'      =>  'HTML',

From 59c2ed49c30fb5be007ff94b6d0e0f6168c649ff Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 12:16:52 +0100
Subject: [PATCH 3/6] simplified :ignore_whitelist option

---
 lib/pandoc-ruby.rb | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index b5b513c..40c2c09 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -101,11 +101,6 @@ def options
     @options ||= []
   end
 
-  attr_writer :ignore_whitelist
-  def ignore_whitelist
-    @ignore_whitelist
-  end
-
   attr_writer :option_string
   def option_string
     @option_string ||= ''
@@ -135,7 +130,6 @@ def initialize(*args)
       self.input_files = args.shift.join(' ')
     end
     self.options = args
-    self.ignore_whitelist = self.options.find { |o| o == :ignore_whitelist }
   end
 
   # Run the conversion. The convert method can take any number of arguments,
@@ -274,8 +268,8 @@ def create_option(flag, argument = nil)
       return '' unless flag
       flag = flag.to_s
       set_pandoc_ruby_options(flag, argument)
-      return '' unless @ignore_whitelist ||
-                       ALLOWED_OPTIONS.include?(flag.gsub('_', '-'))
+      return '' unless ALLOWED_OPTIONS.include?(flag.gsub('_', '-')) ||
+                       @options.find { |o| o == :ignore_whitelist }
       if !argument.nil?
         "#{format_flag(flag)} #{argument}"
       else

From e73c8082f982bab355683f52033ecfdb55d4538c Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 12:22:06 +0100
Subject: [PATCH 4/6] updated documentation

---
 README.md          | 4 ++++
 lib/pandoc-ruby.rb | 4 ++--
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 487250b..36e88ff 100644
--- a/README.md
+++ b/README.md
@@ -121,6 +121,10 @@ the correct header and footer are added.
 PandocRuby.new("# Some title", :standalone).to_rtf
 ```
 
+My default only whitelisted options (those known to Pandoc) are passed on to
+Pandoc. You can send all options to Pandoc by using the `:ignore_whitelist`
+option.
+
 ## Note on Patches/Pull Requests
 
 * Fork the project.
diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index 40c2c09..dd08899 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -58,7 +58,7 @@ class PandocRuby
   WRITERS = STRING_WRITERS.merge(BINARY_WRITERS)
 
   # Options understood by pandoc, taken from http://pandoc.org/MANUAL.html.
-  # Ignore all other options passed to pandoc
+  # Ignore all other options passed to pandoc, unless overriden.
   AVAILABLE_OPTIONS = Set.new %w(from read to write output data-dir strict
     parse-raw smart old-dashes base-header-level indented-code-classes filter
     normalize preserve-tabs tab-stop track-changes extract-media standalone
@@ -263,7 +263,7 @@ def prepare_options(opts = [])
     # Takes a flag and optional argument, uses it to set any relevant options
     # used by the library, and returns string with the option formatted as a
     # command line options. If the option has an argument, it is also included.
-    # Only whitelisted options are sent to pandoc.
+    # Only whitelisted options are sent to pandoc, unless overridden.
     def create_option(flag, argument = nil)
       return '' unless flag
       flag = flag.to_s

From d8f9a19b965c5cb06cb2433a3cc5185f9bcfde6d Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 13:37:25 +0100
Subject: [PATCH 5/6] added pandoc-citeproc options

---
 lib/pandoc-ruby.rb | 32 ++++++++++++++++++--------------
 1 file changed, 18 insertions(+), 14 deletions(-)

diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index dd08899..4771329 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -61,21 +61,25 @@ class PandocRuby
   # Ignore all other options passed to pandoc, unless overriden.
   AVAILABLE_OPTIONS = Set.new %w(from read to write output data-dir strict
     parse-raw smart old-dashes base-header-level indented-code-classes filter
-    normalize preserve-tabs tab-stop track-changes extract-media standalone
-    template metadata variable print-default-template print-default-data-file
-    no-wrap columns toc table-of-contents toc-depth no-highlight
-    highlight-style include-in-header include-before-body include-after-body
-    self-contained offline html5 html-q-tags ascii reference-links atx-headers
-    chapters number-sections number-offsetS no-tex-ligatures listings
-    incremental slide-level section-divs default-image-extension
-    email-obfuscation id-prefix title-prefix css reference-odt reference-docx
-    epub-stylesheet epub-cover-image epub-metadata epub-embed-font
-    epub-chapter-level latex-engine latex-engine-opt bibliography csl
-    citation-abbreviations natbib biblatex latexmathml asciimathml mathml
-    mimetex webtex jsmath mathjax katex katex-stylesheet gladtex trace
-    dump-args ignore-args verbose bash-completion)
+    normalize preserve-tabs tab-stop track-changes file-scope extract-media
+    standalone template metadata variable print-default-template
+    print-default-data-file no-wrap wrap columns toc table-of-contents toc-depth
+    no-highlight highlight-style include-in-header include-before-body
+    include-after-body self-contained offline html5 html-q-tags ascii
+    reference-links reference-location atx-headers chapters top-level-division
+    number-sections number-offsetS no-tex-ligatures listings incremental
+    slide-level section-divs default-image-extension email-obfuscation id-prefix
+    title-prefix css reference-odt reference-docx epub-stylesheet
+    epub-cover-image epub-metadata epub-embed-font epub-chapter-level
+    latex-engine latex-engine-opt bibliography csl citation-abbreviations natbib
+    biblatex latexmathml asciimathml mathml mimetex webtex jsmath mathjax katex
+    katex-stylesheet gladtex trace dump-args ignore-args verbose bash-completion
+    list-input-formats list-output-formats list-extensions
+    list-highlight-languages list-highlight-styles)
+  PANDOC_CITEPROC_OPTIONS = Set.new %w(bib2yaml bib2json reference-section-title
+    suppress-bibliography notes-after-punctuation y j)
   ALIAS_OPTIONS = Set.new %w(f r t w o R S F p s M V D H B A 5 N i T c m)
-  ALLOWED_OPTIONS = AVAILABLE_OPTIONS + ALIAS_OPTIONS
+  ALLOWED_OPTIONS = AVAILABLE_OPTIONS + PANDOC_CITEPROC_OPTIONS + ALIAS_OPTIONS
 
   # To use run the pandoc command with a custom executable path, the path
   # to the pandoc executable can be set here.

From c0ae6aa79c4676268fa29e6a08f9c686fc7af41b Mon Sep 17 00:00:00 2001
From: Martin Fenner <martin.fenner@datacite.org>
Date: Sat, 17 Dec 2016 15:18:50 +0100
Subject: [PATCH 6/6] don't use pandoc-citeproc options

---
 lib/pandoc-ruby.rb | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/lib/pandoc-ruby.rb b/lib/pandoc-ruby.rb
index 4771329..a2b4be7 100644
--- a/lib/pandoc-ruby.rb
+++ b/lib/pandoc-ruby.rb
@@ -76,10 +76,8 @@ class PandocRuby
     katex-stylesheet gladtex trace dump-args ignore-args verbose bash-completion
     list-input-formats list-output-formats list-extensions
     list-highlight-languages list-highlight-styles)
-  PANDOC_CITEPROC_OPTIONS = Set.new %w(bib2yaml bib2json reference-section-title
-    suppress-bibliography notes-after-punctuation y j)
   ALIAS_OPTIONS = Set.new %w(f r t w o R S F p s M V D H B A 5 N i T c m)
-  ALLOWED_OPTIONS = AVAILABLE_OPTIONS + PANDOC_CITEPROC_OPTIONS + ALIAS_OPTIONS
+  ALLOWED_OPTIONS = AVAILABLE_OPTIONS + ALIAS_OPTIONS
 
   # To use run the pandoc command with a custom executable path, the path
   # to the pandoc executable can be set here.