https://github.com/whitequark/parser/issues/134

generate_table.rb

# coding: utf-8

require 'tmpdir'
require 'parser/current'

def in_tmpdir
  Dir.mktmpdir do |tmpdir|
    Dir.chdir(tmpdir) do
      yield
    end
  end
end

def each_sample
  encodings = [
    'utf-8',
    'us-ascii',
    'ascii-8bit',
    'koi8-r'
  ]

  regexps = [
    'abc',
    'À',
    'д',
    '\xff'
  ]

  options = [
    '',
    'u',
    'n'
  ]

  encodings.each do |encoding|
    regexps.each do |regexp|
      options.each do |option|
        source = <<-END.gsub(/^\s+\|/, '').chomp
          |# coding: #{encoding}
          |puts /#{regexp}/#{option}.encoding
        END
        yield encoding, regexp, option, source
      end
    end
  end
end

# Derived from https://github.com/whitequark/parser/blob/v2.1.7/lib/parser/builders/default.rb#L707-L726
def regexp_from_node(regexp_node)
  str_node, opt_node = *regexp_node
  regexp_body, = *str_node
  *regexp_opt  = *opt_node

  regexp_body = case
  when regexp_opt.include?(:u)
    regexp_body.encode(Encoding::UTF_8)
  when regexp_opt.include?(:e)
    regexp_body.encode(Encoding::EUC_JP)
  when regexp_opt.include?(:s)
    regexp_body.encode(Encoding::WINDOWS_31J)
  when regexp_opt.include?(:n)
    regexp_body.encode(Encoding::BINARY)
  else
    regexp_body
  end

  Regexp.new(regexp_body)
end

markdown = ''
markdown << "* Ruby: #{RUBY_ENGINE} #{RUBY_VERSION}\n"
markdown << "* Encoding of the source Parser was run on: #{''.encoding}\n"
markdown << "\n"
markdown << "Magic comment | Regexp | Ruby (Regexp encoding or error) | `Regexp.new().encoding` with Parser node or error | Parser regexp node \n"
markdown << "--------------|--------|---------------------------------|---------------------------------------------------|--------------------\n"

in_tmpdir do
  path = 'sample.rb'

  each_sample do |encoding, regexp, option, source|
    begin
      File.open(path, "w:#{encoding}") do |file|
        file.write(source)
      end
    rescue
      next # Skip if the file encoding does not support the character
    end

    ruby_result = `ruby #{path} 2>&1`

    ast = Parser::CurrentRuby.parse_file(path)
    regexp_node = ast.children[2].children[0]

    begin
      parser_result = regexp_from_node(regexp_node).encoding
    rescue => error
      parser_result = error.message
    end

    markdown << "#{encoding} | `/#{regexp}/#{option}` | #{ruby_result.chomp} | #{parser_result} | `#{regexp_node.inspect.gsub("\n", '')}`\n"
  end
end

File.write('result.md', markdown)

result.md

• Ruby: ruby 2.1.1
• Encoding of the source Parser was run on: UTF-8

Magic comment	Regexp	Ruby (Regexp encoding or error)	Regexp.new().encoding with Parser node or error	Parser regexp node
utf-8	/abc/	US-ASCII	US-ASCII	(regexp (str "abc") (regopt))
utf-8	/abc/u	UTF-8	US-ASCII	(regexp (str "abc") (regopt :u))
utf-8	/abc/n	US-ASCII	US-ASCII	(regexp (str "abc") (regopt :n))
utf-8	/À/	UTF-8	UTF-8	(regexp (str "À") (regopt))
utf-8	/À/u	UTF-8	UTF-8	(regexp (str "À") (regopt :u))
utf-8	/À/n	sample.rb:2: regexp encoding option 'n' differs from source encoding 'UTF-8'	U+00C0 from UTF-8 to ASCII-8BIT	(regexp (str "À") (regopt :n))
utf-8	/д/	UTF-8	UTF-8	(regexp (str "д") (regopt))
utf-8	/д/u	UTF-8	UTF-8	(regexp (str "д") (regopt :u))
utf-8	/д/n	sample.rb:2: regexp encoding option 'n' differs from source encoding 'UTF-8'	U+0434 from UTF-8 to ASCII-8BIT	(regexp (str "д") (regopt :n))
utf-8	/\xff/	sample.rb:2: invalid multibyte escape: /\xff/	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt))
utf-8	/\xff/u	sample.rb:2: invalid multibyte escape: /\xff/	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt :u))
utf-8	/\xff/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\\xff") (regopt :n))
us-ascii	/abc/	US-ASCII	US-ASCII	(regexp (str "abc") (regopt))
us-ascii	/abc/u	UTF-8	US-ASCII	(regexp (str "abc") (regopt :u))
us-ascii	/abc/n	US-ASCII	US-ASCII	(regexp (str "abc") (regopt :n))
us-ascii	/\xff/	ASCII-8BIT	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt))
us-ascii	/\xff/u	sample.rb:2: invalid multibyte escape: /\xff/	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt :u))
us-ascii	/\xff/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\\xff") (regopt :n))
ascii-8bit	/abc/	US-ASCII	US-ASCII	(regexp (str "abc") (regopt))
ascii-8bit	/abc/u	UTF-8	US-ASCII	(regexp (str "abc") (regopt :u))
ascii-8bit	/abc/n	US-ASCII	US-ASCII	(regexp (str "abc") (regopt :n))
ascii-8bit	/À/	ASCII-8BIT	ASCII-8BIT	(regexp (str "\xC3\x80") (regopt))
ascii-8bit	/À/u	sample.rb:2: regexp encoding option 'u' differs from source encoding 'ASCII-8BIT'	"\xC3" from ASCII-8BIT to UTF-8	(regexp (str "\xC3\x80") (regopt :u))
ascii-8bit	/À/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\xC3\x80") (regopt :n))
ascii-8bit	/д/	ASCII-8BIT	ASCII-8BIT	(regexp (str "\xD0\xB4") (regopt))
ascii-8bit	/д/u	sample.rb:2: regexp encoding option 'u' differs from source encoding 'ASCII-8BIT'	"\xD0" from ASCII-8BIT to UTF-8	(regexp (str "\xD0\xB4") (regopt :u))
ascii-8bit	/д/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\xD0\xB4") (regopt :n))
ascii-8bit	/\xff/	ASCII-8BIT	ASCII-8BIT	(regexp (str "\\xff") (regopt))
ascii-8bit	/\xff/u	sample.rb:2: invalid multibyte escape: /\xff/	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt :u))
ascii-8bit	/\xff/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\\xff") (regopt :n))
koi8-r	/abc/	US-ASCII	US-ASCII	(regexp (str "abc") (regopt))
koi8-r	/abc/u	UTF-8	US-ASCII	(regexp (str "abc") (regopt :u))
koi8-r	/abc/n	US-ASCII	US-ASCII	(regexp (str "abc") (regopt :n))
koi8-r	/д/	KOI8-R	UTF-8	(regexp (str "д") (regopt))
koi8-r	/д/u	sample.rb:2: regexp encoding option 'u' differs from source encoding 'KOI8-R'	UTF-8	(regexp (str "д") (regopt :u))
koi8-r	/д/n	sample.rb:2: regexp encoding option 'n' differs from source encoding 'KOI8-R'	U+0434 from UTF-8 to ASCII-8BIT	(regexp (str "д") (regopt :n))
koi8-r	/\xff/	KOI8-R	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt))
koi8-r	/\xff/u	sample.rb:2: invalid multibyte escape: /\xff/	invalid multibyte escape: /\xff/	(regexp (str "\\xff") (regopt :u))
koi8-r	/\xff/n	ASCII-8BIT	ASCII-8BIT	(regexp (str "\\xff") (regopt :n))