class Erubi::Engine

def initialize(input, properties={})

:trim :: Whether to trim leading and trailing whitespace, true by default.
:preamble :: The preamble for the template, by default initializes up the buffer variable.
:postable :: the postamble for the template, by default returns the resulting source code.
:outvar :: Same as bufvar, with lower priority.
:freeze :: Whether to enable frozen string literals in the resulting source code.
:filename :: The filename for the template.
:escape_html :: Same as :escape, with lower priority.
:escape :: Whether to make <%= escape by default, and <%== not escape by default.
:escapefunc :: The function to use for escaping, as a string (default: ::Erubi.h).
:bufvar :: The variable name to use for the buffer variable, as a string.
Initialize a new Erubi::Engine. Options: 
def initialize(input, properties={})
  escape     = properties.fetch(:escape){properties.fetch(:escape_html, false)}
  trim       = properties[:trim] != false
  @filename  = properties[:filename]
  @bufvar = bufvar = properties[:bufvar] || properties[:outvar] || "_buf"
  preamble   = properties[:preamble] || "#{bufvar} = String.new;"
  postamble  = properties[:postamble] || "#{bufvar}.to_s\n"
  @src = src = String.new
  src << "# frozen_string_literal: true\n" if properties[:freeze]
  unless escapefunc = properties[:escapefunc]
    if escape
      escapefunc = '__erubi.h'
      src << "__erubi = ::Erubi;"
    else
      escapefunc = '::Erubi.h'
    end
  end
  src << preamble
  pos = 0
  is_bol = true
  input.scan(/<%(={1,2}|-|\#|%)?(.*?)([-=])?%>([ \t]*\r?\n)?/m) do |indicator, code, tailch, rspace|
    match = Regexp.last_match
    len  = match.begin(0) - pos
    text = input[pos, len]
    pos  = match.end(0)
    ch   = indicator ? indicator[RANGE_FIRST] : nil
    lspace = nil
    unless ch == '='
      if text.empty?
        lspace = "" if is_bol
      elsif text[RANGE_LAST] == "\n"
        lspace = ""
      else
        rindex = text.rindex("\n")
        if rindex
          range = rindex+1..-1
          s = text[range]
          if s =~ /\A[ \t]*\z/
            lspace = s
            text[range] = ''
          end
        else
          if is_bol && text =~ /\A[ \t]*\z/
            lspace = text.dup
            text[RANGE_ALL] = ''
          end
        end
      end
    end
    is_bol = rspace ? true : false
    add_text(text) if text && !text.empty?
    if ch == '='
      rspace = nil if tailch && !tailch.empty?
      add_text(lspace) if lspace
      if ((indicator == '=') ^ escape)
        src << " #{bufvar} << (" << code << ').to_s;'
      else
        src << " #{bufvar} << #{escapefunc}((" << code << '));'
      end
      add_text(rspace) if rspace
    elsif ch == '#'
      n = code.count("\n") + (rspace ? 1 : 0)
      if trim
        add_code("\n" * n)
      else
        add_text(lspace) if lspace
        add_code("\n" * n)
        add_text(rspace) if rspace
      end
    elsif ch == '%'
      add_text("#{lspace}#{prefix||='<%'}#{code}#{tailch}#{postfix||='%>'}#{rspace}")
    else
      if trim
        add_code("#{lspace}#{code}#{rspace}")
      else
        add_text(lspace) if lspace
        add_code(code)
        add_text(rspace) if rspace
      end
    end
  end
  rest = pos == 0 ? input : input[pos..-1]
  add_text(rest)
  src << "\n" unless src[RANGE_LAST] == "\n"
  src << postamble
  freeze
end