module PDF::Core

Constants

ESCAPED_NAME_CHARACTERS

Public Instance Methods

pdf_object(obj, in_content_stream = false) click to toggle source

Serializes Ruby objects to their PDF equivalents. Most primitive objects will work as expected, but please note that Name objects are represented by Ruby Symbol objects and Dictionary objects are represented by Ruby hashes (keyed by symbols)

Examples:

   pdf_object(true)      #=> "true"
   pdf_object(false)     #=> "false"
   pdf_object(1.2124)    #=> "1.2124"
   pdf_object('foo bar') #=> "(foo bar)"
   pdf_object(:Symbol)   #=> "/Symbol"
   pdf_object(['foo',:bar, [1,2]]) #=> "[foo /bar [1 2]]"
# File lib/pdf/core/pdf_object.rb, line 51
def pdf_object(obj, in_content_stream = false)
  case obj
  when NilClass   then 'null'
  when TrueClass  then 'true'
  when FalseClass then 'false'
  when Numeric
    obj = real(obj) unless obj.is_a?(Integer)

    # NOTE: this can fail on huge floating point numbers, but it seems
    # unlikely to ever happen in practice.
    num_string = String(obj)

    # Truncate trailing fraction zeroes
    num_string.sub(/(\d*)((\.0*$)|(\.0*[1-9]*)0*$)/, '\1\4')
  when Array
    "[#{obj.map { |e| pdf_object(e, in_content_stream) }.join(' ')}]"
  when PDF::Core::LiteralString
    obj = obj.gsub(/[\\\n\r\t\b\f()]/) { |m| "\\#{m}" }
    "(#{obj})"
  when Time
    obj = "#{obj.strftime('D:%Y%m%d%H%M%S%z').chop.chop}'00'"
    obj = obj.gsub(/[\\\n\r\t\b\f()]/) { |m| "\\#{m}" }
    "(#{obj})"
  when PDF::Core::ByteString
    "<#{obj.unpack1('H*')}>"
  when String
    obj = utf8_to_utf16(obj) unless in_content_stream
    "<#{string_to_hex(obj)}>"
  when Symbol
    name_string =
      obj.to_s.unpack('C*').map do |n|
        if ESCAPED_NAME_CHARACTERS.include?(n)
          "##{n.to_s(16).upcase}"
        else
          [n].pack('C*')
        end
      end.join
    "/#{name_string}"
  when ::Hash
    output = +'<< '
    obj.each do |k, v|
      unless k.is_a?(String) || k.is_a?(Symbol)
        raise PDF::Core::Errors::FailedObjectConversion,
          'A PDF Dictionary must be keyed by names'
      end
      output << pdf_object(k.to_sym, in_content_stream) << ' ' <<
        pdf_object(v, in_content_stream) << "\n"
    end
    output << '>>'
  when PDF::Core::Reference
    obj.to_s
  when PDF::Core::NameTree::Node
    pdf_object(obj.to_hash)
  when PDF::Core::NameTree::Value
    "#{pdf_object(obj.name)} #{pdf_object(obj.value)}"
  when PDF::Core::OutlineRoot, PDF::Core::OutlineItem
    pdf_object(obj.to_hash)
  else
    raise PDF::Core::Errors::FailedObjectConversion,
      "This object cannot be serialized to PDF (#{obj.inspect})"
  end
end
real(num) click to toggle source
# File lib/pdf/core/pdf_object.rb, line 15
def real(num)
  format('%<number>.5f', number: num).sub(/((?<!\.)0)+\z/, '')
end
real_params(array) click to toggle source
# File lib/pdf/core/pdf_object.rb, line 19
def real_params(array)
  array.map { |e| real(e) }.join(' ')
end
string_to_hex(str) click to toggle source

encodes any string into a hex representation. The result is a string with only 0-9 and a-f characters. That result is valid ASCII so tag it as such to account for behaviour of different ruby VMs

# File lib/pdf/core/pdf_object.rb, line 31
def string_to_hex(str)
  str.unpack1('H*').force_encoding(::Encoding::US_ASCII)
end
utf8_to_utf16(str) click to toggle source
# File lib/pdf/core/pdf_object.rb, line 23
def utf8_to_utf16(str)
  (+"\xFE\xFF").force_encoding(::Encoding::UTF_16BE) +
    str.encode(::Encoding::UTF_16BE)
end