Ruby extract plain text for PDF by wrapping pdftotext shell command.

pdf_text_extractor.rb

# frozen_string_literal: true

##
# Primary responsibility is extracting text from a PDF or confirming if
# text is available in the PDF.
#
# Security note: This simple wrapper assumes that the PDF filename that you give it has been 
# chosen by an internal method, such as a tempfile name. Do not pass unsafe user supplied file names
# into this class.
#
# Copyright 2017 Rietta Inc. BSD Licensed.
#
class PdfTextExtractor
  attr_accessor :pdf_file

  def initialize(pdf_file:)
    unless command?('pdftotext')
      raise 'pdftotext is not installed, but is required.'
    end

    @pdf_file = pdf_file
  end

  # Determine if a command is available on the current Unix system.
  def command?(command)
    system("which #{command} > /dev/null 2>&1")
  end

  def text
    @text ||= `pdftotext '#{@pdf_file}' -`.strip
  end

  def text?
    text != ''
  end

  def as_json(_opts = {})
    {
      filename: @pdf_file,
      text: text
    }
  end
end