ocr a pdf

ocr_pdf.rb

#! /usr/bin/env ruby

require 'pdfshaver'

# brew install ghostscript imagemagick #yikes
# brew install tesseract --HEAD # needs >=3.04

ARGV.each do |pdf|
  puts pdf
  pdf_basename = pdf.gsub(".pdf", '')
  if PDFShaver
    document = PDFShaver::Document.new(pdf)
    document.pages.each{|page| page.render("./#{pdf_basename}-#{page.number}.png") }
  else
    `convert -monochrome -density 300x300 "#{pdf}" -depth 8 "#{pdf_basename}.png"`
  end
  (Dir["#{pdf_basename}-*.png"] + Dir["#{pdf_basename}.png"]).each do |png|
    puts png
     # `tesseract "#{png}" "#{png}" pdf`
     
    `tesseract "#{png}" "#{png}" pdf`
  end
  files = Dir["#{pdf_basename}-*.png.pdf"].sort_by{|pdf| Regexp.new("#{pdf_basename}-([0-9]+).png.pdf").match(pdf)[1].to_i }.join('" "')
  puts files.inspect
  `pdftk "#{files}" cat output "#{pdf_basename}.ocr.pdf"`
end