I have my own web application which uses sphinx as search engine, i'm using Readiris as an OCR and i store the searchibul Pdf into the application Rreadris generates 1.7 pdf version which is indexed So while searching for Arabic words it doesn't show results but when searching for same pdf/xps file genrated by Word Microsoft it shows relevent results In both cases i a using pdf_to_text function to take the text and store it The application is running on Ubento 18
Supposed to give same results whether the producer of pdf dile is Microsoft or Readiris because the word file is even generated by Readiris so it's the same OCR