E.g. oftentimes there is l and I (capital I), this may be an issue for OCR. The perfect case is when there is a PDF document and data embedded as XML data, but unfortunately it is not the case.