IHPDFOCREngine Interface

Abstracts the OCR provider while HotPDF owns page rendering, validation, PDF geometry, Unicode mapping, resource limits, and transactional publication

Declaration

IHPDFOCREngine = interface(IInterface)
  function GetName: AnsiString;
  function Recognize(const Request: THPDFOCRRequest;
    out Words: THPDFOCRWords;
    out Diagnostic: AnsiString): boolean;
  property Name: AnsiString read GetName;
end;

Contract

Recognize returns True after producing zero or more words and returns False with a diagnostic when recognition cannot complete

The bitmap and cancellation token are borrowed for the call, and implementations must obey MaxWords and MaxTextCodeUnits

Implementations may report word boxes without baselines, in which case HotPDF infers a horizontal baseline near the bottom of the box

THPDFOCRRequest Record · THPDFOCRWord Record