Loading a PDF with its content

The IRISOCR™ SDK allows you to load the page content (i.e. the text layer and the graphical zones) from PDFs without the need of running OCR.

For this, two parameters from the CImageLoadOptionsPdf class are available:

Parameter Default value Explanation

LoadPageContent

IDRS_FALSE

This parameter enables/disables the loading of the PDF page content.

AllowIncompleteTextLoading

IDRS_TRUE

The iDRS does not support unicode characters with value higher than U+FFFF.

If set to IDRS_TRUE and such characters are encountered, then they are replaced by U+FFFD (replacement character).

If set to IDRS_FALSE, then an exception with error code IDRS_ERROR_IMAGE_FILE_PDF_UNSUPPORTED_CHARACTER is thrown.
Note

PDF rendering will always be done before loading the text layer. So after loading the page content, OCR can be run.
Limitations

  • When loading the PDF text layer, re-OCR is not done internally. It can be triggered by the user afterwards.

  • Top-to-bottom text is not handled correctly.

  • Right-to-left documents (e.g: Arabic) are not supported yet.
Code Snippet(s)
CIDRS objIdrs = CIDRS::Create();

// Set PDF load options
CImageLoadOptionsPdf objImageLoadOptionsPdf = CImageLoadOptionsPdf::Create();
objImageLoadOptionsPdf.SetLoadPageContent(IDRS_TRUE);

// Do load operation
CImageIO objImageIO = CImageIO::Create(objIdrs);
objImageIO.SetPdfLoadOptions(objImageLoadOptionsPdf);
CPage objPage = objImageIO.LoadPage("myfile.pdf");
CPageContent objPageContent = objPage.GetPageContent();
//Use objPageContent to get the text layer and the graphical zones
using (CIDRS objIdrs = new CIDRS())
{
	using (CImageIO objImageIO = new CImageIO(objIdrs))
	{
		using (CImageLoadOptionsPdf objPdfLoadOptions = new CImageLoadOptionsPdf())
		{
			objPdfLoadOptions.LoadPageContent = true;
			objImageIO.PdfLoadOptions = objPdfLoadOptions;
			CPage objPage = objImageIO.LoadPage("my path");
			CPageContent objPageContent = objPage.PageContent;
			//Use objPageContent2 to get the text layer and the graphical zones
		}
	}
}