package jPDFProcessSamples;

import java.io.IOException;
import java.util.List;

import com.qoppa.pdf.PDFException;
import com.qoppa.pdf.TextPosition;
import com.qoppa.pdf.annotations.Redaction;
import com.qoppa.pdfProcess.PDFDocument;
import com.qoppa.pdfProcess.PDFPage;

/*
*
* This programs finds all instances of social security numbers in a PDF document 
* using a regex expression (pattern matching  “123-12-1234”). The SSN are then covered 
* with redaction annotations and removed from the PDF content when the redaction annotations 
* are applied or burnt, leaving just a black rectangle where the SSN used to be. 
* This process is called permanent redaction.
* 
*/
public class TextRedactionUsingRegex
{

	public static void main(String[] args)
	{
		try 
		{
			// Open the document
			PDFDocument pdfDoc = new PDFDocument("input.pdf", null);
			
			// Regular expression to check valid SSN
			String redactSSN = "^(?!666|000|9\\d{2})\\d{3}-(?!00)\\d{2}-(?!0{4})\\d{4}$";
			
			// per page: search text, create redaction annotations, then apply
			for (int i = 0; i < pdfDoc.getPageCount(); i++) 
			{
				PDFPage pdfPage = pdfDoc.getPage(i);
				
				// Search for the text
				List<TextPosition> searchResults = pdfPage.findTextUsingRegex(redactSSN);
				
				//create redaction annotations
				for (TextPosition textPos : searchResults)
				{
					Redaction redact = pdfDoc.getAnnotationFactory().createRedaction("Redaction sample", textPos.getPDFQuadrilaterals());
					pdfPage.addAnnotation(redact);
				}
				
				//apply ("burn-in") all redaction annotations on the page
				pdfPage.applyRedactionAnnotations();
			}
			
			pdfDoc.saveDocument("output.pdf");
		
		} catch (PDFException ex) {
			ex.printStackTrace();
		}
		catch (IOException ex) {
			ex.printStackTrace();
		}
	}
}