Cherry-pick from PR #143: pdf/a support

openhtmltopdf-pdfbox/src/main/java/com/openhtmltopdf/pdfboxout/PdfBoxRenderer.java

@@ -39,7 +39,6 @@
 import com.openhtmltopdf.render.PageBox;
 import com.openhtmltopdf.render.RenderingContext;
 import com.openhtmltopdf.render.ViewportBox;
-import com.openhtmltopdf.render.AbstractOutputDevice.ClipInfo;
 import com.openhtmltopdf.render.displaylist.DisplayListCollector;
 import com.openhtmltopdf.render.displaylist.DisplayListContainer;
 import com.openhtmltopdf.render.displaylist.DisplayListPainter;
@@ -49,14 +48,20 @@
 import com.openhtmltopdf.util.Configuration;
 import com.openhtmltopdf.util.ThreadCtx;
 import com.openhtmltopdf.util.XRLog;
-
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDDocumentInformation;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.*;
 import org.apache.pdfbox.pdmodel.PDPageContentStream.AppendMode;
+import org.apache.pdfbox.pdmodel.common.PDMetadata;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDMarkInfo;
+import org.apache.pdfbox.pdmodel.documentinterchange.logicalstructure.PDStructureTreeRoot;
 import org.apache.pdfbox.pdmodel.encryption.PDEncryption;
+import org.apache.pdfbox.pdmodel.graphics.color.PDOutputIntent;
+import org.apache.xmpbox.XMPMetadata;
+import org.apache.xmpbox.schema.AdobePDFSchema;
+import org.apache.xmpbox.schema.PDFAIdentificationSchema;
+import org.apache.xmpbox.schema.XMPBasicSchema;
+import org.apache.xmpbox.type.BadFieldValueException;
+import org.apache.xmpbox.xml.XmpSerializer;
 import org.w3c.dom.Document;
 import org.w3c.dom.Element;
 import org.w3c.dom.Node;
@@ -97,7 +102,11 @@ public class PdfBoxRenderer implements Closeable {
 
     // Usually 1.7
     private float _pdfVersion;
-
+
+    private String _pdfAConformance;
+
+    private byte[] _colorProfile;
+
     private boolean _testMode;
 
     private PDFCreationListener _listener;
@@ -122,8 +131,13 @@ public class PdfBoxRenderer implements Closeable {
 
         _producer = state._producer;
 
+
         _svgImpl = state._svgImpl;
         _mathmlImpl = state._mathmlImpl;
+
+        _pdfAConformance = state._pdfAConformance;
+        _colorProfile = state._colorProfile;
+
         _dotsPerPoint = DEFAULT_DOTS_PER_POINT;
         _testMode = state._testMode;
         _useFastMode = state._useFastRenderer;
@@ -621,7 +635,11 @@ private void writePDF(List<PageBox> pages, RenderingContext c, Rectangle2D first
         c.setPageCount(pageCount);
         firePreWrite(pageCount); // opportunity to adjust meta data
         setDidValues(doc); // set PDF header fields from meta data
-
+
+        if (_pdfAConformance != null) {
+            addPdfASchema(doc, _pdfAConformance);
+        }
+
         for (int i = 0; i < pageCount; i++) {
             PageBox currentPage = pages.get(i);
 
@@ -643,10 +661,63 @@ private void writePDF(List<PageBox> pages, RenderingContext c, Rectangle2D first
         _outputDevice.finish(c, _root);
     }
 
+    private void addPdfASchema(PDDocument document, String conformance) {
+        PDDocumentInformation information = document.getDocumentInformation();
+        XMPMetadata metadata = XMPMetadata.createXMPMetadata();
+
+        try {
+            PDFAIdentificationSchema pdfaid = metadata.createAndAddPFAIdentificationSchema();
+            pdfaid.setConformance(conformance);
+            pdfaid.setPart(1);
+
+            AdobePDFSchema pdfSchema = metadata.createAndAddAdobePDFSchema();
+            pdfSchema.setProducer(information.getProducer());
+
+            XMPBasicSchema xmpBasicSchema = metadata.createAndAddXMPBasicSchema();
+            xmpBasicSchema.setCreateDate(information.getCreationDate());
+
+            PDMetadata metadataStream = new PDMetadata(document);
+            PDMarkInfo markInfo = new PDMarkInfo();
+            markInfo.setMarked(true);
+
+            // add to catalog
+            PDDocumentCatalog catalog = document.getDocumentCatalog();
+            catalog.setMetadata(metadataStream);
+            // for pdf/a-1 compliance, add the StructTreeRoot that https://www.pdf-online.com/osa/validate.aspx was
+            // complaining. Based on https://stackoverflow.com/a/46806392
+            catalog.setStructureTreeRoot(new PDStructureTreeRoot());
+            //
+
+            catalog.setMarkInfo(markInfo);
+
+            XmpSerializer serializer = new XmpSerializer();
+            ByteArrayOutputStream baos = new ByteArrayOutputStream();
+            serializer.serialize(metadata, baos, true);
+            metadataStream.importXMPMetadata( baos.toByteArray() );
+
+
+            if (_colorProfile != null) {
+                ByteArrayInputStream colorProfile = new ByteArrayInputStream(_colorProfile);
+                PDOutputIntent oi = new PDOutputIntent(document, colorProfile);
+                oi.setInfo("sRGB IEC61966-2.1");
+                oi.setOutputCondition("sRGB IEC61966-2.1");
+                oi.setOutputConditionIdentifier("sRGB IEC61966-2.1");
+                oi.setRegistryName("http://www.color.org");
+                catalog.addOutputIntent(oi);
+            }
+        } catch (BadFieldValueException e) {
+            throw new RuntimeException(e);
+        } catch (IOException e) {
+            throw new RuntimeException(e);
+        } catch (TransformerException e) {
+            throw new RuntimeException(e);
+        }
+    }
+
     // Sets the document information dictionary values from html metadata
     private void setDidValues(PDDocument doc) {
         PDDocumentInformation info = new PDDocumentInformation();
-        
+
         info.setCreationDate(Calendar.getInstance());
 
         if (_producer == null) {

openhtmltopdf-pdfbox/src/main/java/com/openhtmltopdf/pdfboxout/PdfRendererBuilder.java

@@ -107,6 +107,30 @@ public PdfRendererBuilder usePdfVersion(float version) {
 		state._pdfVersion = version;
 		return this;
 	}
+
+	/**
+	 * Set the PDF/A conformance, typically we use PDF/A-1
+	 *
+	 * @param pdfAConformance
+	 * @return
+	 */
+	public PdfRendererBuilder usePdfAConformance(PdfAConformance pdfAConformance) {
+		this.state._pdfAConformance = pdfAConformance.value;
+		return this;
+	}
+
+	/**
+	 * Sets the color profile, needed for PDF/A conformance.
+	 *
+	 * You can use the sRGB.icc from https://svn.apache.org/viewvc/pdfbox/trunk/examples/src/main/resources/org/apache/pdfbox/resources/pdfa/
+	 *
+	 * @param colorProfile
+	 * @return
+	 */
+	public PdfRendererBuilder useColorProfile(byte[] colorProfile) {
+		this.state._colorProfile = colorProfile;
+		return this;
+	}
 
 	/**
 	 * By default, this project creates an entirely in-memory <code>PDDocument</code>.
@@ -214,5 +238,21 @@ private AddedFont(FSSupplier<InputStream> supplier, File fontFile, Integer weigh
 			this.style = style;
 		}
 	}
+
+	/**
+	 * Various level of PDF/A conformance:
+	 *
+	 * PDF/A-1, PDF/A-2 and PDF/A-3
+	 */
+	public enum PdfAConformance {
+
+		PDF_A_1("A"), PDF_A_2("B"), PDF_A_3("U");
+
+		PdfAConformance(String value) {
+			this.value = value;
+		}
+
+		private final String value;
+	}
 }
 

openhtmltopdf-pdfbox/src/main/java/com/openhtmltopdf/pdfboxout/PdfRendererBuilderState.java

@@ -22,4 +22,6 @@ public class PdfRendererBuilderState extends BaseRendererBuilder.BaseRendererBui
 	public float _pdfVersion = 1.7f;
 	public String _producer;
 	public PDDocument pddocument;
+	public String _pdfAConformance;
+	public byte[] _colorProfile;
 }