Added a WriteOption to insert/update document level xmp metadata stre…

…am based on the info dictionary.

README.md

@@ -7,7 +7,7 @@ An [Apache PDFBox](https://github.com/apache/pdfbox) fork intended to be used as
 
 What's different from PDFBox?
 ---------
-+ Requires JDK 17
++ Requires JDK 21
 + Lazy loading/parsing of PDF objects. Only the document xref table(s)/stream(s) is(are) initially parsed and information to lookup objects are retrieved, when later a PDF object is requested, the object is retrieve/parsed using the lookup information. This allows minimal memory footprint when you only need part of the document (Ex. you only need the information dictionary or the number of pages of the document).
 + Multiple I/O implementations to read from. SAMBox uses [Sejda-io](https://github.com/torakiki/sejda-io) allowing to use one of the provided implementation based on `java.nio.channels.FileChannel`, `java.io.InputStream` and `java.nio.MappedByteBuffer` (buffered or not).
 + Minimized GC through the use of a pool of `java.lang.StringBuilder`.

pom.xml

@@ -1,5 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
-<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
+<project xmlns="http://maven.apache.org/POM/4.0.0"
+         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/maven-v4_0_0.xsd">
     <modelVersion>4.0.0</modelVersion>
     <groupId>org.sejda</groupId>
     <artifactId>sambox</artifactId>
@@ -56,6 +58,7 @@
         <maven.javadoc.failOnError>false</maven.javadoc.failOnError>
         <bouncycastle.version>1.78.1</bouncycastle.version>
         <fontbox.version>2.0.28</fontbox.version>
+        <xmpbox.version>3.0.2</xmpbox.version>
         <sejda.commons.version>2.0.0</sejda.commons.version>
         <sejda.io.version>3.0.1</sejda.io.version>
         <slf4j.version>2.0.13</slf4j.version>
@@ -1529,10 +1532,14 @@
                             <goal>wget</goal>
                         </goals>
                         <configuration>
-                            <url>https://issues.apache.org/jira/secure/attachment/13047577/PDFBOX-5484.ttf</url>
+                            <url>
+                                https://issues.apache.org/jira/secure/attachment/13047577/PDFBOX-5484.ttf
+                            </url>
                             <outputDirectory>${project.build.directory}/fonts</outputDirectory>
                             <outputFileName>PDFBOX-5484.ttf</outputFileName>
-                            <sha512>7c3d8bbc18654315d6341a277dcd5c66218b95c43baf190b6e32f77817d17bab421ef76f2c904b46c97f84c49b00d58525449cff970897010534d6aa2812a4e2</sha512>
+                            <sha512>
+                                7c3d8bbc18654315d6341a277dcd5c66218b95c43baf190b6e32f77817d17bab421ef76f2c904b46c97f84c49b00d58525449cff970897010534d6aa2812a4e2
+                            </sha512>
                         </configuration>
                     </execution>
                 </executions>
@@ -1572,6 +1579,17 @@
                 </exclusion>
             </exclusions>
         </dependency>
+        <dependency>
+            <groupId>org.apache.pdfbox</groupId>
+            <artifactId>xmpbox</artifactId>
+            <version>${xmpbox.version}</version>
+            <exclusions>
+                <exclusion>
+                    <artifactId>commons-logging</artifactId>
+                    <groupId>commons-logging</groupId>
+                </exclusion>
+            </exclusions>
+        </dependency>
         <dependency>
             <groupId>org.bouncycastle</groupId>
             <artifactId>bcmail-jdk18on</artifactId>

src/main/java/module-info.java

@@ -27,6 +27,7 @@
     requires transitive java.xml;
     requires transitive org.apache.fontbox;
     requires transitive org.sejda.io;
+    requires transitive org.apache.xmpbox;
 
     exports org.sejda.sambox;
     exports org.sejda.sambox.contentstream;

src/main/java/org/sejda/sambox/output/WriteOption.java

@@ -18,7 +18,7 @@
 
 /**
  * Options that can be selected when writing a PDF document.
- * 
+ *
  * @author Andrea Vacondio
  */
 public enum WriteOption
@@ -42,5 +42,14 @@ public enum WriteOption
     /**
      * Does not automatically update metadata modified date and producer when saving
      */
-    NO_METADATA_PRODUCER_MODIFIED_DATE_UPDATE
+    NO_METADATA_PRODUCER_MODIFIED_DATE_UPDATE,
+    /**
+     * It creates or updates the document XMP metadata before the document is written.
+     * <ul>
+     *     <li>Creates: if the document XMP metadata does not exist, it creates a new one based on the info dictionary</li>
+     *     <li>Updates: if the document XMP metadata exists, it updates all the values corresponding to the info dictionary (see ISO 32000-2:2020 Chap 14.3.3 Table 349)</li>
+     * </ul>
+     * Note: we currently leave untouched a malformed metadata stream.
+     */
+    UPSERT_DOCUMENT_METADATA_STREAM
 }

src/main/java/org/sejda/sambox/pdmodel/PDDocument.java

@@ -27,6 +27,8 @@
 import java.awt.Point;
 import java.awt.image.DataBuffer;
 import java.awt.image.Raster;
+import java.io.BufferedInputStream;
+import java.io.BufferedOutputStream;
 import java.io.Closeable;
 import java.io.File;
 import java.io.IOException;
@@ -41,6 +43,10 @@
 import java.util.Set;
 
 import org.apache.fontbox.ttf.TrueTypeFont;
+import org.apache.xmpbox.XMPMetadata;
+import org.apache.xmpbox.xml.DomXmpParser;
+import org.apache.xmpbox.xml.XmpParsingException;
+import org.apache.xmpbox.xml.XmpSerializer;
 import org.sejda.commons.util.IOUtils;
 import org.sejda.io.CountingWritableByteChannel;
 import org.sejda.io.SeekableSources;
@@ -61,13 +67,15 @@
 import org.sejda.sambox.output.PDDocumentWriter;
 import org.sejda.sambox.output.PreSaveCOSTransformer;
 import org.sejda.sambox.output.WriteOption;
+import org.sejda.sambox.pdmodel.common.PDMetadata;
 import org.sejda.sambox.pdmodel.encryption.AccessPermission;
 import org.sejda.sambox.pdmodel.encryption.PDEncryption;
 import org.sejda.sambox.pdmodel.encryption.SecurityHandler;
 import org.sejda.sambox.pdmodel.font.Subsettable;
 import org.sejda.sambox.pdmodel.graphics.color.PDDeviceRGB;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
+import javax.xml.transform.TransformerException;
 
 /**
  * This is the in-memory representation of the PDF document.
@@ -535,14 +543,27 @@ private void writeTo(CountingWritableByteChannel output, StandardSecurity securi
     {
         requireOpen();
 
-        if (Arrays.stream(options)
-                .noneMatch(i -> i == WriteOption.NO_METADATA_PRODUCER_MODIFIED_DATE_UPDATE))
+        updateMetadata(options);
+        subsetFonts();
+
+        EncryptionContext encryptionContext = ofNullable(security).map(EncryptionContext::new)
+                .orElse(null);
+        generateFileIdentifier(output.toString().getBytes(StandardCharsets.ISO_8859_1),
+                encryptionContext);
+        try (PDDocumentWriter writer = new PDDocumentWriter(output, encryptionContext,
+                preSaveCOSTransformer, options))
         {
-            // update producer and last modification date only if the write option doesn't state otherwise
-            getDocumentInformation().setProducer(SAMBox.PRODUCER);
-            getDocumentInformation().setModificationDate(Calendar.getInstance());
+            onBeforeWrite.onBeforeWrite();
+            writer.write(this);
         }
+        finally
+        {
+            IOUtils.close(this);
+        }
+    }
 
+    private void subsetFonts()
+    {
         for (Subsettable font : fontsToSubset)
         {
             try
@@ -555,20 +576,57 @@ private void writeTo(CountingWritableByteChannel output, StandardSecurity securi
             }
         }
         fontsToSubset.clear();
-        EncryptionContext encryptionContext = ofNullable(security).map(EncryptionContext::new)
-                .orElse(null);
-        generateFileIdentifier(output.toString().getBytes(StandardCharsets.ISO_8859_1),
-                encryptionContext);
-        try (PDDocumentWriter writer = new PDDocumentWriter(output, encryptionContext,
-                preSaveCOSTransformer, options))
+    }
+
+    private void updateMetadata(WriteOption[] options)
+    {
+        if (Arrays.stream(options)
+                .noneMatch(i -> i == WriteOption.NO_METADATA_PRODUCER_MODIFIED_DATE_UPDATE))
         {
-            onBeforeWrite.onBeforeWrite();
-            writer.write(this);
+            // update producer and last modification date only if the write option doesn't state otherwise
+            getDocumentInformation().setProducer(SAMBox.PRODUCER);
+            getDocumentInformation().setModificationDate(Calendar.getInstance());
         }
-        finally
+        if (Arrays.stream(options).anyMatch(o -> o == WriteOption.UPSERT_DOCUMENT_METADATA_STREAM))
         {
-            IOUtils.close(this);
+            requireMinVersion(V1_4);
+            var metadataStream = new PDMetadata();
+            try (var metadataOutputStream = new BufferedOutputStream(
+                    metadataStream.getCOSObject().createUnfilteredStream()))
+            {
+                new XmpSerializer().serialize(
+                        getDocumentInformation().toXMPMetadata(getOrCreateXmpMetadata(),
+                                getVersion()), metadataOutputStream, true);
+                getDocumentCatalog().setMetadata(metadataStream);
+            }
+            catch (IOException | TransformerException e)
+            {
+                LOG.warn("Unable to set xmp document metadata", e);
+            }
+            catch (XmpParsingException e)
+            {
+                LOG.warn("Unable to parse existing document level xmp metadata", e);
+            }
+        }
+    }
+
+    private XMPMetadata getOrCreateXmpMetadata() throws XmpParsingException, IOException
+    {
+        var metadata = getDocumentCatalog().getMetadata();
+        if (nonNull(metadata))
+        {
+            try
+            {
+                var parser = new DomXmpParser();
+                parser.setStrictParsing(false);
+                return parser.parse(new BufferedInputStream(metadata.createInputStream()));
+            }
+            finally
+            {
+                metadata.getCOSObject().unDecode();
+            }
         }
+        return XMPMetadata.createXMPMetadata();
     }
 
     /**