create mixin fedora:checksum (mixed into our datastream jcr:content n…

…ode) that records a checksum and content size for the jcr:data
fcrepo · Mar 6, 2013 · b4afade · b4afade
1 parent 8f0a94a
commit b4afade
Show file tree

Hide file tree

Showing 5 changed files with 150 additions and 3 deletions.
diff --git a/fcrepo-kernel/pom.xml b/fcrepo-kernel/pom.xml
@@ -73,6 +73,12 @@
       <groupId>com.google.guava</groupId>
       <artifactId>guava</artifactId>
     </dependency>
+    <dependency>
+      <groupId>commons-io</groupId>
+      <artifactId>commons-io</artifactId>
+      <version>2.4</version>
+      <scope>test</scope>
+    </dependency>
   </dependencies>
 
   <build>

diff --git a/fcrepo-kernel/src/main/java/org/fcrepo/Datastream.java b/fcrepo-kernel/src/main/java/org/fcrepo/Datastream.java
@@ -10,6 +10,11 @@
 import static org.modeshape.jcr.api.JcrConstants.NT_RESOURCE;
 
 import java.io.InputStream;
+import java.net.URI;
+import java.net.URISyntaxException;
+import java.security.DigestInputStream;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
 import java.util.Date;
 
 import javax.jcr.Node;
@@ -20,6 +25,8 @@
 import javax.jcr.nodetype.ConstraintViolationException;
 import javax.jcr.version.VersionException;
 
+import org.apache.commons.codec.binary.Hex;
+import org.fcrepo.utils.ContentDigest;
 import org.modeshape.jcr.api.JcrTools;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -32,6 +39,11 @@
  */
 public class Datastream extends JcrTools {
 
+    private static MessageDigest md;
+    private final static String CONTENT_SIZE = "fedora:size";
+    private final static String DIGEST_VALUE = "fedora:digest";
+    private final static String DIGEST_ALGORITHM = "fedora:digestAlgorithm";
+
     private final static Logger logger = LoggerFactory
             .getLogger(Datastream.class);
 
@@ -48,6 +60,18 @@ public Node getNode() {
         return node;
     }
 
+    private MessageDigest getMessageDigest() {
+
+        if(this.md == null) {
+            try {
+                this.md  = MessageDigest.getInstance("SHA-1");
+            } catch (NoSuchAlgorithmException e) {
+                e.printStackTrace();
+            }
+        }
+
+        return md;
+    }
     /**
      * @return The InputStream of content associated with this datastream.
      * @throws RepositoryException
@@ -66,6 +90,13 @@ public InputStream getContent() throws RepositoryException {
     public void setContent(InputStream content) throws RepositoryException {
         final Node contentNode =
                 findOrCreateChild(node, JCR_CONTENT, NT_RESOURCE);
+
+        if(contentNode.canAddMixin("fedora:checksum")) {
+            contentNode.addMixin("fedora:checksum");
+        }
+
+        final DigestInputStream dis = new DigestInputStream(content, getMessageDigest());
+
         logger.debug("Created content node at path: " + contentNode.getPath());
         /*
          * This next line of code deserves explanation. If we chose for the
@@ -84,7 +115,11 @@ public void setContent(InputStream content) throws RepositoryException {
          */
         Property dataProperty =
                 contentNode.setProperty(JCR_DATA, node.getSession()
-                        .getValueFactory().createBinary(content));
+                        .getValueFactory().createBinary(dis));
+
+        contentNode.setProperty(CONTENT_SIZE, dataProperty.getLength());
+        contentNode.setProperty(DIGEST_VALUE,  Hex.encodeHexString(dis.getMessageDigest().digest()));
+        contentNode.setProperty(DIGEST_ALGORITHM, dis.getMessageDigest().getAlgorithm());
         logger.debug("Created data property at path: " + dataProperty.getPath());
 
     }
@@ -100,8 +135,12 @@ public void setContent(InputStream content, String mimeType)
      * @throws RepositoryException
      */
     public long getContentSize() throws RepositoryException {
-        return node.getNode(JCR_CONTENT).getProperty(JCR_DATA).getBinary()
-                .getSize();
+        return node.getNode(JCR_CONTENT).getProperty(CONTENT_SIZE).getLong();
+    }
+
+    public URI getContentDigest() throws RepositoryException {
+        final Node contentNode = node.getNode(JCR_CONTENT);
+        return ContentDigest.asURI(contentNode.getProperty(DIGEST_ALGORITHM).getString(), contentNode.getProperty(DIGEST_VALUE).getString());
     }
 
     /**

diff --git a/fcrepo-kernel/src/main/java/org/fcrepo/utils/ContentDigest.java b/fcrepo-kernel/src/main/java/org/fcrepo/utils/ContentDigest.java
@@ -0,0 +1,27 @@
+package org.fcrepo.utils;
+
+import java.net.URI;
+import java.net.URISyntaxException;
+import java.util.HashMap;
+import java.util.Map;
+
+public class ContentDigest {
+    public final static Map<String, String> algorithmToScheme;
+
+    static
+    {
+        algorithmToScheme = new HashMap<String, String>();
+        algorithmToScheme.put("SHA-1", "urn:sha1");
+    }
+
+    public static URI asURI(String algorithm, String value) {
+        try {
+            String scheme = algorithmToScheme.get(algorithm);
+
+            return new URI(scheme, value, null);
+        } catch (URISyntaxException e) {
+            e.printStackTrace();
+            return null;
+        }
+    }
+}
diff --git a/fcrepo-kernel/src/main/resources/fedora-node-types.cnd b/fcrepo-kernel/src/main/resources/fedora-node-types.cnd
@@ -84,3 +84,11 @@
 	 */
 	- foaf:name (STRING) multiple COPY
 
+
+/*
+ * Some content that can have a checksum
+ */
+[fedora:checksum] mixin
+  - fedora:size (LONG) COPY
+  - fedora:digest (STRING) COPY
+  - fedora:digestAlgorithm (STRING) COPY
diff --git a/fcrepo-kernel/src/test/java/org/fcrepo/DatastreamTest.java b/fcrepo-kernel/src/test/java/org/fcrepo/DatastreamTest.java
@@ -9,13 +9,15 @@
 
 import java.io.ByteArrayInputStream;
 import java.io.IOException;
+import java.io.StringWriter;
 
 import javax.inject.Inject;
 import javax.jcr.Node;
 import javax.jcr.Repository;
 import javax.jcr.RepositoryException;
 import javax.jcr.Session;
 
+import org.apache.commons.io.IOUtils;
 import org.junit.Test;
 import org.springframework.test.context.ContextConfiguration;
 
@@ -57,4 +59,69 @@ public void testCreatedDate() throws RepositoryException, IOException {
         final Datastream ds = getDatastream("testObject", "testDatastreamNode1");
         assertNotNull("Couldn't find created date on datastream!", ds.getCreatedDate());
     }
+
+    @Test
+    public void testDatastreamContent() throws IOException, RepositoryException {
+        Session session = repo.login();
+        createObjectNode(session, "testObject");
+        createDatastreamNode(session,
+                "/objects/testObject/testDatastreamNode1",
+                "application/octet-stream", new ByteArrayInputStream(
+                "asdf".getBytes()));
+
+        session.save();
+
+        final Datastream ds = getDatastream("testObject", "testDatastreamNode1");
+        String contentString = IOUtils.toString(ds.getContent(), "ASCII");
+
+        assertEquals("asdf", contentString);
+
+    }
+
+    @Test
+    public void testDatastreamContentDigestAndLength() throws IOException, RepositoryException {
+        Session session = repo.login();
+        createObjectNode(session, "testObject");
+        createDatastreamNode(session,
+                "/objects/testObject/testDatastreamNode2",
+                "application/octet-stream", new ByteArrayInputStream(
+                "asdf".getBytes()));
+
+
+        session.save();
+
+        final Datastream ds = getDatastream("testObject", "testDatastreamNode2");
+        assertEquals("urn:sha1:3da541559918a808c2402bba5012f6c60b27661c", ds.getContentDigest().toString());
+        assertEquals(4L, ds.getContentSize());
+
+
+        String contentString = IOUtils.toString(ds.getContent(), "ASCII");
+
+        assertEquals("asdf", contentString);
+    }
+
+    @Test
+    public void testModifyDatastreamContentDigestAndLength() throws IOException, RepositoryException {
+        Session session = repo.login();
+        createObjectNode(session, "testObject");
+        createDatastreamNode(session,
+                "/objects/testObject/testDatastreamNode3",
+                "application/octet-stream", new ByteArrayInputStream(
+                "asdf".getBytes()));
+
+
+        session.save();
+
+        final Datastream ds = getDatastream("testObject", "testDatastreamNode3");
+        ds.setContent(new ByteArrayInputStream(
+                "0123456789".getBytes()));
+
+        assertEquals("urn:sha1:87acec17cd9dcd20a716cc2cf67417b71c8a7016", ds.getContentDigest().toString());
+        assertEquals(10L, ds.getContentSize());
+
+
+        String contentString = IOUtils.toString(ds.getContent(), "ASCII");
+
+        assertEquals("0123456789", contentString);
+    }
 }