HDFS-7287. The OfflineImageViewer (OIV) can output invalid XML depending on the filename (Ravi Prakash via Colin P. McCabe)

(cherry picked from commit d33e07dc49e00db138921fb3aa52c4ef00510161)
(cherry picked from commit 112d69408630068f297616f7bac4b106f5eaa402)

hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt

@@ -850,6 +850,9 @@ Release 2.6.0 - UNRELEASED
     HDFS-7128. Decommission slows way down when it gets towards the end.
     (Ming Ma via cnauroth)
 
+    HDFS-7287. The OfflineImageViewer (OIV) can output invalid XML depending on
+    the filename (Ravi Prakash via Colin P. McCabe)
+
   BREAKDOWN OF HDFS-6584 ARCHIVAL STORAGE
 
     HDFS-6677. Change INodeFile and FSImage to support storage policy ID.

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/tools/offlineImageViewer/PBImageXmlWriter.java

@@ -49,6 +49,7 @@ import org.apache.hadoop.hdfs.server.namenode.FsImageProto.SecretManagerSection;
 import org.apache.hadoop.hdfs.server.namenode.FsImageProto.SnapshotDiffSection;
 import org.apache.hadoop.hdfs.server.namenode.FsImageProto.SnapshotSection;
 import org.apache.hadoop.hdfs.server.namenode.FsImageProto.StringTableSection;
+import org.apache.hadoop.hdfs.util.XMLUtils;
 import org.apache.hadoop.io.IOUtils;
 
 import com.google.common.collect.Lists;
@@ -410,7 +411,7 @@ public final class PBImageXmlWriter {
   }
 
   private PBImageXmlWriter o(final String e, final Object v) {
-    out.print("<" + e + ">" + v + "</" + e + ">");
+    out.print("<" + e + ">" + XMLUtils.mangleXmlString(v.toString()) + "</" + e + ">");
     return this;
   }
 }

hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/tools/offlineImageViewer/XmlImageVisitor.java

@@ -20,6 +20,7 @@ package org.apache.hadoop.hdfs.tools.offlineImageViewer;
 import java.io.IOException;
 import java.util.LinkedList;
 
+import org.apache.hadoop.hdfs.util.XMLUtils;
 /**
  * An XmlImageVisitor walks over an fsimage structure and writes out
  * an equivalent XML document that contains the fsimage's components.
@@ -83,6 +84,6 @@ public class XmlImageVisitor extends TextWriterImageVisitor {
   }
 
   private void writeTag(String tag, String value) throws IOException {
-    write("<" + tag + ">" + value + "</" + tag + ">\n");
+    write("<" + tag + ">" + XMLUtils.mangleXmlString(value) + "</" + tag + ">\n");
   }
 }

hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/tools/offlineImageViewer/TestOfflineImageViewer.java

@@ -23,6 +23,7 @@ import static org.junit.Assert.assertTrue;
 import java.io.File;
 import java.io.FileInputStream;
 import java.io.FileOutputStream;
+import java.io.FileWriter;
 import java.io.IOException;
 import java.io.PrintWriter;
 import java.io.RandomAccessFile;
@@ -125,6 +126,10 @@ public class TestOfflineImageViewer {
       hdfs.mkdirs(emptydir);
       writtenFiles.put(emptydir.toString(), hdfs.getFileStatus(emptydir));
 
+      //Create a directory whose name should be escaped in XML
+      Path invalidXMLDir = new Path("/dirContainingInvalidXMLChar\u0000here");
+      hdfs.mkdirs(invalidXMLDir);
+
       // Get delegation tokens so we log the delegation token op
       Token<?>[] delegationTokens = hdfs
           .addDelegationTokens(TEST_RENEWER, null);
@@ -220,7 +225,7 @@ public class TestOfflineImageViewer {
     assertTrue(matcher.find() && matcher.groupCount() == 1);
     int totalDirs = Integer.parseInt(matcher.group(1));
     // totalDirs includes root directory, empty directory, and xattr directory
-    assertEquals(NUM_DIRS + 3, totalDirs);
+    assertEquals(NUM_DIRS + 4, totalDirs);
 
     FileStatus maxFile = Collections.max(writtenFiles.values(),
         new Comparator<FileStatus>() {
@@ -273,7 +278,7 @@ public class TestOfflineImageViewer {
 
       // verify the number of directories
       FileStatus[] statuses = webhdfs.listStatus(new Path("/"));
-      assertEquals(NUM_DIRS + 2, statuses.length); // contains empty and xattr directory
+      assertEquals(NUM_DIRS + 3, statuses.length); // contains empty and xattr directory
 
       // verify the number of files in the directory
       statuses = webhdfs.listStatus(new Path("/dir0"));