From bf5b508074955d8ddcf3da5ac65cf72131526333 Mon Sep 17 00:00:00 2001
From: Tsz-wo Sze <szetszwo@apache.org>
Date: Thu, 5 Apr 2012 18:28:34 +0000
Subject: [PATCH] HADOOP-6963. In FileUtil.getDU(..), neither include the size
 of directories nor follow symbolic links.  Contributed by Ravi Prakash

git-svn-id: https://svn.apache.org/repos/asf/hadoop/common/trunk@1309994 13f79535-47bb-0310-9956-ffa450edef68
---
 hadoop-common-project/hadoop-common/CHANGES.txt |  3 +++
 hadoop-common-project/hadoop-common/pom.xml     |  5 +++++
 .../java/org/apache/hadoop/fs/FileUtil.java     | 11 +++++++++--
 .../java/org/apache/hadoop/fs/TestFileUtil.java | 17 +++++++++++++++++
 4 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/hadoop-common-project/hadoop-common/CHANGES.txt b/hadoop-common-project/hadoop-common/CHANGES.txt
index 8e0f2c3db97..bac3f4a4a21 100644
--- a/hadoop-common-project/hadoop-common/CHANGES.txt
+++ b/hadoop-common-project/hadoop-common/CHANGES.txt
@@ -324,6 +324,9 @@ Release 2.0.0 - UNRELEASED
     HADOOP-8243. Security support broken in CLI (manual) failover controller
     (todd)
 
+    HADOOP-6963. In FileUtil.getDU(..), neither include the size of directories
+    nor follow symbolic links.  (Ravi Prakash via szetszwo)
+
   BREAKDOWN OF HADOOP-7454 SUBTASKS
 
     HADOOP-7455. HA: Introduce HA Service Protocol Interface. (suresh)
diff --git a/hadoop-common-project/hadoop-common/pom.xml b/hadoop-common-project/hadoop-common/pom.xml
index 41305460501..8b97faf37c9 100644
--- a/hadoop-common-project/hadoop-common/pom.xml
+++ b/hadoop-common-project/hadoop-common/pom.xml
@@ -78,6 +78,11 @@
       <artifactId>commons-net</artifactId>
       <scope>compile</scope>
     </dependency>
+    <dependency>
+      <groupId>commons-io</groupId>
+      <artifactId>commons-io</artifactId>
+      <scope>compile</scope>
+    </dependency>
     <dependency>
       <groupId>javax.servlet</groupId>
       <artifactId>servlet-api</artifactId>
diff --git a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileUtil.java b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileUtil.java
index 8e7aa302a61..ba9bb4eafee 100644
--- a/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileUtil.java
+++ b/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/fs/FileUtil.java
@@ -483,11 +483,18 @@ public class FileUtil {
     if (!dir.isDirectory()) {
       return dir.length();
     } else {
-      size = dir.length();
       File[] allFiles = dir.listFiles();
       if(allFiles != null) {
          for (int i = 0; i < allFiles.length; i++) {
-            size = size + getDU(allFiles[i]);
+           boolean isSymLink;
+           try {
+             isSymLink = org.apache.commons.io.FileUtils.isSymlink(allFiles[i]);
+           } catch(IOException ioe) {
+             isSymLink = true;
+           }
+           if(!isSymLink) {
+             size += getDU(allFiles[i]);
+           }
          }
       }
       return size;
diff --git a/hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/fs/TestFileUtil.java b/hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/fs/TestFileUtil.java
index 65c43435b75..0b74baeb6e7 100644
--- a/hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/fs/TestFileUtil.java
+++ b/hadoop-common-project/hadoop-common/src/test/java/org/apache/hadoop/fs/TestFileUtil.java
@@ -93,6 +93,9 @@ public class TestFileUtil {
     // create files in partitioned directories
     createFile(partitioned, "part-r-00000", "foo");
     createFile(partitioned, "part-r-00001", "bar");
+
+    // create a cycle using symlinks. Cycles should be handled
+    FileUtil.symLink(del.toString(), dir1.toString() + "/cycle");
   }
 
   /**
@@ -458,4 +461,18 @@ public class TestFileUtil {
 
     return result;
   }
+
+  /**
+   * Test that getDU is able to handle cycles caused due to symbolic links
+   * and that directory sizes are not added to the final calculated size
+   * @throws IOException
+   */
+  @Test
+  public void testGetDU() throws IOException {
+    setupDirs();
+
+    long du = FileUtil.getDU(TEST_DIR);
+    //Only two files (in partitioned) have 4 bytes each
+    Assert.assertEquals(du, 8);
+  }
 }