Add support for filename-encodings other than UTF8 to <untar>.

PR: 10504


git-svn-id: https://svn.apache.org/repos/asf/ant/core/trunk@274746 13f79535-47bb-0310-9956-ffa450edef68

WHATSNEW

@@ -453,6 +453,10 @@ Other changes:
 * FileUtils#createTempFile will now create temporary files in the 
   directory pointed to by the property java.io.tmpdir
 
+* <unzip> and friends now supports an optional encoding attribute to
+  enable it to expand archives created with filenames using an encoding
+  other than UTF8.  Bugzilla Report 10504.
+
 Changes from Ant 1.5.2 to Ant 1.5.3
 ===================================
 

docs/manual/CoreTasks/unzip.html

@@ -46,11 +46,25 @@ to perform unarchival upon.
   </tr>
   <tr>
     <td valign="top">compression</td>
-    <td valign="top">compression method for untar.  Allowable values are
-       &quot;none&quot;, &quot;gzip&quot; and &quot;bzip2&quot;.  Default is
-       &quot;none&quot;.</td>
+    <td valign="top"><b>Note:</b> This attribute is only available for
+    the <code>untar</code> task.<br>
+    compression method.  Allowable values are &quot;none&quot;,
+    &quot;gzip&quot; and &quot;bzip2&quot;.  Default is
+    &quot;none&quot;.</td>
     <td valign="top" align="center">No</td>
   </tr>
+  <tr>
+    <td valign="top">encoding</td>
+    <td valign="top"><b>Note:</b> This attribute is not available for
+    the <code>untar</code> task.<br>
+    The character encoding that has been used for filenames
+    inside the zip file.  For a list of possible values see <a
+    href="http://java.sun.com/products/jdk/1.2/docs/guide/internat/encoding.doc.html">http://java.sun.com/products/jdk/1.2/docs/guide/internat/encoding.doc.html</a>.<br>
+    Defaults to &quot;UTF8&quot;, use the magic value
+    <code>native-encoding</code> for the platform's default character
+    encoding.</td>
+    <td align="center" valign="top">No</td>
+  </tr>
 
 </table>
 <h3>Examples</h3>
@@ -89,7 +103,7 @@ to perform unarchival upon.
 </pre></p>
 </blockquote>
 <hr>
-<p align="center">Copyright &copy; 2000-2002 Apache Software Foundation. All rights
+<p align="center">Copyright &copy; 2000-2003 Apache Software Foundation. All rights
 Reserved.</p>
 
 </body>

src/etc/testcases/taskdefs/untar.xml

@@ -41,4 +41,7 @@
     <untar src="." dest="." />
   </target>
 
+  <target name="encoding">
+    <untar src="expected/asf-logo.gif.tar" dest="." encoding="foo"/>
+  </target>
 </project>

src/etc/testcases/taskdefs/unzip.xml

@@ -78,4 +78,13 @@
     </unzip>
   </target>
 
+  <!-- Bugzilla Report 10504 -->
+  <target name="encodingTest">
+    <mkdir dir="unziptestin"/>
+    <touch file="unziptestin/foo"/>
+    <zip zipfile="unziptest.zip" basedir="unziptestin" encoding="UTF16"/>
+    <mkdir dir="unziptestout"/>
+    <unzip src="unziptest.zip" dest="unziptestout" encoding="UTF16"/>
+  </target>
+
 </project>

src/main/org/apache/tools/ant/taskdefs/Expand.java

@@ -95,9 +95,10 @@ public class Expand extends Task {
     private boolean overwrite = true;
     private Vector patternsets = new Vector();
     private Vector filesets = new Vector();
-    private static final byte[] ZIPMARKER = {0x50, 0x4b, 0x03, 0x04};
-    private static final int MARKER_SIZE = ZIPMARKER.length;
-    private static final int MAX_LOOKAHEAD = 50 * 1024; // 50K.
+
+    private static final String NATIVE_ENCODING = "native-encoding";
+
+    private String encoding = "UTF8";
 
     /**
      * Do the work.
@@ -155,7 +156,7 @@ public class Expand extends Task {
         log("Expanding: " + srcF + " into " + dir, Project.MSG_INFO);
         ZipFile zf = null;
         try {
-            zf = new ZipFile(srcF, "UTF8");
+            zf = new ZipFile(srcF, encoding);
             Enumeration enum = zf.getEntries();
             while (enum.hasMoreElements()) {
                 ZipEntry ze = (ZipEntry) enum.nextElement();
@@ -324,4 +325,19 @@ public class Expand extends Task {
         filesets.addElement(set);
     }
 
+    /**
+     * Sets the encoding to assume for file names and comments.
+     *
+     * <p>Set to <code>native-encoding</code> if you want your
+     * platform's native encoding, defaults to UTF8.</p>
+     *
+     * @since Ant 1.6
+     */
+    public void setEncoding(String encoding) {
+        if (NATIVE_ENCODING.equals(encoding)) {
+            encoding = null;
+        }
+        this.encoding = encoding;
+    }
+
 }

src/main/org/apache/tools/ant/taskdefs/Untar.java

@@ -1,7 +1,7 @@
 /*
  * The Apache Software License, Version 1.1
  *
- * Copyright (c) 2000-2002 The Apache Software Foundation.  All rights
+ * Copyright (c) 2000-2003 The Apache Software Foundation.  All rights
  * reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -112,6 +112,17 @@ public class Untar extends Expand {
         compression = method;
     }
 
+    /**
+     * No encoding support in Untar.
+     *
+     * @since Ant 1.6
+     */
+    public void setEncoding(String encoding) {
+        throw new BuildException("The " + getTaskName() 
+                                 + " task doesn't support the encoding"
+                                 + " attribute", getLocation());
+    }
+
     protected void expandFile(FileUtils fileUtils, File srcF, File dir) {
         TarInputStream tis = null;
         try {

src/testcases/org/apache/tools/ant/taskdefs/UntarTest.java

@@ -1,7 +1,7 @@
 /*
  * The Apache Software License, Version 1.1
  *
- * Copyright (c) 2001-2002 The Apache Software Foundation.  All rights
+ * Copyright (c) 2001-2003 The Apache Software Foundation.  All rights
  * reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -53,7 +53,6 @@
  */
 package org.apache.tools.ant.taskdefs;
 
-import java.io.File;
 import org.apache.tools.ant.BuildFileTest;
 import org.apache.tools.ant.util.FileUtils;
 
@@ -117,8 +116,15 @@ public class UntarTest extends BuildFileTest {
                                            project.resolveFile("asf-logo.gif")));
     }
 
-    public void testSrcDirTest() throws java.io.IOException {
-        FileUtils fileUtils = FileUtils.newFileUtils();
+    public void testSrcDirTest() {
         expectBuildException("srcDirTest", "Src cannot be a directory.");
     }
+
+    public void testEncoding() {
+        expectSpecificBuildException("encoding", 
+                                     "<untar> overrides setEncoding.",
+                                     "The untar task doesn't support the "
+                                     + "encoding attribute");
+    }
+
 }

src/testcases/org/apache/tools/ant/taskdefs/UnzipTest.java

@@ -153,4 +153,13 @@ public class UnzipTest extends BuildFileTest {
                    getProject().resolveFile("unziptestout/2/bar").exists());
     }
 
+    /*
+     * PR 10504
+     */
+    public void testEncoding() {
+        executeTarget("encodingTest");
+        assertTrue("foo has been properly named",
+                   getProject().resolveFile("unziptestout/foo").exists());
+    }
+
 }