Updated to always add the Info-ZIP Unicode Path Extra Field when creating an

archive using an encoding different from UTF-8 instead of only when a name is
not encodeable. Additionally support that extra field when unarchiving.

This closes #39

Author: ChristianSchulte

src/main/java/org/codehaus/plexus/archiver/zip/AbstractZipArchiver.java

@@ -22,6 +22,7 @@
 import org.apache.commons.compress.archivers.zip.ZipEncoding;
 import org.apache.commons.compress.archivers.zip.ZipEncodingHelper;
 import org.apache.commons.compress.parallel.InputStreamSupplier;
+import org.apache.commons.compress.utils.Charsets;
 import org.codehaus.plexus.archiver.AbstractArchiver;
 import org.codehaus.plexus.archiver.ArchiveEntry;
 import org.codehaus.plexus.archiver.Archiver;
@@ -41,6 +42,7 @@
 import java.io.InputStream;
 import java.io.OutputStream;
 import java.io.SequenceInputStream;
+import java.nio.charset.Charset;
 import java.util.Hashtable;
 import java.util.Stack;
 import java.util.concurrent.ExecutionException;
@@ -208,7 +210,6 @@ public boolean isFilesonly()
         return doFilesonly;
     }
 
-
     protected void execute()
         throws ArchiverException, IOException
     {
@@ -315,8 +316,7 @@ private void createArchiveMain()
             zipArchiveOutputStream =
                 new ZipArchiveOutputStream( bufferedOutputStream( fileOutputStream( zipFile, "zip" ) ) );
             zipArchiveOutputStream.setEncoding( encoding );
-            zipArchiveOutputStream.setCreateUnicodeExtraFields(
-                ZipArchiveOutputStream.UnicodeExtraFieldPolicy.NOT_ENCODEABLE );
+            zipArchiveOutputStream.setCreateUnicodeExtraFields( this.getUnicodeExtraFieldPolicy() );
             zipArchiveOutputStream.setMethod(
                 doCompress ? ZipArchiveOutputStream.DEFLATED : ZipArchiveOutputStream.STORED );
 
@@ -339,6 +339,45 @@ private void createArchiveMain()
         success = true;
     }
 
+    /**
+     * Gets the {@code UnicodeExtraFieldPolicy} to apply.
+     *
+     * @return {@link ZipArchiveOutputStream.UnicodeExtraFieldPolicy.NOT_ENCODEABLE}, if the effective encoding is
+     * UTF-8; {@link ZipArchiveOutputStream.UnicodeExtraFieldPolicy.ALWAYS}, if the effective encoding is not
+     * UTF-8.
+     *
+     * @see #getEncoding()
+     */
+    private ZipArchiveOutputStream.UnicodeExtraFieldPolicy getUnicodeExtraFieldPolicy()
+    {
+        // Copied from ZipEncodingHelper.isUTF8()
+        String effectiveEncoding = this.getEncoding();
+
+        if ( effectiveEncoding == null )
+        {
+            effectiveEncoding = Charset.defaultCharset().name();
+        }
+
+        boolean utf8 = Charsets.UTF_8.name().equalsIgnoreCase( effectiveEncoding );
+
+        if ( !utf8 )
+        {
+            for ( String alias : Charsets.UTF_8.aliases() )
+            {
+                if ( alias.equalsIgnoreCase( effectiveEncoding ) )
+                {
+                    utf8 = true;
+                    break;
+                }
+            }
+        }
+
+        return utf8
+                   ? ZipArchiveOutputStream.UnicodeExtraFieldPolicy.NOT_ENCODEABLE
+                   : ZipArchiveOutputStream.UnicodeExtraFieldPolicy.ALWAYS;
+
+    }
+
     /**
      * Add the given resources.
      *

src/main/java/org/codehaus/plexus/archiver/zip/AbstractZipUnArchiver.java

@@ -20,19 +20,21 @@
 import java.io.File;
 import java.io.IOException;
 import java.io.InputStream;
+import java.io.UnsupportedEncodingException;
 import java.net.URL;
 import java.util.Date;
 import java.util.Enumeration;
+import javax.annotation.Nonnull;
 
+import org.apache.commons.compress.archivers.zip.UnicodePathExtraField;
 import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
 import org.apache.commons.compress.archivers.zip.ZipFile;
 import org.apache.commons.compress.utils.IOUtils;
+
 import org.codehaus.plexus.archiver.AbstractUnArchiver;
 import org.codehaus.plexus.archiver.ArchiverException;
 import org.codehaus.plexus.components.io.resources.PlexusIoResource;
 
-import javax.annotation.Nonnull;
-
 /**
  * @author <a href="mailto:[email protected]">Emmanuel Venisse</a>
  * @version $Id$
@@ -84,7 +86,20 @@ private static class ZipEntryFileInfo
 
         public String getName()
         {
-            return zipEntry.getName();
+            try
+            {
+                final UnicodePathExtraField unicodePath =
+                    (UnicodePathExtraField) zipEntry.getExtraField( UnicodePathExtraField.UPATH_ID );
+
+                return unicodePath != null
+                           ? new String( unicodePath.getUnicodeName(), "UTF-8" )
+                           : zipEntry.getName();
+
+            }
+            catch ( final UnsupportedEncodingException e )
+            {
+                throw new AssertionError( e );
+            }
         }
 
         public boolean isDirectory()
@@ -141,27 +156,30 @@ protected void execute()
         InputStream in = null;
         try
         {
-            zf = new org.apache.commons.compress.archivers.zip.ZipFile( getSourceFile(), encoding );
-            final Enumeration e = zf.getEntries();
+            zf = new org.apache.commons.compress.archivers.zip.ZipFile( getSourceFile(), encoding, true );
+            final Enumeration e = zf.getEntriesInPhysicalOrder();
             while ( e.hasMoreElements() )
             {
                 final ZipArchiveEntry ze = (ZipArchiveEntry) e.nextElement();
                 final ZipEntryFileInfo fileInfo = new ZipEntryFileInfo( zf, ze );
-                if ( isSelected( ze.getName(), fileInfo ) )
+                if ( isSelected( fileInfo.getName(), fileInfo ) )
                 {
-					in = zf.getInputStream( ze );
-					extractFileIfIncluded(getSourceFile(), getDestDirectory(), in, ze.getName(),
-							new Date(ze.getTime()), ze.isDirectory(), ze.getUnixMode() != 0 ? ze.getUnixMode() : null,
-                            resolveSymlink( zf, ze ) );
+                    in = zf.getInputStream( ze );
+
+                    extractFileIfIncluded( getSourceFile(), getDestDirectory(), in, fileInfo.getName(),
+                                           new Date( ze.getTime() ), ze.isDirectory(),
+                                           ze.getUnixMode() != 0 ? ze.getUnixMode() : null,
+                                           resolveSymlink( zf, ze ) );
+
                     in.close();
                     in = null;
-				}
-
-			}
+                }
+            }
 
-            getLogger().debug( "expand complete" );
             zf.close();
             zf = null;
+
+            getLogger().debug( "expand complete" );
         }
         catch ( final IOException ioe )
         {
@@ -197,9 +215,9 @@ protected void execute( final String path, final File outputDirectory )
         InputStream in = null;
         try
         {
-            zipFile = new org.apache.commons.compress.archivers.zip.ZipFile( getSourceFile(), encoding );
+            zipFile = new org.apache.commons.compress.archivers.zip.ZipFile( getSourceFile(), encoding, true );
 
-            final Enumeration e = zipFile.getEntries();
+            final Enumeration e = zipFile.getEntriesInPhysicalOrder();
 
             while ( e.hasMoreElements() )
             {
@@ -213,9 +231,12 @@ protected void execute( final String path, final File outputDirectory )
                 if ( ze.getName().startsWith( path ) )
                 {
                     in = zipFile.getInputStream( ze );
+
                     extractFileIfIncluded( getSourceFile(), outputDirectory, in,
                                            ze.getName(), new Date( ze.getTime() ), ze.isDirectory(),
-                                           ze.getUnixMode() != 0 ? ze.getUnixMode() : null, resolveSymlink( zipFile, ze ) );
+                                           ze.getUnixMode() != 0 ? ze.getUnixMode() : null,
+                                           resolveSymlink( zipFile, ze ) );
+
                     in.close();
                     in = null;
                 }