Changeset 1430 for trunk/docx4j/src/main/java/org/docx4j/openpackaging

Timestamp:

02/21/11 09:51:41 (15 months ago)

Author:

jharrop

Message:

Patch contributed by Alberto. Re the implementation prior to this patch:
"If the part/file is less than 1024 bytes then the capacity of the ByteBuffer? is 1024 (and clear() sets the limit to the capacity). This means, that the file written will be padded
with zero-bytes. In most cases this doesn't matter, but I noticed it while doing the altChunk stuff with a small text-file. In this case, winword will show some garbage after the text. "

File:

• trunk/docx4j/src/main/java/org/docx4j/openpackaging/io/SaveToZipFile.java (modified) (1 diff)

trunk/docx4j/src/main/java/org/docx4j/openpackaging/io/SaveToZipFile.java

@@ -386 +386 @@
                 String resolvedPartUri = part.getPartName().getName().substring(1);
 
-                //InputStream bin = ((BinaryPart)part).getBinaryData();         
-
                 try {
                 // Add ZIP entry to output stream.
                 out.putNextEntry(new ZipEntry(resolvedPartUri));
-                        
-                        // Copy the input stream to the output stream
-//                      byte[] buffer = new byte[8192];
-//                      int amount;
-//                      while (( amount = bin.read(buffer)) >=0 ) {
-//                              log.debug(amount);
-//                              out.write(buffer, 0, amount);                   
-//                      }
+                                
+            java.nio.ByteBuffer bb = ((BinaryPart)part).getBuffer();
+            byte[] bytes = null;
+            bytes = new byte[bb.limit()];
+            bb.get(bytes);              
                 
-                java.nio.ByteBuffer bb = ((BinaryPart)part).getBuffer();
-                bb.clear();
-                byte[] bytes = new byte[bb.capacity()];
-                bb.get(bytes, 0, bytes.length);
-                                
                 out.write( bytes );