Major speed improvements on BZip2 streams.

PR: 24798
Submitted by:	Joerg Wassmer <joerg dot wassmer at web dot de>


git-svn-id: https://svn.apache.org/repos/asf/ant/core/trunk@278137 13f79535-47bb-0310-9956-ffa450edef68

CONTRIBUTORS

@@ -104,6 +104,7 @@ Jerome Lacoste
 Jesse Glick
 Jesse Stockall
 Jim Allers
+Joerg Wassmer
 Jon Dickinson
 Jon S. Stevens
 Jose Alberto Fernandez
@@ -160,6 +161,7 @@ Nick Pellow
 Nicola Ken Barozzi
 Nico Seessle
 Nigel Magnay
+Oliver Merkel
 Oliver Rossmueller
 Patrick C. Beard
 Patrick Chanezon

src/main/org/apache/tools/bzip2/BZip2Constants.java

@@ -1,5 +1,5 @@
 /*
- * Copyright  2001,2004 The Apache Software Foundation
+ * Copyright  2001,2004-2005 The Apache Software Foundation
  *
  *  Licensed under the Apache License, Version 2.0 (the "License");
  *  you may not use this file except in compliance with the License.
@@ -26,7 +26,10 @@ package org.apache.tools.bzip2;
 /**
  * Base class for both the compress and decompress classes.
  * Holds common arrays, and static data.
- *
+ * <p>
+ * This interface is public for historical purposes.
+ * You should have no need to use it.
+ * </p>
  */
 public interface BZip2Constants {
 
@@ -41,6 +44,13 @@ public interface BZip2Constants {
     int MAX_SELECTORS = (2 + (900000 / G_SIZE));
     int NUM_OVERSHOOT_BYTES = 20;
 
+    /**
+     * This array really shouldn't be here.
+     * Again, for historical purposes it is.
+     *
+     * <p>FIXME: This array should be in a private or package private
+     * location, since it could be modified by malicious code.</p>
+     */
     int[] rNums = {
         619, 720, 127, 481, 931, 816, 813, 233, 566, 247,
         985, 724, 205, 454, 863, 491, 741, 242, 949, 214,

src/main/org/apache/tools/bzip2/CRC.java

@@ -1,5 +1,5 @@
 /*
- * Copyright  2001-2002,2004 The Apache Software Foundation
+ * Copyright  2001-2002,2004-2005 The Apache Software Foundation
  *
  *  Licensed under the Apache License, Version 2.0 (the "License");
  *  you may not use this file except in compliance with the License.
@@ -28,8 +28,8 @@ package org.apache.tools.bzip2;
  * of the data.
  *
  */
-class CRC {
-    public static int crc32Table[] = {
+final class CRC {
+    static final int crc32Table[] = {
         0x00000000, 0x04c11db7, 0x09823b6e, 0x0d4326d9,
         0x130476dc, 0x17c56b6b, 0x1a864db2, 0x1e475005,
         0x2608edb8, 0x22c9f00f, 0x2f8ad6d6, 0x2b4bcb61,
@@ -96,7 +96,7 @@ class CRC {
         0xbcb4666d, 0xb8757bda, 0xb5365d03, 0xb1f740b4
     };
 
-    public CRC() {
+    CRC() {
         initialiseCRC();
     }
 
@@ -124,6 +124,17 @@ class CRC {
         globalCrc = (globalCrc << 8) ^ CRC.crc32Table[temp];
     }
 
+    void updateCRC(int inCh, int repeat) {
+        int globalCrc = this.globalCrc;
+        while (repeat-- > 0) {
+            int temp = (globalCrc >> 24) ^ inCh;
+            globalCrc = (globalCrc << 8) ^ crc32Table[(temp >= 0)
+                                                      ? temp
+                                                      : (temp + 256)];
+        }
+        this.globalCrc = globalCrc;
+    }
+
     int globalCrc;
 }
 

src/testcases/org/apache/tools/ant/taskdefs/BZip2Test.java

@@ -19,7 +19,12 @@ package org.apache.tools.ant.taskdefs;
 
 import org.apache.tools.ant.BuildFileTest;
 import org.apache.tools.ant.util.FileUtils;
+import org.apache.tools.bzip2.CBZip2InputStream;
 
+import java.io.BufferedInputStream;
+import java.io.File;
+import java.io.FileInputStream;
+import java.io.InputStream;
 import java.io.IOException;
 
 /**
@@ -44,9 +49,50 @@ public class BZip2Test extends BuildFileTest {
 
     public void testRealTest() throws IOException {
         executeTarget("realTest");
-        assertTrue("File content mismatch",
-            FILE_UTILS.contentEquals(project.resolveFile("expected/asf-logo-huge.tar.bz2"),
-                                    project.resolveFile("asf-logo-huge.tar.bz2")));
+
+        // doesn't work: Depending on the compression engine used,
+        // compressed bytes may differ. False errors would be
+        // reported.
+        // assertTrue("File content mismatch",
+        // FILE_UTILS.contentEquals(project.resolveFile("expected/asf-logo-huge.tar.bz2"),
+        // project.resolveFile("asf-logo-huge.tar.bz2")));
+
+        // We have to compare the decompressed content instead:
+
+        File originalFile =
+            project.resolveFile("expected/asf-logo-huge.tar.bz2");
+        File actualFile   = project.resolveFile("asf-logo-huge.tar.bz2");
+
+        InputStream originalIn =
+            new BufferedInputStream(new FileInputStream(originalFile));
+        assertEquals((byte) 'B', originalIn.read());
+        assertEquals((byte) 'Z', originalIn.read());
+
+        InputStream actualIn =
+            new BufferedInputStream(new FileInputStream(actualFile));
+        assertEquals((byte) 'B', actualIn.read());
+        assertEquals((byte) 'Z', actualIn.read());
+
+        originalIn = new CBZip2InputStream(originalIn);
+        actualIn   = new CBZip2InputStream(actualIn);
+
+        while (true) {
+            int expected = originalIn.read();
+            int actual   = actualIn.read();
+            if (expected >= 0) {
+                if (expected != actual) {
+                    fail("File content mismatch");
+                }
+            } else {
+                if (actual >= 0) {
+                    fail("File content mismatch");
+                }
+                break;
+            }
+        }
+
+        originalIn.close();
+        actualIn.close();
     }
 
     public void testDateCheck(){