PDFBOX-5902: provide singleton instances for heavily used Integer and byte[] values to avoid multiple instances

git-svn-id: https://svn.apache.org/repos/asf/pdfbox/trunk@1922578 13f79535-47bb-0310-9956-ffa450edef68

Author: lehmi

fontbox/src/main/java/org/apache/fontbox/cmap/CMap.java

@@ -332,25 +332,6 @@ private int toCIDFromRanges(byte[] code)
         return 0;
     }
 
-    /**
-     * Convert the given part of a byte array to an integer.
-     * 
-     * @param data   the byte array
-     * @param offset The offset into the byte array.
-     * @param length The length of the data we are getting.
-     * @return the resulting integer
-     */
-    private int getCodeFromArray( byte[] data, int offset, int length )
-    {
-        int code = 0;
-        for( int i=0; i<length; i++ )
-        {
-            code <<= 8;
-            code |= (data[offset+i]+256)%256;
-        }
-        return code;
-    }
-
     /**
      * This will add a character code to Unicode character sequence mapping.
      *
@@ -359,15 +340,15 @@ private int getCodeFromArray( byte[] data, int offset, int length )
      */
     void addCharMapping(byte[] codes, String unicode)
     {
-        unicodeToByteCodes.put(unicode, codes.clone()); // clone needed, bytes is modified later
-        int code = getCodeFromArray(codes, 0, codes.length);
         if (codes.length == 1)
         {
-            charToUnicodeOneByte.put(code, unicode);
+            charToUnicodeOneByte.put(CMapStrings.getIndexValue(codes), unicode);
+            unicodeToByteCodes.put(unicode, CMapStrings.getByteValue(codes));
         }
         else if (codes.length == 2)
         {
-            charToUnicodeTwoBytes.put(code, unicode);
+            charToUnicodeTwoBytes.put(CMapStrings.getIndexValue(codes), unicode);
+            unicodeToByteCodes.put(unicode, CMapStrings.getByteValue(codes));
         }
         else
         {
@@ -376,7 +357,7 @@ else if (codes.length == 2)
         // fixme: ugly little hack
         if (SPACE.equals(unicode))
         {
-            spaceMapping = code;
+            spaceMapping = toInt(codes);
         }
     }
 

fontbox/src/main/java/org/apache/fontbox/cmap/CMapStrings.java

@@ -30,6 +30,10 @@ public class CMapStrings
     private static final List<String> twoByteMappings = new ArrayList<>(256 * 256);
     private static final List<String> oneByteMappings = new ArrayList<>(256);
 
+    private static final List<Integer> indexValues = new ArrayList<>(256 * 256);
+    private static final List<byte[]> oneByteValues = new ArrayList<>(256);
+    private static final List<byte[]> twoByteValues = new ArrayList<>(256 * 256);
+
     static
     {
         // create all mappings when loading the class to avoid concurrency issues
@@ -48,12 +52,15 @@ private static void fillMappings()
             {
                 byte[] bytes = { (byte) i, (byte) j };
                 twoByteMappings.add(new String(bytes, StandardCharsets.UTF_16BE));
+                twoByteValues.add(bytes);
+                indexValues.add((i * 256) + j);
             }
         }
         for (int i = 0; i < 256; i++)
         {
             byte[] bytes = { (byte) i };
             oneByteMappings.add(new String(bytes, StandardCharsets.ISO_8859_1));
+            oneByteValues.add(bytes);
         }
     }
 
@@ -73,4 +80,39 @@ public static String getMapping(byte[] bytes)
         return bytes.length == 1 ? oneByteMappings.get(CMap.toInt(bytes))
                 : twoByteMappings.get(CMap.toInt(bytes));
     }
+
+    /**
+     * Get an Integer instance of the given combination of bytes. Each value is a singleton to avoid multiple instances
+     * for same value. The values are limited to one and two-byte sequences. Any longer byte sequence produces null as
+     * return value.
+     * 
+     * @param bytes the given combination of bytes
+     * @return the Integer representation for the given combination of bytes
+     */
+    public static Integer getIndexValue(byte[] bytes)
+    {
+        if (bytes.length > 2)
+        {
+            return null;
+        }
+        return indexValues.get(CMap.toInt(bytes));
+    }
+
+    /**
+     * Get a singleton instance of the given combination of bytes to avoid multiple instances for same value. The values
+     * are limited to one and two-byte sequences. Any longer byte sequence produces null as return value.
+     * 
+     * @param bytes the given combination of bytes
+     * @return a singleton instance for the given combination of bytes
+     */
+    public static byte[] getByteValue(byte[] bytes)
+    {
+        if (bytes.length > 2)
+        {
+            return null;
+        }
+        return bytes.length == 1 ? oneByteValues.get(CMap.toInt(bytes))
+                : twoByteValues.get(CMap.toInt(bytes));
+    }
+
 }