StringUtils.toUpperEnglish now has a small cache. This should speed up reading…

StringUtils.toUpperEnglish now has a small cache. This should speed up reading from a ResultSet when using the column name.

h2/src/docsrc/html/changelog.html

@@ -17,7 +17,9 @@ Change Log
 <h1>Change Log</h1>
 
 <h2>Next Version (unreleased)</h2>
-<ul><li>MVStore: up to 65535 open transactions are now supported.
+<ul><li>StringUtils.toUpperEnglish now has a small cache.
+    This should speed up reading from a ResultSet when using the column name.
+</li><li>MVStore: up to 65535 open transactions are now supported.
     Previously, the limit was at most 65535 transactions between the oldest open and the
     newest open transaction (which was quite a strange limit).
 </li><li>The default limit for in-place LOB objects was changed from 128 to 256 bytes.

h2/src/main/org/h2/util/StringUtils.java

@@ -23,9 +23,16 @@ public class StringUtils {
     private static SoftReference<String[]> softCache =
             new SoftReference<String[]>(null);
     private static long softCacheCreated;
+    
     private static final char[] HEX = "0123456789abcdef".toCharArray();
     private static final int[] HEX_DECODE = new int['f' + 1];
 
+    // memory used by this cache: 
+    // 4 * 1024 * 2 (strings per pair) * 64 * 2 (bytes per char) = 0.5 MB
+    private static final int TO_UPPER_CACHE_LENGTH = 2 * 1024;
+    private static final int TO_UPPER_CACHE_MAX_ENTRY_LENGTH = 64;
+    private static final String[][] TO_UPPER_CACHE = new String[TO_UPPER_CACHE_LENGTH][];
+
     static {
         for (int i = 0; i < HEX_DECODE.length; i++) {
             HEX_DECODE[i] = -1;
@@ -89,8 +96,21 @@ public class StringUtils {
      * @return the uppercase text
      */
     public static String toUpperEnglish(String s) {
+        if (s.length() > TO_UPPER_CACHE_MAX_ENTRY_LENGTH) {
             return s.toUpperCase(Locale.ENGLISH);
         }
+        int index = s.hashCode() & (TO_UPPER_CACHE_LENGTH - 1);
+        String[] e = TO_UPPER_CACHE[index];
+        if (e != null) {
+            if (e[0].equals(s)) {
+                return e[1];
+            }
+        }
+        String s2 = s.toUpperCase(Locale.ENGLISH);
+        e = new String[] { s, s2 };
+        TO_UPPER_CACHE[index] = e;
+        return s2;
+    }
 
     /**
      * Convert a string to lowercase using the English locale.

h2/src/test/org/h2/test/unit/TestStringCache.java

@@ -5,6 +5,7 @@
  */
 package org.h2.test.unit;
 
+import java.util.Locale;
 import java.util.Random;
 
 import org.h2.test.TestBase;
@@ -31,12 +32,14 @@ public class TestStringCache extends TestBase {
      *
      * @param args the command line parameters
      */
-    public static void main(String... args) {
+    public static void main(String... args) throws Exception {
+        TestBase.createCaller().init().test();
         new TestStringCache().runBenchmark();
     }
     
     @Override
     public void test() throws InterruptedException {
+        testToUpperToLower();
         returnNew = true;
         StringUtils.clearCache();
         testSingleThread(getSize(5000, 20000));
@@ -47,7 +50,61 @@ public class TestStringCache extends TestBase {
         testMultiThreads();
     }
     
+    private void testToUpperCache() {
+        Random r = new Random();
+        String[] test = new String[50];
+        for (int i = 0; i < test.length; i++) {
+            StringBuilder buff = new StringBuilder();
+            for (int a = 0; a < 50; a++) {
+                buff.append((char) r.nextInt());
+            }
+            String a = buff.toString();
+            test[i] = a;
+        }
+        int repeat = 100000;
+        int testLen = 0;
+        long time = System.currentTimeMillis();
+        for (int a = 0; a < repeat; a++) {
+            for (String x : test) {
+                String y = StringUtils.toUpperEnglish(x);
+                testLen += y.length();
+            }
+        }
+        time = System.currentTimeMillis() - time;
+        System.out.println("cache " + time);
+        time = System.currentTimeMillis();
+        for (int a = 0; a < repeat; a++) {
+            for (String x : test) {
+                String y = x.toUpperCase(Locale.ENGLISH);
+                testLen -= y.length();
+            }
+        }
+        time = System.currentTimeMillis() - time;
+        System.out.println("toUpperCase " + time);
+        assertEquals(0, testLen);
+    }
+    
+    private void testToUpperToLower() {
+        Random r = new Random();
+        for (int i = 0; i < 1000; i++) {
+            StringBuilder buff = new StringBuilder();
+            for (int a = 0; a < 100; a++) {
+                buff.append((char) r.nextInt());
+            }
+            String a = buff.toString();
+            String b = StringUtils.toUpperEnglish(a);
+            String c = a.toUpperCase(Locale.ENGLISH);
+            assertEquals(c, b);
+            String d = StringUtils.toLowerEnglish(a);
+            String e = a.toLowerCase(Locale.ENGLISH);
+            assertEquals(e, d);
+        }
+    }
+
     private void runBenchmark() {
+        testToUpperCache();
+        testToUpperCache();
+        testToUpperCache();
         returnNew = false;
         for (int i = 0; i < 6; i++) {
             useIntern = (i % 2) == 0;