SOLR-18315: Optimize UTF conversions by using standard Java APIs

changelog/unreleased/SOLR-18315-optimize-utf-conversion.yml

@@ -0,0 +1,7 @@
+title: Replace custom code for UTF8-UTF16 conversions by standard Java API for better performances.
+type: changed
+authors:
+  - name: Pierre Salagnac
+links:
+  - name: SOLR-18315
+    url: https://issues.apache.org/jira/browse/SOLR-18315

solr/benchmark/src/java/org/apache/solr/bench/search/RequestWriters.java

@@ -60,7 +60,7 @@ public class RequestWriters {
   @State(Scope.Benchmark)
   public static class BenchState {
 
-    @Param({"xml", "binary"})
+    @Param({"xml", "javabin"})
     String type;
 
     @Param({"10", "100", "1000", "10000"})

solr/core/src/test/org/apache/solr/response/TestJavaBinResponseWriter.java

@@ -18,6 +18,7 @@
 
 import java.io.ByteArrayInputStream;
 import java.io.ByteArrayOutputStream;
+import java.nio.charset.StandardCharsets;
 import java.util.Locale;
 import java.util.Map;
 import java.util.UUID;
@@ -26,7 +27,6 @@
 import org.apache.solr.common.SolrDocument;
 import org.apache.solr.common.SolrDocumentList;
 import org.apache.solr.common.params.CommonParams;
-import org.apache.solr.common.util.ByteUtils;
 import org.apache.solr.common.util.JavaBinCodec;
 import org.apache.solr.common.util.NamedList;
 import org.apache.solr.request.SolrQueryRequest;
@@ -57,10 +57,10 @@ public void testBytesRefWriting() {
   }
 
   private void compareStringFormat(String input) {
-    byte[] bytes1 = new byte[1024];
-    int len1 = ByteUtils.UTF16toUTF8(input, 0, input.length(), bytes1, 0);
+    byte[] bytes1 = input.getBytes(StandardCharsets.UTF_8);
+    int len1 = bytes1.length;
     BytesRef bytesref = new BytesRef(input);
-    System.out.println();
+
     assertEquals(len1, bytesref.length);
     for (int i = 0; i < len1; i++) {
       assertEquals(input + " not matching char at :" + i, bytesref.bytes[i], bytes1[i]);

solr/solrj/src/java/org/apache/solr/common/util/ByteArrayUtf8CharSequence.java

@@ -17,13 +17,12 @@
 
 package org.apache.solr.common.util;
 
+import java.nio.charset.StandardCharsets;
 import java.util.AbstractMap;
 import java.util.ArrayList;
 import java.util.Collection;
 import java.util.Map;
 import java.util.Set;
-import java.util.function.Function;
-import org.noggit.CharArr;
 
 /**
  * A mutable byte[] backed Utf8CharSequence. This is quite similar to the BytesRef of Lucene Do not
@@ -36,18 +35,13 @@ public class ByteArrayUtf8CharSequence implements Utf8CharSequence {
   protected int offset;
   protected int hashCode = Integer.MIN_VALUE;
   protected int length;
-  protected volatile String utf16;
-  public Function<ByteArrayUtf8CharSequence, String> stringProvider;
+  protected String utf16;
 
   public ByteArrayUtf8CharSequence(String utf16) {
-    buf = new byte[Math.multiplyExact(utf16.length(), 3)];
+    this.utf16 = utf16;
+    buf = utf16.getBytes(StandardCharsets.UTF_8);
     offset = 0;
-    length = ByteUtils.UTF16toUTF8(utf16, 0, utf16.length(), buf, 0);
-    if (buf.length > length) {
-      byte[] copy = new byte[length];
-      System.arraycopy(buf, 0, copy, 0, length);
-      buf = copy;
-    }
+    length = buf.length;
     assert isValid();
   }
 
@@ -154,15 +148,8 @@ public char charAt(int index) {
   }
 
   private String _getStr() {
-    String utf16 = this.utf16;
     if (utf16 == null) {
-      if (stringProvider != null) {
-        this.utf16 = utf16 = stringProvider.apply(this);
-      } else {
-        CharArr arr = new CharArr();
-        ByteUtils.UTF8toUTF16(buf, offset, length, arr);
-        this.utf16 = utf16 = arr.toString();
-      }
+      utf16 = new String(buf, offset, length, StandardCharsets.UTF_8);
     }
     return utf16;
   }