Avoid unintended garbage collection of raw data in PreparedDictionary…

…Impl (#190)

Motivation:

As described in #189 the garbage collection currently removes the data
from the dictionary, resulting in poor compression ratios, since no
dictionary was actually used.
The code to finalize the rawData already exists.

Modification:
Hold the reference on rawData in PreparedDictionaryImpl

Result:
Fixes #189.

---------

Co-authored-by: Benjamin Wollmer <benni@wollmer.dev>
Co-authored-by: Aayush Atharva <aayush@shieldblaze.com>

brotli4j/src/main/java/com/aayushatharva/brotli4j/encoder/EncoderJNI.java

@@ -36,6 +36,7 @@ private static class PreparedDictionaryImpl implements PreparedDictionary {
 
         private PreparedDictionaryImpl(ByteBuffer data, ByteBuffer rawData) {
             this.data = data;
+            this.rawData = rawData;
         }
 
         @Override

brotli4j/src/test/java/com/aayushatharva/brotli4j/encoder/EncoderTest.java

@@ -17,13 +17,16 @@
 package com.aayushatharva.brotli4j.encoder;
 
 import com.aayushatharva.brotli4j.Brotli4jLoader;
+import com.aayushatharva.brotli4j.common.BrotliCommon;
 import io.netty.buffer.ByteBuf;
 import io.netty.buffer.ByteBufUtil;
 import io.netty.buffer.Unpooled;
 import org.junit.jupiter.api.BeforeAll;
 import org.junit.jupiter.api.Test;
 
+import java.io.ByteArrayOutputStream;
 import java.io.IOException;
+import java.util.ArrayList;
 
 import static org.junit.jupiter.api.Assertions.assertArrayEquals;
 import static org.junit.jupiter.api.Assertions.assertEquals;
@@ -77,4 +80,32 @@ void encodeModeEnumValues() {
         assertEquals(Encoder.Mode.TEXT, Encoder.Mode.of(Encoder.Mode.TEXT.ordinal()));
         assertEquals(Encoder.Mode.GENERIC, Encoder.Mode.of(Encoder.Mode.GENERIC.ordinal()));
     }
+
+
+    @Test
+    void ensureDictionaryDataRemainsAfterGC() throws IOException, InterruptedException {
+        // We hard code the compressed data, since the dictionary could also be collected just before our first compression
+        final byte[] expectedCompression = new byte[]{27, 43, 0, -8, 37, 0, -62, -104, -40, -63, 0};
+        final String dictionaryData = "This is some data to be used as a dictionary";
+        final byte[] rawBytes = dictionaryData.getBytes(); // Use dictionary also as data to keep it small
+        final PreparedDictionary dic = Encoder.prepareDictionary(BrotliCommon.makeNative(dictionaryData.getBytes()), 0);
+
+        // Create gc pressure to trigger potential collection of dictionary data
+        ArrayList<Integer> hashes = new ArrayList<>();
+        for (int i = 0; i < 1_000_000; i++) {
+            String obj = String.valueOf(Math.random());
+            hashes.add(obj.hashCode());
+        }
+        hashes = null;
+        System.gc();
+
+        try (ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream();
+             BrotliOutputStream brotliOutputStream = new BrotliOutputStream(byteArrayOutputStream)) {
+            brotliOutputStream.attachDictionary(dic);
+            brotliOutputStream.write(rawBytes);
+            brotliOutputStream.close();
+            byteArrayOutputStream.close();
+            assertArrayEquals(expectedCompression, byteArrayOutputStream.toByteArray());  // Otherwise the GC already cleared the data
+        }
+    }
 }