AVRO-3918: enforce network byte order

As stated in RFC 4122 section 4.1.2, UUIDs are in network byte order. Also added a test for string based UUID conversion.
apache · Jan 24, 2024 · 992e3c2 · 992e3c2
1 parent c9f4dea
commit 992e3c2
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 48 deletions.
diff --git a/lang/java/avro/src/main/java/org/apache/avro/Conversions.java b/lang/java/avro/src/main/java/org/apache/avro/Conversions.java
@@ -45,16 +45,6 @@ public class Conversions {
 
   public static class UUIDConversion extends Conversion<UUID> {
 
-    private final boolean isBigEndian;
-
-    public UUIDConversion() {
-      this(ByteOrder.nativeOrder() == ByteOrder.BIG_ENDIAN);
-    }
-
-    public UUIDConversion(final boolean isBigEndian) {
-      this.isBigEndian = isBigEndian;
-    }
-
     @Override
     public Class<UUID> getConvertedType() {
       return UUID.class;
@@ -82,36 +72,18 @@ public CharSequence toCharSequence(UUID value, Schema schema, LogicalType type)
 
     @Override
     public UUID fromFixed(final GenericFixed value, final Schema schema, final LogicalType type) {
-      long mostSigBits = 0;
-      long leastSigBits = 0;
-      byte[] bytes = value.bytes();
-      for (int i = 0; i < Long.BYTES; i++) {
-        mostSigBits |= ((long) (bytes[i] & 255)) << (Byte.SIZE * i);
-        leastSigBits |= ((long) (bytes[i + Long.BYTES] & 255)) << (Byte.SIZE * i);
-      }
-
-      return new UUID(this.convert(mostSigBits), this.convert(leastSigBits));
-    }
-
-    private long convert(long value) {
-      if (this.isBigEndian) {
-        return value;
-      } else {
-        return Long.reverseBytes(value);
-      }
+      ByteBuffer buffer = ByteBuffer.wrap(value.bytes());
+      long mostSigBits = buffer.getLong();
+      long leastSigBits = buffer.getLong();
+      return new UUID(mostSigBits, leastSigBits);
     }
 
     @Override
     public GenericFixed toFixed(final UUID value, final Schema schema, final LogicalType type) {
-      final long mostSigBits = this.convert(value.getMostSignificantBits());
-      final long leastSigBits = this.convert(value.getLeastSignificantBits());
-      byte[] result = new byte[2 * Long.BYTES];
-      for (int i = 0; i < Long.BYTES; i++) {
-        result[i] = (byte) ((mostSigBits >> (i * Byte.SIZE)) & 255);
-        result[i + Long.BYTES] = (byte) ((leastSigBits >> (i * Byte.SIZE)) & 255);
-      }
-
-      return new GenericData.Fixed(schema, result);
+      ByteBuffer buffer = ByteBuffer.allocate(2 * Long.BYTES);
+      buffer.putLong(value.getMostSignificantBits());
+      buffer.putLong(value.getLeastSignificantBits());
+      return new GenericData.Fixed(schema, buffer.array());
     }
   }
 

diff --git a/lang/java/avro/src/test/java/org/apache/avro/TestUuidConversions.java b/lang/java/avro/src/test/java/org/apache/avro/TestUuidConversions.java
@@ -24,6 +24,7 @@
 import org.junit.jupiter.params.provider.Arguments;
 import org.junit.jupiter.params.provider.MethodSource;
 
+import java.math.BigInteger;
 import java.util.UUID;
 import java.util.stream.Stream;
 
@@ -32,29 +33,34 @@ public class TestUuidConversions {
   private Conversions.UUIDConversion uuidConversion = new Conversions.UUIDConversion();
 
   private Schema fixed = Schema.createFixed("fixed", "doc", "", Long.BYTES * 2);
-  private Schema fixedUUid = LogicalTypes.uuid().addToSchema(fixed);
+  private Schema fixedUuid = LogicalTypes.uuid().addToSchema(fixed);
+
+  private Schema string = Schema.createFixed("fixed", "doc", "", Long.BYTES * 2);
+  private Schema stringUuid = LogicalTypes.uuid().addToSchema(string);
 
   @ParameterizedTest
   @MethodSource("uuidData")
   void uuidFixed(UUID uuid) {
-    GenericFixed value = uuidConversion.toFixed(uuid, fixedUUid, LogicalTypes.uuid());
-    UUID uuid1 = uuidConversion.fromFixed(value, fixedUUid, LogicalTypes.uuid());
+    GenericFixed value = uuidConversion.toFixed(uuid, fixedUuid, LogicalTypes.uuid());
+
+    byte[] b = new byte[Long.BYTES];
+    System.arraycopy(value.bytes(), 0, b, 0, b.length);
+    Assertions.assertEquals(uuid.getMostSignificantBits(), new BigInteger(b).longValue());
+    System.arraycopy(value.bytes(), Long.BYTES, b, 0, b.length);
+    Assertions.assertEquals(uuid.getLeastSignificantBits(), new BigInteger(b).longValue());
+
+    UUID uuid1 = uuidConversion.fromFixed(value, fixedUuid, LogicalTypes.uuid());
     Assertions.assertEquals(uuid, uuid1);
   }
 
   @ParameterizedTest
   @MethodSource("uuidData")
-  void uuidFixedReverse(UUID uuid) {
-    Conversions.UUIDConversion bigEndianConversion = new Conversions.UUIDConversion(true);
-    Conversions.UUIDConversion littleEndianConversion = new Conversions.UUIDConversion(false);
+  void uuidCharSequence(UUID uuid) {
+    CharSequence value = uuidConversion.toCharSequence(uuid, stringUuid, LogicalTypes.uuid());
 
-    // simulate uuid from other endian source, assume it's big endian (nevermind if
-    // current jvm is big endian, as we simulate).
-    UUID reverserUUID = new UUID(Long.reverseBytes(uuid.getMostSignificantBits()),
-        Long.reverseBytes(uuid.getLeastSignificantBits()));
-    GenericFixed value = bigEndianConversion.toFixed(reverserUUID, fixedUUid, LogicalTypes.uuid());
+    Assertions.assertEquals(uuid.toString(), value.toString());
 
-    UUID uuid1 = littleEndianConversion.fromFixed(value, fixedUUid, LogicalTypes.uuid());
+    UUID uuid1 = uuidConversion.fromCharSequence(value, stringUuid, LogicalTypes.uuid());
     Assertions.assertEquals(uuid, uuid1);
   }