feat: Extract Codec & friends from Util into separate modules (#5727)

This fixes #5726 by removing ObjectCodec and ObjectDecoder, their
implementations, as well as CodecCache and it's exception from the Util
module and into their own separate packages, codec-base, codec-builtin
and codec-cache.

codec/api/build.gradle

@@ -0,0 +1,10 @@
+plugins {
+ id 'java-library'
+ id 'io.deephaven.project.register'
+}
+
+description 'Codec API: The base package for column codecs'
+
+dependencies {
+ compileOnly libs.jetbrains.annotations
+}

codec/api/gradle.properties

@@ -0,0 +1 @@
+io.deephaven.project.ProjectType=JAVA_PUBLIC

Util/src/main/java/io/deephaven/util/codec/ObjectCodec.java → codec/api/src/main/java/io/deephaven/util/codec/ObjectCodec.java

@@ -24,14 +24,12 @@ public interface ObjectCodec<TYPE> extends ObjectDecoder<TYPE> {
 
  /**
  * Encode the specified input as an array of bytes. Note that it is up to the implementation how to encode null
- * inputs. The use of a zero-length byte array (e.g.
- * {@link io.deephaven.datastructures.util.CollectionUtil#ZERO_LENGTH_BYTE_ARRAY}) is strongly encouraged.
+ * inputs. The use of a zero-length byte array is strongly encouraged.
  *
  * @param input The input object, possibly null
  * @return The output byte array
  */
- @NotNull
- byte[] encode(@Nullable TYPE input);
+ byte @NotNull [] encode(@Nullable TYPE input);
 
  /**
  * Does this codec support encoding of null values?

Util/src/main/java/io/deephaven/util/codec/ObjectDecoder.java → codec/api/src/main/java/io/deephaven/util/codec/ObjectDecoder.java

@@ -3,7 +3,6 @@
 //
 package io.deephaven.util.codec;
 
-import io.deephaven.base.verify.Assert;
 import org.jetbrains.annotations.NotNull;
 import org.jetbrains.annotations.Nullable;
 
@@ -36,7 +35,7 @@ public interface ObjectDecoder<TYPE> {
  * @return The output object, possibly null
  */
  @Nullable
- TYPE decode(@NotNull byte[] input, int offset, int length);
+ TYPE decode(byte @NotNull [] input, int offset, int length);
 
  /**
  * Decode an object from a ByteBuffer. The position of the input buffer may or may not be modified by this method.
@@ -72,6 +71,9 @@ default TYPE decode(@NotNull final ByteBuffer buffer) {
  */
  default void checkWidth(int actualWidth) throws IllegalArgumentException {
  final int expectedWidth = expectedObjectWidth();
- Assert.eq(expectedWidth, "expectedWidth", actualWidth, "actualWidth");
+ if (expectedWidth != actualWidth) {
+ throw new IllegalArgumentException(
+ "Expected width `" + expectedWidth + "` does not match actual width `" + actualWidth + "`");
+ }
  }
 }

codec/builtin/build.gradle

@@ -0,0 +1,17 @@
+plugins {
+ id 'java-library'
+ id 'io.deephaven.project.register'
+}
+
+description 'Codec Builtin: Deephaven builtin codec implementations'
+
+dependencies {
+ api project(":codec-api")
+
+ implementation project(":Base")
+ implementation project(":engine-query-constants")
+
+ compileOnly libs.jetbrains.annotations
+
+ testImplementation libs.junit4
+}

codec/builtin/gradle.properties

@@ -0,0 +1 @@
+io.deephaven.project.ProjectType=JAVA_PUBLIC

Util/src/main/java/io/deephaven/util/codec/BigDecimalCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/BigDecimalCodec.java

@@ -3,7 +3,6 @@
 //
 package io.deephaven.util.codec;
 
-import io.deephaven.datastructures.util.CollectionUtil;
 import org.jetbrains.annotations.NotNull;
 import org.jetbrains.annotations.Nullable;
 
@@ -73,18 +72,18 @@ public BigDecimalCodec(@Nullable String arguments) {
  try {
  int _precision = 0, _scale = 0; // zero indicates unlimited precision/scale, variable width encoding
  boolean _strict = true;
- if (arguments != null && arguments.trim().length() > 0) {
+ if (arguments != null && !arguments.trim().isEmpty()) {
  final String[] tokens = arguments.split(",");
- if (tokens.length > 0 && tokens[0].trim().length() > 0) {
+ if (tokens.length > 0 && !tokens[0].trim().isEmpty()) {
  _precision = Integer.parseInt(tokens[0].trim());
  if (_precision < 1) {
  throw new IllegalArgumentException("Specified precision must be >= 1");
  }
  }
- if (tokens.length > 1 && tokens[1].trim().length() > 0) {
+ if (tokens.length > 1 && !tokens[1].trim().isEmpty()) {
  _scale = Integer.parseInt(tokens[1].trim());
  }
- if (tokens.length > 2 && tokens[2].trim().length() > 0) {
+ if (tokens.length > 2 && !tokens[2].trim().isEmpty()) {
  String mode = tokens[2].trim();
  switch (mode.toLowerCase()) {
  case "allowrounding":
@@ -141,13 +140,12 @@ private void init() {
  final byte[] unscaledZero = BigDecimal.ZERO.unscaledValue().toByteArray();
  zeroBytes = new byte[Integer.BYTES + unscaledZero.length];
  Arrays.fill(zeroBytes, (byte) 0);
- nullBytes = CollectionUtil.ZERO_LENGTH_BYTE_ARRAY;
+ nullBytes = CodecUtil.ZERO_LENGTH_BYTE_ARRAY;
  }
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable final BigDecimal input) {
+ public byte @NotNull [] encode(@Nullable final BigDecimal input) {
  if (input == null) {
  return nullBytes;
  }
@@ -173,7 +171,7 @@ public byte[] encode(@Nullable final BigDecimal input) {
  // (i.e. too high a scale requires reducing precision to "make room")
  if ((value.precision() > this.precision || value.scale() > scale)) {
  if (strict) {
- throw new IllegalArgumentException("Unable to encode value " + value.toString() + " with precision "
+ throw new IllegalArgumentException("Unable to encode value " + value + " with precision "
  + precision + " scale " + scale);
  }
  final int targetPrecision = Math.min(precision, value.precision() - Math.max(0, value.scale() - scale));
@@ -194,7 +192,7 @@ public byte[] encode(@Nullable final BigDecimal input) {
  // copy unscaled bytes to proper size array
  final byte[] unscaledValue = value.unscaledValue().toByteArray();
  if (unscaledValue.length >= bytes.length) { // unscaled value must be at most one less than length of our buffer
- throw new IllegalArgumentException("Value " + input.toString() + " is too large to encode with precision "
+ throw new IllegalArgumentException("Value " + input + " is too large to encode with precision "
  + precision + " and scale " + scale);
  }
 
@@ -213,7 +211,7 @@ public byte[] encode(@Nullable final BigDecimal input) {
 
  @Nullable
  @Override
- public BigDecimal decode(@NotNull final byte[] input, final int offset, final int length) {
+ public BigDecimal decode(final byte @NotNull [] input, final int offset, final int length) {
 
  // variable size value
  if (precision == 0) {

Util/src/main/java/io/deephaven/util/codec/BigIntegerCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/BigIntegerCodec.java

@@ -32,7 +32,7 @@ public BigIntegerCodec(@Nullable String arguments) {
  // noinspection ConstantConditions
  try {
  int _precision = 0; // zero indicates unlimited precision, variable width encoding
- if (arguments != null && arguments.trim().length() > 0) {
+ if (arguments != null && !arguments.trim().isEmpty()) {
  _precision = Integer.parseInt(arguments.trim());
  if (_precision < 1) {
  throw new IllegalArgumentException("Specified precision must be >= 1");
@@ -61,17 +61,16 @@ public int getScale() {
  return 0;
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable final BigInteger input) {
+ public byte @NotNull [] encode(@Nullable final BigInteger input) {
  return input == null
  ? codec.encodedNullValue()
  : codec.encode(new BigDecimal(input));
  }
 
  @Nullable
  @Override
- public BigInteger decode(@NotNull final byte[] input, final int offset, final int length) {
+ public BigInteger decode(final byte @NotNull [] input, final int offset, final int length) {
  final BigDecimal bd = codec.decode(input, offset, length);
  return bd == null ? null : bd.toBigInteger();
  }

Util/src/main/java/io/deephaven/util/EncodingUtil.java → codec/builtin/src/main/java/io/deephaven/util/codec/CodecUtil.java

@@ -1,74 +1,23 @@
 //
 // Copyright (c) 2016-2024 Deephaven Data Labs and Patent Pending
 //
-package io.deephaven.util;
+package io.deephaven.util.codec;
 
 import io.deephaven.base.string.EncodingInfo;
-import org.apache.commons.io.ByteOrderMark;
 import org.jetbrains.annotations.NotNull;
 
 import java.nio.BufferOverflowException;
 import java.nio.BufferUnderflowException;
 import java.nio.ByteBuffer;
 import java.nio.CharBuffer;
 import java.nio.charset.CharacterCodingException;
-import java.nio.charset.Charset;
 import java.nio.charset.CharsetDecoder;
 import java.nio.charset.CharsetEncoder;
-import java.util.Arrays;
 
-public class EncodingUtil {
- @SuppressWarnings("WeakerAccess")
- public static final ByteOrderMark[] EMPTY_BOM_ARRAY = new ByteOrderMark[0];
+class CodecUtil {
+ public static final byte[] ZERO_LENGTH_BYTE_ARRAY = new byte[0];
 
- /**
- * Get the {@link EncodingInfo} associated with a particular {@link Charset}
- *
- * @param charSet The charset
- * @return the matching {@link EncodingInfo}
- * @throws IllegalArgumentException if there is no associated encoding
- */
- @NotNull
- public static EncodingInfo getEncodingInfoForCharset(@NotNull Charset charSet) throws IllegalArgumentException {
- return getEncodingInfoForCharset(charSet.name());
- }
-
- /**
- * Get the {@link EncodingInfo} associated with a particular charset name
- *
- * @param charsetName the charset
- * @return the matching {@link EncodingInfo}
- * @throws IllegalArgumentException if there is no associated encoding
- */
- public static EncodingInfo getEncodingInfoForCharset(@NotNull String charsetName) {
- return Arrays.stream(EncodingInfo.values())
- .filter(info -> info.getCharset().name().equals(charsetName))
- .findFirst()
- .orElseThrow(() -> new IllegalArgumentException("No EncodingInfo for " + charsetName));
- }
-
- /**
- * Get an array containing the possible {@link ByteOrderMark byte order marks} that could be present within a file
- * of the specified encoding. This is intended for use with {@link org.apache.commons.io.input.BOMInputStream}
- *
- * @param encoding The encoding.
- * @return An array containing the possible {@link ByteOrderMark BOMs} for the encoding.
- */
- @NotNull
- public static ByteOrderMark[] getBOMsForEncoding(EncodingInfo encoding) {
- switch (encoding) {
- case UTF_8:
- return new ByteOrderMark[] {ByteOrderMark.UTF_8};
- case UTF_16BE:
- return new ByteOrderMark[] {ByteOrderMark.UTF_16BE};
- case UTF_16LE:
- return new ByteOrderMark[] {ByteOrderMark.UTF_16LE};
- case UTF_16:
- return new ByteOrderMark[] {ByteOrderMark.UTF_16BE, ByteOrderMark.UTF_16LE};
- }
-
- return EMPTY_BOM_ARRAY;
- }
+ private CodecUtil() {}
 
  /**
  * Encode the given string in UTF-8 format into the given ByteBuffer. The string is encoded as an int length

Util/src/main/java/io/deephaven/util/codec/ExternalizableCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/ExternalizableCodec.java

@@ -7,6 +7,7 @@
 import org.jetbrains.annotations.Nullable;
 
 import java.io.*;
+import java.lang.reflect.InvocationTargetException;
 
 public class ExternalizableCodec<T extends Externalizable> implements ObjectCodec<T> {
 
@@ -21,9 +22,8 @@ public ExternalizableCodec(String className) {
  }
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable T input) {
+ public byte @NotNull [] encode(@Nullable T input) {
  if (input == null) {
  throw new UnsupportedOperationException(getClass() + " does not support null input");
  }
@@ -55,16 +55,17 @@ public int getScale() {
 
  @Nullable
  @Override
- public T decode(@NotNull byte[] input, int offset, int length) {
+ public T decode(byte @NotNull [] input, int offset, int length) {
  try {
  final ByteArrayInputStream byteInput = new ByteArrayInputStream(input, offset, length);
  final ObjectInputStream objectInput = new ObjectInputStream(byteInput);
- T result = externalizableClass.newInstance();
+ T result = externalizableClass.getDeclaredConstructor().newInstance();
  result.readExternal(objectInput);
  return result;
  } catch (IOException e) {
  throw new UncheckedIOException(e);
- } catch (IllegalAccessException | InstantiationException | ClassNotFoundException e) {
+ } catch (IllegalAccessException | InstantiationException | ClassNotFoundException | InvocationTargetException
+ | NoSuchMethodException e) {
  throw new RuntimeException(e);
  }
  }

Util/src/main/java/io/deephaven/util/codec/LocalDateCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/LocalDateCodec.java

@@ -104,9 +104,8 @@ public int getScale() {
  return 0;
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable final LocalDate input) {
+ public byte @NotNull [] encode(@Nullable final LocalDate input) {
  if (input == null) {
  if (nullBytes != null) {
  return nullBytes;
@@ -157,7 +156,7 @@ public byte[] encode(@Nullable final LocalDate input) {
 
  @Nullable
  @Override
- public LocalDate decode(@NotNull final byte[] input, final int offset, final int length) {
+ public LocalDate decode(final byte @NotNull [] input, final int offset, final int length) {
  final int year, month, dayOfMonth;
  if (input[offset] == NULL_INDICATOR) {
  return null;

Util/src/main/java/io/deephaven/util/codec/LocalTimeCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/LocalTimeCodec.java

@@ -99,9 +99,8 @@ public int getScale() {
  return 0;
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable final LocalTime input) {
+ public byte @NotNull [] encode(@Nullable final LocalTime input) {
 
  if (input == null) {
  if (nullBytes != null) {
@@ -130,7 +129,7 @@ public byte[] encode(@Nullable final LocalTime input) {
 
  @Nullable
  @Override
- public LocalTime decode(@NotNull final byte[] input, final int offset, final int length) {
+ public LocalTime decode(final byte @NotNull [] input, final int offset, final int length) {
  // test for null indicator (leading bit)
  if ((input[offset] & NULL_INDICATOR) != 0) {
  return null;

Util/src/main/java/io/deephaven/util/codec/MapCodec.java → codec/builtin/src/main/java/io/deephaven/util/codec/MapCodec.java

@@ -3,7 +3,6 @@
 //
 package io.deephaven.util.codec;
 
-import io.deephaven.datastructures.util.CollectionUtil;
 import org.jetbrains.annotations.NotNull;
 import org.jetbrains.annotations.Nullable;
 
@@ -23,7 +22,6 @@
  */
 @SuppressWarnings("unused")
 public abstract class MapCodec<K, V> implements ObjectCodec<Map<K, V>> {
- private static final byte[] nullBytes = CollectionUtil.ZERO_LENGTH_BYTE_ARRAY;
  private static final byte[] zeroBytes = new byte[4];
 
  private static final int MINIMUM_SCRATCH_CAPACITY = 4096;
@@ -47,13 +45,12 @@ public int getScale() {
  return 0;
  }
 
- @NotNull
  @Override
- public byte[] encode(@Nullable final Map<K, V> input) {
+ public byte @NotNull [] encode(@Nullable final Map<K, V> input) {
  if (input == null) {
- return nullBytes;
+ return CodecUtil.ZERO_LENGTH_BYTE_ARRAY;
  }
- if (input.size() == 0) {
+ if (input.isEmpty()) {
  return zeroBytes;
  }
 
@@ -137,7 +134,7 @@ public Map<K, V> decode(@NotNull final ByteBuffer byteBuffer) {
 
  @Nullable
  @Override
- public Map<K, V> decode(@NotNull final byte[] input, final int offset, final int length) {
+ public Map<K, V> decode(final byte @NotNull [] input, final int offset, final int length) {
  if (input.length == 0) {
  return null;
  }