netty5/buffer/src/main/java/io/netty/buffer/ByteBufUtil.java

693 lines
25 KiB
Java
Raw Normal View History

/*
* Copyright 2012 The Netty Project
*
* The Netty Project licenses this file to you under the Apache License,
* version 2.0 (the "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at:
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations
* under the License.
*/
package io.netty.buffer;
import static io.netty.util.internal.ObjectUtil.checkNotNull;
import io.netty.util.ByteProcessor;
import io.netty.util.ByteString;
import io.netty.util.CharsetUtil;
import io.netty.util.Recycler;
import io.netty.util.Recycler.Handle;
import io.netty.util.internal.PlatformDependent;
import io.netty.util.internal.SystemPropertyUtil;
import io.netty.util.internal.logging.InternalLogger;
import io.netty.util.internal.logging.InternalLoggerFactory;
import java.nio.ByteBuffer;
import java.nio.ByteOrder;
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.nio.charset.CoderResult;
import java.util.Arrays;
import java.util.Locale;
/**
* A collection of utility methods that is related with handling {@link ByteBuf},
* such as the generation of hex dump and swapping an integer's byte order.
*/
Revamp the core API to reduce memory footprint and consumption The API changes made so far turned out to increase the memory footprint and consumption while our intention was actually decreasing them. Memory consumption issue: When there are many connections which does not exchange data frequently, the old Netty 4 API spent a lot more memory than 3 because it always allocates per-handler buffer for each connection unless otherwise explicitly stated by a user. In a usual real world load, a client doesn't always send requests without pausing, so the idea of having a buffer whose life cycle if bound to the life cycle of a connection didn't work as expected. Memory footprint issue: The old Netty 4 API decreased overall memory footprint by a great deal in many cases. It was mainly because the old Netty 4 API did not allocate a new buffer and event object for each read. Instead, it created a new buffer for each handler in a pipeline. This works pretty well as long as the number of handlers in a pipeline is only a few. However, for a highly modular application with many handlers which handles connections which lasts for relatively short period, it actually makes the memory footprint issue much worse. Changes: All in all, this is about retaining all the good changes we made in 4 so far such as better thread model and going back to the way how we dealt with message events in 3. To fix the memory consumption/footprint issue mentioned above, we made a hard decision to break the backward compatibility again with the following changes: - Remove MessageBuf - Merge Buf into ByteBuf - Merge ChannelInboundByte/MessageHandler and ChannelStateHandler into ChannelInboundHandler - Similar changes were made to the adapter classes - Merge ChannelOutboundByte/MessageHandler and ChannelOperationHandler into ChannelOutboundHandler - Similar changes were made to the adapter classes - Introduce MessageList which is similar to `MessageEvent` in Netty 3 - Replace inboundBufferUpdated(ctx) with messageReceived(ctx, MessageList) - Replace flush(ctx, promise) with write(ctx, MessageList, promise) - Remove ByteToByteEncoder/Decoder/Codec - Replaced by MessageToByteEncoder<ByteBuf>, ByteToMessageDecoder<ByteBuf>, and ByteMessageCodec<ByteBuf> - Merge EmbeddedByteChannel and EmbeddedMessageChannel into EmbeddedChannel - Add SimpleChannelInboundHandler which is sometimes more useful than ChannelInboundHandlerAdapter - Bring back Channel.isWritable() from Netty 3 - Add ChannelInboundHandler.channelWritabilityChanges() event - Add RecvByteBufAllocator configuration property - Similar to ReceiveBufferSizePredictor in Netty 3 - Some existing configuration properties such as DatagramChannelConfig.receivePacketSize is gone now. - Remove suspend/resumeIntermediaryDeallocation() in ByteBuf This change would have been impossible without @normanmaurer's help. He fixed, ported, and improved many parts of the changes.
2013-05-28 13:40:19 +02:00
public final class ByteBufUtil {
private static final InternalLogger logger = InternalLoggerFactory.getInstance(ByteBufUtil.class);
private static final char[] HEXDUMP_TABLE = new char[256 * 4];
static final ByteBufAllocator DEFAULT_ALLOCATOR;
private static final int THREAD_LOCAL_BUFFER_SIZE;
static {
final char[] DIGITS = "0123456789abcdef".toCharArray();
for (int i = 0; i < 256; i ++) {
2012-11-09 09:26:11 +01:00
HEXDUMP_TABLE[ i << 1 ] = DIGITS[i >>> 4 & 0x0F];
HEXDUMP_TABLE[(i << 1) + 1] = DIGITS[i & 0x0F];
}
String allocType = SystemPropertyUtil.get(
"io.netty.allocator.type", PlatformDependent.isAndroid() ? "unpooled" : "pooled");
allocType = allocType.toLowerCase(Locale.US).trim();
ByteBufAllocator alloc;
if ("unpooled".equals(allocType)) {
alloc = UnpooledByteBufAllocator.DEFAULT;
logger.debug("-Dio.netty.allocator.type: {}", allocType);
} else if ("pooled".equals(allocType)) {
alloc = PooledByteBufAllocator.DEFAULT;
logger.debug("-Dio.netty.allocator.type: {}", allocType);
} else {
alloc = PooledByteBufAllocator.DEFAULT;
logger.debug("-Dio.netty.allocator.type: pooled (unknown: {})", allocType);
}
DEFAULT_ALLOCATOR = alloc;
THREAD_LOCAL_BUFFER_SIZE = SystemPropertyUtil.getInt("io.netty.threadLocalDirectBufferSize", 64 * 1024);
logger.debug("-Dio.netty.threadLocalDirectBufferSize: {}", THREAD_LOCAL_BUFFER_SIZE);
}
/**
* Returns a <a href="http://en.wikipedia.org/wiki/Hex_dump">hex dump</a>
* of the specified buffer's readable bytes.
*/
public static String hexDump(ByteBuf buffer) {
return hexDump(buffer, buffer.readerIndex(), buffer.readableBytes());
}
/**
* Returns a <a href="http://en.wikipedia.org/wiki/Hex_dump">hex dump</a>
* of the specified buffer's sub-region.
*/
public static String hexDump(ByteBuf buffer, int fromIndex, int length) {
if (length < 0) {
throw new IllegalArgumentException("length: " + length);
}
if (length == 0) {
return "";
}
int endIndex = fromIndex + length;
char[] buf = new char[length << 1];
int srcIdx = fromIndex;
int dstIdx = 0;
for (; srcIdx < endIndex; srcIdx ++, dstIdx += 2) {
System.arraycopy(
HEXDUMP_TABLE, buffer.getUnsignedByte(srcIdx) << 1,
buf, dstIdx, 2);
}
return new String(buf);
}
/**
* Returns a <a href="http://en.wikipedia.org/wiki/Hex_dump">hex dump</a>
* of the specified byte array.
*/
public static String hexDump(byte[] array) {
return hexDump(array, 0, array.length);
}
/**
* Returns a <a href="http://en.wikipedia.org/wiki/Hex_dump">hex dump</a>
* of the specified byte array's sub-region.
*/
public static String hexDump(byte[] array, int fromIndex, int length) {
if (length < 0) {
throw new IllegalArgumentException("length: " + length);
}
if (length == 0) {
return "";
}
int endIndex = fromIndex + length;
char[] buf = new char[length << 1];
int srcIdx = fromIndex;
int dstIdx = 0;
for (; srcIdx < endIndex; srcIdx ++, dstIdx += 2) {
System.arraycopy(HEXDUMP_TABLE, (array[srcIdx] & 0xFF) << 1, buf, dstIdx, 2);
}
return new String(buf);
}
/**
* Calculates the hash code of the specified buffer. This method is
* useful when implementing a new buffer type.
*/
public static int hashCode(ByteBuf buffer) {
final int aLen = buffer.readableBytes();
final int intCount = aLen >>> 2;
final int byteCount = aLen & 3;
int hashCode = 1;
int arrayIndex = buffer.readerIndex();
if (buffer.order() == ByteOrder.BIG_ENDIAN) {
for (int i = intCount; i > 0; i --) {
hashCode = 31 * hashCode + buffer.getInt(arrayIndex);
arrayIndex += 4;
}
} else {
for (int i = intCount; i > 0; i --) {
hashCode = 31 * hashCode + swapInt(buffer.getInt(arrayIndex));
arrayIndex += 4;
}
}
for (int i = byteCount; i > 0; i --) {
hashCode = 31 * hashCode + buffer.getByte(arrayIndex ++);
}
if (hashCode == 0) {
hashCode = 1;
}
return hashCode;
}
/**
* Returns {@code true} if and only if the two specified buffers are
* identical to each other for {@code length} bytes starting at {@code aStartIndex}
* index for the {@code a} buffer and {@code bStartIndex} index for the {@code b} buffer.
* A more compact way to express this is:
* <p>
* {@code a[aStartIndex : aStartIndex + length] == b[bStartIndex : bStartIndex + length]}
*/
public static boolean equals(ByteBuf a, int aStartIndex, ByteBuf b, int bStartIndex, int length) {
if (aStartIndex < 0 || bStartIndex < 0 || length < 0) {
throw new IllegalArgumentException("All indexes and lengths must be non-negative");
}
if (a.writerIndex() - length < aStartIndex || b.writerIndex() - length < bStartIndex) {
return false;
}
final int longCount = length >>> 3;
final int byteCount = length & 7;
if (a.order() == b.order()) {
for (int i = longCount; i > 0; i --) {
if (a.getLong(aStartIndex) != b.getLong(bStartIndex)) {
return false;
}
aStartIndex += 8;
bStartIndex += 8;
}
} else {
for (int i = longCount; i > 0; i --) {
if (a.getLong(aStartIndex) != swapLong(b.getLong(bStartIndex))) {
return false;
}
aStartIndex += 8;
bStartIndex += 8;
}
}
for (int i = byteCount; i > 0; i --) {
if (a.getByte(aStartIndex) != b.getByte(bStartIndex)) {
return false;
}
aStartIndex ++;
bStartIndex ++;
}
return true;
}
/**
* Returns {@code true} if and only if the two specified buffers are
* identical to each other as described in {@link ByteBuf#equals(Object)}.
* This method is useful when implementing a new buffer type.
*/
public static boolean equals(ByteBuf bufferA, ByteBuf bufferB) {
final int aLen = bufferA.readableBytes();
if (aLen != bufferB.readableBytes()) {
return false;
}
return equals(bufferA, bufferA.readerIndex(), bufferB, bufferB.readerIndex(), aLen);
}
/**
* Compares the two specified buffers as described in {@link ByteBuf#compareTo(ByteBuf)}.
* This method is useful when implementing a new buffer type.
*/
public static int compare(ByteBuf bufferA, ByteBuf bufferB) {
final int aLen = bufferA.readableBytes();
final int bLen = bufferB.readableBytes();
final int minLength = Math.min(aLen, bLen);
final int uintCount = minLength >>> 2;
final int byteCount = minLength & 3;
int aIndex = bufferA.readerIndex();
int bIndex = bufferB.readerIndex();
if (bufferA.order() == bufferB.order()) {
for (int i = uintCount; i > 0; i --) {
long va = bufferA.getUnsignedInt(aIndex);
long vb = bufferB.getUnsignedInt(bIndex);
if (va > vb) {
return 1;
2012-11-12 01:31:40 +01:00
}
if (va < vb) {
return -1;
}
aIndex += 4;
bIndex += 4;
}
} else {
for (int i = uintCount; i > 0; i --) {
long va = bufferA.getUnsignedInt(aIndex);
long vb = swapInt(bufferB.getInt(bIndex)) & 0xFFFFFFFFL;
if (va > vb) {
return 1;
2012-11-12 01:31:40 +01:00
}
if (va < vb) {
return -1;
}
aIndex += 4;
bIndex += 4;
}
}
for (int i = byteCount; i > 0; i --) {
short va = bufferA.getUnsignedByte(aIndex);
short vb = bufferB.getUnsignedByte(bIndex);
if (va > vb) {
return 1;
2012-11-12 01:31:40 +01:00
}
if (va < vb) {
return -1;
}
aIndex ++;
bIndex ++;
}
return aLen - bLen;
}
/**
* The default implementation of {@link ByteBuf#indexOf(int, int, byte)}.
* This method is useful when implementing a new buffer type.
*/
public static int indexOf(ByteBuf buffer, int fromIndex, int toIndex, byte value) {
if (fromIndex <= toIndex) {
return firstIndexOf(buffer, fromIndex, toIndex, value);
} else {
return lastIndexOf(buffer, fromIndex, toIndex, value);
}
}
/**
* Toggles the endianness of the specified 16-bit short integer.
*/
public static short swapShort(short value) {
return Short.reverseBytes(value);
}
/**
* Toggles the endianness of the specified 24-bit medium integer.
*/
public static int swapMedium(int value) {
int swapped = value << 16 & 0xff0000 | value & 0xff00 | value >>> 16 & 0xff;
if ((swapped & 0x800000) != 0) {
swapped |= 0xff000000;
}
return swapped;
}
/**
* Toggles the endianness of the specified 32-bit integer.
*/
public static int swapInt(int value) {
return Integer.reverseBytes(value);
}
/**
* Toggles the endianness of the specified 64-bit long integer.
*/
public static long swapLong(long value) {
return Long.reverseBytes(value);
}
/**
* Read the given amount of bytes into a new {@link ByteBuf} that is allocated from the {@link ByteBufAllocator}.
*/
public static ByteBuf readBytes(ByteBufAllocator alloc, ByteBuf buffer, int length) {
boolean release = true;
ByteBuf dst = alloc.buffer(length);
try {
buffer.readBytes(dst);
release = false;
return dst;
} finally {
if (release) {
dst.release();
}
}
}
private static int firstIndexOf(ByteBuf buffer, int fromIndex, int toIndex, byte value) {
fromIndex = Math.max(fromIndex, 0);
if (fromIndex >= toIndex || buffer.capacity() == 0) {
return -1;
}
return buffer.forEachByte(fromIndex, toIndex - fromIndex, new ByteProcessor.IndexOfProcessor(value));
}
private static int lastIndexOf(ByteBuf buffer, int fromIndex, int toIndex, byte value) {
fromIndex = Math.min(fromIndex, buffer.capacity());
if (fromIndex < 0 || buffer.capacity() == 0) {
return -1;
}
return buffer.forEachByteDesc(toIndex, fromIndex - toIndex, new ByteProcessor.IndexOfProcessor(value));
}
Provide helper methods in ByteBufUtil to write UTF-8/ASCII CharSequences. Related to [#909] Motivation: We expose no methods in ByteBuf to directly write a CharSequence into it. This leads to have the user either convert the CharSequence first to a byte array or use CharsetEncoder. Both cases have some overheads and we can do a lot better for well known Charsets like UTF-8 and ASCII. Modifications: Add ByteBufUtil.writeAscii(...) and ByteBufUtil.writeUtf8(...) which can do the task in an optimized way. This is especially true if the passed in ByteBuf extends AbstractByteBuf which is true for all of our implementations which not wrap another ByteBuf. Result: Writing an ASCII and UTF-8 CharSequence into a AbstractByteBuf is a lot faster then what the user could do by himself as we can make use of some package private methods and so eliminate reference and range checks. When the Charseq is not ASCII or UTF-8 we can still do a very good job and are on par in most of the cases with what the user would do. The following benchmark shows the improvements: Result: 2456866.966 ?(99.9%) 59066.370 ops/s [Average] Statistics: (min, avg, max) = (2297025.189, 2456866.966, 2586003.225), stdev = 78851.914 Confidence interval (99.9%): [2397800.596, 2515933.336] Benchmark Mode Samples Score Score error Units i.n.m.b.ByteBufUtilBenchmark.writeAscii thrpt 50 9398165.238 131503.098 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiString thrpt 50 9695177.968 176684.821 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiStringViaArray thrpt 50 4788597.415 83181.549 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiStringViaArrayWrapped thrpt 50 4722297.435 98984.491 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiStringWrapped thrpt 50 4028689.762 66192.505 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiViaArray thrpt 50 3234841.565 91308.009 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiViaArrayWrapped thrpt 50 3311387.474 39018.933 ops/s i.n.m.b.ByteBufUtilBenchmark.writeAsciiWrapped thrpt 50 3379764.250 66735.415 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8 thrpt 50 5671116.821 101760.081 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8String thrpt 50 5682733.440 111874.084 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8StringViaArray thrpt 50 3564548.995 55709.512 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8StringViaArrayWrapped thrpt 50 3621053.671 47632.820 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8StringWrapped thrpt 50 2634029.071 52304.876 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8ViaArray thrpt 50 3397049.332 57784.119 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8ViaArrayWrapped thrpt 50 3318685.262 35869.562 ops/s i.n.m.b.ByteBufUtilBenchmark.writeUtf8Wrapped thrpt 50 2473791.249 46423.114 ops/s Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 1,387.417 sec - in io.netty.microbench.buffer.ByteBufUtilBenchmark Results : Tests run: 1, Failures: 0, Errors: 0, Skipped: 0 Results : Tests run: 1, Failures: 0, Errors: 0, Skipped: 0 The *ViaArray* benchmarks are basically doing a toString().getBytes(Charset) which the others are using ByteBufUtil.write*(...).
2014-09-04 16:04:33 +02:00
/**
* Encode a {@link CharSequence} in <a href="http://en.wikipedia.org/wiki/UTF-8">UTF-8</a> and write
* it to a {@link ByteBuf}.
*
* This method returns the actual number of bytes written.
*/
public static int writeUtf8(ByteBuf buf, CharSequence seq) {
if (buf == null) {
throw new NullPointerException("buf");
}
if (seq == null) {
throw new NullPointerException("seq");
}
// UTF-8 uses max. 3 bytes per char, so calculate the worst case.
final int len = seq.length();
final int maxSize = len * 3;
buf.ensureWritable(maxSize);
if (buf instanceof AbstractByteBuf) {
// Fast-Path
AbstractByteBuf buffer = (AbstractByteBuf) buf;
int oldWriterIndex = buffer.writerIndex;
int writerIndex = oldWriterIndex;
// We can use the _set methods as these not need to do any index checks and reference checks.
// This is possible as we called ensureWritable(...) before.
for (int i = 0; i < len; i++) {
char c = seq.charAt(i);
if (c < 0x80) {
buffer._setByte(writerIndex++, (byte) c);
} else if (c < 0x800) {
buffer._setByte(writerIndex++, (byte) (0xc0 | (c >> 6)));
buffer._setByte(writerIndex++, (byte) (0x80 | (c & 0x3f)));
} else {
buffer._setByte(writerIndex++, (byte) (0xe0 | (c >> 12)));
buffer._setByte(writerIndex++, (byte) (0x80 | ((c >> 6) & 0x3f)));
buffer._setByte(writerIndex++, (byte) (0x80 | (c & 0x3f)));
}
}
// update the writerIndex without any extra checks for performance reasons
buffer.writerIndex = writerIndex;
return writerIndex - oldWriterIndex;
} else {
// Maybe we could also check if we can unwrap() to access the wrapped buffer which
// may be an AbstractByteBuf. But this may be overkill so let us keep it simple for now.
byte[] bytes = seq.toString().getBytes(CharsetUtil.UTF_8);
buf.writeBytes(bytes);
return bytes.length;
}
}
/**
* Encode a {@link CharSequence} in <a href="http://en.wikipedia.org/wiki/ASCII">ASCII</a> and write it
* to a {@link ByteBuf}.
*
* This method returns the actual number of bytes written.
*/
public static int writeAscii(ByteBuf buf, CharSequence seq) {
if (buf == null) {
throw new NullPointerException("buf");
}
if (seq == null) {
throw new NullPointerException("seq");
}
// ASCII uses 1 byte per char
final int len = seq.length();
buf.ensureWritable(len);
if (buf instanceof AbstractByteBuf) {
// Fast-Path
AbstractByteBuf buffer = (AbstractByteBuf) buf;
int writerIndex = buffer.writerIndex;
// We can use the _set methods as these not need to do any index checks and reference checks.
// This is possible as we called ensureWritable(...) before.
for (int i = 0; i < len; i++) {
buffer._setByte(writerIndex++, (byte) seq.charAt(i));
}
// update the writerIndex without any extra checks for performance reasons
buffer.writerIndex = writerIndex;
} else {
// Maybe we could also check if we can unwrap() to access the wrapped buffer which
// may be an AbstractByteBuf. But this may be overkill so let us keep it simple for now.
buf.writeBytes(seq.toString().getBytes(CharsetUtil.US_ASCII));
}
return len;
}
/**
* Encode the given {@link CharBuffer} using the given {@link Charset} into a new {@link ByteBuf} which
* is allocated via the {@link ByteBufAllocator}.
*/
public static ByteBuf encodeString(ByteBufAllocator alloc, CharBuffer src, Charset charset) {
return encodeString0(alloc, false, src, charset);
}
static ByteBuf encodeString0(ByteBufAllocator alloc, boolean enforceHeap, CharBuffer src, Charset charset) {
final CharsetEncoder encoder = CharsetUtil.getEncoder(charset);
int length = (int) ((double) src.remaining() * encoder.maxBytesPerChar());
boolean release = true;
final ByteBuf dst;
if (enforceHeap) {
dst = alloc.heapBuffer(length);
} else {
dst = alloc.buffer(length);
}
try {
final ByteBuffer dstBuf = dst.internalNioBuffer(0, length);
final int pos = dstBuf.position();
CoderResult cr = encoder.encode(src, dstBuf, true);
if (!cr.isUnderflow()) {
cr.throwException();
}
cr = encoder.flush(dstBuf);
if (!cr.isUnderflow()) {
cr.throwException();
}
2013-11-15 15:07:19 +01:00
dst.writerIndex(dst.writerIndex() + dstBuf.position() - pos);
release = false;
return dst;
} catch (CharacterCodingException x) {
throw new IllegalStateException(x);
} finally {
if (release) {
dst.release();
}
}
}
static String decodeString(ByteBuffer src, Charset charset) {
final CharsetDecoder decoder = CharsetUtil.getDecoder(charset);
final CharBuffer dst = CharBuffer.allocate(
(int) ((double) src.remaining() * decoder.maxCharsPerByte()));
try {
CoderResult cr = decoder.decode(src, dst, true);
if (!cr.isUnderflow()) {
cr.throwException();
}
cr = decoder.flush(dst);
if (!cr.isUnderflow()) {
cr.throwException();
}
} catch (CharacterCodingException x) {
throw new IllegalStateException(x);
}
return dst.flip().toString();
}
/**
* Returns a cached thread-local direct buffer, if available.
*
* @return a cached thread-local direct buffer, if available. {@code null} otherwise.
*/
public static ByteBuf threadLocalDirectBuffer() {
if (THREAD_LOCAL_BUFFER_SIZE <= 0) {
return null;
}
if (PlatformDependent.hasUnsafe()) {
return ThreadLocalUnsafeDirectByteBuf.newInstance();
} else {
return ThreadLocalDirectByteBuf.newInstance();
}
}
/**
* Create a copy of the underlying storage from {@link value} into a byte array.
* The copy will start at {@link ByteBuf#readerIndex()} and copy {@link ByteBuf#readableBytes()} bytes.
*/
public static byte[] getBytes(ByteBuf buf) {
return getBytes(buf, checkNotNull(buf, "buf").readerIndex(), buf.readableBytes());
}
/**
* Create a copy of the underlying storage from {@link buf} into a byte array.
* The copy will start at {@code start} and copy {@code length} bytes.
*/
public static byte[] getBytes(ByteBuf buf, int start, int length) {
return getBytes(buf, start, length, true);
}
/**
* Return an array of the underlying storage from {@link buf} into a byte array.
* The copy will start at {@code start} and copy {@code length} bytes.
* If {@code copy} is true a copy will be made of the memory.
* If {@code copy} is false the underlying storage will be shared, if possible.
*/
public static byte[] getBytes(ByteBuf buf, int start, int length, boolean copy) {
if (start < 0 || length > checkNotNull(buf, "buf").capacity() - start) {
throw new IndexOutOfBoundsException("expected: " + "0 <= start(" + start + ") <= start + length(" + length
+ ") <= " + "buf.capacity(" + buf.capacity() + ')');
}
if (buf.hasArray()) {
if (copy || start != 0 || length != buf.capacity()) {
int baseOffset = buf.arrayOffset() + start;
return Arrays.copyOfRange(buf.array(), baseOffset, baseOffset + length);
} else {
return buf.array();
}
}
byte[] v = new byte[length];
buf.getBytes(start, v);
return v;
}
/**
* Copies the content of {@code src} to a {@link ByteBuf} using {@link ByteBuf#writeBytes(byte[], int, int)}.
* @param src The source of the data to copy.
* @param srcIdx the starting offset of characters to copy.
* @param dst the destination byte array.
* @param dstIdx the starting offset in the destination byte array.
* @param length the number of characters to copy.
*/
public static void copy(ByteString src, int srcIdx, ByteBuf dst, int dstIdx, int length) {
final int thisLen = src.length();
if (srcIdx < 0 || length > thisLen - srcIdx) {
throw new IndexOutOfBoundsException("expected: " + "0 <= srcIdx(" + srcIdx + ") <= srcIdx + length("
+ length + ") <= srcLen(" + thisLen + ')');
}
checkNotNull(dst, "dst").setBytes(dstIdx, src.array(), srcIdx + src.arrayOffset(), length);
}
/**
* Copies the content of {@code src} to a {@link ByteBuf} using {@link ByteBuf#writeBytes(byte[], int, int)}.
* @param src The source of the data to copy.
* @param srcIdx the starting offset of characters to copy.
* @param dst the destination byte array.
* @param length the number of characters to copy.
*/
public static void copy(ByteString src, int srcIdx, ByteBuf dst, int length) {
final int thisLen = src.length();
if (srcIdx < 0 || length > thisLen - srcIdx) {
throw new IndexOutOfBoundsException("expected: " + "0 <= srcIdx(" + srcIdx + ") <= srcIdx + length("
+ length + ") <= srcLen(" + thisLen + ')');
}
checkNotNull(dst, "dst").writeBytes(src.array(), srcIdx + src.arrayOffset(), length);
}
static final class ThreadLocalUnsafeDirectByteBuf extends UnpooledUnsafeDirectByteBuf {
private static final Recycler<ThreadLocalUnsafeDirectByteBuf> RECYCLER =
new Recycler<ThreadLocalUnsafeDirectByteBuf>() {
@Override
protected ThreadLocalUnsafeDirectByteBuf newObject(Handle handle) {
return new ThreadLocalUnsafeDirectByteBuf(handle);
}
};
static ThreadLocalUnsafeDirectByteBuf newInstance() {
ThreadLocalUnsafeDirectByteBuf buf = RECYCLER.get();
buf.setRefCnt(1);
return buf;
}
private final Handle handle;
private ThreadLocalUnsafeDirectByteBuf(Handle handle) {
super(UnpooledByteBufAllocator.DEFAULT, 256, Integer.MAX_VALUE);
this.handle = handle;
}
@Override
protected void deallocate() {
if (capacity() > THREAD_LOCAL_BUFFER_SIZE) {
super.deallocate();
} else {
clear();
RECYCLER.recycle(this, handle);
}
}
}
static final class ThreadLocalDirectByteBuf extends UnpooledDirectByteBuf {
private static final Recycler<ThreadLocalDirectByteBuf> RECYCLER = new Recycler<ThreadLocalDirectByteBuf>() {
@Override
protected ThreadLocalDirectByteBuf newObject(Handle handle) {
return new ThreadLocalDirectByteBuf(handle);
}
};
static ThreadLocalDirectByteBuf newInstance() {
ThreadLocalDirectByteBuf buf = RECYCLER.get();
buf.setRefCnt(1);
return buf;
}
private final Handle handle;
private ThreadLocalDirectByteBuf(Handle handle) {
super(UnpooledByteBufAllocator.DEFAULT, 256, Integer.MAX_VALUE);
this.handle = handle;
}
@Override
protected void deallocate() {
if (capacity() > THREAD_LOCAL_BUFFER_SIZE) {
super.deallocate();
} else {
clear();
RECYCLER.recycle(this, handle);
}
}
}
Revamp the core API to reduce memory footprint and consumption The API changes made so far turned out to increase the memory footprint and consumption while our intention was actually decreasing them. Memory consumption issue: When there are many connections which does not exchange data frequently, the old Netty 4 API spent a lot more memory than 3 because it always allocates per-handler buffer for each connection unless otherwise explicitly stated by a user. In a usual real world load, a client doesn't always send requests without pausing, so the idea of having a buffer whose life cycle if bound to the life cycle of a connection didn't work as expected. Memory footprint issue: The old Netty 4 API decreased overall memory footprint by a great deal in many cases. It was mainly because the old Netty 4 API did not allocate a new buffer and event object for each read. Instead, it created a new buffer for each handler in a pipeline. This works pretty well as long as the number of handlers in a pipeline is only a few. However, for a highly modular application with many handlers which handles connections which lasts for relatively short period, it actually makes the memory footprint issue much worse. Changes: All in all, this is about retaining all the good changes we made in 4 so far such as better thread model and going back to the way how we dealt with message events in 3. To fix the memory consumption/footprint issue mentioned above, we made a hard decision to break the backward compatibility again with the following changes: - Remove MessageBuf - Merge Buf into ByteBuf - Merge ChannelInboundByte/MessageHandler and ChannelStateHandler into ChannelInboundHandler - Similar changes were made to the adapter classes - Merge ChannelOutboundByte/MessageHandler and ChannelOperationHandler into ChannelOutboundHandler - Similar changes were made to the adapter classes - Introduce MessageList which is similar to `MessageEvent` in Netty 3 - Replace inboundBufferUpdated(ctx) with messageReceived(ctx, MessageList) - Replace flush(ctx, promise) with write(ctx, MessageList, promise) - Remove ByteToByteEncoder/Decoder/Codec - Replaced by MessageToByteEncoder<ByteBuf>, ByteToMessageDecoder<ByteBuf>, and ByteMessageCodec<ByteBuf> - Merge EmbeddedByteChannel and EmbeddedMessageChannel into EmbeddedChannel - Add SimpleChannelInboundHandler which is sometimes more useful than ChannelInboundHandlerAdapter - Bring back Channel.isWritable() from Netty 3 - Add ChannelInboundHandler.channelWritabilityChanges() event - Add RecvByteBufAllocator configuration property - Similar to ReceiveBufferSizePredictor in Netty 3 - Some existing configuration properties such as DatagramChannelConfig.receivePacketSize is gone now. - Remove suspend/resumeIntermediaryDeallocation() in ByteBuf This change would have been impossible without @normanmaurer's help. He fixed, ported, and improved many parts of the changes.
2013-05-28 13:40:19 +02:00
private ByteBufUtil() { }
}