ByteBuffer

A kotlin multiplatform library that allows you to allocate and modify byte[] natively using an API similar to Java's ByteBuffer API.


Report Bug · Request Feature

Table of Contents

1. About The Project
   • Runtime Dependencies
   • Supported Platforms
2. Installation
3. Usage
   • Allocate a new platform agnostic buffer
   • Allocation Zones
   • Byte order
   • Relative write data into platform agnostic buffer
   • Absolute write data into platform agnostic buffer
   • Relative read data into platform agnostic buffer
   • Absolute read data into platform agnostic buffer
4. Building Locally
5. Getting Started
6. Roadmap
7. Contributing
8. License

About The Project

Allocating and managing a chunk of memory can be slightly different based on each platform. This project aims to make it easier to manage buffers in a cross platform way using kotlin multiplatform. This was originally created as a side project for a kotlin multiplatform mqtt data sync solution.

Implementation notes:

• JVM + Android delegate to direct ByteBuffers to avoid memory copies when possible.
• Apple targets use NSData or NSMutableData
• JS targets use Uint8Array.
• Native platforms use standard byte arrays to manage memory.

Runtime Dependencies

• None

Supported Platforms

• All Kotlin Multiplatform supported OS's.

Platform	Wrapped Type
JVM 1.8	ByteBuffer
Node.js	Uint8Array including SharedArrayBuffer
Browser (Chrome)	Uint8Array including SharedArrayBuffer
Android	ByteBuffer including SharedMemory
iOS	NSData
WatchOS	NSData
TvOS	NSData
MacOS	NSData
Linux X64	kotlin ByteArray
Windows X64	TODO

Installation

Gradle

• Add implementation("com.ditchoom:buffer:$version") to your build.gradle dependencies

Usage

Allocate a new platform agnostic buffer

val buffer = PlatformBuffer.allocate(
    byteSize,
    zone = AllocationZone.Direct,
    byteOrder = ByteOrder.BIG_ENDIAN
)

Wrap an existing byte array into a platform agnostic buffer

val byteArray = byteArrayOf(1, 2, 3, 4, 5)
val buffer = PlatformBuffer.wrap(byteArray, byteOrder = ByteOrder.BIG_ENDIAN)

Allocation Zones

Allocation zones allow you to change where the buffer is allocated.

• AllocationZone.Custom -> Allows you to override the underlying buffer. This can be helpful for memory mapped structures.
• AllocationZone.Heap -> On JVM platforms, allocates a HeapByteBuffer, otherwise a native byte array
• AllocationZone.Direct -> On JVM platforms, allocates a DirectByteBuffer, otherwise a native byte array
• AllocationZone.SharedMemory -> On JS Platforms this will populate the sharedArrayBuffer parameter in JsBuffer. On API 27+ it allocates a Shared Memory instance, otherwise will pipe the data during parcel using ParcelFileDescriptor and java.nio.Channel api. For JS platforms it will allocate a SharedArrayBuffer. If the proper security requirements are not set, it will fallback to a standard ArrayBuffer.

Android: All JvmBuffers are Parcelable. To avoid extra memory copies when using IPC, always choose AllocationZone.SharedMemory.

Browser JS: To enable SharedArrayBuffer, add the appropriate headers for the JS server in a gradle project by adding any file to a directory webpack.config.d next to the src directory containing:

if (config.devServer != null) {
   config.devServer.headers = {
       "Cross-Origin-Opener-Policy": "same-origin",
       "Cross-Origin-Embedder-Policy": "require-corp"
   }
}

Byte order

Byte order defaults to big endian but can be specified when creating the buffer with ByteOrder.BIG_ENDIAN or ByteOrder.LITTLE_ENDIAN

The byte order of a buffer can be checked with buffer.byteOrder

Relative write data into platform agnostic buffer

val buffer: WriteBuffer
// write signed byte
buffer.writeByte(5.toByte())
// write unsigned byte
buffer.writeUByte(5.toUByte())
// write short
buffer.writeShort(5.toShort())
// write unsigned short
buffer.writeUShort(5.toUShort())
// write int
buffer.writeInt(5)
// write unsigned int
buffer.writeUInt(5.toUInt())
// write long
buffer.writeLong(5L)
// write unsigned long
buffer.writeULong(5uL)
// write float
buffer.writeFloat(123.456f)
// write double
buffer.writeDouble(123.456)
// write text
buffer.writeString("5", Charset.UTF8)
// copy buffer into this one
buffer.write(otherBuffer)
// write byte array
buffer.writeBytes(byteArrayOf(1, 2, 3, 4))
// write partial byte array
buffer.writeBytes(byteArrayOf(1, 2, 3, 4, 5), offset, length)

Absolute write data into platform agnostic buffer

val buffer: WriteBuffer
// set signed byte
buffer[index] = 5.toByte()
// set unsigned byte
buffer[index] = 5.toUByte()
// set short
buffer[index] = 5.toByte()
// set unsigned short
buffer[index] = 5.toUShort()
// set int
buffer[index] = 5
// set unsigned int
buffer[index] = 5.toUInt()
// set long
buffer[index] = 5L
// set unsigned long
buffer[index] = 5uL
// set float
buffer[index] = 123.456f
// set double
buffer[index] = 123.456

Relative read data into platform agnostic buffer

val buffer: ReadBuffer
// read signed byte
val b = buffer.readByte()
// read unsigned byte
val uByte = buffer.readUnsignedByte()
// read short
val short = buffer.readShort()
// read unsigned short
val uShort = buffer.readUnsignedShort()
// read int
val intValue = buffer.readInt()
// read unsigned int
val uIntValue = buffer.readUnsignedInt()
// read long
val longValue = buffer.readLong()
// read unsigned long
val uLongValue = buffer.readUnsignedLong()
// read float
val float = buffer.readFloat()
// read double
val double = buffer.readDouble()
// read text
val string = buffer.readUtf8(numOfBytesToRead)
// read byte array
val byteArray = buffer.readByteArray(numOfBytesToRead)
// read a shared subsequence read buffer (changes to the original reflect here)
val readBuffer = buffer.readBytes(numOfBytesForBuffer)

Absolute read data into platform agnostic buffer

val buffer: ReadBuffer
// get signed byte
val b = buffer.get(index) // or buffer[index]
// get unsigned byte
val uByte = buffer.getUnsignedByte(index)
// get short
val short = buffer.getShort(index)
// get unsigned short
val uShort = buffer.getUnsignedShort(index)
// get int
val intValue = buffer.getInt(index)
// get unsigned int
val uIntValue = buffer.getUnsignedInt(index)
// get long
val longValue = buffer.getLong(index)
// get unsigned long
val uLongValue = buffer.getUnsignedLong(index)
// get float
val float = buffer.getFloat(index)
// get double
val double = buffer.getDouble(index)
// slice the buffer without adjusting the position or limit (changes to the original reflect here)
val slicedBuffer = buffer.slice()

Building Locally

• git clone [email protected]:DitchOoM/buffer.git
• Open cloned directory with Intellij IDEA.
  • Be sure to open with gradle

Roadmap

See the open issues for a list of proposed features ( and known issues).

Contributing

Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

Distributed under the Apache 2.0 License. See LICENSE for more information.