Buffer

Kotlin Multiplatform byte buffers — same code on JVM, Android, iOS, macOS, Linux, JS, and WASM with zero-copy performance.

Buffer gives you one ReadBuffer/WriteBuffer API that delegates to platform-native types — ByteBuffer on JVM, NSData on Apple, malloc on Linux, Uint8Array on JS — so data never copies between your code and the OS.

Why Buffer?

• Zero-copy I/O: Delegates directly to platform-native buffers — no copying between your code and the OS
• Buffer Pooling: Allocate once, reuse for every network request — no GC pauses, predictable latency
• Stream Processing: Parse protocols that span chunk boundaries without manual accumulator code
• Compression: compress()/decompress() on any ReadBuffer across all platforms
• Flow Extensions: mapBuffer(), asStringFlow(), lines() — compose streaming transforms with Kotlin Flow
• SIMD-Accelerated Operations: Bulk comparison, search (indexOf), fill, and XOR masking
• 7 Platforms: Same API on JVM, Android, iOS, macOS, Linux, JS, WASM

Quick Example

// Works identically on all platforms
val buffer = PlatformBuffer.allocate(1024)
buffer.writeInt(42)
buffer.writeString("Hello, Buffer!")
buffer.resetForRead()

val number = buffer.readInt()      // 42
val text = buffer.readString(14)   // "Hello, Buffer!"

High-Value Patterns

Buffer pooling — no allocation per request:

withPool(defaultBufferSize = 8192) { pool ->
    pool.withBuffer(1024) { buffer ->
        buffer.writeInt(requestId)
        buffer.writeString(payload)
        buffer.resetForRead()
        sendToNetwork(buffer)
    } // buffer returned to pool
}

Streaming pipeline — decompress, decode, split lines:

bufferFlow
    .mapBuffer { decompress(it, Gzip).getOrThrow() }
    .asStringFlow()
    .lines()
    .collect { line -> process(line) }

Platform Implementations

Platform	Native Type	Notes
JVM	java.nio.ByteBuffer	Direct buffers for zero-copy I/O
Android	ByteBuffer + SharedMemory	IPC via Parcelable
iOS/macOS	NSData / NSMutableData	Foundation integration
JavaScript	Uint8Array	SharedArrayBuffer support
WASM	LinearBuffer / ByteArrayBuffer	Zero-copy JS interop
Linux	NativeBuffer (malloc) / ByteArrayBuffer	Zero-copy io_uring I/O

Part of the DitchOoM Stack

Buffer is the foundation for the Socket library — cross-platform TCP + TLS that streams these same ReadBuffer/WriteBuffer types.

┌─────────────────────────────┐
│  Your Protocol              │
├─────────────────────────────┤
│  socket (TCP + TLS)         │  ← com.ditchoom:socket
├─────────────────────────────┤
│  buffer-compression         │  ← com.ditchoom:buffer-compression
├─────────────────────────────┤
│  buffer-flow                │  ← com.ditchoom:buffer-flow
├─────────────────────────────┤
│  buffer                     │  ← com.ditchoom:buffer
└─────────────────────────────┘

Installation

Add to your build.gradle.kts (see Maven Central for the latest version):

dependencies {
    // Core buffer library
    implementation("com.ditchoom:buffer:<latest-version>")

    // Optional: Compression (gzip, deflate)
    implementation("com.ditchoom:buffer-compression:<latest-version>")

    // Optional: Flow extensions (lines, mapBuffer, asStringFlow)
    implementation("com.ditchoom:buffer-flow:<latest-version>")
}

Next Steps

• Getting Started - Installation and first buffer
• Core Concepts - Understanding buffers, positions, and limits
• Recipes - Common patterns and examples
• Performance - Optimization tips