snappy-java/CLAUDE.md
Taro L. Saito e41d1f133a Update CLAUDE.md development workflow
- Fix typo in 'commit'
- Add PR guidelines and common development tasks
- Update native code build command to use 'make clean-native native'
2025-07-19 09:47:44 -07:00

3.7 KiB

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

snappy-java is a Java port of Google's Snappy compression library, providing fast compression/decompression with JNI bindings for native performance across multiple platforms.

Build Commands

Using sbt (primary build tool)

# Enter sbt console
./sbt

# Run tests
./sbt test

# Run tests matching a pattern
./sbt "testOnly *BitShuffleTest"

# Create jar file
./sbt package

# Publish to local Maven repository
./sbt publishM2

# Run tests continuously on code changes
./sbt "~test"

Using make (for native library compilation)

# Build native libraries
make

# Clean build artifacts
make clean

# Platform-specific builds (when cross-compiling)
make native-all

Architecture

Core Components

  1. Java API Layer (src/main/java/org/xerial/snappy/)

    • Snappy.java: Main API facade providing high-level compression/decompression methods
    • SnappyNative.java: JNI interface to native Snappy library
    • SnappyLoader.java: Handles platform-specific native library loading
  2. Native Layer (src/main/java/org/xerial/snappy/)

    • SnappyNative.cpp: JNI implementation bridging Java and C++ Snappy
    • BitShuffleNative.cpp: JNI implementation for BitShuffle algorithm
  3. Stream API (src/main/java/org/xerial/snappy/)

    • SnappyOutputStream/SnappyInputStream: Block-based compression streams
    • SnappyFramedOutputStream/SnappyFramedInputStream: Framing format implementation
    • SnappyHadoopCompatibleOutputStream: Hadoop-compatible format
  4. Memory Management (src/main/java/org/xerial/snappy/buffer/ and /pool/)

    • Buffer allocation and pooling for efficient memory usage
    • Direct ByteBuffer management for zero-copy operations

Platform Support

The project includes pre-built native libraries for multiple platforms in src/main/resources/org/xerial/snappy/native/:

  • Windows (x86, x86_64, aarch64)
  • macOS (x86, x86_64, aarch64)
  • Linux (x86, x86_64, aarch64, arm, armv6, armv7, ppc64, ppc64le, s390x, riscv64, loongarch64)
  • AIX (ppc, ppc64)
  • SunOS (x86, x86_64, sparc)
  • Android (arm, aarch64)

Cross-compilation

The project uses Docker-based cross-compilation toolchains (see docker/ directory) for building native libraries across different architectures.

Testing

# Run all tests with debug logging
./sbt "testOnly * -- -l debug"

# Run specific test class
./sbt "testOnly org.xerial.snappy.SnappyTest"

# The project uses AirSpec (for Scala tests) and JUnit (for Java tests)
# Tests are located in src/test/java/org/xerial/snappy/

Important Notes

  1. Native Library Loading: The project automatically detects the platform and loads the appropriate native library from resources
  2. Memory Safety: Uses direct ByteBuffers for efficient memory operations - be aware of buffer boundaries
  3. Thread Safety: Snappy compression/decompression methods are thread-safe as they don't maintain state
  4. OSGi Support: The project includes OSGi bundle configuration in build.sbt
  5. Compatibility: Multiple stream formats are supported - ensure you use matching read/write formats (see compatibility matrix in README.md)

Development Workflow

Pull Request Guidelines

  • Use squashed commits when merging PRs to maintain clean commit history
  • Run tests before creating PR: ./sbt test
  • Format code: ./sbt scalafmtAll

Common Development Tasks

  • When making changes to native code, rebuild with make clean-native native
  • For Java/Scala changes, use ./sbt ~test for continuous testing
  • Always test on multiple JDK versions if possible (8, 11, 17, 21)

Native Code Testing

  • Use make clean-native native for testing native code