v0.2.0

Bump version to 0.2.0 and add binary target for CLI installation via cargo install.

v0.1.2 - Line-Level Diffs

What's New

Features

• feat(diff): Add line-level diff functionality with unified format output
  • Git-like unified diff format showing exactly what changed between checkpoints
  • Line-by-line comparison with context lines
  • Support for ignoring whitespace changes
  • Configurable context lines and file size limits
  • Binary file detection and handling

Improvements

• chore: Replace hardcoded version strings in tests with env!("CARGO_PKG_VERSION")
• docs: Update version strings in examples and documentation

API Additions

• Added diff_file() method for line-level comparison of individual files
• Added diff_detailed() method for complete checkpoint diffs with line-level changes
• New DiffOptions struct for configuring diff behavior
• New FileDiff, DiffHunk, and LineChange types for diff results

v0.1.1 - Documentation and Compatibility Improvements

What's Changed

Improvements

• refactor: Make gxhash optional for docs.rs compatibility
• docs: Improve CPU intrinsics compatibility explanation in collections module

Technical Details

• Added gxhash as an optional feature to ensure docs.rs builds successfully
• Enhanced documentation for better clarity on CPU intrinsics requirements

v0.1.0 - Initial Release

Initial Release

Features

• Incremental Snapshots: Only changed files are stored between checkpoints
• Content-Addressable Storage: Automatic deduplication using SHA-256 hashing
• LZ4 Compression: Extreme-speed compression with adaptive strategies
• Parallel Processing: Multi-threaded file scanning leveraging all CPU cores
• Merkle Tree Verification: Cryptographic integrity checking for tamper detection
• Timeline Branching: Git-like branching model supporting multiple timelines
• Atomic Operations: All checkpoint and restore operations are atomic
• Memory Efficient: Streaming architecture for large files
• Line-Level Diffs: Git-like unified diff output

Technical Specifications

• Compression: LZ4 via lz4_flex crate with 4+ GB/s decompression speeds
• Hashing: SHA-256 for content identification and verification
• Serialization: Bincode for efficient binary storage
• Concurrency: Rayon-based parallel processing
• Storage: Sharded object storage with reference counting