https://github.com/homer6/automerge-cpp
A modern C++23 implementation of Automerge — a conflict-free replicated data type (CRDT) library for building collaborative applications.
https://github.com/homer6/automerge-cpp
cpp cpp23-library crdts
Last synced: 20 days ago
JSON representation
A modern C++23 implementation of Automerge — a conflict-free replicated data type (CRDT) library for building collaborative applications.
- Host: GitHub
- URL: https://github.com/homer6/automerge-cpp
- Owner: homer6
- License: mit
- Created: 2026-02-18T00:05:21.000Z (26 days ago)
- Default Branch: main
- Last Pushed: 2026-02-19T06:01:31.000Z (25 days ago)
- Last Synced: 2026-02-20T03:03:46.849Z (24 days ago)
- Topics: cpp, cpp23-library, crdts
- Language: C++
- Homepage: https://homer6.github.io/automerge-cpp/
- Size: 696 KB
- Stars: 3
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# automerge-cpp
A modern C++23 implementation of [Automerge](https://automerge.org/) — a conflict-free
replicated data type (CRDT) library for building collaborative applications.
This is a **from-scratch** reimplementation, not a wrapper. It mirrors the upstream
Automerge semantics while embracing idiomatic C++23: algebraic types, ranges pipelines,
strong types, and APIs that make illegal states unrepresentable.
**281 tests passing** across 8 implementation phases.
## Quick Example
```cpp
#include
namespace am = automerge_cpp;
int main() {
auto doc1 = am::Document{};
am::ObjId list_id;
doc1.transact([&](auto& tx) {
tx.put(am::root, "title", std::string{"Shopping List"});
list_id = tx.put_object(am::root, "items", am::ObjType::list);
tx.insert(list_id, 0, std::string{"Milk"});
tx.insert(list_id, 1, std::string{"Eggs"});
});
// Fork and make concurrent edits
auto doc2 = doc1.fork();
doc1.transact([&](auto& tx) {
tx.insert(list_id, 2, std::string{"Bread"});
});
doc2.transact([&](auto& tx) {
tx.insert(list_id, 2, std::string{"Butter"});
});
// Merge — both edits preserved, no data loss
doc1.merge(doc2);
// list now contains: Milk, Eggs, Bread, Butter (deterministic order)
// Save to binary and reload
auto bytes = doc1.save();
auto loaded = am::Document::load(bytes);
// Time travel — read past state
auto heads_v1 = doc1.get_heads();
auto past = doc1.get_at(am::root, "title", heads_v1);
}
```
## Features
### Implemented
- **CRDT data types**: Map, List, Text, Counter
- **Conflict-free merging**: concurrent edits merge deterministically (RGA for lists/text)
- **Fork and merge**: create independent document copies, merge them back
- **Binary serialization**: `save()` / `load()` with upstream-compatible columnar encoding
- **Sync protocol**: bloom filter-based peer-to-peer synchronization
- **Patches**: incremental change notifications via `transact_with_patches()`
- **Time travel**: read document state at any historical point (`get_at()`, `text_at()`, etc.)
- **Cursors**: stable positions in lists/text that survive edits and merges
- **Rich text marks**: range annotations (bold, italic, links) anchored by identity, not index
- **Strong types**: `ActorId`, `ObjId`, `ChangeHash`, `OpId` never implicitly convert
- **Type-safe values**: `std::variant`-based `ScalarValue` and `Value` types
- **Columnar encoding**: upstream-compatible columnar op encoding with RLE, delta, and boolean encoders
- **DEFLATE compression**: raw DEFLATE (no zlib/gzip header) for columns exceeding 256 bytes, matching upstream Rust format
- **SHA-256 checksums**: chunk envelope with SHA-256 integrity validation
- **Backward compatibility**: v1 format loading with automatic format detection
- **Fuzz testing**: libFuzzer targets for `Document::load()`, LEB128 decode, and change chunk parsing
- **Static analysis**: clang-tidy CI with `bugprone-*`, `performance-*`, and `clang-analyzer-*` checks
- **Sanitizer CI**: Address Sanitizer + Undefined Behavior Sanitizer on all tests
- **Thread safety**: `Document` is thread-safe via `std::shared_mutex` — N concurrent readers, exclusive writers
- **Thread pool**: built-in work-stealing thread pool (Barak Shoshany's BS::thread_pool), shared across documents
- **Lock-free reads**: `set_read_locking(false)` eliminates shared_mutex contention for read-heavy workloads (13.5x parallel scaling on 30 cores)
- **Performance caching**: change hash cache (22x faster time travel), actor table cache (1.2x faster save)
- **Doxygen documentation**: auto-generated API docs with GitHub Pages deployment
## Performance
Release-build highlights (Intel Xeon Platinum 8358, 30 cores, Linux, GCC 13.3, `-O3 -march=native`):
### Single-Threaded
| Operation | Throughput |
|-----------|------------|
| Map put (batched) | 4.3 M ops/s |
| Map get | 19.8 M ops/s |
| Sync round trip | 23.6 K ops/s |
| Time travel get_at | 2.9 M ops/s |
| Merge (10+10 puts) | 241 K ops/s |
| Cursor resolve | 1.8 M ops/s |
| Save (100 keys) | 30.6 K ops/s |
### Parallel Scaling (30 cores)
| Operation | Sequential | Parallel | Speedup |
|-----------|-----------|----------|---------|
| Get (100K keys, lock-free) | 7.1 M ops/s | **95.0 M ops/s** | **13.5x** |
| Get (1M keys, lock-free) | 5.7 M ops/s | **55.0 M ops/s** | **9.6x** |
| Put (100K keys, sharded) | 2.0 M ops/s | **13.5 M ops/s** | **6.9x** |
| Put (1M keys, sharded) | 1.7 M ops/s | **8.4 M ops/s** | **4.8x** |
| Save 500 docs | 95 K docs/s | **806 K docs/s** | **8.4x** |
| Load 500 docs | 48 K docs/s | **190 K docs/s** | **3.9x** |
v0.4.0 optimizations: **22x** faster time travel, **5.6x** faster sync (hash/actor table caching),
**13.5x** parallel read scaling (lock-free reads eliminate shared_mutex contention).
See [docs/benchmark-results.md](docs/benchmark-results.md) for full results, platform comparison,
and perf analysis.
## Design Philosophy
Inspired by Ben Deane's approach to modern C++:
- **Make illegal states unrepresentable** — algebraic types model the domain precisely
- **Algorithms over raw loops** — `std::ranges` pipelines, folds, transforms
- **CRDTs are monoids** — merge is associative, commutative, and idempotent
- **Strong types prevent mixups** — `ActorId`, `ObjId`, `ChangeHash` never interconvert
- **Value semantics** — immutable outside transactions, explicit mutation boundaries
## Building
### Requirements
- C++23 compiler (GCC 14+, Clang 18+, MSVC 19.38+)
- CMake 3.28+
- zlib (for raw DEFLATE compression)
### Build
```bash
cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build
```
### Build with tests, examples, and benchmarks
```bash
cmake -B build \
-DCMAKE_BUILD_TYPE=Release \
-DAUTOMERGE_CPP_BUILD_TESTS=ON \
-DAUTOMERGE_CPP_BUILD_EXAMPLES=ON \
-DAUTOMERGE_CPP_BUILD_BENCHMARKS=ON
cmake --build build
ctest --test-dir build --output-on-failure
```
### Run benchmarks
```bash
./build/benchmarks/automerge_cpp_benchmarks
```
## Project Structure
```
automerge-cpp/
include/automerge-cpp/ # public headers
automerge.hpp # umbrella header
document.hpp # Document class
transaction.hpp # Transaction class
types.hpp # ActorId, ObjId, OpId, ChangeHash, Prop
value.hpp # ScalarValue, Value, ObjType
change.hpp # Change struct
op.hpp # Op, OpType
sync_state.hpp # SyncState, SyncMessage
patch.hpp # Patch, PatchAction types
cursor.hpp # Cursor (stable position)
mark.hpp # Mark (rich text annotation)
error.hpp # Error, ErrorKind
thread_pool.hpp # BS::thread_pool (header-only)
src/ # implementation
document.cpp # Document methods
transaction.cpp # Transaction methods
doc_state.hpp # internal: DocState, ObjectState
crypto/ # SHA-256 (vendored)
encoding/ # LEB128, RLE, delta, boolean codecs
storage/ # columnar binary format
columns/ # column spec, op encoding, value encoding, compression
sync/ # bloom filter for sync protocol
tests/ # unit and integration tests
examples/ # example programs
benchmarks/ # performance benchmarks
docs/ # documentation
api.md # API reference
style.md # coding style guide
plans/ # architecture and roadmap
upstream/automerge/ # upstream Rust reference (git submodule)
```
## Examples
Six example programs in `examples/`:
| Example | Description |
|---------|-------------|
| `basic_usage` | Create doc, put/get values, counters, save/load |
| `collaborative_todo` | Two actors concurrently editing a shared todo list |
| `text_editor` | Text editing with patches, cursors, and time travel |
| `sync_demo` | Peer-to-peer sync with SyncState |
| `thread_safe_demo` | Multi-threaded concurrent reads and writes on a single Document |
| `parallel_perf_demo` | Monoid-powered fork/merge parallelism, parallel save/load/sync |
```bash
./build/examples/basic_usage
./build/examples/collaborative_todo
./build/examples/text_editor
./build/examples/sync_demo
./build/examples/thread_safe_demo
./build/examples/parallel_perf_demo
```
## Documentation
- [API Reference](docs/api.md) — every public type, method, and usage examples
- [Benchmark Results](docs/benchmark-results.md) — performance measurements
- [Style Guide](docs/style.md) — coding style and conventions (Ben Deane's modern C++ principles)
- [Architecture Plan](docs/plans/architecture.md) — design principles, core types, module layout
- [Implementation Roadmap](docs/plans/roadmap.md) — phased development plan with status
## License
MIT License. See [LICENSE](LICENSE).
## Acknowledgments
- [Automerge](https://automerge.org/) — the original CRDT library by Martin Kleppmann et al.
- [automerge-rs](https://github.com/automerge/automerge) — the upstream Rust implementation