The Swim Core framework implements a dependency-free foundation on which the rest of the Swim backend stack is built. It provides:
- Incremental Codecs: I/O encoders, decoders, parsers, and writer.
- Immutable Collections: Hash tries, finger tries, and b-trees.
- Structural Data Models: Format-agnostic structural data types.
- Serialization Formats: JSON, XML, Recon, Protobuf, Avro, CSV, and URI codecs.
- Security Engine: JWT, JWS, and JWK implementations.
- Concurrency Primitives: Atomic tasks, timers, and continuation passing style interfaces.
- Streaming State Machines: Stateful dataflow combinators.
- Data Compression: Incremental Deflate transcoder.
- Concurrent Datastore: Eventually persistent trees, optimized for high rate updates.
- Network Protocols: Flow-controlled HTTP, WebSocket, WARP, and MQTT implementations.
- Low-Level Web Server: HTTP routing API.