pamoja-sync 0.1.13

Offline-first store-and-forward buffering for pamoja: durable queues behind the core Store trait.
Documentation
<!-- Generated by `cargo xtask docs` from this crate's lib.rs; edit the crate doc, not this file. -->

# pamoja-sync

Offline-first store-and-forward buffering for pamoja: durable queues behind the core Store trait.

<a href="https://crates.io/crates/pamoja-sync"><img height="28" alt="crates.io" src="https://raw.githubusercontent.com/molexxxx/pamoja/main/.github/badges/btn-cratesio.svg"></a>
<a href="https://docs.rs/pamoja-sync"><img height="28" alt="docs.rs" src="https://raw.githubusercontent.com/molexxxx/pamoja/main/.github/badges/btn-docsrs.svg"></a>

Offline-first store-and-forward buffering for pamoja.

Devices on intermittent links append records while offline and drain them in
order when a link returns, so a node that is disconnected for hours or days
loses nothing. This crate provides durable first-in first-out implementations
of the core [`Store`](pamoja_core::Store) trait:

- [`MemoryStore`] - a fast in-memory queue, optionally capacity-bounded so a
  full queue becomes an explicit backpressure signal.
- [`FileStore`] - a crash-safe on-disk queue that survives power loss, for the
  power-loss-safe field logging the mission depends on.

Both buffer raw bytes, so an application pairs a [`Store`](pamoja_core::Store)
with a [`Codec`](https://docs.rs/pamoja-codec) to persist encoded payloads and
a [`Transport`](pamoja_core::Transport) to forward them when a link appears.
[`drain_to`] is the forward half: it publishes buffered records onto a
transport in order, removing each only after it is sent, so a failed link
loses nothing.

## License

MIT - part of the [pamoja](https://github.com/molexxxx/pamoja) workspace: one memory-safe Rust core with bindings for every language.