moq-json 0.5.7

//! JSON publishing over [`moq-net`](moq_net) tracks, in three modes:
//!
//! - [`snapshot`]: **lossy**. One JSON value updated over time; a consumer only gets the most
//!   recent value. Intermediate updates are collapsed and older groups are dropped.
//! - [`stream`]: **lossless**. An ordered append-log of self-contained records; every record is
//!   preserved and delivered in order, nothing is ever superseded.
//! - [`window`]: **bounded**. An ordered run of records appended to the back and dropped from the
//!   front, which a reader can join at any point.
//!
//! Pick [`snapshot`] when consumers care about "what is the value now" (a catalog, a status
//! document), [`stream`] when they care about every record of an unbounded log, and [`window`] when
//! the publisher retires old records and a late reader should start from what is still retained.
//!
//! Each mode comes in two layers. `Producer`/`Consumer` own a [`moq_net`] track and manage its
//! groups. `Encoder`/`Decoder` are the same logic without the track: values in, frame payloads out
//! (and back), with the encoder saying where the group boundaries fall. Reach for the codec layer
//! when something else already owns the track, such as a `moq_mux::container::Producer` also
//! managing a timeline and a catalog estimate.

mod diff;
mod merge;
pub mod snapshot;
pub mod stream;
pub mod window;

pub use crate::diff::{Diff, diff};

/// How a JSON track compresses its frames.
#[derive(Debug, Clone, Copy, Default, PartialEq, Eq)]
pub enum Compression {
	/// Uncompressed JSON frames.
	#[default]
	None,

	/// Group-scoped raw DEFLATE, sync-flushed at each frame boundary.
	Deflate,
}

impl Compression {
	pub(crate) const fn is_deflate(self) -> bool {
		matches!(self, Self::Deflate)
	}
}

/// Errors produced while publishing or consuming JSON.
#[derive(thiserror::Error, Debug, Clone)]
#[non_exhaustive]
pub enum Error {
	/// An error from the underlying track.
	#[error(transparent)]
	Net(#[from] moq_net::Error),

	/// A value failed to serialize, deserialize, or apply as a merge patch.
	///
	/// Stored as a string since [`serde_json::Error`] is not [`Clone`].
	#[error("json: {0}")]
	Json(String),

	/// A compressed frame could not be decoded (malformed, truncated, or oversized).
	#[error(transparent)]
	Flate(#[from] moq_flate::Error),

	/// A merge patch arrived with no snapshot to apply it to.
	///
	/// Every group opens with a full snapshot, so this means frames reached
	/// [`snapshot::Decoder`] out of order, or a group's first frame was routed as a delta.
	#[error("delta before snapshot")]
	MissingSnapshot,

	/// A compressed [`stream`] frame was encoded but never written, so the shared DEFLATE window is
	/// ahead of what the consumer holds and nothing later in this group can be decoded.
	///
	/// Unlike [`snapshot`], a stream has no keyframe to resynchronize on, so the encoder refuses to
	/// continue rather than emit frames that cannot be read. Recover by rolling a new group and
	/// calling [`stream::Encoder::reset`].
	#[error("compression desynchronized: a frame was encoded but never written")]
	Desync,

	/// A [`stream`] track carried a second group, which a lossless log cannot do.
	///
	/// A stream is a single group by construction: a publisher that cannot write a record ends the
	/// track rather than rolling. A second group therefore means the records that would have
	/// completed the first one are gone, so the read fails instead of presenting the remainder as
	/// a continuous log.
	#[error("stream rolled to a second group")]
	Rolled,

	/// A decoded frame or reconstructed snapshot exceeded its configured byte budget.
	#[error("decoded JSON exceeded {0} bytes")]
	TooLarge(usize),
}

impl From<serde_json::Error> for Error {
	fn from(err: serde_json::Error) -> Self {
		Error::Json(err.to_string())
	}
}

/// A [`Result`](std::result::Result) using this crate's [`Error`].
pub type Result<T> = std::result::Result<T, Error>;