Trait Encoder

Source

pub trait Encoder {
    // Required methods
    fn encode(&self, writer: &mut BytesMut) -> Result<()>;
    fn is_default(&self) -> bool;

    // Provided method
    fn pack(&self, writer: &mut BytesMut) -> Result<()> { ... }
}

Expand description

Trait for types that can be encoded into the senax binary format.

Implement this trait for your type to enable serialization. Most users should use #[derive(Encode)] instead of manual implementation.

§Errors

Returns EncoderError if the value cannot be encoded.

Required Methods§

Source

fn encode(&self, writer: &mut BytesMut) -> Result<()>

Encode the value into the given buffer with schema evolution support.

This method includes field IDs and type tags for forward/backward compatibility. Use this when you need schema evolution support.

§Arguments

writer - The buffer to write the encoded bytes into.

Source

fn is_default(&self) -> bool

Returns true if this value equals its default value. Used by #[senax(skip_default)] attribute to skip encoding default values.

Provided Methods§

Source

fn pack(&self, writer: &mut BytesMut) -> Result<()>

Pack the value into the given buffer without schema evolution support.

This method omits field IDs and some type tags for compact encoding. Fields are encoded in declaration order. Use this when you need maximum performance and compact size, but don’t need schema evolution.

§Arguments

writer - The buffer to write the packed bytes into.

§Note

For structs: Field IDs are omitted, fields are encoded in declaration order
For enums: Variant IDs are still included, but field IDs within variants are omitted
For primitives: Some type tags may be omitted for maximum compactness

Implementations on Foreign Types§

Source §

impl Encoder for bool

Encodes a bool as a single tag byte: TAG_ZERO for false, TAG_ONE for true.

Source §

fn encode(&self, writer: &mut BytesMut) -> Result<()>

Source §

fn is_default(&self) -> bool

Source §

impl Encoder for f32

Encodes an f32 as a tag and 4 bytes (little-endian IEEE 754).

Source §

fn pack(&self, writer: &mut BytesMut) -> Result<()>

Packs an f32 as 4 bytes (little-endian IEEE 754) without a type tag.

Source §

fn encode(&self, writer: &mut BytesMut) -> Result<()>

Source §

fn is_default(&self) -> bool

Source §

impl Encoder for f64

Encodes an f64 as a tag and 8 bytes (little-endian IEEE 754).

Source §

fn pack(&self, writer: &mut BytesMut) -> Result<()>

Packs an f64 as 8 bytes (little-endian IEEE 754) without a type tag.

Source §

fn encode(&self, writer: &mut BytesMut) -> Result<()>

Source §

fn is_default(&self) -> bool

Source §

impl Encoder for i8

Encodes signed integers using bit-inverted encoding for negative values.

Non-negative values (>= 0) are encoded as unsigned integers
Negative values use TAG_NEGATIVE and bit-inverted encoding

Source §

impl Encoder for u8

Encodes unsigned integers using a compact variable-length format.

Values 0/1 are encoded as TAG_ZERO/TAG_ONE (1 byte)
2..=127 are encoded as a single tag byte (1 byte)
Larger values use TAG_U8, TAG_U16, TAG_U32, TAG_U64, or TAG_U128 with the value in little-endian
The encoding is stable and compatible across platforms