Struct Decoder 

pub struct Decoder(/* private fields */);

Reed-Solomon decoder using DefaultEngine and DefaultRate.

Implementations

impl Decoder

pub fn add_original_shard<T: AsRef<[u8]>>( &mut self, index: usize, original_shard: T, ) -> Result<(), Error>

Adds one original shard to the decoder.

• Shards can be added in any order.
• The index must be the same one used in encoding.

See basic usage for an example.

pub fn add_recovery_shard<T: AsRef<[u8]>>( &mut self, index: usize, recovery_shard: T, ) -> Result<(), Error>

Adds one recovery shard to the decoder.

• Shards can be added in any order.
• The index must be the same one used in encoding.

See basic usage for an example.

pub fn decode(&mut self) -> Result<Option<DecoderResult<'_>>, Error>

Decodes the added shards, returning Some(DecoderResult) with the restored original shards, or None if every original shard was already provided (nothing to reconstruct).

When the returned DecoderResult is dropped the decoder is automatically reset and ready for a new round of decoding. When every original shard was supplied, this returns None and clears the received shards immediately. A result that is leaked instead of dropped (for example with core::mem::forget) leaves the decoder reporting a complete round, and the next decode would transform the previous round’s buffers again. Call reset before reusing a decoder whose result was leaked.

See basic usage for an example.

pub fn decode_with_plan( &mut self, plan: &Plan, ) -> Result<Option<DecoderResult<'_>>, Error>

Like decode, but reads erasure coefficients from plan instead of computing them.

plan must match this decoder’s shard counts and received shard indices. Shard byte length is not part of the plan. A mismatched plan returns Error::PlanMismatch and leaves the added shards in place, so the decoder can retry with the matching plan or decode.

pub fn decode_with_recovery( &mut self, ) -> Result<Option<RecoveryDecoderResult<'_>>, Error>

Like decode, but also reconstructs the missing recovery shards, returning Some(RecoveryDecoderResult) that additionally exposes them via RecoveryDecoderResult::recovery / recovery_iter. This adds a multiplication of each missing recovery shard by a field element, so prefer decode when only the original data is needed.

pub fn decode_with_recovery_plan( &mut self, plan: &Plan, ) -> Result<Option<RecoveryDecoderResult<'_>>, Error>

Like decode_with_recovery, but reads erasure coefficients from plan as decode_with_plan does.

pub fn new( original_count: usize, recovery_count: usize, shard_bytes: usize, ) -> Result<Self, Error>

Creates a new decoder with the given configuration and allocates the required working space.

See working space for its size and the configurations a coder accepts, and basic usage for an example.

pub fn reset( &mut self, original_count: usize, recovery_count: usize, shard_bytes: usize, ) -> Result<(), Error>

Resets the decoder to the given configuration.

• Added shards are forgotten.
• Existing working space is reused if it is large enough, or reallocated otherwise.

pub fn supports(original_count: usize, recovery_count: usize) -> bool

Returns true if the given original_count / recovery_count combination is supported.

Examples

use commonware_cryptography::reed_solomon::Decoder;

assert_eq!(Decoder::supports(60_000, 4_000), true);
assert_eq!(Decoder::supports(60_000, 5_000), false);