Trait sp_consensus::SelectChain

pub trait SelectChain<Block: BlockT>: Sync + Send + Clone {
    fn leaves(&self) -> Result<Vec<<Block as BlockT>::Hash>, Error>;
    fn best_chain(&self) -> Result<<Block as BlockT>::Header, Error>;

    fn finality_target(
        &self, 
        target_hash: <Block as BlockT>::Hash, 
        _maybe_max_number: Option<NumberFor<Block>>
    ) -> Result<Option<<Block as BlockT>::Hash>, Error> { ... }
}

The SelectChain trait defines the strategy upon which the head is chosen if multiple forks are present for an opaque definition of “best” in the specific chain build.

The Strategy can be customized for the two use cases of authoring new blocks upon the best chain or which fork to finalize. Unless implemented differently by default finalization methods fall back to use authoring, so as a minimum _authoring-functions must be implemented.

Any particular user must make explicit, however, whether they intend to finalize or author through the using the right function call, as these might differ in some implementations.

Non-deterministically finalizing chains may only use the _authoring functions.

Required methods

fn leaves(&self) -> Result<Vec<<Block as BlockT>::Hash>, Error>

Get all leaves of the chain: block hashes that have no children currently. Leaves that can never be finalized will not be returned.

fn best_chain(&self) -> Result<<Block as BlockT>::Header, Error>

Among those leaves deterministically pick one chain as the generally best chain to author new blocks upon and probably finalize.

Provided methods

fn finality_target(
    &self,
    target_hash: <Block as BlockT>::Hash,
    _maybe_max_number: Option<NumberFor<Block>>
) -> Result<Option<<Block as BlockT>::Hash>, Error>

Get the best descendent of target_hash that we should attempt to finalize next, if any. It is valid to return the given target_hash itself if no better descendent exists.

Implementors