Skip to main content

Cache

vantage_live::cache

Trait Cache 

Source 
pub trait Cache: Send + Sync {
    // Required methods
    fn get<'life0, 'life1, 'async_trait>(
        &'life0 self,
        key: &'life1 str,
    ) -> Pin<Box<dyn Future<Output = Result<Option<CachedRows>>> + Send + 'async_trait>>
       where Self: 'async_trait,
             'life0: 'async_trait,
             'life1: 'async_trait;
    fn put<'life0, 'life1, 'async_trait>(
        &'life0 self,
        key: &'life1 str,
        rows: CachedRows,
    ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>
       where Self: 'async_trait,
             'life0: 'async_trait,
             'life1: 'async_trait;
    fn invalidate_prefix<'life0, 'life1, 'async_trait>(
        &'life0 self,
        prefix: &'life1 str,
    ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>
       where Self: 'async_trait,
             'life0: 'async_trait,
             'life1: 'async_trait;
}
Expand description

Cache backend interface. Implementations must be safe to share across the read path, the write-queue worker, and the live-event consumer.

Required Methods§

Source

fn get<'life0, 'life1, 'async_trait>( &'life0 self, key: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<Option<CachedRows>>> + Send + 'async_trait>>
where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Look up key; Ok(None) is a normal miss, Err is reserved for real backend faults (disk error, redb panic recovery, etc.).

Source

fn put<'life0, 'life1, 'async_trait>( &'life0 self, key: &'life1 str, rows: CachedRows, ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>
where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Store rows under key, replacing whatever was there.

Source

fn invalidate_prefix<'life0, 'life1, 'async_trait>( &'life0 self, prefix: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<()>> + Send + 'async_trait>>
where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Drop every entry whose key starts with prefix. v1 callers always pass the bare cache_key — every page suffix below it goes. Implementations should accept any prefix (so finer-grained invalidation can be added later without breaking the trait).

Implementors§