pub trait DataSourceProvider: Send + Sync {
// Required method
fn fetch<'life0, 'async_trait>(
&'life0 self,
request: FetchRequest,
) -> Pin<Box<dyn Future<Output = Result<FetchResult, FetchError>> + Send + 'async_trait>>
where Self: 'async_trait,
'life0: 'async_trait;
// Provided method
fn shutdown<'life0, 'async_trait>(
&'life0 self,
) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>
where Self: 'async_trait,
'life0: 'async_trait { ... }
}Expand description
Provider dispatch + fetch trait. Host apps implement this for their
custom data sources (BigQuery, Snowflake, internal REST APIs, …) and
register them via ChartML::register_provider(kind, provider).
?Send on WASM matches TransformMiddleware so single-threaded
browser environments can hold non-Send state inside provider impls.
Required Methods§
Sourcefn fetch<'life0, 'async_trait>(
&'life0 self,
request: FetchRequest,
) -> Pin<Box<dyn Future<Output = Result<FetchResult, FetchError>> + Send + 'async_trait>>where
Self: 'async_trait,
'life0: 'async_trait,
fn fetch<'life0, 'async_trait>(
&'life0 self,
request: FetchRequest,
) -> Pin<Box<dyn Future<Output = Result<FetchResult, FetchError>> + Send + 'async_trait>>where
Self: 'async_trait,
'life0: 'async_trait,
Fetch one source. The resolver handles caching + dedup + parallelism; providers only see this single call per actual upstream invocation.
Provided Methods§
Sourcefn shutdown<'life0, 'async_trait>(
&'life0 self,
) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>where
Self: 'async_trait,
'life0: 'async_trait,
fn shutdown<'life0, 'async_trait>(
&'life0 self,
) -> Pin<Box<dyn Future<Output = ()> + Send + 'async_trait>>where
Self: 'async_trait,
'life0: 'async_trait,
Optional graceful shutdown hook. Called by ChartML::shutdown() on
SSR request end / tab close. Default no-op so providers that hold no
pooled resources don’t have to implement it.