Skip to main content

Crate pondrs

Crate pondrs 

Source
Expand description

A pipeline execution library for building data processing workflows.

Pipelines are composed of Nodes that read from and write to Datasets. Nodes are grouped into Steps (tuples) and executed by a Runner. The App struct ties everything together with CLI dispatch, YAML config loading, and hook-based lifecycle events.

Works in no_std environments (with CellDataset + SequentialRunner) and scales up to parallel execution with std.

Re-exports§

pub use app::App;
pub use catalog_indexer::CatalogIndex;
pub use catalog_indexer::index_catalog;
pub use catalog_indexer::index_catalog_with_params;
pub use datasets::Dataset;
pub use datasets::DatasetMeta;
pub use datasets::TemplatedCatalog;
pub use error::CheckError;
pub use error::PondError;
pub use graph::PipelineGraph;
pub use graph::build_pipeline_graph;
pub use hooks::Hook;
pub use hooks::Hooks;
pub use pipeline::DatasetEvent;
pub use pipeline::DatasetRef;
pub use pipeline::Ident;
pub use pipeline::IntoNodeResult;
pub use pipeline::Node;
pub use pipeline::Pipeline;
pub use pipeline::PipelineFn;
pub use pipeline::PipelineInfo;
pub use pipeline::RunnableStep;
pub use pipeline::StepInfo;
pub use pipeline::Steps;
pub use pipeline::StepVec;
pub use pipeline::Split;
pub use pipeline::Join;
pub use runners::ParallelRunner;
pub use runners::Runner;
pub use runners::Runners;
pub use runners::SequentialRunner;

Modules§

app
Application framework for building pipeline executables.
catalog_indexer
Catalog indexer: maps dataset pointer IDs to human-readable field names.
datasets
Dataset types for loading and saving data.
error
Error types for the pipeline framework.
graph
Pipeline graph: a pre-computed DAG representation of the pipeline.
hooks
Hooks for pipeline execution events.
pipeline
Core types and traits for pipelines.
runners
Pipeline runners.
viz
Pipeline visualization: axum server, serialization, and VizHook.