Pipeline Executor

@ucdjs/pipeline-executor is the runtime that executes pipeline definitions against source files.

Role

• Orders pipelines and routes, resolves versions, and executes matched files.
• Applies caching, tracing, logging, and output materialization.
• Turns pure pipeline definitions into concrete execution results.

Related Docs

Mental Model

The executor has two layers:

• createPipelineExecutor() orders multiple pipelines and wraps execution in runtime logging
• run() executes one pipeline version-by-version and route-by-route

The critical runtime responsibilities are:

• source listing
• route matching
• parse/transform/resolve execution
• cache lookup and writeback
• output materialization and summary creation

Method Flows

createPipelineExecutor().run()

Single pipeline execution

Matched file execution

Design Notes

• Pipeline ordering happens before execution so pipeline-output dependencies can cross pipeline boundaries.
• Route ordering happens inside a pipeline through the DAG layers from pipeline-core.
• The runtime abstraction owns spans, logging, and output capture so execution remains portable.
• Errors are accumulated into execution results instead of crashing the whole batch whenever possible.

Testing Use

• pipeline ordering and circular dependency failures
• route-layer execution and concurrency behavior
• cache hit/miss behavior
• fallback and strict unmatched-file handling
• summary accounting, tracing, and failure accumulation