Parallel Execution

Barnum has three ways to run work concurrently: all for a fixed set of tasks, Iterator.map for dynamic fan-out over arrays, and Iterator.flatMap for fan-out with flattening.

all — fixed parallel tasks

all runs multiple actions concurrently with the same input and collects their results as a tuple.

From demos/analyze-file/run.ts:

runPipeline(
  all(analyzeClassComponents, analyzeImpossibleStates, analyzeErrorHandling),
  "source/index.ts",
);

All three analyzers receive "source/index.ts" as input and run in parallel. The output is a tuple [ResultA, ResultB, ResultC].

Iterator.map — dynamic fan-out

.iterate() wraps an array as an Iterator. .map(action) applies an action to each element concurrently. .collect() gathers results back into an array.

From demos/simple-workflow/run.ts:

runPipeline(
  listFiles
    .iterate()
    .map(
      implementRefactor
        .then(typeCheckFiles)
        .then(fixTypeErrors)
        .then(commitChanges)
        .then(createPullRequest),
    )
    .collect(),
);

listFiles returns string[]. .iterate() enters Iterator, .map() fans out — each filename flows through the full pipeline independently and concurrently. .collect() gathers results.

Fan-out with aggregation

Chain .collect() into a follow-up step to aggregate results after parallel work completes.

From demos/convert-folder-to-ts/run.ts:

runPipeline(
  setup
    .then(
      listFiles.iterate().map(migrate({ to: "Typescript" })).collect().drop(),
    )
    .then(typeCheckFix),
);

All files are migrated in parallel. After every migration finishes, .collect() gathers results, .drop() discards them, and typeCheckFix runs once — a single type-check pass over the entire project, not per file.