Parallel Execution

Barnum has three ways to run work concurrently: all for a fixed set of tasks, forEach for dynamic fan-out over arrays, and forEach + pipe for fan-out with a follow-up aggregation step.

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].

forEach — dynamic fan-out

forEach maps an action over each element of an array, processing all elements concurrently.

From demos/simple-workflow/run.ts:

runPipeline(
  listFiles
    .forEach(
      pipe(
        implementRefactor,
        typeCheckFiles,
        fixTypeErrors,
        commitChanges,
        createPullRequest,
      ),
    ),
);

listFiles returns string[]. Each filename flows through the full pipeline independently and concurrently.

Fan-out with aggregation

Chain forEach into a follow-up step to aggregate results after parallel work completes.

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

runPipeline(
  pipe(
    setup,
    listFiles
      .forEach(migrate({ to: "Typescript" }))
      .drop(),
    typeCheckFix,
  ),
);

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