Documentation

Choose what you need

Learn the model in order, solve a concrete ETL problem, or look up the exact contract of one operator.

New to Exstream?

Start with the quick start if you have not run a pipeline yet. Then read the pipeline model and backpressure. Together they explain when work starts, who controls the pace, and why the pipeline stays bounded.

source → transformations → terminal consumer
   ↑                              │
   └──────────── demand ──────────┘

Solving a problem?

Guides begin with a real task and end with a complete pipeline. The first one fetches and transforms a public CSV in the browser without a server or upload.

More guides will cover bounded enrichment, multiple writers, paginated APIs, dead-letter flows, and hot event sources. They belong here only when their code can be verified against a released package.

Looking up an operator?

Use the operator index. Reference pages state the details production code depends on: input and output, sync or async execution, order, concurrency, buffering, backpressure, errors, cancellation, and runtime support.

Deciding whether to adopt it?

Read when to use Exstream. Native arrays, a small for await loop, Web Streams, or a columnar engine can all be the better answer. Exstream earns its place when the pipeline as a whole becomes the difficult part.

Runtime baseline

• Node.js 22 or newer
• ESM and CommonJS entry points
• Portable core for modern browsers
• Zero runtime dependencies
• TypeScript declarations included

The default import selects the Node.js or browser implementation through package exports. Explicit entry points are available as exstream.js/node, exstream.js/core, and exstream.js/web.