Streams API

Streaming involves breaking a resource that you want to receive over a network down into small chunks, then processing it bit by bit. This is something browsers do anyway when receiving assets to be shown on webpages — videos buffer and more is gradually available to play, and sometimes you'll see images display gradually as more is loaded.

But this has never been available to JavaScript before. Previously, if we wanted to process a resource of some kind (be it a video, or a text file, etc.), we'd have to download the entire file, wait for it to be deserialized into a suitable format, then process the whole lot after it is fully received.

With Streams being available to JavaScript, this all changes — you can now start processing raw data with JavaScript bit by bit as soon as it is available on the client-side, without needing to generate a buffer, string, or blob.

There are more advantages too — you can detect when streams start or end, chain streams together, handle errors and cancel streams as required, and react to the speed the stream is being read at.

The basic usage of Streams hinges around making responses available as streams. For example, the response body returned by a successful fetch request can be exposed as a ReadableStream, and you can then read it using a reader created with ReadableStream.getReader(), cancel it with ReadableStream.cancel(), etc.

More complicated uses involve creating your own stream using the ReadableStream() constructor, for example to process data inside a service worker.

You can also write data to streams using WritableStream.

ReadableStream

Represents a readable stream of data. It can be used to handle response streams of the Fetch API, or developer-defined streams (e.g. a custom ReadableStream() constructor).

ReadableStreamDefaultReader

Represents a default reader that can be used to read stream data supplied from a network (e.g. a fetch request).

ReadableStreamDefaultController

Represents a controller allowing control of a ReadableStream's state and internal queue. Default controllers are for streams that are not byte streams.

WritableStream

Provides a standard abstraction for writing streaming data to a destination, known as a sink. This object comes with built-in backpressure and queuing.

WritableStreamDefaultWriter

Represents a default writable stream writer that can be used to write chunks of data to a writable stream.

WritableStreamDefaultController

Represents a controller allowing control of a WritableStream's state. When constructing a WritableStream, the underlying sink is given a corresponding WritableStreamDefaultController instance to manipulate.

TransformStream

Represents a set of transformable data.

TransformStreamDefaultController

Provides methods to manipulate the ReadableStream and WritableStream associated with a transform stream.

ByteLengthQueuingStrategy

Provides a built-in byte length queuing strategy that can be used when constructing streams.

CountQueuingStrategy

Provides a built-in chunk counting queuing strategy that can be used when constructing streams.

Request

When a new Request object is constructed, you can pass it a ReadableStream in the body property of its RequestInit dictionary. This Request could then be passed to a fetch() to commence fetching the stream.

Response.body

The response body returned by a successful fetch request is exposed by default as a ReadableStream, and can have a reader attached to it, etc.

ReadableStreamBYOBReader

Represents a BYOB ("bring your own buffer") reader that can be used to read stream data supplied by the developer (e.g. a custom ReadableStream() constructor).

ReadableByteStreamController

Represents a controller allowing control of a ReadableStream's state and internal queue. Byte stream controllers are for byte streams.

ReadableStreamBYOBRequest

Represents a pull into request in a ReadableByteStreamController.

We have created a directory of examples to go along with the Streams API documentation — see mdn/dom-examples/streams. The examples are as follows:

• Simple stream pump: This example shows how to consume a ReadableStream and pass its data to another.
• Grayscale a PNG: This example shows how a ReadableStream of a PNG can be turned into grayscale.
• Simple random stream: This example shows how to use a custom stream to generate random strings, enqueue them as chunks, and then read them back out again.
• Simple tee example: This example extends the Simple random stream example, showing how a stream can be teed and both resulting streams can be read independently.
• Simple writer: This example shows how to write to a writable stream, then decode the stream and write the contents to the UI.
• Unpack chunks of a PNG: This example shows how pipeThrough() can be used to transform a ReadableStream into a stream of other data types by transforming a data of a PNG file into a stream of PNG chunks.

Examples from other developers:

• Progress Indicators with Streams, Service Workers, & Fetch.

• Streams API concepts
• Using readable streams
• Using readable byte streams
• Using writable streams