Configuration.fork()

Description

Create a single sub-pipeline

When fork() has no parameters, only one sub-pipeline will be created for each input stream. The sub-pipeline shares the same context from the fork filter's pipeline, aka. the main pipeline. That is, all states in the main pipeline is visible to the sub-pipeline, and all changes to the states in the sub-pipeline will be reflected in the main pipeline as well.

Create multiple sub-pipelines

When fork() is given an argument, it is supposed to be an array, or a function that returns an array. The length of the array is the number of sub-pipelines to create for every input stream. Each sub-pipeline is given one of the array elements as the startup parameter. These sub-pipelines can obtain their startup parameters by the parameter to a callback function in onStart(). Sub-pipelines created this way also get their own new contexts cloned from the main pipeline (the fork filter's pipeline).

pipy({
  _index: 0,
})

  .task()
  .onStart(() => new Message)
  // Create 10 sub-pipelines, each having
  // numbers 1 to 10 for their _index values
  .fork(
    new Array(10).fill().map(
      (_, i) => i
    )
  ).to(
    $=>$
    .onStart(i => _index = i)
    .dump(() => _index) // Dump all events with tag _index
  )

Syntax

pipy()
  .pipeline()
  .fork().to(
    subPipelineLayout
  )

pipy()
  .pipeline()
  .fork([ ...startupValues ]).to(
    subPipelineLayout
  )

pipy()
  .pipeline()
  .fork(() => getStartupValues()).to(
    subPipelineLayout
  )

Parameters

fork(startupValues?)

Example

pipy()

  .listen(8000)
  // Clone all data to a sub-pipeline
  .fork().to(
    // Send the cloned data to localhost:8008
    $=>$.connect('localhost:8008')
  )
  // Respond with 'Hello!' to all requests
  .serveHTTP(
    new Message('Hello!')
  )

See Also

• Configuration
• link()
• use()