Concepts

Undercut helps constructing pipelines for data processing. Instead of creating new concepts it leverages existing JavaScript protocols and features like Iteration protocols and Generators.

Pipelines is just a term for ordered sequences of operations (just like a conveyor on a factory). Imagine you want to process a bunch of source items and put the result items into a target.

source ----> [ op_0 | op_1 | ... | op_N ] ----> target                       pipeline

Depending on the fact whether the source items are or aren't yet available you will have two different approaches: pull and push.

Pull#

If source items are available at the moment of execution, we can pull items from it one by one and process them synchronously. Undercut call this a Pull Line. It is created by combining a source and a pipeline into an Iterable. This Iterable may be passed around and used by any native ES construct like for-of loop to read the result. Undercut also provides a bunch of target helpers for one-line extracting results out of Iterables into common structures like arrays/objects/maps/etc.

Pull Line -- pull items from an Iterable:

const source = [0, 1, 2, 3];const pipeline = [    append(4, 5),    compact(),    skip(2)];
// source + pipeline = Iterableconst iterable = pullLine(pipeline, source);

Push#

If source items are not available at the moment of execution, we have to process items as they appear. We can do this by pushing items into a Push Line. It is created by combining a pipeline and a target into an Observer. Observers are convenient in use cases like reacting on a button click and may be passed around or used by several producers.

Push line -- push items into an Observer:

const target = toArray();const pipeline = [    append(4, 5),    compact(),    skip(2)];
// pipeline + target = Observerconst observer = pushLine(pipeline, target);

Of course, you can process synchronous items with Push Lines too, but Pull Lines are easier to operate and write operations for.