Converts various types of streams to Kefir streams. This is intended for use by libraries which use Kefir internally, but want to be able to accept streams from other libraries or other versions of Kefir as arguments given by the application.
- Supports converting RxJS (v2-v7) Observables into Kefir streams.
- Supports converting Bacon.js (v0.7-v3) streams into Kefir streams.
- Supports converting a Kefir stream created by one instance of the Kefir library into a stream usable by a different instance of a Kefir library. (This smooths over possible issues if the application uses a different version of Kefir.)
- Casts non-streams into a Kefir stream of one item by using Kefir.constant().
KefirCast is intended for use in nodejs and in browsers via CommonJS bundlers like Browserify. This project is in NPM and can be installed with
npm install kefir-cast
The related project BaconCast exists to convert streams to Bacon.js streams.
Suppose you have a library that exports a single function doStuff
, which can
take a stream as an argument. By using KefirCast, you can support any RxJS
streams, Kefir streams, Bacon.js streams, or constants that your users might
pass to you.
var Kefir = require('kefir');
var kefirCast = require('kefir-cast');
module.exports = function doStuff(input) {
var inputStream = kefirCast(Kefir, input);
// Log anything that comes through the stream for 5 seconds.
inputStream.takeUntilBy(Kefir.later(5000)).onValue(function(value) {
console.log('doStuff received value', value);
});
};
If you did not use KefirCast, then your users would be required to use the same version of Kefir as you, you wouldn't support RxJS or Bacon streams without more work, and you would have to handle non-stream constant values specially.
kefirCast(Kefir, input)
takes your Kefir library instance as the first
argument, and the input stream or constant to convert into a Kefir stream as
the second argument. A Kefir stream compatible with the given Kefir library
will be returned.
Both TypeScript and Flow type definitions for this module are included! The type definitions won't require any configuration to use.