What is Streamz?

Streamz was a project made purely as a playground to look for ideas and ways to implement actual lazy evaluation of collection manipulations. It turns out this is a fairly usable implementation and could actually make a nice library.

Streamz is mainly inspired by C#'s LINQ, Java 8's Stream API, Rust's iter, Kotlin's sequences, C++'s Range-v3 and finally sequency.

How to install Streamz?

First you need to install it via NPM : npm i -S @voltra/streamz

Streamz does not provide a direct "plug into browser" file yet so you'll be required to use a build system (if you are targeting browsers).

Once installed, you can simply require/import it :

import { Stream } from "@voltra/streamz"

What does Streamz provide?

Streamz was fully developed using typescript, therefore you have access to the entire source code in the src directory.

The entry point dist/index.js provides the following :

import {
    Stream,
    Packer,
    KeyGen,
    ValueGen,
    emptyPayloadValue,
    isPayloadValueEmpty,
    streamIsValidValue,
    invalidStreamIteratorPayload
} from "@voltra/streamz"

• Stream is the class used for collection manipulation, it provides factories
• Packer is used to transform a Stream into a collection (e.g. an array, an object, a Map, a Set, etc...)
• KeyGen is an helper object that contains functions destined to be used in Packer#toObject and similar methods
• ValueGen is an helper object that contains functions destined to be used in Packer#toObject and similar methods
• emptyPayloadValue is a global Symbol use to mark invalid payloads
• isPayloadValueEmpty is a function that checks whether or not the given value is one of an invalid payload
• streamIsValidValue is a function that checks whether or not the given value is valid (not from an invalid payload)
• invalidStreamIteratorPayload is a factory for invalid payloads

src/index.ts provides the following :

import {
    Stream,
    Packer,
    KeyGen,
    ValueGen,
    emptyPayloadValue,
    isPayloadValueEmpty,
    streamIsValidValue,
    invalidStreamIteratorPayload,
    BaseStreamIterator,
    EndStreamIterator,
    StreamIteratorPayload
} from "@voltra/streamz/src/index"

• BaseStreamIterator is an interface that represents both first and intermediate iterators in the iterator chain (i.e. operation chain)
• EndStreamIterator is an interface that represents the final iterator in the iterator chain (i.e. operation chain)
• StreamIteratorPayload is an interface that represents the payload retrieved when calling next on an iterator

Helper types are provided under src/types (function.d.ts and helpers.d.ts).

Examples

import { Stream, KeyGen, ValueGen } from "@voltra/streamz"

Stream.of(1, 2, 3, 4) //Stream{1, 2, 3, 4}
.map(x => x+1) //Stream{2, 3, 4, 5}
.filter(x => x%2) //Stream{3, 5}
.pack.toArray(); //[3, 5]


Stream.fromObject({hello: "world", wombo: "combo"})
.map(([k, v]) => ([`key(${k})`, `value(${v})`])) 
.pack.toObject(KeyGen.entries, ValueGen.entries); 
//Stream{["hello", "world"], ["wombo", "combo"]}
//Stream{["key(hello)", "value(world)"], ["key(wombo)", "value(combo)"]}
//{"key(wombo)": "value(combo)", "key(hello)": "value(world)"}


Stream.of(1, 2, 1, 1, 2, 1) //Stream{1, 2, 1, 1, 2, 1}
.map(x => x+1) //Stream{2, 3, 2, 2, 3, 2}
.unique() //Stream{2, 3}
.pack.toArray(); //[2, 3]


Stream.of(2, 4, 6, 8) //Stream{2, 4, 6, 8}
.map(x => x+1) //Stream{3, 5, 7, 9}
.all(x => x%2); //true

import { Stream } from "@voltra/streamz"

const X = Stream.of(1, 2, 3, 4)
.map(x => x+1)
.filter(x => x%2)
.pack.toArray();
console.log(X);

//Is roughly equivalent to

const X = [];
for(const it in [1, 2, 3, 4]){
    const x = it+1;
    if(x % 2)
        X.push(x);
}
console.log(X);

Changes

A complete JSDoc will be provided once ready.

v1.0.2

extend in src/extend.ts (also exported in src/index.ts and therefore in dist/index.js) :

• Installs global prototype extensions :
  • Array#stream() is Stream.from(this)
  • Object#stream() is Stream.fromObject(this)
  • Set#stream() is Stream.fromSet(this)
  • Map#stream() is Stream.fromMap(this)
• And class methods extensions
  • Object.fromStreams is Stream.zipToObject
  • Number.range is Stream.range
  • Number.infiniteRange is Stream.infinite

Compare in src/utils.ts (also exported in src/index.ts and therefore in dist/index.js) :

• Compare.asc(lhs, rhs) standard comparison function (ascending)
• Compare.desc(lhs, rhs) standard comparison function (descending)
• Compare.mapped "namespace" for mapped comparison functions
  • Compare.mapped.asc(mapper) crafts a Compare#asc where elements were mapped using mapper
  • Compare.mapped.desc(mapper) crafts a Compare#desc where elements were mapped using mapper

Stream initial operations :

• Stream.fromMap(map) creates a stream from a Map
• Stream.fromSet(set) creates a stream from a Set

Stream intermediate operations :

• Regular
  • Stream#chunked(size = 3) groups items in chunks whose size is at most size
• Zipping
  • Stream#zip(stream) combines this stream (makes a pair using one item from each stream)
  • Stream.zip(lhs, rhs) same as lhs.zip(rhs)
  • Stream#zipBy(stream, mapper) combines this stream using the mapper function to craft the new item
  • Stream.zipBy(lhs, rhs, mapper) same as lhs.zipBy(rhs, mapper)
• Filtering
  • Stream#nonNull() filters out any null
  • Stream#nonFalsy() filters out any falsy value (0, null, undefined, etc...)
  • Stream#filterNot(predicate) filters element that satisfy the predicate
  • Stream#filterOut(predicate) alias for Stream#filterNot(predicate)
  • Stream#filterIsInstance(class) filters out elements that are not instance of class
• Index manipulation
  • Stream#takeWhile(predicate) keeps item until predicate is not satisfied
  • Stream#takeUntil(predicate) keeps item while predicate is not satisfied

Stream terminal operations :

• Regular
  • Stream#count(predicate = (_ => true)) counts the elements that satisfy the predicate
• Predicate tests
  • Stream#contains(elem) checks whether or not elem is an item of this stream
• Index based
  • Stream#atIndex(index) retrieves the element at the given index (or null if there is none)
  • Stream#atIndexOr(index, default) retrieves the element or get back default
  • Stream#first() retrieves the first element (or null)
  • Stream#firstOr(default) retrieves the first element (or default)
• Zipping
  • Stream#zipToObject(stream) zips into an object using this as keys and stream as values
  • Stream.zipToObject(keys, values) equivalent to keys.zipToObject(values)
  • Stream#unzip() unzips a stream of pairs into a pair of arrays
  • Stream#unzipBy(firstGen, lastGen) unzips a stream into a pair of arrays applyingfirstGen for the first element of the pair and lastGen for the last element of the pair
  • Stream#unzipVia(mapper) convenience method to unzip a stream of pairs into a pair of arrays using a single mapper function that returns a pair
• Sorting
  • Stream#sortedWith(comparator) packs to an array sorted using comparator as the comparison function
  • Stream#sortedBy(mapper) maps the objects and sort in ascending order via regular comparison operators
  • Stream#sortedDescBy(mapper) same as sortedBy but in descending order
  • Stream#sorted() same as sortedBy but without mapping
  • Stream#sortedDesc() same as sorted but in descending order

v1.0.1 and inferior

Factories (creators)

Streamz is not a usual object, it doesn't really expect you to call its constructor, it provides a lot of static factory methods in order for you to get the most fluent experience possible :

• Stream.of(...args) takes any amount of parameters and makes a stream out of it
• Stream.from(args) takes an array and makes a stream out of it
• Stream.fromObject(obj) takes an object and makes a stream out of it
• Stream.range(higher) takes a number used as its lower bound makes the range [0 ; higher) as a stream
• Stream.range(lower, higher, step = 1) takes both bounds, the step and makes the range [lower ; higher) as a stream (uses step as increment instead of 1)
• Stream.infinite(lower = 0) makes a stream for the range [lower ; Infinity)

Intermediate computations (intermediary)

Streamz offers a lot of intermediate computations, these are not effective until a blocking computation is requested :

• Stream#map(mapper) maps each element using the provided mapper function
• Stream#filter(predicate) only gives back elements that satisfy the provided predicate
• Stream#peek(functor) applies the function on each element and passes it
• Stream#unique() removes duplicates (uses a Set behind the scenes to check for duplicates)
• Stream#take(amount = 10) only keeps the first x ≤ amount elements
• Stream#skip(amount = 10) skips the first x ≤ amount elements
• Stream#between(begin = 0, end = 10, excludeRight = false) takes the element whose index is in the range [begin ; end] if excludeRight == false otherwise from [begin ; end)

Blocking computations (terminators)

Operations

Common

• Stream#forEach(functor) applies the provided function on each element
• Stream#reduce(reducer, acc) computes the reduced value by getting a new acc by applying the reducer function and the current acc and the current element

Predicative

• Stream#all(predicate) checks whether or not every element satisfies the given predicate
• Stream#any(predicate) checks whether or not any element satisfies the given predicate
• Stream#none(predicate) checks whether or not none of the elements satisfies the given predicate

Packing

Stream#pack is an instance of Packer and exposes the following :

• Packer#toArray() converts the stream into an array
• Packer#toObjectBy(keyGen) converts the stream into an object whose keys are generated using keyGen on the current element and whose values are the elements of the stream
• Packer#toObject(keyGen, valueGen) converts the stream into an object whose keys and values are generated using the current element on keyGen and valueGen
• Packer#toSet() converts the stream into a Set
• Packer#toMapBy(keyGen) converts the stream into a Map whose keys are generated using keyGen on the current element and whose values are the elements of the stream
• Packer#toMap(keyGen, valueGen) converts the stream into a Map whose keys and values are generated using the current element on keyGen and valueGen