Seq

Seq describes a lazy operation, allowing them to efficiently chain use of all the higher-order collection methods (such as map and filter) by not creating intermediate collections.

type Seq<K, V> extends Collection<K, V>

Seq is immutable — Once a Seq is created, it cannot be changed, appended to, rearranged or otherwise modified. Instead, any mutative method called on a Seq will return a new Seq.

Seq is lazy — Seq does as little work as necessary to respond to any method call. Values are often created during iteration, including implicit iteration when reducing or converting to a concrete data structure such as a List or JavaScript Array.

For example, the following performs no work, because the resulting Seq's values are never iterated:

import { Seq } from 'immutable';
const oddSquares = Seq([1, 2, 3, 4, 5, 6, 7, 8])
  .filter((x) => x % 2 !== 0)
  .map((x) => x * x);

Once the Seq is used, it performs only the work necessary. In this example, no intermediate arrays are ever created, filter is called three times, and map is only called once:

oddSquares.get(1); // 9

Any collection can be converted to a lazy Seq with Seq().

import { Map } from 'immutable';

const map = Map({ a: 1, b: 2, c: 3 });
const lazySeq = Seq(map);

Seq allows for the efficient chaining of operations, allowing for the expression of logic that can otherwise be very tedious:

lazySeq
  .flip()
  .map((key) => key.toUpperCase())
  .flip();
// Seq { A: 1, B: 1, C: 1 }

As well as expressing logic that would otherwise seem memory or time limited, for example Range is a special kind of Lazy sequence.

Seq is often used to provide a rich collection API to JavaScript Object.

Construction

Creates a Seq.

function Seq<S extends Seq>(seq: S): S;
function Seq(collection: Collection.Keyed<K, V>): Seq.Keyed<K, V>;
function Seq(collection: Collection.Set<T>): Seq.Set<T>;
function Seq(collection: Collection.Indexed<T> | Iterable<T> | ArrayLike<T>): Seq.Indexed<T>;
function Seq(obj: { [key: string]: V }): Seq.Keyed<string, V>;

Returns a particular kind of Seq based on the input.

• If a Seq, that same Seq.
• If an Collection, a Seq of the same kind (Keyed, Indexed, or Set).
• If an Array-like, an Seq.Indexed.
• If an Iterable Object, an Seq.Indexed.
• If an Object, a Seq.Keyed.

Note: An Iterator itself will be treated as an object, becoming a Seq.Keyed, which is usually not what you want. You should turn your Iterator Object into an iterable object by defining a Symbol.iterator (or @@iterator) method which returns this.

Note: Seq is a conversion function and not a class, and does not use the new keyword during construction.

Static methods

function isSeq(maybeSeq: unknown): maybeSeq is Seq.Indexed | Seq.Keyed | Seq.Set;

Members

Some Seqs can describe their size lazily. When this is the case, size will be an integer. Otherwise it will be undefined.

readonly size: number | undefined;

For example, Seqs returned from map or reverse preserve the size of the original Seq while filter does not.

Note: Range(),

Repeat() and Seqs made from Lists and Maps will always have a size.

Force evaluation

Because Sequences are lazy and designed to be chained together, they do not cache their results. For example, this map function is called a total of 6 times, as each join iterates the Seq of three values.

cacheResult(): this;

var squares = Seq([1, 2, 3]).map((x) => x * x);
squares.join() + squares.join();

If you know a Seq will be used multiple times, it may be more efficient to first cache it in memory. Here, the map function is called only 3 times.

var squares = Seq([1, 2, 3])
  .map((x) => x * x)
  .cacheResult();
squares.join() + squares.join();

Use this method judiciously, as it must fully evaluate a Seq which can be a burden on memory and possibly performance.

Note: after calling cacheResult, a Seq will always have a size.

Sequence algorithms

Returns a new Seq with values passed through a mapper function.

map<M>(mapper: (value: V, key: K, iter: this) => M, context?: unknown): Seq<K, M>;

Note: map always returns a new instance, even if it produced the same value at every step. Note: used only for sets.

Flat-maps the Seq, returning a Seq of the same type.

flatMap<M>(mapper: (value: V, key: K, iter: this) => Iterable<M>, context?: unknown): Seq<K, M>;

Similar to map(...).flatten(true).

Note: Used only for sets.

Returns a new Seq with only the values for which the predicate function returns true.

filter(predicate: (value: V, key: K, iter: this) => unknown, context?: unknown): this;

Note: filter always returns a new instance, even if it results in not filtering out any values.

Returns a new Seq with the values for which the predicate function returns false and another for which is returns true.

partition<F extends V>(
  predicate: (this, value: V, key: K, iter) => value is F,
  context
): [Seq<K, V>, Seq<K, F>];

Returns a new Seq of the same type with other values and collection-like concatenated to this one.

concat(...valuesOrCollections): Seq;

All entries will be present in the resulting Seq, even if they have the same key.

Returns a new Collection of the same type with only the entries for which the predicate function returns false.

filterNot(predicate: (value: V, key: K, iter: this) => boolean, context): this;

Note: filterNot() always returns a new instance, even if it results in not filtering out any values.

Returns a new Collection of the same type in reverse order.

reverse(): this;

Returns a new Collection of the same type which includes the same entries, stably sorted by using a comparator.

sort(comparator?: Comparator<V>): this;

If a comparator is not provided, a default comparator uses < and >.

comparator(valueA, valueB):

• Returns 0 if the elements should not be swapped.
• Returns -1 (or any negative number) if valueA comes before valueB
• Returns 1 (or any positive number) if valueA comes after valueB
• Alternatively, can return a value of the PairSorting enum type
• Is pure, i.e. it must always return the same value for the same pair of values.

When sorting collections which have no defined order, their ordered equivalents will be returned. e.g. map.sort() returns OrderedMap.

Note: sort() Always returns a new instance, even if the original was already sorted.

Note: This is always an eager operation.

Like sort, but also accepts a comparatorValueMapper which allows for sorting by more sophisticated means.

sortBy<C>(comparatorValueMapper: (value: V, key: K, iter: this) => C, comparator?: Comparator<C>): this;

Note: sortBy() Always returns a new instance, even if the original was already sorted.

Note: This is always an eager operation.

Returns a Map of Collection, grouped by the return value of the grouper function.

groupBy<G>(
  grouper: (value: V, key: K, iter: this) => G,
  context?: unknown
): Map<G, this>

Note: This is always an eager operation.

Value equality

True if this and the other Collection have value equality, as defined by Immutable.is().

equals(other: unknown): boolean;

Note: This is equivalent to Immutable.is(this, other), but provided to allow for chained expressions.

Computes and returns the hashed identity for this Collection.

hashCode(): number;

The hashCode of a Collection is used to determine potential equality, and is used when adding this to a Set or as a key in a Map, enabling lookup via a different instance.

import { Seq, Set } from 'immutable';

const a = Seq([1, 2, 3]);
const b = Seq([1, 2, 3]);
assert.notStrictEqual(a, b); // different instances
const set = Set([a]);
assert.equal(set.has(b), true);

Note: hashCode() MUST return a Uint32 number. The easiest way to guarantee this is to return myHash | 0 from a custom implementation. If two values have the same hashCode, they are not guaranteed to be equal. If two values have different hashCodes, they must not be equal.

Note: hashCode() is not guaranteed to always be called before equals(). Most but not all Immutable.js collections use hash codes to organize their internal data structures, while all Immutable.js collections use equality during lookups.

Reading values

Returns the value associated with the provided key, or notSetValue if the Collection does not contain this key.

Note: it is possible a key may be associated with an undefined value, so if notSetValue is not provided and this method returns undefined, that does not guarantee the key was not found.

get<NSV>(key: K, notSetValue: NSV): V | NSV;

get(key: K): V | undefined;

True if a key exists within this Collection, using Immutable.is to determine equality.

has(key: K): boolean;

True if a value exists within this Collection, using Immutable.is to determine equality.

includes(value: V): boolean;

In case the Collection is not empty returns the first element of the Collection. In case the Collection is empty returns the optional default value if provided, if no default value is provided returns undefined.

first<NSV>(notSetValue: NSV): V | NSV;

first(): V | undefined;

In case the Collection is not empty returns the last element of the Collection. In case the Collection is empty returns the optional default value if provided, if no default value is provided returns undefined.

last<NSV>(notSetValue: NSV): V | NSV;

last(): V | undefined;

Reading deep values

Returns the value found by following a path of keys or indices through nested Collections.

getIn(searchKeyPath: Iterable<unknown>, notSetValue?: unknown): unknown;

Plain JavaScript Object or Arrays may be nested within an Immutable.js Collection, and getIn() can access those values as well:

True if the result of following a path of keys or indices through nested Collections results in a set value.

hasIn(searchKeyPath: Iterable<unknown>): boolean;

Persistent changes

This can be very useful as a way to "chain" a normal function into a sequence of methods. RxJS calls this "let" and lodash calls it "thru".

update<R>(updater: (value: this) => R): R;

For example, to sum a Seq after mapping and filtering:

Conversion to JavaScript types

Deeply converts this Seq to equivalent native JavaScript Array or Object.

toJS(): Array<V> | { [key: string]: V };

Collection.Indexed, and Collection.Set become Array, while Collection.Keyed become Object, converting keys to Strings.

Shallowly converts this Seq to equivalent native JavaScript Array or Object.

toJSON(): Array<V> | { [key: string]: V };

Collection.Indexed, and Collection.Set become Array, while Collection.Keyed become Object, converting keys to Strings.

Shallowly converts this collection to an Array.

toArray(): Array<V>;

Collection.Indexed, and Collection.Set produce an Array of values. Collection.Keyed produce an Array of [key, value] tuples.

Shallowly converts this Collection to an Object.

toObject(): { [key: string]: V };

Converts keys to Strings.

Conversion to Collections

Converts this Collection to a Map, Throws if keys are not hashable.

toMap(): Map<K, V>;

Note: This is equivalent to Map(this.toKeyedSeq()), but provided for convenience and to allow for chained expressions.

Converts this Collection to a Map, maintaining the order of iteration.

toOrderedMap(): OrderedMap<K, V>;

Note: This is equivalent to OrderedMap(this.toKeyedSeq()), but provided for convenience and to allow for chained expressions.

Converts this Collection to a Set, discarding keys. Throws if values are not hashable.

toSet(): Set<V>;

Note: This is equivalent to Set(this), but provided to allow for chained expressions.

Converts this Collection to a Set, maintaining the order of iteration and discarding keys.

toOrderedSet(): OrderedSet<V>;

Note: This is equivalent to OrderedSet(this.valueSeq()), but provided for convenience and to allow for chained expressions.

Converts this Collection to a List, discarding keys.

toList(): List<V>;

This is similar to List(collection), but provided to allow for chained expressions. However, when called on Map or other keyed collections, collection.toList() discards the keys and creates a list of only the values, whereas List(collection) creates a list of entry tuples.

Converts this Collection to a Stack, discarding keys. Throws if values are not hashable.

toStack(): Stack<V>;

Note: This is equivalent to Stack(this), but provided to allow for chained expressions.

Conversion to Seq

Converts this Collection to a Seq of the same kind (indexed, keyed, or set).

toSeq(): Seq<K, V>;

Returns a Seq.Keyed from this Collection where indices are treated as keys.

toKeyedSeq(): Seq.Keyed<K, V>;

This is useful if you want to operate on an Collection.Indexed and preserve the [index, value] pairs.

The returned Seq will have identical iteration order as this Collection.

import { Seq } from 'immutable';

const indexedSeq = Seq(['A', 'B', 'C']);
// Seq [ "A", "B", "C" ]
indexedSeq.filter((v) => v === 'B');
// Seq [ "B" ]
const keyedSeq = indexedSeq.toKeyedSeq();
// Seq { 0: "A", 1: "B", 2: "C" }
keyedSeq.filter((v) => v === 'B');
// Seq { 1: "B" }

Returns an Seq.Indexed of the values of this Collection, discarding keys.

toIndexedSeq(): Seq.Indexed<V>;

Returns a Seq.Set of the values of this Collection, discarding keys.

toSetSeq(): Seq.Set<V>;

Iterators

An iterator of this Collection's keys.

keys(): IterableIterator<K>;

Note: this will return an ES6 iterator which does not support Immutable.js sequence algorithms. Use keySeq instead, if this is what you want.

An iterator of this Collection's values.

Note: this will return an ES6 iterator which does not support Immutable.js sequence algorithms. Use valueSeq instead, if this is what you want.

values(): IterableIterator<V>;

An iterator of this Collection's entries as [ key, value ] tuples.

Note: this will return an ES6 iterator which does not support Immutable.js sequence algorithms. Use entrySeq instead, if this is what you want.

entries(): IterableIterator<[K, V]>;

Collections (Seq)

Returns a new Seq.Indexed of the keys of this Collection, discarding values.

keySeq(): Seq.Indexed<K>;

Returns an Seq.Indexed of the values of this Collection, discarding keys.

valueSeq(): Seq.Indexed<V>;

Returns a new Seq.Indexed of [key, value] tuples.

entrySeq(): Seq.Indexed<[K, V]>;

Side effects

The sideEffect is executed for every entry in the Seq.

forEach(sideEffect: (value: V, key: K, iter: this) => unknown, context?: unknown): number;

Unlike Array#forEach, if any call of sideEffect returns false, the iteration will stop. Returns the number of entries iterated (including the last iteration which returned false).

Creating subsets

Returns a new Seq of the same type containing entries from begin up to but not including end.

If begin is negative, it is offset from the end of the Seq. If end is negative, it is also offset from the end of the Seq. If end is not provided, it will default to the size of the Seq. If the requested slice is empty, returns the same type of empty Seq.

slice(begin?: number, end?: number): this;

Returns a new Seq of the same type containing all entries except the first.

rest(): this;

Returns a new Seq of the same type containing all entries except the last.

butLast(): this;

Returns a new Seq of the same type containing all entries except the first amount.

skip(amount: number): this;

Returns a new Seq of the same type containing all entries except the last amount.

skipLast(amount: number): this;

Returns a new Seq of the same type containing entries from the first entry for which predicate returns false.

skipWhile(predicate: (value: V, key: K, iter: this) => boolean, context?: unknown): this;

Returns a new Seq of the same type containing entries from the first entry for which predicate returns true.

skipUntil(predicate: (value: V, key: K, iter: this) => boolean, context?: unknown): this;

Returns a new Seq of the same type containing the first amount entries.

take(amount: number): this;

Returns a new Seq of the same type containing the last amount entries.

takeLast(amount: number): this;

Returns a new Seq of the same type containing entries from the start until predicate returns false.

takeWhile(predicate: (value: V, key: K, iter: this) => boolean, context?: unknown): this;

Returns a new Seq of the same type containing entries from the start until predicate returns true.

takeUntil(predicate: (value: V, key: K, iter: this) => boolean, context?: unknown): this;

Combination

Flattens nested Collections.

Will deeply flatten the Collection by default, returning a Collection of the same type, but a depth can be provided in the form of a number or boolean (where true means to shallowly flatten one level). A depth of 0 (or shallow: false) will deeply flatten.

Flattens only other Collections, not Arrays or Objects.

Note: flatten(true) operates on Collection<unknown, Collection<K, V>> and returns Collection<K, V>.

flatten(depth?: number): Collection;

flatten(shallow?: boolean): Collection;

Reduces the Collection to a value by calling the reducer for every entry in the Collection and passing along the reduced value.

If initialReduction is not provided, the first item in the Collection will be used.

@see Array#reduce.

reduce<R>(reducer: (reduction, value, key, iter: this) => R): R;

Reduces the Collection in reverse (from the right side).

Note: Similar to this.reverse().reduce(), and provided for parity with Array#reduceRight.

reduceRight<R>(reducer: (reduction, value, key, iter: this) => R): R;

True if predicate returns true for all entries in the Collection.

every(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown
): boolean;

True if predicate returns true for any entry in the Collection.

some(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown
): boolean;

Joins values together as a string, inserting a separator between each. The default separator is ",".

join(separator?: string): string;

Returns true if this Collection includes no values.

For some lazy Seq, isEmpty might need to iterate to determine emptiness. At most one iteration will occur.

isEmpty(): boolean;

Returns the size of this Collection.

Regardless of if this Collection can describe its size lazily (some Seqs cannot), this method will always return the correct size. E.g. it evaluates a lazy Seq if necessary.

If predicate is provided, then this returns the count of entries in the Collection for which the predicate returns true.

count(): number;

count(predicate: (value: V, key: K, iter: this) => boolean, context?: unknown): number;

Returns a Seq.Keyed of counts, grouped by the return value of the grouper function.

countBy<G>(grouper: (value: V, key: K, iter: this) => G, context?: unknown): Map<G, number>;

Note: This is not a lazy operation.

Search for value

Returns the first value for which the predicate returns true.

find(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown,
  notSetValue?: V
): V | undefined;

Returns the last value for which the predicate returns true.

Note: predicate will be called for each entry in reverse.

findLast(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown,
  notSetValue?: V
): V | undefined;

Returns the first [key, value] entry for which the predicate returns true.

findEntry(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown,
  notSetValue?: V
): [K, V] | undefined;

Returns the last [key, value] entry for which the predicate returns true.

Note: predicate will be called for each entry in reverse.

findLastEntry(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown,
  notSetValue?: V
): [K, V] | undefined;

Returns the key for which the predicate returns true.

findKey(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown
): K | undefined;

Returns the last key for which the predicate returns true.

findLastKey(
  predicate: (value: V, key: K, iter: this) => boolean,
  context?: unknown
): K | undefined;

Note: predicate will be called for each entry in reverse.

Returns the key associated with the search value, or undefined.

keyOf(searchValue: V): K | undefined;

Returns the last key associated with the search value, or undefined.

lastKeyOf(searchValue: V): K | undefined;

Returns the maximum value in this collection. If any values are comparatively equivalent, the first one found will be returned.

max(comparator?: Comparator<V>): V | undefined;

The comparator is used in the same way as sort. If it is not provided, the default comparator is >.

When two values are considered equivalent, the first encountered will be returned. Otherwise, max will operate independent of the order of input as long as the comparator is commutative. The default comparator > is commutative only when types do not differ.

If comparator returns 0 and either value is NaN, undefined, or null, that value will be returned.

Like max, but also accepts a comparatorValueMapper which allows for comparing by more sophisticated means:

maxBy<C>(
  comparatorValueMapper: (value: V, key: K, iter: this) => C,
  comparator?: Comparator<C>
): V | undefined;

Returns the minimum value in this collection. If any values are comparatively equivalent, the first one found will be returned.

min(comparator?: Comparator<V>): V | undefined;

The comparator is used in the same way as sort. If it is not provided, the default comparator is <.

When two values are considered equivalent, the first encountered will be returned. Otherwise, min will operate independent of the order of input as long as the comparator is commutative. The default comparator < is commutative only when types do not differ.

If comparator returns 0 and either value is NaN, undefined, or null, that value will be returned.

Like min, but also accepts a comparatorValueMapper which allows for comparing by more sophisticated means:

minBy<C>(
  comparatorValueMapper: (value: V, key: K, iter: this) => C,
  comparator?: Comparator<C>
): V | undefined;

Comparison

True if iter includes every value in this Collection.

isSubset(iter: Iterable<V>): boolean;

True if this Collection includes every value in iter.

isSuperset(iter: Iterable<V>): boolean;