immutable

package module
v0.0.0-...-de859e0 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Nov 16, 2022 License: MIT Imports: 3 Imported by: 0

README

Immutable release test coverage license

This repository contains generic immutable collection types for Go. It includes List, Map, and SortedMap implementations. Immutable collections can provide efficient, lock free sharing of data by requiring that edits to the collections return new collections.

The collection types in this library are meant to mimic Go built-in collections such asslice and map. The primary usage difference between Go collections and immutable collections is that immutable collections always return a new collection on mutation so you will need to save the new reference.

Immutable collections are not for every situation, however, as they can incur additional CPU and memory overhead. Please evaluate the cost/benefit for your particular project.

Special thanks to the Immutable.js team as the List & Map implementations are loose ports from that project.

List

The List type represents a sorted, indexed collection of values and operates similarly to a Go slice. It supports efficient append, prepend, update, and slice operations.

Adding list elements

Elements can be added to the end of the list with the Append() method or added to the beginning of the list with the Prepend() method. Unlike Go slices, prepending is as efficient as appending.

// Create a list with 3 elements.
l := immutable.NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Prepend("baz")

fmt.Println(l.Len())  // 3
fmt.Println(l.Get(0)) // "baz"
fmt.Println(l.Get(1)) // "foo"
fmt.Println(l.Get(2)) // "bar"

Note that each change to the list results in a new list being created. These lists are all snapshots at that point in time and cannot be changed so they are safe to share between multiple goroutines.

Updating list elements

You can also overwrite existing elements by using the Set() method. In the following example, we'll update the third element in our list and return the new list to a new variable. You can see that our old l variable retains a snapshot of the original value.

l := immutable.NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
newList := l.Set(2, "baz")

fmt.Println(l.Get(1))       // "bar"
fmt.Println(newList.Get(1)) // "baz"
Deriving sublists

You can create a sublist by using the Slice() method. This method works with the same rules as subslicing a Go slice:

l = l.Slice(0, 2)

fmt.Println(l.Len())  // 2
fmt.Println(l.Get(0)) // "baz"
fmt.Println(l.Get(1)) // "foo"

Please note that since List follows the same rules as slices, it will panic if you try to Get(), Set(), or Slice() with indexes that are outside of the range of the List.

Iterating lists

Iterators provide a clean, simple way to iterate over the elements of the list in order. This is more efficient than simply calling Get() for each index.

Below is an example of iterating over all elements of our list from above:

itr := l.Iterator()
for !itr.Done() {
	index, value, _ := itr.Next()
	fmt.Printf("Index %d equals %v\n", index, value)
}

// Index 0 equals baz
// Index 1 equals foo

By default iterators start from index zero, however, the Seek() method can be used to jump to a given index.

Efficiently building lists

If you are building large lists, it is significantly more efficient to use the ListBuilder. It uses nearly the same API as List except that it updates a list in-place until you are ready to use it. This can improve bulk list building by 10x or more.

b := immutable.NewListBuilder[string]()
b.Append("foo")
b.Append("bar")
b.Set(2, "baz")

l := b.List()
fmt.Println(l.Get(0)) // "foo"
fmt.Println(l.Get(1)) // "baz"

Builders are invalid after the call to List().

Map

The Map represents an associative array that maps unique keys to values. It is implemented to act similarly to the built-in Go map type. It is implemented as a Hash-Array Mapped Trie.

Maps require a Hasher to hash keys and check for equality. There are built-in hasher implementations for most primitive types such as int, uint, and string keys. You may pass in a nil hasher to NewMap() if you are using one of these key types.

Currently []byte keys are not supported since Go doesn't included them in constraints.Ordered unline strings. For now Ordered constraint it used even for Map where key order is unimportant but this could be relaxed in the future.

Setting map key/value pairs

You can add a key/value pair to the map by using the Set() method. It will add the key if it does not exist or it will overwrite the value for the key if it does exist.

Values may be fetched for a key using the Get() method. This method returns the value as well as a flag indicating if the key existed. The flag is useful to check if a nil value was set for a key versus a key did not exist.

m := immutable.NewMap[string,int](nil)
m = m.Set("jane", 100)
m = m.Set("susy", 200)
m = m.Set("jane", 300) // overwrite

fmt.Println(m.Len())   // 2

v, ok := m.Get("jane")
fmt.Println(v, ok)     // 300 true

v, ok = m.Get("susy")
fmt.Println(v, ok)     // 200, true

v, ok = m.Get("john")
fmt.Println(v, ok)     // nil, false
Removing map keys

Keys may be removed from the map by using the Delete() method. If the key does not exist then the original map is returned instead of a new one.

m := immutable.NewMap[string,int](nil)
m = m.Set("jane", 100)
m = m.Delete("jane")

fmt.Println(m.Len())   // 0

v, ok := m.Get("jane")
fmt.Println(v, ok)     // nil false
Iterating maps

Maps are unsorted, however, iterators can be used to loop over all key/value pairs in the collection. Unlike Go maps, iterators are deterministic when iterating over key/value pairs.

m := immutable.NewMap[string,int](nil)
m = m.Set("jane", 100)
m = m.Set("susy", 200)

itr := m.Iterator()
for !itr.Done() {
	k, v := itr.Next()
	fmt.Println(k, v)
}

// susy 200
// jane 100

Note that you should not rely on two maps with the same key/value pairs to iterate in the same order. Ordering can be insertion order dependent when two keys generate the same hash.

Efficiently building maps

If you are executing multiple mutations on a map, it can be much more efficient to use the MapBuilder. It uses nearly the same API as Map except that it updates a map in-place until you are ready to use it.

b := immutable.NewMapBuilder[string,int](nil)
b.Set("foo", 100)
b.Set("bar", 200)
b.Set("foo", 300)

m := b.Map()
fmt.Println(m.Get("foo")) // "300"
fmt.Println(m.Get("bar")) // "200"

Builders are invalid after the call to Map().

Implementing a custom Hasher

If you need to use a key type besides int, uint, or string then you'll need to create a custom Hasher implementation and pass it to NewMap() on creation.

Hashers are fairly simple. They only need to generate hashes for a given key and check equality given two keys.

type Hasher[K constraints.Ordered] interface {
	Hash(key K) uint32
	Equal(a, b K) bool
}

Please see the internal intHasher, uintHasher, stringHasher, and byteSliceHasher for examples.

Sorted Map

The SortedMap represents an associative array that maps unique keys to values. Unlike the Map, however, keys can be iterated over in-order. It is implemented as a B+tree.

Sorted maps require a Comparer to sort keys and check for equality. There are built-in comparer implementations for int, uint, and string keys. You may pass a nil comparer to NewSortedMap() if you are using one of these key types.

Currently []byte keys are not supported since Go doesn't included them in constraints.Ordered unline strings.

The API is identical to the Map implementation. The sorted map also has a companion SortedMapBuilder for more efficiently building maps.

Implementing a custom Comparer

If you need to use a key type besides int, uint, or string then you'll need to create a custom Comparer implementation and pass it to NewSortedMap() on creation.

Comparers on have one method—Compare(). It works the same as the strings.Compare() function. It returns -1 if a is less than b, returns 1 if a is greater than b, and returns 0 if a is equal to b.

type Comparer[K constraints.Ordered] interface {
	Compare(a, b K) int
}

Please see the internal intComparer, uintComparer, and stringComparer for examples.

Contributing

The goal of immutable is to provide stable, reasonably performant, immutable collections library for Go that has a simple, idiomatic API. As such, additional features and minor performance improvements will generally not be accepted. If you have a suggestion for a clearer API or substantial performance improvement, please open an issue first to discuss. All pull requests without a related issue will be closed immediately.

Please submit issues relating to bugs & documentation improvements.

Documentation

Overview

Package immutable provides immutable collection types.

Introduction

Immutable collections provide an efficient, safe way to share collections of data while minimizing locks. The collections in this package provide List, Map, and SortedMap implementations. These act similarly to slices and maps, respectively, except that altering a collection returns a new copy of the collection with that change.

Because collections are unable to change, they are safe for multiple goroutines to read from at the same time without a mutex. However, these types of collections come with increased CPU & memory usage as compared with Go's built-in collection types so please evaluate for your specific use.

Collection Types

The List type provides an API similar to Go slices. They allow appending, prepending, and updating of elements. Elements can also be fetched by index or iterated over using a ListIterator.

The Map & SortedMap types provide an API similar to Go maps. They allow values to be assigned to unique keys and allow for the deletion of keys. Values can be fetched by key and key/value pairs can be iterated over using the appropriate iterator type. Both map types provide the same API. The SortedMap, however, provides iteration over sorted keys while the Map provides iteration over unsorted keys. Maps improved performance and memory usage as compared to SortedMaps.

Hashing and Sorting

Map types require the use of a Hasher implementation to calculate hashes for their keys and check for key equality. SortedMaps require the use of a Comparer implementation to sort keys in the map.

These collection types automatically provide built-in hasher and comparers for int, string, and byte slice keys. If you are using one of these key types then simply pass a nil into the constructor. Otherwise you will need to implement a custom Hasher or Comparer type. Please see the provided implementations for reference.

Index

Examples

Constants

This section is empty.

Variables

This section is empty.

Functions

This section is empty.

Types

type Hasher 

type Hasher[K comparable] interface {
	// Computes a hash for key.
	Hash(key K) uint32

	// Returns true if a and b are equal.
	Equal(a, b K) bool
}

Hasher hashes keys and checks them for equality.

func NewHasher 

func NewHasher[K comparable](key K) Hasher[K]

NewHasher returns the built-in hasher for a given key type.

type List 

type List[T comparable] struct {
	// contains filtered or unexported fields
}

List is a dense, ordered, indexed collections. They are analogous to slices in Go. They can be updated by appending to the end of the list, prepending values to the beginning of the list, or updating existing indexes in the list.

func NewList 

func NewList[T comparable]() *List[T]

NewList returns a new empty instance of List.

func (*List[T]) Append 

func (l *List[T]) Append(value T) *List[T]

Append returns a new list with value added to the end of the list.

Example 
l := NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Append("baz")

fmt.Println(l.Get(0))
fmt.Println(l.Get(1))
fmt.Println(l.Get(2))

Output:

foo
bar
baz

func (*List[T]) Get 

func (l *List[T]) Get(index int) T

Get returns the value at the given index. Similar to slices, this method will panic if index is below zero or is greater than or equal to the list size.

func (*List[T]) Iterator 

func (l *List[T]) Iterator() *ListIterator[T]

Iterator returns a new iterator for this list positioned at the first index.

Example 
l := NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Append("baz")

itr := l.Iterator()
for !itr.Done() {
	i, v := itr.Next()
	fmt.Println(i, v)
}

Output:

0 foo
1 bar
2 baz
Example (Reverse) 
l := NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Append("baz")

itr := l.Iterator()
itr.Last()
for !itr.Done() {
	i, v := itr.Prev()
	fmt.Println(i, v)
}

Output:

2 baz
1 bar
0 foo

func (*List[T]) Len 

func (l *List[T]) Len() int

Len returns the number of elements in the list.

func (*List[T]) Prepend 

func (l *List[T]) Prepend(value T) *List[T]

Prepend returns a new list with value added to the beginning of the list.

Example 
l := NewList[string]()
l = l.Prepend("foo")
l = l.Prepend("bar")
l = l.Prepend("baz")

fmt.Println(l.Get(0))
fmt.Println(l.Get(1))
fmt.Println(l.Get(2))

Output:

baz
bar
foo

func (*List[T]) Set 

func (l *List[T]) Set(index int, value T) *List[T]

Set returns a new list with value set at index. Similar to slices, this method will panic if index is below zero or if the index is greater than or equal to the list size.

Example 
l := NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Set(1, "baz")

fmt.Println(l.Get(0))
fmt.Println(l.Get(1))

Output:

foo
baz

func (*List[T]) Slice 

func (l *List[T]) Slice(start, end int) *List[T]

Slice returns a new list of elements between start index and end index. Similar to slices, this method will panic if start or end are below zero or greater than the list size. A panic will also occur if start is greater than end.

Unlike Go slices, references to inaccessible elements will be automatically removed so they can be garbage collected.

Example 
l := NewList[string]()
l = l.Append("foo")
l = l.Append("bar")
l = l.Append("baz")
l = l.Slice(1, 3)

fmt.Println(l.Get(0))
fmt.Println(l.Get(1))

Output:

bar
baz

type ListBuilder 

type ListBuilder[T comparable] struct {
	// contains filtered or unexported fields
}

ListBuilder represents an efficient builder for creating new Lists.

func NewListBuilder 

func NewListBuilder[T comparable]() *ListBuilder[T]

NewListBuilder returns a new instance of ListBuilder.

func (*ListBuilder[T]) Append 

func (b *ListBuilder[T]) Append(value T)

Append adds value to the end of the list.

Example 
b := NewListBuilder[string]()
b.Append("foo")
b.Append("bar")
b.Append("baz")

l := b.List()
fmt.Println(l.Get(0))
fmt.Println(l.Get(1))
fmt.Println(l.Get(2))

Output:

foo
bar
baz

func (*ListBuilder[T]) Get 

func (b *ListBuilder[T]) Get(index int) T

Get returns the value at the given index. Similar to slices, this method will panic if index is below zero or is greater than or equal to the list size.

func (*ListBuilder[T]) Iterator 

func (b *ListBuilder[T]) Iterator() *ListIterator[T]

Iterator returns a new iterator for the underlying list.

func (*ListBuilder[T]) Len 

func (b *ListBuilder[T]) Len() int

Len returns the number of elements in the underlying list.

func (*ListBuilder[T]) List 

func (b *ListBuilder[T]) List() *List[T]

List returns the current copy of the list. The builder should not be used again after the list after this call.

func (*ListBuilder[T]) Prepend 

func (b *ListBuilder[T]) Prepend(value T)

Prepend adds value to the beginning of the list.

Example 
b := NewListBuilder[string]()
b.Prepend("foo")
b.Prepend("bar")
b.Prepend("baz")

l := b.List()
fmt.Println(l.Get(0))
fmt.Println(l.Get(1))
fmt.Println(l.Get(2))

Output:

baz
bar
foo

func (*ListBuilder[T]) Set 

func (b *ListBuilder[T]) Set(index int, value T)

Set updates the value at the given index. Similar to slices, this method will panic if index is below zero or if the index is greater than or equal to the list size.

Example 
b := NewListBuilder[string]()
b.Append("foo")
b.Append("bar")
b.Set(1, "baz")

l := b.List()
fmt.Println(l.Get(0))
fmt.Println(l.Get(1))

Output:

foo
baz

func (*ListBuilder[T]) Slice 

func (b *ListBuilder[T]) Slice(start, end int)

Slice updates the list with a sublist of elements between start and end index. See List.Slice() for more details.

Example 
b := NewListBuilder[string]()
b.Append("foo")
b.Append("bar")
b.Append("baz")
b.Slice(1, 3)

l := b.List()
fmt.Println(l.Get(0))
fmt.Println(l.Get(1))

Output:

bar
baz

type ListIterator 

type ListIterator[T comparable] struct {
	// contains filtered or unexported fields
}

ListIterator represents an ordered iterator over a list.

func (*ListIterator[T]) Done 

func (itr *ListIterator[T]) Done() bool

Done returns true if no more elements remain in the iterator.

func (*ListIterator[T]) First 

func (itr *ListIterator[T]) First()

First positions the iterator on the first index. If source list is empty then no change is made.

func (*ListIterator[T]) Last 

func (itr *ListIterator[T]) Last()

Last positions the iterator on the last index. If source list is empty then no change is made.

func (*ListIterator[T]) Next 

func (itr *ListIterator[T]) Next() (index int, value T)

Next returns the current index and its value & moves the iterator forward. Returns an index of -1 if the there are no more elements to return.

func (*ListIterator[T]) Prev 

func (itr *ListIterator[T]) Prev() (index int, value T)

Prev returns the current index and value and moves the iterator backward. Returns an index of -1 if the there are no more elements to return.

func (*ListIterator[T]) Seek 

func (itr *ListIterator[T]) Seek(index int)

Seek moves the iterator position to the given index in the list. Similar to Go slices, this method will panic if index is below zero or if the index is greater than or equal to the list size.

type Map 

type Map[K comparable, V any] struct {
	// contains filtered or unexported fields
}

Map represents an immutable hash map implementation. The map uses a Hasher to generate hashes and check for equality of key values.

It is implemented as an Hash Array Mapped Trie.

func NewMap 

func NewMap[K comparable, V any](hasher Hasher[K]) *Map[K, V]

NewMap returns a new instance of Map. If hasher is nil, a default hasher implementation will automatically be chosen based on the first key added. Default hasher implementations only exist for int, string, and byte slice types.

func (*Map[K, V]) Delete 

func (m *Map[K, V]) Delete(key K) *Map[K, V]

Delete returns a map with the given key removed. Removing a non-existent key will cause this method to return the same map.

Example 
m := NewMap[string, any](nil)
m = m.Set("foo", "bar")
m = m.Set("baz", 100)
m = m.Delete("baz")

v, ok := m.Get("foo")
fmt.Println("foo", v, ok)

v, ok = m.Get("baz")
fmt.Println("baz", v, ok)

Output:

foo bar true
baz <nil> false

func (*Map[K, V]) Get 

func (m *Map[K, V]) Get(key K) (value V, ok bool)

Get returns the value for a given key and a flag indicating whether the key exists. This flag distinguishes a nil value set on a key versus a non-existent key in the map.

func (*Map[K, V]) Iterator 

func (m *Map[K, V]) Iterator() *MapIterator[K, V]

Iterator returns a new iterator for the map.

Example 
m := NewMap[string, int](nil)
m = m.Set("apple", 100)
m = m.Set("grape", 200)
m = m.Set("kiwi", 300)
m = m.Set("mango", 400)
m = m.Set("orange", 500)
m = m.Set("peach", 600)
m = m.Set("pear", 700)
m = m.Set("pineapple", 800)
m = m.Set("strawberry", 900)

itr := m.Iterator()
for !itr.Done() {
	k, v, _ := itr.Next()
	fmt.Println(k, v)
}

Output:

mango 400
pear 700
pineapple 800
grape 200
orange 500
strawberry 900
kiwi 300
peach 600
apple 100

func (*Map[K, V]) Len 

func (m *Map[K, V]) Len() int

Len returns the number of elements in the map.

func (*Map[K, V]) Set 

func (m *Map[K, V]) Set(key K, value V) *Map[K, V]

Set returns a map with the key set to the new value. A nil value is allowed.

This function will return a new map even if the updated value is the same as the existing value because Map does not track value equality.

Example 
m := NewMap[string, any](nil)
m = m.Set("foo", "bar")
m = m.Set("baz", 100)

v, ok := m.Get("foo")
fmt.Println("foo", v, ok)

v, ok = m.Get("baz")
fmt.Println("baz", v, ok)

v, ok = m.Get("bat") // does not exist
fmt.Println("bat", v, ok)

Output:

foo bar true
baz 100 true
bat <nil> false

type MapBuilder 

type MapBuilder[K comparable, V any] struct {
	// contains filtered or unexported fields
}

MapBuilder represents an efficient builder for creating Maps.

func NewMapBuilder 

func NewMapBuilder[K comparable, V any](hasher Hasher[K]) *MapBuilder[K, V]

NewMapBuilder returns a new instance of MapBuilder.

func (*MapBuilder[K, V]) Delete 

func (b *MapBuilder[K, V]) Delete(key K)

Delete removes the given key. See Map.Delete() for additional details.

Example 
b := NewMapBuilder[string, any](nil)
b.Set("foo", "bar")
b.Set("baz", 100)
b.Delete("baz")

m := b.Map()
v, ok := m.Get("foo")
fmt.Println("foo", v, ok)

v, ok = m.Get("baz")
fmt.Println("baz", v, ok)

Output:

foo bar true
baz <nil> false

func (*MapBuilder[K, V]) Get 

func (b *MapBuilder[K, V]) Get(key K) (value V, ok bool)

Get returns the value for the given key.

func (*MapBuilder[K, V]) Iterator 

func (b *MapBuilder[K, V]) Iterator() *MapIterator[K, V]

Iterator returns a new iterator for the underlying map.

func (*MapBuilder[K, V]) Len 

func (b *MapBuilder[K, V]) Len() int

Len returns the number of elements in the underlying map.

func (*MapBuilder[K, V]) Map 

func (b *MapBuilder[K, V]) Map() *Map[K, V]

Map returns the underlying map. Only call once. Builder is invalid after call. Will panic on second invocation.

func (*MapBuilder[K, V]) Set 

func (b *MapBuilder[K, V]) Set(key K, value V)

Set sets the value of the given key. See Map.Set() for additional details.

Example 
b := NewMapBuilder[string, any](nil)
b.Set("foo", "bar")
b.Set("baz", 100)

m := b.Map()
v, ok := m.Get("foo")
fmt.Println("foo", v, ok)

v, ok = m.Get("baz")
fmt.Println("baz", v, ok)

v, ok = m.Get("bat") // does not exist
fmt.Println("bat", v, ok)

Output:

foo bar true
baz 100 true
bat <nil> false

type MapIterator 

type MapIterator[K comparable, V any] struct {
	// contains filtered or unexported fields
}

MapIterator represents an iterator over a map's key/value pairs. Although map keys are not sorted, the iterator's order is deterministic.

func (*MapIterator[K, V]) Done 

func (itr *MapIterator[K, V]) Done() bool

Done returns true if no more elements remain in the iterator.

func (*MapIterator[K, V]) First 

func (itr *MapIterator[K, V]) First()

First resets the iterator to the first key/value pair.

func (*MapIterator[K, V]) Next 

func (itr *MapIterator[K, V]) Next() (key K, value V, ok bool)

Next returns the next key/value pair. Returns a nil key when no elements remain.

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.