collections-d 0.4.7

Advanced containers for D

To use this package, run the following command in your project's root directory:

Manual usage
Put the following dependency into your project's dependences section:


Advanced data containers for D.

Currently available containers

Name:Based on:Accessed By:Adding:Ordered by:Theo. Insertion Complexity:Theo. Access Complexity:Foreach:Key:Array:
SortedListArray with sort per ins.Indexput onlyElement val.O(n)O(1)Fwd+BckwdNoneYes
SortedSetPrev. + check for duplicatesIndexput onlyElement val.O(n)O(1)Fwd+BckwdNoneYes
TreeMapSelf-balancing AVL TreeKeyBy keyKeyO(log n) + N (balancing)O(log n)Fwd+Bckwd,K+E,EYesNo
TreeSetSelf-balancing AVL TreeKeyBy keyKeyO(log n) + N (balancing)O(log n)Fwd+Bckwd,K+E,EYesNo
HashMapTreeMap storing hash valuesKey or HashBy keyHashcode's num. valueO(log n) + N (balancing)O(log n)Fwd,HC+E,ENot preservedNo
HashSetTreeSet storing hash valueshas only, K/HBy keyHashcode's num. valueO(log n) + N (balancing)O(log n)NoNot preservedNo
LinkedListTwo-way linked listIndexBy index, can insert at exact positionNoneO(n)O(n)Fwd+BckwdNoneNo
LinkedSetPrev. + check for duplicatesIndexput onlySimilar to FIFO, order can be changedO(n)O(n)Fwd+BckwdNoneNo
LinkedMapTwo-way linked list with keysKeyBy keyFIFOO(n)O(n)Fwd+BckwdYesNo
LinkedHashMapTwo-way linked list with hash keysKey or HashBy keyFIFOO(n)O(n)Fwd+Bckwd, Depending on key preservationPreserved optionallyNo
LinkedHashSetLinkedSet storing hash valueshas only, K/HBy keyFIFOO(n)O(n)Only with key preservationPreserved optionallyNo


An array-backed data structure that enforces the elements stored within itself to be in order. This eliminates the need for frequent reordering if a regular array was used. Has some optimizations to reduce the search and insertion times.

Has random access for indexing, elements can be only put into the list with the put() function.

alias cmp can be used both to set how the elements are compared and the way of ordering

Using as a sorted set

The template of the SortedList can disable duplicates, which will also enable certain functions, like removal by element.

TreeMap and TreeSet

Implements an AVL-tree backed associative array. Automatically rebalances itself after each insertion and removal, if imbalance is greater or equal than 2. The more elements the tree has, the better the optimization will be.

Accessing is done through keys, and accessing an n-th element needs some computation.

Tree traversal done through opApply() and opApplyReverse() functions, accessing elements in order when iterated with foreach or foreach_reverse.

alias less can be used to both set how the keys are compared and whether lesser or greater elements are on the left hand side.

Behavior is changed if E set to void, in that case keys will be the only way the TreeMap stores data, like a regular AVL-tree, or a treeset.


Implements a hashmap using TreeMap as a backend, and MurMurHash3/32 as the default hashing algorithm. Only can be used with hashing algorithms that can be compared similarly to an integer.


Implements a hashset, similarly to hashmap. Operation is done through set operators only, as it only stores the keys of the hashed elements. Only can be used with hashing algorithms that can be compared similarly to an integer.


A list with pretty low insertion cost. Ideal for applications where frequent reordering is needed. Can work as a set too by setting allowDuplicates false.


Very similar to D's own associative array.

Differentiates keys by checking whether they're equal or not. Has worse access times than TreeMap, but doesn't need optimization during insertion. Keeps the elements in insertion order, resulting in an ordered map, and behavior in case of key matches can be changed with template parameter ovrwrtBhvr.


Implements a hashmap, using LinkedMap as a backend, and MurMurHash3/128 as the default hashing algorithm. Can be used with more complex keys. The backend's ovrwrtBhvr template parameter is exposed.

By default keys are not kept, but retainKeys changes that, alongside with adding ability of storing multiple keys with matching hash codes, and foreach capability with keys.


Implements a hashset, using LinkedList in set mode as a backend, and MurMurHash3/128 as the default hashing algorithm. Can be used with more complex keys. Operation is done through set operators only, as it only stores the keys of the hashed elements.

Code example

    //I recommend using an alias to shorten the declaration of types.
    alias SortedIntList = SortedList!(int, "a < b", true);
    SortedIntList sil;
	assert(sil.arrayOf == [5], sil.toString);
	assert(sil.arrayOf == [5, 3], sil.toString);
	assert(sil.arrayOf == [7, 5, 3], sil.toString);
	assert(sil.arrayOf == [7, 5, 3, -1], sil.toString);
	assert(sil.arrayOf == [7, 5, -1], sil.toString);
	assert(sil.arrayOf == [7 ,5, 2, -1], sil.toString);
    //This will print out the numbers in order, beginning with the greatest
    foreach (int i ; sil)

Set operators

The library has some standardized set operators, repurposing preexisting operators found in D.

  • | and ~: union
  • & and *: intersection
  • - and /: complement
  • ^: difference
  • in: returns true if key found
    alias IntSet = TreeMap!(int, void, true);
    alias SortedIntSet = SortedList!(int, "a < b", false);
    IntSet a = IntSet([5, 8, 9, 1]);
    SortedIntSet b = SortedIntSet([6, 8, 9, 3]);
    //Set operators work between different types of sets.
    //The LHS type will decide what the result type will be
    IntSet intr = a & b;
    SortedIntSet diff = b ^ a;
    //RHS range can be of a non-set type, e.g. an array
    IntSet uni = a | [6, 5, 4, -1];

Known issues

  • Foreach capabilities of many containers are provided through opApply operators, which have a complicated relationship with attributes. A workaround was provided through string mixins and CTFE trickery, but it looks janky. I'll try to talk with the D community about possible solutions.
  • laszl
0.5.1 2023-Oct-18
0.5.0 2023-Apr-16
0.4.7 2022-Oct-11
0.4.6 2022-Oct-03
0.4.5 2021-Sep-21
Show all 20 versions
Download Stats:
  • 0 downloads today

  • 0 downloads this week

  • 0 downloads this month

  • 195 downloads total

Short URL: