]add CircularList
It is essentially a doubly linked list.
- Adding a new node is O(1)
- Delete an existing node is O(1)
- Ability to handle millions of nodes
To construct a circular list, you must start with at least 1 datum element.
h = circularlist(0) # CircularList.List(0)
h = circularlist([1,2]) # CircularList.List(1,2)
When inserting new data, the new node becomes the head.
h = circularlist(0) # CircularList.List(0)
insert!(h, 1) # CircularList.List(1,0)
insert!(h, 2) # CircularList.List(2,0,1)
insert!(h, 3) # CircularList.List(3,0,1,2)
When deleting the current node, the previous node becomes the head:
delete!(h) # CircularList.List(2,0,1)
You can move the head pointer in any direction:
forward!(h) # CircularList.List(0,1,2)
backward!(h) # CircularList.List(2,0,1)
shift!(h, 2, :forward) # CircularList.List(1,2,0)
shift!(h, 2, :backward) # CircularList.List(2,0,1)
Or, if you have a reference to a specific node, you can jump to that node directly and that node becomes the head!
jump!(h, node)
You can get the current head and tail node as follows:
head(h) # CircularList.Node(2)
tail(h) # CircularList.Node(1)
You can peek at the data of current, previous, or next nodes:
current(h) # 2
previous(h) # 1
next(h) # 0
Most methods returns CircularList.List so they are highly chainable:
julia> using Lazy
julia> @> h = circularlist(0) insert!(1) insert!(2) insert!(3) forward!
CircularList.List(0,1,2,3)
It is iteration friendly:
[x for x in h] # Int[2,0,1]
sum(x for x in h) # 3
Ingestion is fairly linear.
julia> @btime addmany(1000);
24.908 μs (1003 allocations: 54.97 KiB)
julia> @btime addmany(10000);
255.348 μs (10004 allocations: 547.11 KiB)
julia> @btime addmany(100000);
2.852 ms (100004 allocations: 5.34 MiB)
julia> @btime addmany(1000000);
31.802 ms (1000004 allocations: 53.41 MiB)