DictDotDot is just a copy of core Dict fully exposing
two internal types (so Dict(..) and NColor(..)):
type Dict k v
= RBNode_elm_builtin NColor k v (Dict k v) (Dict k v)
| RBEmpty_elm_builtin
type NColor
= Red
| BlackThe goal is to open up internals to folks who want to
- teach Dict internals
- learn Dict internals
- test/validate Dict internal (example: are the trees well-balanced?)
There's a helper function called DictInternal.toInternalRepresentation
that produces data like the following:
[{ color = "B", k = 1, path = "l", v = "one" },
{ color = "B", k = 2, path = "", v = "two" },
{ color = "R", k = 3, path = "rl", v = "three" },
{ color = "B", k = 4, path = "r", v = "four" }]That format's pretty easy to manipulate for many use
cases, and it is a complete representation of the tree,
but if that doesn't work for you, just read the
source for DictInternal.
You can also read/make/run Main.elm to see this in action for
a tiny dict.
To play with DictDotDot in apps with minimal changes, you can
do import DictDotDot as Dict.