A rust based graph system which can handle many nodes quiet fast.
Since I am kind of a idealist I moved the whole project including the sub crates to Gitlab: https://gitlab.com/Siebencorgie/jakar-engine
The system is based on the idea of an "registry" system. You got a tree which can hold several nodes in a hierachy.
The tree struct has a BTreeMap which holds the path to a node, keyed by its name. This way, if you want to retrieve a node,
the system can go the "perfect" way to the node.
You can assign jobs to a node. A job (like "move 50 units on the x axis") is executed when the update() function is called on
this node or tree. A job get distributed to the children, which means, if you move the parent 50 units, all children will move the 50
units as well and After that execute their own jobs.
Well, I tested the system with ~50 nodes at two levels. The average getting process per node was around 500 nano seconds (on a Ryzen 1700x @ 3.4 GHz), which is okay I guess, at least for a single threaded system.
I am currently using it in a game engine. However, the base structs and traits (node and tree) can be used everywhere.
Most of the internal stuff uses generics, so it should be no problem to implement the system for use somewhere else as well.
The easiest way is to clone the repository and run in either debug or release mode:
$ git clone https://github.com/SiebenCorgie/jakar-tree.git
$ cd jakar-tree/
for release mode do:
$ cargo run --example simple --release
for debug mode do:
cargo run --example simple
There is also a speed test with a bit bigger tree and the use of tick function
cargo run --example speed