HashOctree structure acts just like a normal octree structure but also saves a lot of memory by recognizing which sub-octrees are the same and holding them in a single sub-octree instance. Worst case scenario if every sub-octree is different it takes the same amount of memory as classic octree structure implemented in a 64-bit system.

• Installation
• Usage
• How does it work
• Release notes
• License

• git
• cmake version 3.9 or higher

1. Open bash
2. Go to directory from which you want the library to be referenced
3. Execute:

   git clone https://github.com/lpiekarski/hash-octree.git/
   cd hash-octree
   cmake .
   make hash-octree

4. Link libhash-octree.a file to your project
5. Use #include "{DIR}/hash-octree/include/hashoctree.h" in your code, where {DIR} is path to directory in which you called git clone

If you are looking for a guide on how to use this library check out the wiki pages. Most common use case is to create a HashOctree instance, insert some data into it using HashOctree::addDataShape and HashOctree::addDataPoint and then mutate contained data or test it for collision against some other objects.

HashOctree structure contains a root sub-octree reference and a dictionary consisting of (parent key -> children keys) bindings. Keys are generated by a specific hashing function satisfying the following: keys of 2 different sub-octrees which have the same children' keys and store the same data are equal.

There are currently no releases.

If you plan to use this library make sure you're familiar with its license terms.