vulkan-object 1.4.329

An easy-to-use Python wrapper for the Vulkan API registry.

Vulkan Object Python Package

Parsing the vk.xml in Vulkan is easy, processing it is hard!

It is very easy for people to mess up trying to process the vk.xml file, so we created VulkanObject

VulkanObject is just a python dataclass that is defined in Vulkan-Headers/registry/vulkan_object.py. It uses the reg.py framework that the Vulkan Spec is generated with in order to populate the VulkanObject data structure.

This python package makes it super easy to get going.

pip install vulkan-object

and then it is as simple as

#!/usr/bin/env python3
from vulkan_object import get_vulkan_object

# This builds the VulkanObject that is populated and ready to be used
vk = get_vulkan_object()

print(f'There are now {len([x for x in vk.extensions.values()])} extensions in Vulkan')

print(f'Built with the {vk.headerVersion} headers')

longest_count = 0
for struct in vk.structs.values():
    if len(struct.name) > longest_count:
        longest_count = len(struct.name)
        longest_struct = struct
print(f'Longest Struct name is {longest_struct.name} at {longest_count} characters')

Quick query from the terminal

Just run the following and you can quickly use vk in your terminal for some quick query

python -i -c "from vulkan_object import get_vulkan_object;vk = get_vulkan_object()"

More example

example.py has more in depth ways to use this

What is this package/repo actually?

The Vulkan-Headers contain a bunch of scripts in the Vulkan-Headers/registry/ directory that repos can use to help generate code.

The main issue is the delivery mechanism for python projects. This package/repo just grabs the Vulkan-Headers scripts (not the C header files themselves!) and make it easier to integrate with projects.

Updating this repo

run python update.py /path/to/Vulkan-Headers and it will do everything to sync this package up with the new headers