- I want to learn Odin because I like philosophy behind it and syntax. I also tried Zig for few weeks, but it's syntax is not compelling to me.
- I don't like cgltf implementation in vendor:cgltf and it doesn't work on *Nix based systems (for now), only on Windows.
- Odin has built-in many great native packages like core/encoding/json so why not write glTF file format package that uses Odin types and remove C hint fields from cgltf wrapper?
- Learning how to implement specification document into working code. In this case it's glTF2
- Because it's fun.
✔️ - fully implemented ➕ - partially implemented ❌ - not implemented
- Download glTF2 package from github and put in your source folder.
- Load entire glTF2 file into memory (simplest way).
- You can change floating point precision (by default it's 32bit) to 64bit, by setting flag to Odin compiler like this: -define:GLTF_DOUBLE_PRECISION=true
import "gltf2"
main :: proc() {
// This procedure sets file format from file extension [gltf/glb]
data, error := gltf2.load_from_file("file_name.[gltf/glb]")
switch err in error {
case gltf2.JSON_Error: // handle json parsing errors
case gltf2.GLTF_Error: // handle gltf2 parsing errors
}
// if there are no errors we want to free memory when we are done with processing gltf/glb file.
defer gltf2.unload(data)
// do stuff with 'data'
...
// Iterate over buffer elements using accessor:
for it := gltf2.buf_iter_make([3]f32, &data.accessors[0], data); it.idx < it.count; it.idx += 1 {
fmt.printf("Index: %v = %v\n", it.idx, gltf2.buf_iter_elem(&it))
}
}- Load parts of file into memory and parse itself. It can be handy if you can't load entire file into memory.
import "gltf2"
main :: proc () {
// Set options for gltf2 parser.
// is_glb must be set to true if file is in binary format. Most likely it will have "*.glb" suffix. By default it's gltf or JSON file format.
// delete_content set to true will delete bytes provided in procedure call. This is what 'load_from_file' does.
options := gltf2.Options{ is_glb = [true/false(default)], delete_content = [true/false(default)] }
// Load some part of file that is valid JSON object
data, error := gltf2.parse(bytes, options)
switch err in error {
case gltf2.JSON_Error: // Handle JSON parsing errors
case gltf2.GLTF_Error: // Handle GLTF2 parsing errors
}
// If there are no errors we want to free memory when we are done with processing gltf/glb file.
defer gltf2.unload(data)
}- Implement missing functionality (package is still missing some stuff).
- Remove bugs (they are for sure).
- Optimize code.