Skip to content

Instantly share code, notes, and snippets.

@gingerBill
Created August 26, 2016 11:11
Show Gist options
  • Save gingerBill/d1f99bab3d0b8bde3481ac69434ae3db to your computer and use it in GitHub Desktop.
Save gingerBill/d1f99bab3d0b8bde3481ac69434ae3db to your computer and use it in GitHub Desktop.
First Odin Demo
// Demo 001
#load "basic.odin"
#load "game.odin"
main :: proc() {
_ = hellope();
procedures();
variables();
constants();
types();
data_control();
using_fields();
// run_game();
}
hellope :: proc() -> int {
print_string("Hellope, 世界\n");
return 1;
}
// Line comment
/*
Block Comment
*/
/*
Nested /*
Block /*
Comment
*/
*/
*/
apple, banana, carrot: bool;
box, carboard: bool = true, false;
// hellope_value: int = hellope(); // The procedure is ran just before `main`
variables :: proc() {
i: int; // initialized with zero value
j: int = 1;
x, y: int = 1, 2;
// Type inference
apple, banana, 世界 := true, 123, "world";
// Basic Types of the Language
//
// bool
//
// i8 i16 i32 i64 i128
// u8 u16 u32 u64 u128
//
// f32 f64
//
// int uint (size_of(int) == size_of(uint) == size_of(rawptr))
//
// rawptr (equivalent to void * in C/C++)
//
// string
//
// byte - alias for u8
// rune - alias for i32 // Unicode Codepoint
//
// "untyped" types can implicitly convert to any of the "typed" types
// Default Type
// untyped bool - bool
// untyped integer - int
// untyped float - f64
// untyped pointer - rawptr
// untyped string - string
// untyped rune - rune/i32
// Zero values
zero_numeric := 0;
zero_boolean := false;
zero_pointer := null;
zero_string1 := ""; // Escaped string
zero_string2 := ``; // Raw string
// Compound types have a different kind of zero value
// Unary operators
// +a
// -a
// ~a
// !a
// Binary operators
// a + b add
// a - b sub
// a ~ b xor
// a | b or
// a * b mul
// a / b quo
// a % b mod
// a & b and
// a &~ b bitclear == a & (~b)
// a << b shl
// a >> b shr
// a as Type // Type cast
// a transmute Type // Bit cast
// a == b eq
// a != b ne
// a < b lt
// a > b gt
// a <= b le
// a >= b ge
}
procedures :: proc() {
add :: proc(x: int, y: int) -> int {
return x + y;
}
print_int(add(3, 4)); // 7
print_nl();
add_v2 :: proc(x, y: int) -> int {
return x + y;
}
fibonacci :: proc(n: int) -> int {
if n < 2 {
return n;
}
return fibonacci(n-1) + fibonacci(n-2);
}
print_int(fibonacci(12)); nl();
swap_strings :: proc(x, y: string) -> (string, string) {
return y, x;
}
a, b := swap_strings("Hellope\n", "World\n");
print_string(a);
print_string(b);
a, b = b, a; // Quirk of grammar the of multiple assignments
// Swap variables
print_string(a);
print_string(b);
// Not a hint like C/C++, it's mandatory (unless it cannot do it but it will warn)
proc1 :: proc(a, b: int) #inline {
print_int(a + b);
}
proc2 :: proc(a, b: int) #no_inline {
print_int(a + b);
}
print_int(3 ''add 4); // Infix style
print_nl();
print_int(12 'fibonacci); // Postfix style
print_nl();
}
TAU :: 6.28318530718;
constants :: proc() {
TAU :: 6.28318530718; // untyped float
WORLD_JAPANESE :: "世界"; // untyped string
TAU_32 : f32 : 6.28318530718;
TAU_AS_32 :: 6.28318530718 as f32;
PI :: TAU / 2;
CLOSE_TO_PI :: 3;
DIFF :: (PI - CLOSE_TO_PI) / PI; // Evaluated at compile time
a := TAU; // the constant's value becomes typed as f32
b := CLOSE_TO_PI; // the constant's value becomes typed as int
c := DIFF;
}
nl :: proc() { print_nl(); }
types :: proc() {
x: int = 123;
y := x; // y: int = x;
// z: f32 = x; // invalid
z: f32 = x as f32;
ptr_z := ^z; // Pascal notation
ptr_z^ = 123; // Derefence Notation
w: f32 = ptr_z^; // 123
print_f32(z); nl();
// ^z - pointer to z
// z^ - z from pointer
// Implicit conversion to and from rawptr
r_ptr: rawptr = ptr_z;
ptr_z = r_ptr;
f32_array: [12]f32; // Array of 12 f32
f32_array[0] = 2;
f32_array[1] = 3;
// f32_array[-1] = 2; // Error - compile time check
// f32_array[13] = 2; // Error - compile time check
f32_array_len := len(f32_array); // builtin procedure
f32_array_cap := cap(f32_array); // == len(f32_array)
mda: [2][3][4]int; // Column-major
// mda[x][y][z]
api: [2]^f32;
papi: ^[2]^f32;
f32_slice: []f32; // Slice / Array reference
f32_slice = f32_array[0:5];
f32_slice = f32_array[:5];
f32_slice = f32_array[:]; // f32_array[0:len(f32_array)-1];
f32_slice = f32_array[1:5:7]; // low:1, high:5, max:7
// len: 5-1 == 4
// cap: 7-1 == 6
append_success := append(^f32_slice, 1);
_ = append(^f32_slice, 2);
_ = copy(f32_array[0:2], f32_array[2:4]); // You can use memcpy/memmove if you want
s := "Hellope World";
sub_string: string = s[5:10];
v0: {4}f32; // Vector of 4 f32
v0[0] = 1;
v0[1] = 3;
v0[2] = 6;
v0[3] = 10;
v1 := v0 + v0; // Simd Arithmetic
v1 = v1 - v0;
v1 *= v0; // i.e. hadamard product
v1 /= v0;
// builtin procedure
v2 := swizzle(v0, 3, 2, 1, 0); // {10, 6, 3, 1}
v3: {4}bool = v0 == v2;
// LLVM rant?
Vec4 :: type {4}f32;
Array3Int :: type [3]int;
Vec3 :: type struct {
x, y, z: f32
}
BinaryNode :: type struct {
left, right: ^BinaryNode; // same format as procedure argument
data: rawptr;
}
AddProc :: type proc(a, b: int) -> int
Packed :: type struct #packed {
a: u8;
b: u16;
c: u32;
}
static_assert(size_of(Packed) == 7); // builtin procedure
{
a, b: ^BinaryNode;
a = alloc(size_of(BinaryNode)) as ^BinaryNode;
b = alloc(size_of(BinaryNode)) as ^BinaryNode;
c := BinaryNode{a, b, null};
c.left^.data = null;
c.left.data = null; // No need to deference
dealloc(a);
dealloc(b);
}
{
MyInt :: type int;
x: int = 1;
y: MyInt = 2;
// z := x + y; // Failure - types cannot implicit convert*
z := x as MyInt + y; // Type cast using `as`
}
{
// From: Quake III Arena
Q_rsqrt :: proc(number: f32) -> f32 {
i: i32;
x2, y: f32;
THREE_HALFS :: 1.5;
x2 = number * 0.5;
y = number;
i = (^y as ^i32)^; // evil floating point bit level hacking
i = 0x5f3759df - i>>1; // what the fuck?
y = (^i as ^f32)^;
y = y * (THREE_HALFS - (x2 * y *y)); // 1st iteration
// y = y * (THREE_HALFS - (x2 * y *y)); // 2nd iteration, this can be removed
return y;
}
Q_rsqrt_v2 :: proc(number: f32) -> f32 {
THREE_HALFS :: 1.5;
x2 := number * 0.5;
y := number;
i := y transmute i32; // evil floating point bit level hacking
i = 0x5f3759df - i>>1; // what the fuck?
y = i transmute f32;
y = y * (THREE_HALFS - (x2 * y *y)); // 1st iteration
// y = y * (THREE_HALFS - (x2 * y *y)); // 2nd iteration, this can be removed
return y;
}
// NOTE(bill): transmute only works if the size of the types are equal
/*
// in C
union {
i32 i;
f32 y;
};
*/
}
{ // Enumeration
Thing :: type enum {
APPLE,
FROG,
TREE,
TOMB,
}
a := Thing.APPLE;
Sized :: type enum u64 {
APPLE,
FROG,
TREE,
TOMB,
}
static_assert(size_of(Sized) == size_of(u64));
Certain :: type enum {
APPLE = 3,
FROG,
TREE = 7,
TOMB,
}
static_assert(Certain.TOMB == 8);
}
{ // Untagged union
BitHack :: type union {
i: i32;
f: f32;
}
b: BitHack;
b.f = 123;
print_int(b.i as int); print_nl();
// Manually tagged union
EntityKind :: type enum {
Invalid,
Constant,
Variable,
TypeName,
Procedure,
Builtin,
Count,
}
Entity :: type struct {
kind: EntityKind;
guid: u64;
// Other data
/*using*/
data: union {
constant: struct{};
variable: struct{
visited, is_field, used, anonymous: bool;
};
procedure: struct { used: bool };
buitlin: struct { id: i32 };
};
}
// NOTE(bill): Tagged unions are not added yet but are planned
}
{ // Compound Literals
a := [3]int{1, 2, 3};
b := [3]int{};
c := [..]int{1, 2, 3};
d := []int{1, 2, 3}; // slice
e := {4}f32{1, 2, 3, 4};
f := {4}f32{1}; // broadcasts to all
// g := {4}f32{1, 2}; // require either 1 or 4 elements
Vec2 :: type {2}f32;
h := Vec2{1, 2};
i := Vec2{5} * h; // For strong type safety
// FORENOTE: 5 * h was originally allowed but it was an edge case in the
// compiler I didn't think it was enough to justify have it it.
print_f32(i[0]); print_rune(#rune ",");
print_f32(i[1]); print_nl();
}
{ // First class procedures
do_thing :: proc(p: proc(a, b: int) -> int) {
print_int(p(3, 4)); nl();
}
add :: proc(a, b: int) -> int {
return a + b;
}
add_lambda := proc(a, b: int) -> int {
return a - b;
}; // note semicolon
do_thing(add);
do_thing(add_lambda);
do_thing(proc(a, b: int) -> int { // Anonymous
return a * b;
});
}
{ // strings and runes
escaped := "Hellope World\n";
raw := `Hellope World\n`;
print_string(escaped);
print_string(raw); nl();
// Crap shader example
shader_string :=
`#version 410
layout (location = 0) in vec3 a_position;
layout (location = 1) in vec3 a_normal;
layout (location = 2) in vec2 a_tex_coord;
out vec3 v_position;
out vec3 v_normal;
out vec2 v_tex_coord;
uniform mat4 u_model_view;
uniform mat3 u_normal;
uniform mat4 u_proj;
uniform mat4 u_mvp;
void main() {
v_tex_coord = a_tex_coord;
v_normal = normalize(u_normal * a_normal);
v_position = vec3(u_model_view * vec4(a_position, 1.0));
gl_Position = u_mvp * vec4(a_position, 1.0);
}`;
hearts1 := #rune "💕";
hearts2 := #rune "\U0001f495"; // 32 bit
hearts3 := #rune "\xf0\x9f\x92\x95";
㐒 := #rune "㐒";
㐒16 := #rune "\u4db5"; // 16 bit but will be `rune`
// String ideas "nicked" from Go, so far. I think I might change how some of it works later.
}
{ // size, align, offset
Thing :: type struct {
a: u8;
b: u16;
c, d, e: u32;
}
s := size_of(Thing);
a := align_of(Thing);
o := offset_of(Thing, b);
t: Thing;
sv := size_of_val(t);
av := align_of_val(t);
ov := offset_of_val(t.b);
}
}
data_control :: proc() {
sum := 0;
for i := 0; i < 12; i++ {
sum += 1;
}
print_string("sum = "); print_int(sum); nl();
sum = 1;
for ; sum < 1000000; {
sum += sum;
}
print_string("sum = "); print_int(sum); nl();
sum = 1;
for sum < 1000000 {
sum += sum;
}
print_string("sum = "); print_int(sum); nl();
// loop
// for { } == for true {}
// Question: Should I separate all these concepts and rename it?
//
// range - iterable
// for - c style
// while
// loop - while true
// Notes:
// conditions _must_ a boolean expression
// i++ and i-- are statements, not expressions
x := 2;
if x < 3 {
print_string("x < 2\n");
}
// Unified initializer syntax - same as for statements
if x := 2; x < 3 {
print_string("x < 2\n");
}
if x := 4; x < 3 {
print_string("Never called\n");
} else {
print_string("This is called\n");
}
{ // String comparison
a := "Hellope";
b := "World";
if a < b {
print_string("a < b\n");
}
if a != b {
print_string("a != b\n");
}
}
{ // Defer statement
defer print_string("日本語\n");
print_string("Japanese\n");
}
{
defer print_string("1\n");
defer print_string("2\n");
defer print_string("3\n");
}
{
prev_allocator := context.allocator;
context.allocator = __default_allocator();
defer context.allocator = prev_allocator;
File :: type struct { filename: string };
FileError :: type int;
open_file :: proc(filename: string) -> (File, FileError) {
return File{}, 0;
}
close_file :: proc(f: ^File) {}
f, err := open_file("Test");
if err != 0 {
// handle error
}
defer close_file(^f);
}
for i := 0; i < 100; i++ {
blah := alloc(100 * size_of(int)) as ^int;
defer {
defer print_string("!");
defer print_string("dealloc");
dealloc(blah);
}
if i == 3 {
// defers called
continue;
}
if i == 5 {
// defers called
return; // End of procedure
}
if i == 8 {
// defers called
break; // never happens
}
}
defer print_string("It'll never happen, mate 1");
print_string("It'll never happen, mate 2");
print_string("It'll never happen, mate 3");
}
using_fields :: proc() {
{ // Everyday stuff
Vec3 :: type struct { x, y, z: f32; }
Entity :: type struct {
name: string;
using pos: Vec3;
vel: Vec3;
}
t: Entity;
t.y = 456;
print_f32(t.y); print_nl();
print_f32(t.pos.y); print_nl();
print_f32(t.vel.y); print_nl();
Frog :: type struct { // Subtype (kind of)
using entity: Entity;
colour: u32;
jump_height: f32;
}
f: Frog;
f.y = 1337;
print_f32(f.y); print_nl();
print_f32(f.pos.y); print_nl();
print_f32(f.vel.y); print_nl();
Buffalo :: type struct {
using entity: Entity;
speed: f32;
noise_level: f32;
}
}
{ // Crazy Shit
Vec2 :: type union {
using _xy: struct {x, y: f32};
e: [2]f32;
v: {2}f32;
}
Entity :: type struct {
using pos: ^Vec2;
name: string;
}
t: Entity;
t.pos = alloc(size_of(Vec2)) as ^Vec2; // TODO(bill): make an alloc type? i.e. new(Type)?
t.x = 123;
print_f32(t._xy.x); print_nl();
print_f32(t.pos.x); print_nl();
print_f32(t.pos._xy.x); print_nl();
}
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment