index | The index of the element. |
T
A ref readonly
to a value of type T
.
Provides read-only access to ReadOnly elements by index.
This method is dangerous because it returns reference to the memory that can be destroyed (with Dispose, for instance) and
if the ref local
is accessed that might lead to undefined results including crashes.
public readonly struct BigStruct { public readonly long a; public readonly long b; public readonly long c; public readonly long d; }
static void ProcessByte(byte b) { ... }
static void ProcessBigStructWithoutCopy(in BigStruct bigStruct) // see 'in' modificator { ... }
static void Example() { const int n = 32;
var nativeArrayOfBytes = new NativeArray<byte>(n, Allocator.Temp); var nativeArrayOfBigStructures = new NativeArray<FixedString4096Bytes>(n, Allocator.Temp);
// ... fill the arrays with some data ...
var readOnlyBytes = nativeArrayOfBytes.AsReadOnly(); for (var i = 0; i < n; ++i) { ProcessByte(readOnlyBytes[i]); //ProcessByte(readOnlyBytes.UnsafeElementAt(i)); is more expensive, since pointer on x64 platforms is 8 times bigger than a byte }
var readOnlyBigStructures = nativeArrayOfBigStructures.AsReadOnly(); for (var i = 0; i < n; ++i) { //ProcessBigStructWithoutCopy(readOnlyBigStructures[i]); copy - expensive in this case ProcessBigStructWithoutCopy(readOnlyBigStructures.UnsafeElementAt(i)); }
// dangerous part ref var element = ref readOnlyBigStructures.UnsafeElementAt(4);
// element is valid here ProcessBigStructWithoutCopy(element);
nativeArrayOfBigStructures.Dispose();
// access to element here is undefined, can lead to a crash or wrong results // ProcessBigStructWithoutCopy(element); <-- do not do this! }