Union SmallBytes 

pub union SmallBytes {
    /* private fields */
}

A 24-byte owned byte string with inline small-string optimization.

Strings of up to 22 bytes live entirely inside the value (no allocation, no pointer chase); larger strings spill to a heap buffer. The discriminator is a single byte at offset 23 (the tag, which doubles as the inline length 0..=22 OR equals 0xFF when the heap variant is active).

See the crate root for layout details.

Examples

Short values live inline; longer ones move to the heap. The API does not change, but heap_bytes reports which happened, which is what the keyspace’s memory accounting reads.

use kevy_bytes::SmallBytes;
let short = SmallBytes::from_slice(b"hello");
assert_eq!(short.as_slice(), b"hello");
assert_eq!(short.len(), 5);
assert_eq!(short.heap_bytes(), 0, "a short value allocates nothing");

let long = SmallBytes::from_slice(&[b'x'; 100]);
assert_eq!(long.len(), 100);
assert!(long.heap_bytes() >= 100, "a long value is on the heap");

use kevy_bytes::SmallBytes;
assert!(SmallBytes::from_slice(b"").is_empty());

Implementations

impl SmallBytes

pub const fn new() -> Self

Empty inline SmallBytes (zero allocation).

Examples

const, so it can seed a static or an array without a run-time initialiser:

use kevy_bytes::SmallBytes;
static EMPTY: SmallBytes = SmallBytes::new();
assert!(EMPTY.is_empty());
assert_eq!(EMPTY.heap_bytes(), 0);

pub fn from_slice(bytes: &[u8]) -> Self

Construct from a byte slice — inline if bytes.len() <= 22, else heap.

Examples

Twenty-two is the boundary, and it is exact:

use kevy_bytes::SmallBytes;
assert_eq!(SmallBytes::from_slice(&[b'x'; 22]).heap_bytes(), 0);
assert_eq!(SmallBytes::from_slice(&[b'x'; 23]).heap_bytes(), 23);

pub fn from_vec(vec: Vec<u8>) -> Self

Take ownership of a Vec<u8> — inline if vec.len() <= 22, else reuse the vec’s allocation (no copy on the heap path).

Examples

The heap path keeps the vec’s own buffer, so a value that arrived as a Vec is stored without a second copy:

use kevy_bytes::SmallBytes;
let v = vec![b'z'; 64];
let addr = v.as_ptr();
let b = SmallBytes::from_vec(v);
assert_eq!(b.as_slice().as_ptr(), addr, "same allocation, not a copy");

A short vec goes inline instead, and its allocation is released:

use kevy_bytes::SmallBytes;
assert_eq!(SmallBytes::from_vec(vec![b'a'; 4]).heap_bytes(), 0);

pub fn len(&self) -> usize

Number of bytes stored.

Examples

The same answer either side of the inline boundary — which is the point of the type: where the bytes live is not the caller’s problem.

use kevy_bytes::SmallBytes;
assert_eq!(SmallBytes::from_slice(&[0u8; 22]).len(), 22);
assert_eq!(SmallBytes::from_slice(&[0u8; 23]).len(), 23);

pub fn is_empty(&self) -> bool

Whether len() == 0.

Examples

use kevy_bytes::SmallBytes;
assert!(SmallBytes::from_slice(b"").is_empty());
assert!(!SmallBytes::from_slice(b"\0").is_empty(), "a NUL byte is a byte");

pub fn heap_bytes(&self) -> usize

Bytes this value holds on the heap (0 when inline). Lets memory-accounting callers (e.g. maxmemory enforcement) charge only the off-stack footprint without re-deriving the inline-length threshold.

Examples

This is what maxmemory charges, so an inline value must cost zero — it is already inside the entry the keyspace has counted:

use kevy_bytes::SmallBytes;
assert_eq!(SmallBytes::from_slice(b"user:1").heap_bytes(), 0);
assert_eq!(SmallBytes::from_slice(&[b'x'; 1000]).heap_bytes(), 1000);

pub fn as_slice(&self) -> &[u8]

Borrow the bytes (no allocation; same for inline and heap variants).

Examples

use kevy_bytes::SmallBytes;
let b = SmallBytes::from_slice(b"GET");
assert_eq!(b.as_slice(), b"GET");
assert_eq!(SmallBytes::new().as_slice(), b"");

pub fn to_vec(&self) -> Vec<u8>

Copy into a fresh Vec<u8> (clone semantics).

Examples

use kevy_bytes::SmallBytes;
let b = SmallBytes::from_slice(b"copy me");
assert_eq!(b.to_vec(), b"copy me");
assert_eq!(b.as_slice(), b"copy me", "the original still holds them");

pub fn into_vec(self) -> Vec<u8>

Consume self and return an owned Vec<u8>. The heap path reuses the existing allocation; the inline path copies into a new vec.

Examples

A heap value hands its buffer straight back, so a round trip through SmallBytes costs no allocation at either end:

use kevy_bytes::SmallBytes;
let v = vec![b'q'; 128];
let addr = v.as_ptr();
assert_eq!(SmallBytes::from_vec(v).into_vec().as_ptr(), addr);

use kevy_bytes::SmallBytes;
assert_eq!(SmallBytes::from_slice(b"short").into_vec(), b"short");

Trait Implementations

impl AsRef<[u8]> for SmallBytes

fn as_ref(&self) -> &[u8]

Converts this type into a shared reference of the (usually inferred) input type.

impl Borrow<[u8]> for SmallBytes

fn borrow(&self) -> &[u8]

Immutably borrows from an owned value.

impl Clone for SmallBytes

fn clone(&self) -> Self

Specialised clone that bypasses as_slice → from_slice → alloc_heap’s two layered length checks. Inline variant is a bitwise union copy (no branch through the slice path); heap variant goes straight to a single alloc + memcpy keyed on the already-known heap length.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SmallBytes

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for SmallBytes

fn default() -> Self

Returns the “default value” for a type.

impl Drop for SmallBytes

fn drop(&mut self)

Executes the destructor for this type.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.

impl Eq for SmallBytes

impl From<&[u8]> for SmallBytes

fn from(bytes: &[u8]) -> Self

Converts to this type from the input type.

impl From<Vec<u8>> for SmallBytes

fn from(vec: Vec<u8>) -> Self

Converts to this type from the input type.

impl Hash for SmallBytes

fn hash<H: Hasher>(&self, state: &mut H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl KevyHash for SmallBytes

KevyHash agrees with the byte-slice impl, so a KevyMap<SmallBytes, V> can be queried with &[u8] (via Borrow<[u8]>) and the hash matches.

fn kevy_hash(&self) -> u64

Compute the final mixed 64-bit hash of self in one call.

impl Ord for SmallBytes

fn cmp(&self, other: &Self) -> Ordering

This method returns an Ordering between self and other.

fn max(self, other: Self) -> Self

where Self: Sized,

Compares and returns the maximum of two values.

fn min(self, other: Self) -> Self

where Self: Sized,

Compares and returns the minimum of two values.

fn clamp(self, min: Self, max: Self) -> Self

where Self: Sized,

Restrict a value to a certain interval.

fn clamp_to<R>(self, range: R) -> Self

where Self: Sized, R: ClampBounds<Self>,

🔬This is a nightly-only experimental API. (clamp_to)

Restrict a value to a certain range.

impl PartialEq for SmallBytes

fn eq(&self, other: &Self) -> bool

Specialised over the slice form (as_slice == as_slice) by branching on variant once and reading the relevant length / pointer pair directly. Same-variant cases (inline/inline + heap/heap, which are the only ones produced by a single allocator) skip a redundant as_slice dispatch on each side; the mixed case falls back to the slice form.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl PartialOrd for SmallBytes

fn partial_cmp(&self, other: &Self) -> Option<Ordering>

This method returns an ordering between self and other values if one exists.

fn lt(&self, other: &Rhs) -> bool

Tests less than (for self and other) and is used by the < operator.

fn le(&self, other: &Rhs) -> bool

Tests less than or equal to (for self and other) and is used by the <= operator.

fn gt(&self, other: &Rhs) -> bool

Tests greater than (for self and other) and is used by the > operator.

fn ge(&self, other: &Rhs) -> bool

Tests greater than or equal to (for self and other) and is used by the >= operator.

impl Send for SmallBytes

impl Sync for SmallBytes