Crates.io | bitfield-struct |
lib.rs | bitfield-struct |
version | 0.9.2 |
source | src |
created_at | 2021-12-21 16:00:17.890695 |
updated_at | 2024-10-15 14:19:05.268203 |
description | Struct-like procedural macro for bitfields. |
homepage | |
repository | https://github.com/wrenger/bitfield-struct-rs.git |
max_upload_size | |
id | 501255 |
size | 60,725 |
Procedural macro for bitfields that allows specifying bitfields as structs.
As this library provides a procedural macro, it has no runtime dependencies and works for no-std
environments.
Default
, fmt::Debug
, or defmt::Format
traitsAdd this to your Cargo.toml
:
[dependencies]
bitfield-struct = "0.9"
Let's begin with a simple example. Suppose we want to store multiple data inside a single Byte, as shown below:
7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
P | Level | S | Kind |
This crate generates a nice wrapper type that makes it easy to do this:
use bitfield_struct::bitfield;
/// Define your type like this with the bitfield attribute
#[bitfield(u8)]
struct MyByte {
/// The first field occupies the least significant bits
#[bits(4)]
kind: usize,
/// Booleans are 1 bit large
system: bool,
/// The bits attribute specifies the bit size of this field
#[bits(2)]
level: usize,
/// The last field spans over the most significant bits
present: bool
}
// The macro creates three accessor functions for each field:
// <name>, with_<name> and set_<name>
let my_byte = MyByte::new()
.with_kind(15)
.with_system(false)
.with_level(3)
.with_present(true);
assert!(my_byte.present());
Additionally, this crate has a few useful features, which are shown here in more detail.
The example below shows how attributes are carried over and how signed integers, padding, and custom types are handled.
use bitfield_struct::bitfield;
/// A test bitfield with documentation
#[bitfield(u64)]
#[derive(PartialEq, Eq)] // <- Attributes after `bitfield` are carried over
struct MyBitfield {
/// Defaults to 16 bits for u16
int: u16,
/// Interpreted as 1 bit flag, with a custom default value
#[bits(default = true)]
flag: bool,
/// Custom bit size
#[bits(1)]
tiny: u8,
/// Sign extend for signed integers
#[bits(13)]
negative: i16,
/// Supports any type with `into_bits`/`from_bits` functions
#[bits(16)]
custom: CustomEnum,
/// Public field -> public accessor functions
#[bits(10)]
pub public: usize,
/// Also supports read-only fields
#[bits(1, access = RO)]
read_only: bool,
/// And write-only fields
#[bits(1, access = WO)]
write_only: bool,
/// Padding
#[bits(5)]
__: u8,
}
/// A custom enum
#[derive(Debug, PartialEq, Eq)]
#[repr(u16)]
enum CustomEnum {
A = 0,
B = 1,
C = 2,
}
impl CustomEnum {
// This has to be a const fn
const fn into_bits(self) -> u16 {
self as _
}
const fn from_bits(value: u16) -> Self {
match value {
0 => Self::A,
1 => Self::B,
_ => Self::C,
}
}
}
// Usage:
let mut val = MyBitfield::new()
.with_int(3 << 15)
.with_tiny(1)
.with_negative(-3)
.with_custom(CustomEnum::B)
.with_public(2)
// .with_read_only(true) <- Would not compile
.with_write_only(false);
println!("{val:?}");
let raw: u64 = val.into();
println!("{raw:b}");
assert_eq!(val.int(), 3 << 15);
assert_eq!(val.flag(), true);
assert_eq!(val.negative(), -3);
assert_eq!(val.tiny(), 1);
assert_eq!(val.custom(), CustomEnum::B);
assert_eq!(val.public(), 2);
assert_eq!(val.read_only(), false);
// const members
assert_eq!(MyBitfield::FLAG_BITS, 1);
assert_eq!(MyBitfield::FLAG_OFFSET, 16);
val.set_negative(1);
assert_eq!(val.negative(), 1);
The macro generates three accessor functions for each field. Each accessor also inherits the documentation of its field.
The signatures for int
are:
// generated struct
struct MyBitfield(u64);
impl MyBitfield {
const fn new() -> Self { Self(0) }
const INT_BITS: usize = 16;
const INT_OFFSET: usize = 0;
const fn int(&self) -> u16 { todo!() }
const fn with_int(self, value: u16) -> Self { todo!() }
const fn with_int_checked(self, value: u16) -> Result<Self, ()> { todo!() }
fn set_int(&mut self, value: u16) { todo!() }
fn set_int_checked(&mut self, value: u16) -> Result<(), ()> { todo!() }
// other field ...
}
// Also generates From<u64>, Into<u64>, Default, and Debug implementations...
Hint: You can use the rust-analyzer "Expand macro recursively" action to view the generated code.
The macro supports any types that are convertible into the underlying bitfield type. This can be enums like in the following example or any other struct.
The conversion and default values can be specified with the following #[bits]
parameters:
from
: Function converting from raw bits into the custom type, defaults to <ty>::from_bits
into
: Function converting from the custom type into raw bits, defaults to <ty>::into_bits
default
: Custom expression, defaults to calling <ty>::from_bits(0)
use bitfield_struct::bitfield;
#[bitfield(u16)]
#[derive(PartialEq, Eq)]
struct Bits {
/// Supports any convertible type
#[bits(8, default = CustomEnum::B, from = CustomEnum::my_from_bits)]
custom: CustomEnum,
/// And nested bitfields
#[bits(8)]
nested: Nested,
}
#[derive(Debug, PartialEq, Eq)]
#[repr(u8)]
enum CustomEnum {
A = 0,
B = 1,
C = 2,
}
impl CustomEnum {
// This has to be a const fn
const fn into_bits(self) -> u8 {
self as _
}
const fn my_from_bits(value: u8) -> Self {
match value {
0 => Self::A,
1 => Self::B,
_ => Self::C,
}
}
}
/// Bitfields implement the conversion functions automatically
#[bitfield(u8)]
struct Nested {
#[bits(4)]
lo: u8,
#[bits(4)]
hi: u8,
}
The optional order
macro argument determines the layout of the bits, with the default being
Lsb (least significant bit) first:
use bitfield_struct::bitfield;
#[bitfield(u8, order = Lsb)]
struct MyLsbByte {
/// The first field occupies the *least* significant bits
#[bits(4)]
kind: usize,
system: bool,
#[bits(2)]
level: usize,
present: bool
}
let my_byte_lsb = MyLsbByte::new()
.with_kind(10)
.with_system(false)
.with_level(2)
.with_present(true);
// .- present
// | .- level
// | | .- system
// | | | .- kind
assert_eq!(my_byte_lsb.0, 0b1_10_0_1010);
The macro generates the reverse order when Msb (most significant bit) is specified:
use bitfield_struct::bitfield;
#[bitfield(u8, order = Msb)]
struct MyMsbByte {
/// The first field occupies the *most* significant bits
#[bits(4)]
kind: usize,
system: bool,
#[bits(2)]
level: usize,
present: bool
}
let my_byte_msb = MyMsbByte::new()
.with_kind(10)
.with_system(false)
.with_level(2)
.with_present(true);
// .- kind
// | .- system
// | | .- level
// | | | .- present
assert_eq!(my_byte_msb.0, 0b1010_0_10_1);
The macro supports custom types for the representation of the bitfield struct.
This can be an endian-defining type like in the following examples (from endian-num
) or any other struct that can be converted to and from the main bitfield type.
The representation and its conversion functions can be specified with the following #[bitfield]
parameters:
repr
specifies the bitfield's representation in memoryfrom
to specify a conversion function from repr to the bitfield's integer typeinto
to specify a conversion function from the bitfield's integer type to reprThis example has a little-endian byte order even on big-endian machines:
use bitfield_struct::bitfield;
use endian_num::le16;
#[bitfield(u16, repr = le16, from = le16::from_ne, into = le16::to_ne)]
struct MyLeBitfield {
#[bits(4)]
first_nibble: u8,
#[bits(12)]
other: u16,
}
let my_be_bitfield = MyLeBitfield::new()
.with_first_nibble(0x1)
.with_other(0x234);
assert_eq!(my_be_bitfield.into_bits().to_le_bytes(), [0x41, 0x23]);
This example has a big-endian byte order even on little-endian machines:
use bitfield_struct::bitfield;
use endian_num::be16;
#[bitfield(u16, repr = be16, from = be16::from_ne, into = be16::to_ne)]
struct MyBeBitfield {
#[bits(4)]
first_nibble: u8,
#[bits(12)]
other: u16,
}
let my_be_bitfield = MyBeBitfield::new()
.with_first_nibble(0x1)
.with_other(0x234);
assert_eq!(my_be_bitfield.into_bits().to_be_bytes(), [0x23, 0x41]);
Besides implementing Clone
and Copy
, this macro automatically creates a suitable fmt::Debug
and Default
implementations similar to the ones created for normal structs by #[derive(Debug, Default)]
.
You can disable this with the extra debug
and default
arguments.
use std::fmt::{Debug, Formatter, Result};
use bitfield_struct::bitfield;
#[bitfield(u64, debug = false, default = false)]
struct CustomDebug {
data: u64
}
impl Debug for CustomDebug {
fn fmt(&self, f: &mut Formatter<'_>) -> Result {
write!(f, "0x{:x}", self.data())
}
}
impl Default for CustomDebug {
fn default() -> Self {
Self(123)
}
}
let val = CustomDebug::default();
println!("{val:?}")
defmt::Format
This macro can automatically implement a defmt::Format
that mirrors the default fmt::Debug
implementation by passing the extra defmt
argument.
This implementation requires the defmt crate to be available as defmt
, and has the same rules and caveats as #[derive(defmt::Format)]
.
use bitfield_struct::bitfield;
#[bitfield(u64, defmt = true)]
struct DefmtExample {
data: u64
}
new
/Debug
/Default
/defmt::Format
Instead of booleans, you can specify cfg(...)
attributes for new
, debug
, default
and defmt
:
use bitfield_struct::bitfield;
#[bitfield(u64, debug = cfg(test), default = cfg(feature = "foo"))]
struct CustomDebug {
data: u64
}