summaryrefslogtreecommitdiff
path: root/vendor/fixedbitset/src
diff options
context:
space:
mode:
authormo khan <mo@mokhan.ca>2025-07-02 18:36:06 -0600
committermo khan <mo@mokhan.ca>2025-07-02 18:36:06 -0600
commit8cdfa445d6629ffef4cb84967ff7017654045bc2 (patch)
tree22f0b0907c024c78d26a731e2e1f5219407d8102 /vendor/fixedbitset/src
parent4351c74c7c5f97156bc94d3a8549b9940ac80e3f (diff)
chore: add vendor directory
Diffstat (limited to 'vendor/fixedbitset/src')
-rw-r--r--vendor/fixedbitset/src/block/avx.rs92
-rw-r--r--vendor/fixedbitset/src/block/avx2.rs88
-rw-r--r--vendor/fixedbitset/src/block/default.rs70
-rw-r--r--vendor/fixedbitset/src/block/mod.rs114
-rw-r--r--vendor/fixedbitset/src/block/sse2.rs104
-rw-r--r--vendor/fixedbitset/src/block/wasm.rs80
-rw-r--r--vendor/fixedbitset/src/lib.rs1711
-rw-r--r--vendor/fixedbitset/src/range.rs45
-rw-r--r--vendor/fixedbitset/src/serde_impl.rs150
9 files changed, 2454 insertions, 0 deletions
diff --git a/vendor/fixedbitset/src/block/avx.rs b/vendor/fixedbitset/src/block/avx.rs
new file mode 100644
index 00000000..33ba100f
--- /dev/null
+++ b/vendor/fixedbitset/src/block/avx.rs
@@ -0,0 +1,92 @@
+#[cfg(target_arch = "x86")]
+use core::arch::x86::*;
+#[cfg(target_arch = "x86_64")]
+use core::arch::x86_64::*;
+use core::ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Not};
+
+#[derive(Copy, Clone, Debug)]
+#[repr(transparent)]
+pub struct Block(pub(super) __m256d);
+
+impl Block {
+ #[inline]
+ pub fn is_empty(self) -> bool {
+ unsafe {
+ let value = _mm256_castpd_si256(self.0);
+ _mm256_testz_si256(value, value) == 1
+ }
+ }
+
+ #[inline]
+ pub fn andnot(self, other: Self) -> Self {
+ unsafe { Self(_mm256_andnot_pd(other.0, self.0)) }
+ }
+}
+
+impl Not for Block {
+ type Output = Block;
+ #[inline]
+ fn not(self) -> Self::Output {
+ unsafe { Self(_mm256_xor_pd(self.0, Self::ALL.0)) }
+ }
+}
+
+impl BitAnd for Block {
+ type Output = Block;
+ #[inline]
+ fn bitand(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_and_pd(self.0, other.0)) }
+ }
+}
+
+impl BitAndAssign for Block {
+ #[inline]
+ fn bitand_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm256_and_pd(self.0, other.0);
+ }
+ }
+}
+
+impl BitOr for Block {
+ type Output = Block;
+ #[inline]
+ fn bitor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_or_pd(self.0, other.0)) }
+ }
+}
+
+impl BitOrAssign for Block {
+ #[inline]
+ fn bitor_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm256_or_pd(self.0, other.0);
+ }
+ }
+}
+
+impl BitXor for Block {
+ type Output = Block;
+ #[inline]
+ fn bitxor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_xor_pd(self.0, other.0)) }
+ }
+}
+
+impl BitXorAssign for Block {
+ #[inline]
+ fn bitxor_assign(&mut self, other: Self) {
+ unsafe { self.0 = _mm256_xor_pd(self.0, other.0) }
+ }
+}
+
+impl PartialEq for Block {
+ #[inline]
+ fn eq(&self, other: &Self) -> bool {
+ unsafe {
+ let new = _mm256_xor_pd(self.0, other.0);
+ let neq = _mm256_castpd_si256(new);
+ _mm256_testz_si256(neq, neq) == 1
+ }
+ }
+}
diff --git a/vendor/fixedbitset/src/block/avx2.rs b/vendor/fixedbitset/src/block/avx2.rs
new file mode 100644
index 00000000..b3593773
--- /dev/null
+++ b/vendor/fixedbitset/src/block/avx2.rs
@@ -0,0 +1,88 @@
+#[cfg(target_arch = "x86")]
+use core::arch::x86::*;
+#[cfg(target_arch = "x86_64")]
+use core::arch::x86_64::*;
+use core::ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Not};
+
+#[derive(Copy, Clone, Debug)]
+#[repr(transparent)]
+pub struct Block(pub(super) __m256i);
+
+impl Block {
+ #[inline]
+ pub fn is_empty(self) -> bool {
+ unsafe { _mm256_testz_si256(self.0, self.0) == 1 }
+ }
+
+ #[inline]
+ pub fn andnot(self, other: Self) -> Self {
+ Self(unsafe { _mm256_andnot_si256(other.0, self.0) })
+ }
+}
+
+impl Not for Block {
+ type Output = Block;
+ #[inline]
+ fn not(self) -> Self::Output {
+ unsafe { Self(_mm256_xor_si256(self.0, Self::ALL.0)) }
+ }
+}
+
+impl BitAnd for Block {
+ type Output = Block;
+ #[inline]
+ fn bitand(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_and_si256(self.0, other.0)) }
+ }
+}
+
+impl BitAndAssign for Block {
+ #[inline]
+ fn bitand_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm256_and_si256(self.0, other.0);
+ }
+ }
+}
+
+impl BitOr for Block {
+ type Output = Block;
+ #[inline]
+ fn bitor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_or_si256(self.0, other.0)) }
+ }
+}
+
+impl BitOrAssign for Block {
+ #[inline]
+ fn bitor_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm256_or_si256(self.0, other.0);
+ }
+ }
+}
+
+impl BitXor for Block {
+ type Output = Block;
+ #[inline]
+ fn bitxor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm256_xor_si256(self.0, other.0)) }
+ }
+}
+
+impl BitXorAssign for Block {
+ #[inline]
+ fn bitxor_assign(&mut self, other: Self) {
+ unsafe { self.0 = _mm256_xor_si256(self.0, other.0) }
+ }
+}
+
+impl PartialEq for Block {
+ #[inline]
+ fn eq(&self, other: &Self) -> bool {
+ unsafe {
+ let neq = _mm256_xor_si256(self.0, other.0);
+ _mm256_testz_si256(neq, neq) == 1
+ }
+ }
+}
diff --git a/vendor/fixedbitset/src/block/default.rs b/vendor/fixedbitset/src/block/default.rs
new file mode 100644
index 00000000..7fc460fb
--- /dev/null
+++ b/vendor/fixedbitset/src/block/default.rs
@@ -0,0 +1,70 @@
+use core::ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Not};
+
+#[derive(Copy, Clone, PartialEq, Debug)]
+#[repr(transparent)]
+pub struct Block(pub(super) usize);
+
+impl Block {
+ #[inline]
+ pub const fn is_empty(self) -> bool {
+ self.0 == Self::NONE.0
+ }
+
+ #[inline]
+ pub fn andnot(self, other: Self) -> Self {
+ Self(!other.0 & self.0)
+ }
+}
+
+impl Not for Block {
+ type Output = Block;
+ #[inline]
+ fn not(self) -> Self::Output {
+ Self(self.0.not())
+ }
+}
+
+impl BitAnd for Block {
+ type Output = Block;
+ #[inline]
+ fn bitand(self, other: Self) -> Self::Output {
+ Self(self.0.bitand(other.0))
+ }
+}
+
+impl BitAndAssign for Block {
+ #[inline]
+ fn bitand_assign(&mut self, other: Self) {
+ self.0.bitand_assign(other.0);
+ }
+}
+
+impl BitOr for Block {
+ type Output = Block;
+ #[inline]
+ fn bitor(self, other: Self) -> Self::Output {
+ Self(self.0.bitor(other.0))
+ }
+}
+
+impl BitOrAssign for Block {
+ #[inline]
+ fn bitor_assign(&mut self, other: Self) {
+ self.0.bitor_assign(other.0)
+ }
+}
+
+impl BitXor for Block {
+ type Output = Block;
+ #[inline]
+ fn bitxor(self, other: Self) -> Self::Output {
+ Self(self.0.bitxor(other.0))
+ }
+}
+
+impl BitXorAssign for Block {
+ #[inline]
+ fn bitxor_assign(&mut self, other: Self) {
+ self.0.bitxor_assign(other.0)
+ }
+}
diff --git a/vendor/fixedbitset/src/block/mod.rs b/vendor/fixedbitset/src/block/mod.rs
new file mode 100644
index 00000000..ae7c2220
--- /dev/null
+++ b/vendor/fixedbitset/src/block/mod.rs
@@ -0,0 +1,114 @@
+#![allow(clippy::undocumented_unsafe_blocks)]
+#![allow(dead_code)]
+// TODO: Remove once the transmutes are fixed
+#![allow(unknown_lints)]
+#![allow(clippy::missing_transmute_annotations)]
+
+use core::cmp::Ordering;
+use core::hash::{Hash, Hasher};
+
+#[cfg(all(
+ not(all(target_family = "wasm", target_feature = "simd128")),
+ not(target_feature = "sse2"),
+ not(target_feature = "avx"),
+ not(target_feature = "avx2"),
+))]
+mod default;
+#[cfg(all(
+ not(all(target_family = "wasm", target_feature = "simd128")),
+ not(target_feature = "sse2"),
+ not(target_feature = "avx"),
+ not(target_feature = "avx2"),
+))]
+pub use self::default::*;
+
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "sse2",
+ not(target_feature = "avx"),
+ not(target_feature = "avx2"),
+))]
+mod sse2;
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "sse2",
+ not(target_feature = "avx"),
+ not(target_feature = "avx2"),
+))]
+pub use self::sse2::*;
+
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "avx",
+ not(target_feature = "avx2")
+))]
+mod avx;
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "avx",
+ not(target_feature = "avx2")
+))]
+pub use self::avx::*;
+
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "avx2"
+))]
+mod avx2;
+#[cfg(all(
+ any(target_arch = "x86", target_arch = "x86_64"),
+ target_feature = "avx2"
+))]
+pub use self::avx2::*;
+
+#[cfg(all(target_family = "wasm", target_feature = "simd128"))]
+mod wasm;
+#[cfg(all(target_family = "wasm", target_feature = "simd128"))]
+pub use self::wasm::*;
+
+impl Block {
+ pub const USIZE_COUNT: usize = core::mem::size_of::<Self>() / core::mem::size_of::<usize>();
+ pub const NONE: Self = Self::from_usize_array([0; Self::USIZE_COUNT]);
+ pub const ALL: Self = Self::from_usize_array([usize::MAX; Self::USIZE_COUNT]);
+ pub const BITS: usize = core::mem::size_of::<Self>() * 8;
+
+ #[inline]
+ pub fn into_usize_array(self) -> [usize; Self::USIZE_COUNT] {
+ unsafe { core::mem::transmute(self.0) }
+ }
+
+ #[inline]
+ pub const fn from_usize_array(array: [usize; Self::USIZE_COUNT]) -> Self {
+ Self(unsafe { core::mem::transmute(array) })
+ }
+}
+
+impl Eq for Block {}
+
+impl PartialOrd for Block {
+ #[inline]
+ fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
+ Some(self.cmp(other))
+ }
+}
+
+impl Ord for Block {
+ #[inline]
+ fn cmp(&self, other: &Self) -> Ordering {
+ self.into_usize_array().cmp(&other.into_usize_array())
+ }
+}
+
+impl Default for Block {
+ #[inline]
+ fn default() -> Self {
+ Self::NONE
+ }
+}
+
+impl Hash for Block {
+ #[inline]
+ fn hash<H: Hasher>(&self, hasher: &mut H) {
+ Hash::hash_slice(&self.into_usize_array(), hasher);
+ }
+}
diff --git a/vendor/fixedbitset/src/block/sse2.rs b/vendor/fixedbitset/src/block/sse2.rs
new file mode 100644
index 00000000..6db08f7e
--- /dev/null
+++ b/vendor/fixedbitset/src/block/sse2.rs
@@ -0,0 +1,104 @@
+#![allow(clippy::undocumented_unsafe_blocks)]
+
+#[cfg(target_arch = "x86")]
+use core::arch::x86::*;
+#[cfg(target_arch = "x86_64")]
+use core::arch::x86_64::*;
+use core::ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Not};
+
+#[derive(Copy, Clone, Debug)]
+#[repr(transparent)]
+pub struct Block(pub(super) __m128i);
+
+impl Block {
+ #[inline]
+ pub fn is_empty(self) -> bool {
+ #[cfg(not(target_feature = "sse4.1"))]
+ {
+ self == Self::NONE
+ }
+ #[cfg(target_feature = "sse4.1")]
+ {
+ unsafe { _mm_test_all_zeros(self.0, self.0) == 1 }
+ }
+ }
+
+ #[inline]
+ pub fn andnot(self, other: Self) -> Self {
+ Self(unsafe { _mm_andnot_si128(other.0, self.0) })
+ }
+}
+
+impl Not for Block {
+ type Output = Block;
+ #[inline]
+ fn not(self) -> Self::Output {
+ unsafe { Self(_mm_xor_si128(self.0, Self::ALL.0)) }
+ }
+}
+
+impl BitAnd for Block {
+ type Output = Block;
+ #[inline]
+ fn bitand(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm_and_si128(self.0, other.0)) }
+ }
+}
+
+impl BitAndAssign for Block {
+ #[inline]
+ fn bitand_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm_and_si128(self.0, other.0);
+ }
+ }
+}
+
+impl BitOr for Block {
+ type Output = Block;
+ #[inline]
+ fn bitor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm_or_si128(self.0, other.0)) }
+ }
+}
+
+impl BitOrAssign for Block {
+ #[inline]
+ fn bitor_assign(&mut self, other: Self) {
+ unsafe {
+ self.0 = _mm_or_si128(self.0, other.0);
+ }
+ }
+}
+
+impl BitXor for Block {
+ type Output = Block;
+ #[inline]
+ fn bitxor(self, other: Self) -> Self::Output {
+ unsafe { Self(_mm_xor_si128(self.0, other.0)) }
+ }
+}
+
+impl BitXorAssign for Block {
+ #[inline]
+ fn bitxor_assign(&mut self, other: Self) {
+ unsafe { self.0 = _mm_xor_si128(self.0, other.0) }
+ }
+}
+
+impl PartialEq for Block {
+ #[inline]
+ fn eq(&self, other: &Self) -> bool {
+ unsafe {
+ #[cfg(not(target_feature = "sse4.1"))]
+ {
+ _mm_movemask_epi8(_mm_cmpeq_epi8(self.0, other.0)) == 0xffff
+ }
+ #[cfg(target_feature = "sse4.1")]
+ {
+ let neq = _mm_xor_si128(self.0, other.0);
+ _mm_test_all_zeros(neq, neq) == 1
+ }
+ }
+ }
+}
diff --git a/vendor/fixedbitset/src/block/wasm.rs b/vendor/fixedbitset/src/block/wasm.rs
new file mode 100644
index 00000000..f823d86a
--- /dev/null
+++ b/vendor/fixedbitset/src/block/wasm.rs
@@ -0,0 +1,80 @@
+use core::{
+ arch::wasm32::*,
+ ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Not},
+};
+
+#[derive(Copy, Clone, Debug)]
+#[repr(transparent)]
+pub struct Block(pub(super) v128);
+
+impl Block {
+ #[inline]
+ pub fn is_empty(self) -> bool {
+ !v128_any_true(self.0)
+ }
+
+ #[inline]
+ pub fn andnot(self, other: Self) -> Self {
+ Self(v128_andnot(self.0, other.0))
+ }
+}
+
+impl Not for Block {
+ type Output = Block;
+ #[inline]
+ fn not(self) -> Self::Output {
+ Self(v128_xor(self.0, Self::ALL.0))
+ }
+}
+
+impl BitAnd for Block {
+ type Output = Block;
+ #[inline]
+ fn bitand(self, other: Self) -> Self::Output {
+ Self(v128_and(self.0, other.0))
+ }
+}
+
+impl BitAndAssign for Block {
+ #[inline]
+ fn bitand_assign(&mut self, other: Self) {
+ self.0 = v128_and(self.0, other.0);
+ }
+}
+
+impl BitOr for Block {
+ type Output = Block;
+ #[inline]
+ fn bitor(self, other: Self) -> Self::Output {
+ Self(v128_or(self.0, other.0))
+ }
+}
+
+impl BitOrAssign for Block {
+ #[inline]
+ fn bitor_assign(&mut self, other: Self) {
+ self.0 = v128_or(self.0, other.0);
+ }
+}
+
+impl BitXor for Block {
+ type Output = Block;
+ #[inline]
+ fn bitxor(self, other: Self) -> Self::Output {
+ Self(v128_xor(self.0, other.0))
+ }
+}
+
+impl BitXorAssign for Block {
+ #[inline]
+ fn bitxor_assign(&mut self, other: Self) {
+ self.0 = v128_xor(self.0, other.0)
+ }
+}
+
+impl PartialEq for Block {
+ #[inline]
+ fn eq(&self, other: &Self) -> bool {
+ !v128_any_true(v128_xor(self.0, other.0))
+ }
+}
diff --git a/vendor/fixedbitset/src/lib.rs b/vendor/fixedbitset/src/lib.rs
new file mode 100644
index 00000000..f9d12f41
--- /dev/null
+++ b/vendor/fixedbitset/src/lib.rs
@@ -0,0 +1,1711 @@
+//! `FixedBitSet` is a simple fixed size set of bits.
+//!
+//! ### Crate features
+//!
+//! - `std` (default feature)
+//! Disabling this feature disables using std and instead uses crate alloc.
+//!
+//! ### SIMD Acceleration
+//! `fixedbitset` is written with SIMD in mind. The backing store and set operations will use aligned SIMD data types and instructions when compiling
+//! for compatible target platforms. The use of SIMD generally enables better performance in many set and batch operations (i.e. intersection/union/inserting a range).
+//!
+//! When SIMD is not available on the target, the crate will gracefully fallback to a default implementation. It is intended to add support for other SIMD architectures
+//! once they appear in stable Rust.
+//!
+//! Currently only SSE2/AVX/AVX2 on x86/x86_64 and wasm32 SIMD are supported as this is what stable Rust supports.
+#![no_std]
+#![deny(clippy::undocumented_unsafe_blocks)]
+
+extern crate alloc;
+use alloc::{vec, vec::Vec};
+
+mod block;
+mod range;
+
+#[cfg(feature = "serde")]
+extern crate serde;
+#[cfg(feature = "serde")]
+mod serde_impl;
+
+use core::fmt::Write;
+use core::fmt::{Binary, Display, Error, Formatter};
+
+use core::cmp::Ordering;
+use core::hash::Hash;
+use core::iter::{Chain, FusedIterator};
+use core::mem::ManuallyDrop;
+use core::mem::MaybeUninit;
+use core::ops::{BitAnd, BitAndAssign, BitOr, BitOrAssign, BitXor, BitXorAssign, Index};
+use core::ptr::NonNull;
+pub use range::IndexRange;
+
+pub(crate) const BITS: usize = core::mem::size_of::<Block>() * 8;
+#[cfg(feature = "serde")]
+pub(crate) const BYTES: usize = core::mem::size_of::<Block>();
+
+use block::Block as SimdBlock;
+pub type Block = usize;
+
+#[inline]
+fn div_rem(x: usize, denominator: usize) -> (usize, usize) {
+ (x / denominator, x % denominator)
+}
+
+fn vec_into_parts<T>(vec: Vec<T>) -> (NonNull<T>, usize, usize) {
+ let mut vec = ManuallyDrop::new(vec);
+ (
+ // SAFETY: A Vec's internal pointer is always non-null.
+ unsafe { NonNull::new_unchecked(vec.as_mut_ptr()) },
+ vec.capacity(),
+ vec.len(),
+ )
+}
+
+/// `FixedBitSet` is a simple fixed size set of bits that each can
+/// be enabled (1 / **true**) or disabled (0 / **false**).
+///
+/// The bit set has a fixed capacity in terms of enabling bits (and the
+/// capacity can grow using the `grow` method).
+///
+/// Derived traits depend on both the zeros and ones, so [0,1] is not equal to
+/// [0,1,0].
+#[derive(Debug, Eq)]
+pub struct FixedBitSet {
+ pub(crate) data: NonNull<MaybeUninit<SimdBlock>>,
+ capacity: usize,
+ /// length in bits
+ pub(crate) length: usize,
+}
+
+// SAFETY: FixedBitset contains no thread-local state and can be safely sent between threads
+unsafe impl Send for FixedBitSet {}
+// SAFETY: FixedBitset does not provide simultaneous unsynchronized mutable access to the
+// underlying buffer.
+unsafe impl Sync for FixedBitSet {}
+
+impl FixedBitSet {
+ /// Create a new empty **FixedBitSet**.
+ pub const fn new() -> Self {
+ FixedBitSet {
+ data: NonNull::dangling(),
+ capacity: 0,
+ length: 0,
+ }
+ }
+
+ /// Create a new **FixedBitSet** with a specific number of bits,
+ /// all initially clear.
+ pub fn with_capacity(bits: usize) -> Self {
+ let (mut blocks, rem) = div_rem(bits, SimdBlock::BITS);
+ blocks += (rem > 0) as usize;
+ Self::from_blocks_and_len(vec![SimdBlock::NONE; blocks], bits)
+ }
+
+ #[inline]
+ fn from_blocks_and_len(data: Vec<SimdBlock>, length: usize) -> Self {
+ let (data, capacity, _) = vec_into_parts(data);
+ FixedBitSet {
+ data: data.cast(),
+ capacity,
+ length,
+ }
+ }
+
+ /// Create a new **FixedBitSet** with a specific number of bits,
+ /// initialized from provided blocks.
+ ///
+ /// If the blocks are not the exact size needed for the capacity
+ /// they will be padded with zeros (if shorter) or truncated to
+ /// the capacity (if longer).
+ ///
+ /// For example:
+ /// ```
+ /// let data = vec![4];
+ /// let bs = fixedbitset::FixedBitSet::with_capacity_and_blocks(4, data);
+ /// assert_eq!(format!("{:b}", bs), "0010");
+ /// ```
+ pub fn with_capacity_and_blocks<I: IntoIterator<Item = Block>>(bits: usize, blocks: I) -> Self {
+ let mut bitset = Self::with_capacity(bits);
+ for (subblock, value) in bitset.as_mut_slice().iter_mut().zip(blocks.into_iter()) {
+ *subblock = value;
+ }
+ bitset
+ }
+
+ /// Grow capacity to **bits**, all new bits initialized to zero
+ #[inline]
+ pub fn grow(&mut self, bits: usize) {
+ #[cold]
+ #[track_caller]
+ #[inline(never)]
+ fn do_grow(slf: &mut FixedBitSet, bits: usize) {
+ // SAFETY: The provided fill is initialized to NONE.
+ unsafe { slf.grow_inner(bits, MaybeUninit::new(SimdBlock::NONE)) };
+ }
+
+ if bits > self.length {
+ do_grow(self, bits);
+ }
+ }
+
+ /// # Safety
+ /// If `fill` is uninitialized, the memory must not be accessed and must be immediately
+ /// written over
+ #[inline(always)]
+ unsafe fn grow_inner(&mut self, bits: usize, fill: MaybeUninit<SimdBlock>) {
+ // SAFETY: The data pointer and capacity were created from a Vec initially. The block
+ // len is identical to that of the original.
+ let mut data = unsafe {
+ Vec::from_raw_parts(self.data.as_ptr(), self.simd_block_len(), self.capacity)
+ };
+ let (mut blocks, rem) = div_rem(bits, SimdBlock::BITS);
+ blocks += (rem > 0) as usize;
+ data.resize(blocks, fill);
+ let (data, capacity, _) = vec_into_parts(data);
+ self.data = data;
+ self.capacity = capacity;
+ self.length = bits;
+ }
+
+ #[inline]
+ unsafe fn get_unchecked(&self, subblock: usize) -> &Block {
+ &*self.data.as_ptr().cast::<Block>().add(subblock)
+ }
+
+ #[inline]
+ unsafe fn get_unchecked_mut(&mut self, subblock: usize) -> &mut Block {
+ &mut *self.data.as_ptr().cast::<Block>().add(subblock)
+ }
+
+ #[inline]
+ fn usize_len(&self) -> usize {
+ let (mut blocks, rem) = div_rem(self.length, BITS);
+ blocks += (rem > 0) as usize;
+ blocks
+ }
+
+ #[inline]
+ fn simd_block_len(&self) -> usize {
+ let (mut blocks, rem) = div_rem(self.length, SimdBlock::BITS);
+ blocks += (rem > 0) as usize;
+ blocks
+ }
+
+ #[inline]
+ fn batch_count_ones(blocks: impl IntoIterator<Item = Block>) -> usize {
+ blocks.into_iter().map(|x| x.count_ones() as usize).sum()
+ }
+
+ #[inline]
+ fn as_simd_slice(&self) -> &[SimdBlock] {
+ // SAFETY: The slice constructed is within bounds of the underlying allocation. This function
+ // is called with a read-only borrow so no other write can happen as long as the returned borrow lives.
+ unsafe { core::slice::from_raw_parts(self.data.as_ptr().cast(), self.simd_block_len()) }
+ }
+
+ #[inline]
+ fn as_mut_simd_slice(&mut self) -> &mut [SimdBlock] {
+ // SAFETY: The slice constructed is within bounds of the underlying allocation. This function
+ // is called with a mutable borrow so no other read or write can happen as long as the returned borrow lives.
+ unsafe { core::slice::from_raw_parts_mut(self.data.as_ptr().cast(), self.simd_block_len()) }
+ }
+
+ #[inline]
+ fn as_simd_slice_uninit(&self) -> &[MaybeUninit<SimdBlock>] {
+ // SAFETY: The slice constructed is within bounds of the underlying allocation. This function
+ // is called with a read-only borrow so no other write can happen as long as the returned borrow lives.
+ unsafe { core::slice::from_raw_parts(self.data.as_ptr(), self.simd_block_len()) }
+ }
+
+ #[inline]
+ fn as_mut_simd_slice_uninit(&mut self) -> &mut [MaybeUninit<SimdBlock>] {
+ // SAFETY: The slice constructed is within bounds of the underlying allocation. This function
+ // is called with a mutable borrow so no other read or write can happen as long as the returned borrow lives.
+ unsafe { core::slice::from_raw_parts_mut(self.data.as_ptr(), self.simd_block_len()) }
+ }
+
+ /// Grows the internal size of the bitset before inserting a bit
+ ///
+ /// Unlike `insert`, this cannot panic, but may allocate if the bit is outside of the existing buffer's range.
+ ///
+ /// This is faster than calling `grow` then `insert` in succession.
+ #[inline]
+ pub fn grow_and_insert(&mut self, bits: usize) {
+ self.grow(bits + 1);
+
+ let (blocks, rem) = div_rem(bits, BITS);
+ // SAFETY: The above grow ensures that the block is inside the Vec's allocation.
+ unsafe {
+ *self.get_unchecked_mut(blocks) |= 1 << rem;
+ }
+ }
+
+ /// The length of the [`FixedBitSet`] in bits.
+ ///
+ /// Note: `len` includes both set and unset bits.
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let bitset = FixedBitSet::with_capacity(10);
+ /// // there are 0 set bits, but 10 unset bits
+ /// assert_eq!(bitset.len(), 10);
+ /// ```
+ /// `len` does not return the count of set bits. For that, use
+ /// [`bitset.count_ones(..)`](FixedBitSet::count_ones) instead.
+ #[inline]
+ pub fn len(&self) -> usize {
+ self.length
+ }
+
+ /// `true` if the [`FixedBitSet`] is empty.
+ ///
+ /// Note that an "empty" `FixedBitSet` is a `FixedBitSet` with
+ /// no bits (meaning: it's length is zero). If you want to check
+ /// if all bits are unset, use [`FixedBitSet::is_clear`].
+ ///
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let bitset = FixedBitSet::with_capacity(10);
+ /// assert!(!bitset.is_empty());
+ ///
+ /// let bitset = FixedBitSet::with_capacity(0);
+ /// assert!(bitset.is_empty());
+ /// ```
+ #[inline]
+ pub fn is_empty(&self) -> bool {
+ self.len() == 0
+ }
+
+ /// `true` if all bits in the [`FixedBitSet`] are unset.
+ ///
+ /// As opposed to [`FixedBitSet::is_empty`], which is `true` only for
+ /// sets without any bits, set or unset.
+ ///
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let mut bitset = FixedBitSet::with_capacity(10);
+ /// assert!(bitset.is_clear());
+ ///
+ /// bitset.insert(2);
+ /// assert!(!bitset.is_clear());
+ /// ```
+ ///
+ /// This is equivalent to [`bitset.count_ones(..) == 0`](FixedBitSet::count_ones).
+ #[inline]
+ pub fn is_clear(&self) -> bool {
+ self.as_simd_slice().iter().all(|block| block.is_empty())
+ }
+
+ /// Finds the lowest set bit in the bitset.
+ ///
+ /// Returns `None` if there aren't any set bits.
+ ///
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let mut bitset = FixedBitSet::with_capacity(10);
+ /// assert_eq!(bitset.minimum(), None);
+ ///
+ /// bitset.insert(2);
+ /// assert_eq!(bitset.minimum(), Some(2));
+ /// bitset.insert(8);
+ /// assert_eq!(bitset.minimum(), Some(2));
+ /// ```
+ #[inline]
+ pub fn minimum(&self) -> Option<usize> {
+ let (block_idx, block) = self
+ .as_simd_slice()
+ .iter()
+ .enumerate()
+ .find(|&(_, block)| !block.is_empty())?;
+ let mut inner = 0;
+ let mut trailing = 0;
+ for subblock in block.into_usize_array() {
+ if subblock != 0 {
+ trailing = subblock.trailing_zeros() as usize;
+ break;
+ } else {
+ inner += BITS;
+ }
+ }
+ Some(block_idx * SimdBlock::BITS + inner + trailing)
+ }
+
+ /// Finds the highest set bit in the bitset.
+ ///
+ /// Returns `None` if there aren't any set bits.
+ ///
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let mut bitset = FixedBitSet::with_capacity(10);
+ /// assert_eq!(bitset.maximum(), None);
+ ///
+ /// bitset.insert(8);
+ /// assert_eq!(bitset.maximum(), Some(8));
+ /// bitset.insert(2);
+ /// assert_eq!(bitset.maximum(), Some(8));
+ /// ```
+ #[inline]
+ pub fn maximum(&self) -> Option<usize> {
+ let (block_idx, block) = self
+ .as_simd_slice()
+ .iter()
+ .rev()
+ .enumerate()
+ .find(|&(_, block)| !block.is_empty())?;
+ let mut inner = 0;
+ let mut leading = 0;
+ for subblock in block.into_usize_array().iter().rev() {
+ if *subblock != 0 {
+ leading = subblock.leading_zeros() as usize;
+ break;
+ } else {
+ inner += BITS;
+ }
+ }
+ let max = self.simd_block_len() * SimdBlock::BITS;
+ Some(max - block_idx * SimdBlock::BITS - inner - leading - 1)
+ }
+
+ /// `true` if all bits in the [`FixedBitSet`] are set.
+ ///
+ /// ```
+ /// # use fixedbitset::FixedBitSet;
+ /// let mut bitset = FixedBitSet::with_capacity(10);
+ /// assert!(!bitset.is_full());
+ ///
+ /// bitset.insert_range(..);
+ /// assert!(bitset.is_full());
+ /// ```
+ ///
+ /// This is equivalent to [`bitset.count_ones(..) == bitset.len()`](FixedBitSet::count_ones).
+ #[inline]
+ pub fn is_full(&self) -> bool {
+ self.contains_all_in_range(..)
+ }
+
+ /// Return **true** if the bit is enabled in the **FixedBitSet**,
+ /// **false** otherwise.
+ ///
+ /// Note: bits outside the capacity are always disabled.
+ ///
+ /// Note: Also available with index syntax: `bitset[bit]`.
+ #[inline]
+ pub fn contains(&self, bit: usize) -> bool {
+ (bit < self.length)
+ // SAFETY: The above check ensures that the block and bit are within bounds.
+ .then(|| unsafe { self.contains_unchecked(bit) })
+ .unwrap_or(false)
+ }
+
+ /// Return **true** if the bit is enabled in the **FixedBitSet**,
+ /// **false** otherwise.
+ ///
+ /// Note: unlike `contains`, calling this with an invalid `bit`
+ /// is undefined behavior.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn contains_unchecked(&self, bit: usize) -> bool {
+ let (block, i) = div_rem(bit, BITS);
+ (self.get_unchecked(block) & (1 << i)) != 0
+ }
+
+ /// Clear all bits.
+ #[inline]
+ pub fn clear(&mut self) {
+ for elt in self.as_mut_simd_slice().iter_mut() {
+ *elt = SimdBlock::NONE
+ }
+ }
+
+ /// Enable `bit`.
+ ///
+ /// **Panics** if **bit** is out of bounds.
+ #[inline]
+ pub fn insert(&mut self, bit: usize) {
+ assert!(
+ bit < self.length,
+ "insert at index {} exceeds fixedbitset size {}",
+ bit,
+ self.length
+ );
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ self.insert_unchecked(bit);
+ }
+ }
+
+ /// Enable `bit` without any length checks.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn insert_unchecked(&mut self, bit: usize) {
+ let (block, i) = div_rem(bit, BITS);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ *self.get_unchecked_mut(block) |= 1 << i;
+ }
+ }
+
+ /// Disable `bit`.
+ ///
+ /// **Panics** if **bit** is out of bounds.
+ #[inline]
+ pub fn remove(&mut self, bit: usize) {
+ assert!(
+ bit < self.length,
+ "remove at index {} exceeds fixedbitset size {}",
+ bit,
+ self.length
+ );
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ self.remove_unchecked(bit);
+ }
+ }
+
+ /// Disable `bit` without any bounds checking.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn remove_unchecked(&mut self, bit: usize) {
+ let (block, i) = div_rem(bit, BITS);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ *self.get_unchecked_mut(block) &= !(1 << i);
+ }
+ }
+
+ /// Enable `bit`, and return its previous value.
+ ///
+ /// **Panics** if **bit** is out of bounds.
+ #[inline]
+ pub fn put(&mut self, bit: usize) -> bool {
+ assert!(
+ bit < self.length,
+ "put at index {} exceeds fixedbitset size {}",
+ bit,
+ self.length
+ );
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe { self.put_unchecked(bit) }
+ }
+
+ /// Enable `bit`, and return its previous value without doing any bounds checking.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn put_unchecked(&mut self, bit: usize) -> bool {
+ let (block, i) = div_rem(bit, BITS);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ let word = self.get_unchecked_mut(block);
+ let prev = *word & (1 << i) != 0;
+ *word |= 1 << i;
+ prev
+ }
+ }
+
+ /// Toggle `bit` (inverting its state).
+ ///
+ /// ***Panics*** if **bit** is out of bounds
+ #[inline]
+ pub fn toggle(&mut self, bit: usize) {
+ assert!(
+ bit < self.length,
+ "toggle at index {} exceeds fixedbitset size {}",
+ bit,
+ self.length
+ );
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ self.toggle_unchecked(bit);
+ }
+ }
+
+ /// Toggle `bit` (inverting its state) without any bounds checking.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn toggle_unchecked(&mut self, bit: usize) {
+ let (block, i) = div_rem(bit, BITS);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ *self.get_unchecked_mut(block) ^= 1 << i;
+ }
+ }
+
+ /// Sets a bit to the provided `enabled` value.
+ ///
+ /// **Panics** if **bit** is out of bounds.
+ #[inline]
+ pub fn set(&mut self, bit: usize, enabled: bool) {
+ assert!(
+ bit < self.length,
+ "set at index {} exceeds fixedbitset size {}",
+ bit,
+ self.length
+ );
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe {
+ self.set_unchecked(bit, enabled);
+ }
+ }
+
+ /// Sets a bit to the provided `enabled` value without doing any bounds checking.
+ ///
+ /// # Safety
+ /// `bit` must be less than `self.len()`
+ #[inline]
+ pub unsafe fn set_unchecked(&mut self, bit: usize, enabled: bool) {
+ let (block, i) = div_rem(bit, BITS);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ let elt = unsafe { self.get_unchecked_mut(block) };
+ if enabled {
+ *elt |= 1 << i;
+ } else {
+ *elt &= !(1 << i);
+ }
+ }
+
+ /// Copies boolean value from specified bit to the specified bit.
+ ///
+ /// If `from` is out-of-bounds, `to` will be unset.
+ ///
+ /// **Panics** if **to** is out of bounds.
+ #[inline]
+ pub fn copy_bit(&mut self, from: usize, to: usize) {
+ assert!(
+ to < self.length,
+ "copy to index {} exceeds fixedbitset size {}",
+ to,
+ self.length
+ );
+ let enabled = self.contains(from);
+ // SAFETY: The above assertion ensures that the block is inside the Vec's allocation.
+ unsafe { self.set_unchecked(to, enabled) };
+ }
+
+ /// Copies boolean value from specified bit to the specified bit.
+ ///
+ /// Note: unlike `copy_bit`, calling this with an invalid `from`
+ /// is undefined behavior.
+ ///
+ /// # Safety
+ /// `to` must both be less than `self.len()`
+ #[inline]
+ pub unsafe fn copy_bit_unchecked(&mut self, from: usize, to: usize) {
+ // SAFETY: Caller must ensure that `from` is within bounds.
+ let enabled = self.contains_unchecked(from);
+ // SAFETY: Caller must ensure that `to` is within bounds.
+ self.set_unchecked(to, enabled);
+ }
+
+ /// Count the number of set bits in the given bit range.
+ ///
+ /// This function is potentially much faster than using `ones(other).count()`.
+ /// Use `..` to count the whole content of the bitset.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn count_ones<T: IndexRange>(&self, range: T) -> usize {
+ Self::batch_count_ones(Masks::new(range, self.length).map(|(block, mask)| {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ unsafe { *self.get_unchecked(block) & mask }
+ }))
+ }
+
+ /// Count the number of unset bits in the given bit range.
+ ///
+ /// This function is potentially much faster than using `zeroes(other).count()`.
+ /// Use `..` to count the whole content of the bitset.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn count_zeroes<T: IndexRange>(&self, range: T) -> usize {
+ Self::batch_count_ones(Masks::new(range, self.length).map(|(block, mask)| {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ unsafe { !*self.get_unchecked(block) & mask }
+ }))
+ }
+
+ /// Sets every bit in the given range to the given state (`enabled`)
+ ///
+ /// Use `..` to set the whole bitset.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn set_range<T: IndexRange>(&mut self, range: T, enabled: bool) {
+ if enabled {
+ self.insert_range(range);
+ } else {
+ self.remove_range(range);
+ }
+ }
+
+ /// Enables every bit in the given range.
+ ///
+ /// Use `..` to make the whole bitset ones.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn insert_range<T: IndexRange>(&mut self, range: T) {
+ for (block, mask) in Masks::new(range, self.length) {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ let block = unsafe { self.get_unchecked_mut(block) };
+ *block |= mask;
+ }
+ }
+
+ /// Disables every bit in the given range.
+ ///
+ /// Use `..` to make the whole bitset ones.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn remove_range<T: IndexRange>(&mut self, range: T) {
+ for (block, mask) in Masks::new(range, self.length) {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ let block = unsafe { self.get_unchecked_mut(block) };
+ *block &= !mask;
+ }
+ }
+
+ /// Toggles (inverts) every bit in the given range.
+ ///
+ /// Use `..` to toggle the whole bitset.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn toggle_range<T: IndexRange>(&mut self, range: T) {
+ for (block, mask) in Masks::new(range, self.length) {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ let block = unsafe { self.get_unchecked_mut(block) };
+ *block ^= mask;
+ }
+ }
+
+ /// Checks if the bitset contains every bit in the given range.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn contains_all_in_range<T: IndexRange>(&self, range: T) -> bool {
+ for (block, mask) in Masks::new(range, self.length) {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ let block = unsafe { self.get_unchecked(block) };
+ if block & mask != mask {
+ return false;
+ }
+ }
+ true
+ }
+
+ /// Checks if the bitset contains at least one set bit in the given range.
+ ///
+ /// **Panics** if the range extends past the end of the bitset.
+ #[inline]
+ pub fn contains_any_in_range<T: IndexRange>(&self, range: T) -> bool {
+ for (block, mask) in Masks::new(range, self.length) {
+ // SAFETY: Masks cannot return a block index that is out of range.
+ let block = unsafe { self.get_unchecked(block) };
+ if block & mask != 0 {
+ return true;
+ }
+ }
+ false
+ }
+
+ /// View the bitset as a slice of `Block` blocks
+ #[inline]
+ pub fn as_slice(&self) -> &[Block] {
+ // SAFETY: The bits from both usize and Block are required to be reinterprettable, and
+ // neither have any padding or alignment issues. The slice constructed is within bounds
+ // of the underlying allocation. This function is called with a read-only borrow so
+ // no other write can happen as long as the returned borrow lives.
+ unsafe {
+ let ptr = self.data.as_ptr().cast::<Block>();
+ core::slice::from_raw_parts(ptr, self.usize_len())
+ }
+ }
+
+ /// View the bitset as a mutable slice of `Block` blocks. Writing past the bitlength in the last
+ /// will cause `contains` to return potentially incorrect results for bits past the bitlength.
+ #[inline]
+ pub fn as_mut_slice(&mut self) -> &mut [Block] {
+ // SAFETY: The bits from both usize and Block are required to be reinterprettable, and
+ // neither have any padding or alignment issues. The slice constructed is within bounds
+ // of the underlying allocation. This function is called with a mutable borrow so
+ // no other read or write can happen as long as the returned borrow lives.
+ unsafe {
+ let ptr = self.data.as_ptr().cast::<Block>();
+ core::slice::from_raw_parts_mut(ptr, self.usize_len())
+ }
+ }
+
+ /// Iterates over all enabled bits.
+ ///
+ /// Iterator element is the index of the `1` bit, type `usize`.
+ #[inline]
+ pub fn ones(&self) -> Ones {
+ match self.as_slice().split_first() {
+ Some((&first_block, rem)) => {
+ let (&last_block, rem) = rem.split_last().unwrap_or((&0, rem));
+ Ones {
+ bitset_front: first_block,
+ bitset_back: last_block,
+ block_idx_front: 0,
+ block_idx_back: (1 + rem.len()) * BITS,
+ remaining_blocks: rem.iter(),
+ }
+ }
+ None => Ones {
+ bitset_front: 0,
+ bitset_back: 0,
+ block_idx_front: 0,
+ block_idx_back: 0,
+ remaining_blocks: [].iter(),
+ },
+ }
+ }
+
+ /// Iterates over all enabled bits.
+ ///
+ /// Iterator element is the index of the `1` bit, type `usize`.
+ /// Unlike `ones`, this function consumes the `FixedBitset`.
+ pub fn into_ones(self) -> IntoOnes {
+ let ptr = self.data.as_ptr().cast();
+ let len = self.simd_block_len() * SimdBlock::USIZE_COUNT;
+ // SAFETY:
+ // - ptr comes from self.data, so it is valid;
+ // - self.data is valid for self.data.len() SimdBlocks,
+ // which is exactly self.data.len() * SimdBlock::USIZE_COUNT usizes;
+ // - we will keep this slice around only as long as self.data is,
+ // so it won't become dangling.
+ let slice = unsafe { core::slice::from_raw_parts(ptr, len) };
+ // SAFETY: The data pointer and capacity were created from a Vec initially. The block
+ // len is identical to that of the original.
+ let data: Vec<SimdBlock> = unsafe {
+ Vec::from_raw_parts(
+ self.data.as_ptr().cast(),
+ self.simd_block_len(),
+ self.capacity,
+ )
+ };
+ let mut iter = slice.iter().copied();
+
+ core::mem::forget(self);
+
+ IntoOnes {
+ bitset_front: iter.next().unwrap_or(0),
+ bitset_back: iter.next_back().unwrap_or(0),
+ block_idx_front: 0,
+ block_idx_back: len.saturating_sub(1) * BITS,
+ remaining_blocks: iter,
+ _buf: data,
+ }
+ }
+
+ /// Iterates over all disabled bits.
+ ///
+ /// Iterator element is the index of the `0` bit, type `usize`.
+ #[inline]
+ pub fn zeroes(&self) -> Zeroes {
+ match self.as_slice().split_first() {
+ Some((&block, rem)) => Zeroes {
+ bitset: !block,
+ block_idx: 0,
+ len: self.len(),
+ remaining_blocks: rem.iter(),
+ },
+ None => Zeroes {
+ bitset: !0,
+ block_idx: 0,
+ len: self.len(),
+ remaining_blocks: [].iter(),
+ },
+ }
+ }
+
+ /// Returns a lazy iterator over the intersection of two `FixedBitSet`s
+ pub fn intersection<'a>(&'a self, other: &'a FixedBitSet) -> Intersection<'a> {
+ Intersection {
+ iter: self.ones(),
+ other,
+ }
+ }
+
+ /// Returns a lazy iterator over the union of two `FixedBitSet`s.
+ pub fn union<'a>(&'a self, other: &'a FixedBitSet) -> Union<'a> {
+ Union {
+ iter: self.ones().chain(other.difference(self)),
+ }
+ }
+
+ /// Returns a lazy iterator over the difference of two `FixedBitSet`s. The difference of `a`
+ /// and `b` is the elements of `a` which are not in `b`.
+ pub fn difference<'a>(&'a self, other: &'a FixedBitSet) -> Difference<'a> {
+ Difference {
+ iter: self.ones(),
+ other,
+ }
+ }
+
+ /// Returns a lazy iterator over the symmetric difference of two `FixedBitSet`s.
+ /// The symmetric difference of `a` and `b` is the elements of one, but not both, sets.
+ pub fn symmetric_difference<'a>(&'a self, other: &'a FixedBitSet) -> SymmetricDifference<'a> {
+ SymmetricDifference {
+ iter: self.difference(other).chain(other.difference(self)),
+ }
+ }
+
+ /// In-place union of two `FixedBitSet`s.
+ ///
+ /// On calling this method, `self`'s capacity may be increased to match `other`'s.
+ pub fn union_with(&mut self, other: &FixedBitSet) {
+ if other.len() >= self.len() {
+ self.grow(other.len());
+ }
+ self.as_mut_simd_slice()
+ .iter_mut()
+ .zip(other.as_simd_slice().iter())
+ .for_each(|(x, y)| *x |= *y);
+ }
+
+ /// In-place intersection of two `FixedBitSet`s.
+ ///
+ /// On calling this method, `self`'s capacity will remain the same as before.
+ pub fn intersect_with(&mut self, other: &FixedBitSet) {
+ let me = self.as_mut_simd_slice();
+ let other = other.as_simd_slice();
+ me.iter_mut().zip(other.iter()).for_each(|(x, y)| {
+ *x &= *y;
+ });
+ let mn = core::cmp::min(me.len(), other.len());
+ for wd in &mut me[mn..] {
+ *wd = SimdBlock::NONE;
+ }
+ }
+
+ /// In-place difference of two `FixedBitSet`s.
+ ///
+ /// On calling this method, `self`'s capacity will remain the same as before.
+ pub fn difference_with(&mut self, other: &FixedBitSet) {
+ self.as_mut_simd_slice()
+ .iter_mut()
+ .zip(other.as_simd_slice().iter())
+ .for_each(|(x, y)| {
+ *x &= !*y;
+ });
+
+ // There's no need to grow self or do any other adjustments.
+ //
+ // * If self is longer than other, the bits at the end of self won't be affected since other
+ // has them implicitly set to 0.
+ // * If other is longer than self, the bits at the end of other are irrelevant since self
+ // has them set to 0 anyway.
+ }
+
+ /// In-place symmetric difference of two `FixedBitSet`s.
+ ///
+ /// On calling this method, `self`'s capacity may be increased to match `other`'s.
+ pub fn symmetric_difference_with(&mut self, other: &FixedBitSet) {
+ if other.len() >= self.len() {
+ self.grow(other.len());
+ }
+ self.as_mut_simd_slice()
+ .iter_mut()
+ .zip(other.as_simd_slice().iter())
+ .for_each(|(x, y)| {
+ *x ^= *y;
+ });
+ }
+
+ /// Computes how many bits would be set in the union between two bitsets.
+ ///
+ /// This is potentially much faster than using `union(other).count()`. Unlike
+ /// other methods like using [`union_with`] followed by [`count_ones`], this
+ /// does not mutate in place or require separate allocations.
+ #[inline]
+ pub fn union_count(&self, other: &FixedBitSet) -> usize {
+ let me = self.as_slice();
+ let other = other.as_slice();
+ let count = Self::batch_count_ones(me.iter().zip(other.iter()).map(|(x, y)| (*x | *y)));
+ match other.len().cmp(&me.len()) {
+ Ordering::Greater => count + Self::batch_count_ones(other[me.len()..].iter().copied()),
+ Ordering::Less => count + Self::batch_count_ones(me[other.len()..].iter().copied()),
+ Ordering::Equal => count,
+ }
+ }
+
+ /// Computes how many bits would be set in the intersection between two bitsets.
+ ///
+ /// This is potentially much faster than using `intersection(other).count()`. Unlike
+ /// other methods like using [`intersect_with`] followed by [`count_ones`], this
+ /// does not mutate in place or require separate allocations.
+ #[inline]
+ pub fn intersection_count(&self, other: &FixedBitSet) -> usize {
+ Self::batch_count_ones(
+ self.as_slice()
+ .iter()
+ .zip(other.as_slice())
+ .map(|(x, y)| (*x & *y)),
+ )
+ }
+
+ /// Computes how many bits would be set in the difference between two bitsets.
+ ///
+ /// This is potentially much faster than using `difference(other).count()`. Unlike
+ /// other methods like using [`difference_with`] followed by [`count_ones`], this
+ /// does not mutate in place or require separate allocations.
+ #[inline]
+ pub fn difference_count(&self, other: &FixedBitSet) -> usize {
+ Self::batch_count_ones(
+ self.as_slice()
+ .iter()
+ .zip(other.as_slice().iter())
+ .map(|(x, y)| (*x & !*y)),
+ )
+ }
+
+ /// Computes how many bits would be set in the symmetric difference between two bitsets.
+ ///
+ /// This is potentially much faster than using `symmetric_difference(other).count()`. Unlike
+ /// other methods like using [`symmetric_difference_with`] followed by [`count_ones`], this
+ /// does not mutate in place or require separate allocations.
+ #[inline]
+ pub fn symmetric_difference_count(&self, other: &FixedBitSet) -> usize {
+ let me = self.as_slice();
+ let other = other.as_slice();
+ let count = Self::batch_count_ones(me.iter().zip(other.iter()).map(|(x, y)| (*x ^ *y)));
+ match other.len().cmp(&me.len()) {
+ Ordering::Greater => count + Self::batch_count_ones(other[me.len()..].iter().copied()),
+ Ordering::Less => count + Self::batch_count_ones(me[other.len()..].iter().copied()),
+ Ordering::Equal => count,
+ }
+ }
+
+ /// Returns `true` if `self` has no elements in common with `other`. This
+ /// is equivalent to checking for an empty intersection.
+ pub fn is_disjoint(&self, other: &FixedBitSet) -> bool {
+ self.as_simd_slice()
+ .iter()
+ .zip(other.as_simd_slice())
+ .all(|(x, y)| (*x & *y).is_empty())
+ }
+
+ /// Returns `true` if the set is a subset of another, i.e. `other` contains
+ /// at least all the values in `self`.
+ pub fn is_subset(&self, other: &FixedBitSet) -> bool {
+ let me = self.as_simd_slice();
+ let other = other.as_simd_slice();
+ me.iter()
+ .zip(other.iter())
+ .all(|(x, y)| x.andnot(*y).is_empty())
+ && me.iter().skip(other.len()).all(|x| x.is_empty())
+ }
+
+ /// Returns `true` if the set is a superset of another, i.e. `self` contains
+ /// at least all the values in `other`.
+ pub fn is_superset(&self, other: &FixedBitSet) -> bool {
+ other.is_subset(self)
+ }
+}
+
+impl Hash for FixedBitSet {
+ fn hash<H: core::hash::Hasher>(&self, state: &mut H) {
+ self.length.hash(state);
+ self.as_simd_slice().hash(state);
+ }
+}
+
+impl PartialEq for FixedBitSet {
+ fn eq(&self, other: &Self) -> bool {
+ self.length == other.length && self.as_simd_slice().eq(other.as_simd_slice())
+ }
+}
+
+impl PartialOrd for FixedBitSet {
+ fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
+ Some(self.cmp(other))
+ }
+}
+
+impl Ord for FixedBitSet {
+ fn cmp(&self, other: &Self) -> Ordering {
+ self.length
+ .cmp(&other.length)
+ .then_with(|| self.as_simd_slice().cmp(other.as_simd_slice()))
+ }
+}
+
+impl Default for FixedBitSet {
+ fn default() -> Self {
+ Self::new()
+ }
+}
+
+impl Drop for FixedBitSet {
+ fn drop(&mut self) {
+ // SAFETY: The data pointer and capacity were created from a Vec initially. The block
+ // len is identical to that of the original.
+ drop(unsafe {
+ Vec::from_raw_parts(self.data.as_ptr(), self.simd_block_len(), self.capacity)
+ });
+ }
+}
+
+impl Binary for FixedBitSet {
+ fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error> {
+ if f.alternate() {
+ f.write_str("0b")?;
+ }
+
+ for i in 0..self.length {
+ if self[i] {
+ f.write_char('1')?;
+ } else {
+ f.write_char('0')?;
+ }
+ }
+
+ Ok(())
+ }
+}
+
+impl Display for FixedBitSet {
+ fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error> {
+ Binary::fmt(&self, f)
+ }
+}
+
+/// An iterator producing elements in the difference of two sets.
+///
+/// This struct is created by the [`FixedBitSet::difference`] method.
+pub struct Difference<'a> {
+ iter: Ones<'a>,
+ other: &'a FixedBitSet,
+}
+
+impl<'a> Iterator for Difference<'a> {
+ type Item = usize;
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ self.iter.by_ref().find(|&nxt| !self.other.contains(nxt))
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ self.iter.size_hint()
+ }
+}
+
+impl<'a> DoubleEndedIterator for Difference<'a> {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ self.iter
+ .by_ref()
+ .rev()
+ .find(|&nxt| !self.other.contains(nxt))
+ }
+}
+
+// Difference will continue to return None once it first returns None.
+impl<'a> FusedIterator for Difference<'a> {}
+
+/// An iterator producing elements in the symmetric difference of two sets.
+///
+/// This struct is created by the [`FixedBitSet::symmetric_difference`] method.
+pub struct SymmetricDifference<'a> {
+ iter: Chain<Difference<'a>, Difference<'a>>,
+}
+
+impl<'a> Iterator for SymmetricDifference<'a> {
+ type Item = usize;
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ self.iter.next()
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ self.iter.size_hint()
+ }
+}
+
+impl<'a> DoubleEndedIterator for SymmetricDifference<'a> {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ self.iter.next_back()
+ }
+}
+
+// SymmetricDifference will continue to return None once it first returns None.
+impl<'a> FusedIterator for SymmetricDifference<'a> {}
+
+/// An iterator producing elements in the intersection of two sets.
+///
+/// This struct is created by the [`FixedBitSet::intersection`] method.
+pub struct Intersection<'a> {
+ iter: Ones<'a>,
+ other: &'a FixedBitSet,
+}
+
+impl<'a> Iterator for Intersection<'a> {
+ type Item = usize; // the bit position of the '1'
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ self.iter.by_ref().find(|&nxt| self.other.contains(nxt))
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ self.iter.size_hint()
+ }
+}
+
+impl<'a> DoubleEndedIterator for Intersection<'a> {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ self.iter
+ .by_ref()
+ .rev()
+ .find(|&nxt| self.other.contains(nxt))
+ }
+}
+
+// Intersection will continue to return None once it first returns None.
+impl<'a> FusedIterator for Intersection<'a> {}
+
+/// An iterator producing elements in the union of two sets.
+///
+/// This struct is created by the [`FixedBitSet::union`] method.
+pub struct Union<'a> {
+ iter: Chain<Ones<'a>, Difference<'a>>,
+}
+
+impl<'a> Iterator for Union<'a> {
+ type Item = usize;
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ self.iter.next()
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ self.iter.size_hint()
+ }
+}
+
+impl<'a> DoubleEndedIterator for Union<'a> {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ self.iter.next_back()
+ }
+}
+
+// Union will continue to return None once it first returns None.
+impl<'a> FusedIterator for Union<'a> {}
+
+struct Masks {
+ first_block: usize,
+ first_mask: usize,
+ last_block: usize,
+ last_mask: usize,
+}
+
+impl Masks {
+ #[inline]
+ fn new<T: IndexRange>(range: T, length: usize) -> Masks {
+ let start = range.start().unwrap_or(0);
+ let end = range.end().unwrap_or(length);
+ assert!(
+ start <= end && end <= length,
+ "invalid range {}..{} for a fixedbitset of size {}",
+ start,
+ end,
+ length
+ );
+
+ let (first_block, first_rem) = div_rem(start, BITS);
+ let (last_block, last_rem) = div_rem(end, BITS);
+
+ Masks {
+ first_block,
+ first_mask: usize::MAX << first_rem,
+ last_block,
+ last_mask: (usize::MAX >> 1) >> (BITS - last_rem - 1),
+ // this is equivalent to `MAX >> (BITS - x)` with correct semantics when x == 0.
+ }
+ }
+}
+
+impl Iterator for Masks {
+ type Item = (usize, usize);
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ match self.first_block.cmp(&self.last_block) {
+ Ordering::Less => {
+ let res = (self.first_block, self.first_mask);
+ self.first_block += 1;
+ self.first_mask = !0;
+ Some(res)
+ }
+ Ordering::Equal => {
+ let mask = self.first_mask & self.last_mask;
+ let res = if mask == 0 {
+ None
+ } else {
+ Some((self.first_block, mask))
+ };
+ self.first_block += 1;
+ res
+ }
+ Ordering::Greater => None,
+ }
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ (self.first_block..=self.last_block).size_hint()
+ }
+}
+
+// Masks will continue to return None once it first returns None.
+impl FusedIterator for Masks {}
+
+// Masks's size_hint implementation is exact. It never returns an
+// unbounded value and always returns an exact number of values.
+impl ExactSizeIterator for Masks {}
+
+/// An iterator producing the indices of the set bit in a set.
+///
+/// This struct is created by the [`FixedBitSet::ones`] method.
+pub struct Ones<'a> {
+ bitset_front: usize,
+ bitset_back: usize,
+ block_idx_front: usize,
+ block_idx_back: usize,
+ remaining_blocks: core::slice::Iter<'a, usize>,
+}
+
+impl<'a> Ones<'a> {
+ #[inline]
+ pub fn last_positive_bit_and_unset(n: &mut usize) -> usize {
+ // Find the last set bit using x & -x
+ let last_bit = *n & n.wrapping_neg();
+
+ // Find the position of the last set bit
+ let position = last_bit.trailing_zeros();
+
+ // Unset the last set bit
+ *n &= *n - 1;
+
+ position as usize
+ }
+
+ #[inline]
+ fn first_positive_bit_and_unset(n: &mut usize) -> usize {
+ /* Identify the first non zero bit */
+ let bit_idx = n.leading_zeros();
+
+ /* set that bit to zero */
+ let mask = !((1_usize) << (BITS as u32 - bit_idx - 1));
+ n.bitand_assign(mask);
+
+ bit_idx as usize
+ }
+}
+
+impl<'a> DoubleEndedIterator for Ones<'a> {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ while self.bitset_back == 0 {
+ match self.remaining_blocks.next_back() {
+ None => {
+ if self.bitset_front != 0 {
+ self.bitset_back = 0;
+ self.block_idx_back = self.block_idx_front;
+ return Some(
+ self.block_idx_front + BITS
+ - Self::first_positive_bit_and_unset(&mut self.bitset_front)
+ - 1,
+ );
+ } else {
+ return None;
+ }
+ }
+ Some(next_block) => {
+ self.bitset_back = *next_block;
+ self.block_idx_back -= BITS;
+ }
+ };
+ }
+
+ Some(
+ self.block_idx_back - Self::first_positive_bit_and_unset(&mut self.bitset_back) + BITS
+ - 1,
+ )
+ }
+}
+
+impl<'a> Iterator for Ones<'a> {
+ type Item = usize; // the bit position of the '1'
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ while self.bitset_front == 0 {
+ match self.remaining_blocks.next() {
+ Some(next_block) => {
+ self.bitset_front = *next_block;
+ self.block_idx_front += BITS;
+ }
+ None => {
+ if self.bitset_back != 0 {
+ // not needed for iteration, but for size_hint
+ self.block_idx_front = self.block_idx_back;
+ self.bitset_front = 0;
+
+ return Some(
+ self.block_idx_back
+ + Self::last_positive_bit_and_unset(&mut self.bitset_back),
+ );
+ } else {
+ return None;
+ }
+ }
+ };
+ }
+
+ Some(self.block_idx_front + Self::last_positive_bit_and_unset(&mut self.bitset_front))
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ (
+ 0,
+ (Some(self.block_idx_back - self.block_idx_front + 2 * BITS)),
+ )
+ }
+}
+
+// Ones will continue to return None once it first returns None.
+impl<'a> FusedIterator for Ones<'a> {}
+
+/// An iterator producing the indices of the set bit in a set.
+///
+/// This struct is created by the [`FixedBitSet::ones`] method.
+pub struct Zeroes<'a> {
+ bitset: usize,
+ block_idx: usize,
+ len: usize,
+ remaining_blocks: core::slice::Iter<'a, usize>,
+}
+
+impl<'a> Iterator for Zeroes<'a> {
+ type Item = usize; // the bit position of the '1'
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ while self.bitset == 0 {
+ self.bitset = !*self.remaining_blocks.next()?;
+ self.block_idx += BITS;
+ }
+ let t = self.bitset & (0_usize).wrapping_sub(self.bitset);
+ let r = self.bitset.trailing_zeros() as usize;
+ self.bitset ^= t;
+ let bit = self.block_idx + r;
+ // The remaining zeroes beyond the length of the bitset must be excluded.
+ if bit < self.len {
+ Some(bit)
+ } else {
+ None
+ }
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ (0, Some(self.len))
+ }
+}
+
+// Zeroes will stop returning Some when exhausted.
+impl<'a> FusedIterator for Zeroes<'a> {}
+
+impl Clone for FixedBitSet {
+ #[inline]
+ fn clone(&self) -> Self {
+ Self::from_blocks_and_len(Vec::from(self.as_simd_slice()), self.length)
+ }
+
+ #[inline]
+ fn clone_from(&mut self, source: &Self) {
+ if self.length < source.length {
+ // SAFETY: `fill` is uninitialized, but is immediately initialized from `source`.
+ unsafe { self.grow_inner(source.length, MaybeUninit::uninit()) };
+ }
+ let me = self.as_mut_simd_slice_uninit();
+ let them = source.as_simd_slice_uninit();
+ match me.len().cmp(&them.len()) {
+ Ordering::Greater => {
+ let (head, tail) = me.split_at_mut(them.len());
+ head.copy_from_slice(them);
+ tail.fill(MaybeUninit::new(SimdBlock::NONE));
+ }
+ Ordering::Equal => me.copy_from_slice(them),
+ // The grow_inner above ensures that self is at least as large as source.
+ // so this branch is unreachable.
+ Ordering::Less => {}
+ }
+ self.length = source.length;
+ }
+}
+
+/// Return **true** if the bit is enabled in the bitset,
+/// or **false** otherwise.
+///
+/// Note: bits outside the capacity are always disabled, and thus
+/// indexing a FixedBitSet will not panic.
+impl Index<usize> for FixedBitSet {
+ type Output = bool;
+
+ #[inline]
+ fn index(&self, bit: usize) -> &bool {
+ if self.contains(bit) {
+ &true
+ } else {
+ &false
+ }
+ }
+}
+
+/// Sets the bit at index **i** to **true** for each item **i** in the input **src**.
+impl Extend<usize> for FixedBitSet {
+ fn extend<I: IntoIterator<Item = usize>>(&mut self, src: I) {
+ let iter = src.into_iter();
+ for i in iter {
+ if i >= self.len() {
+ self.grow(i + 1);
+ }
+ self.put(i);
+ }
+ }
+}
+
+/// Return a FixedBitSet containing bits set to **true** for every bit index in
+/// the iterator, other bits are set to **false**.
+impl FromIterator<usize> for FixedBitSet {
+ fn from_iter<I: IntoIterator<Item = usize>>(src: I) -> Self {
+ let mut fbs = FixedBitSet::with_capacity(0);
+ fbs.extend(src);
+ fbs
+ }
+}
+
+pub struct IntoOnes {
+ bitset_front: Block,
+ bitset_back: Block,
+ block_idx_front: usize,
+ block_idx_back: usize,
+ remaining_blocks: core::iter::Copied<core::slice::Iter<'static, usize>>,
+ // Keep buf along so that `remaining_blocks` remains valid.
+ _buf: Vec<SimdBlock>,
+}
+
+impl IntoOnes {
+ #[inline]
+ pub fn last_positive_bit_and_unset(n: &mut Block) -> usize {
+ // Find the last set bit using x & -x
+ let last_bit = *n & n.wrapping_neg();
+
+ // Find the position of the last set bit
+ let position = last_bit.trailing_zeros();
+
+ // Unset the last set bit
+ *n &= *n - 1;
+
+ position as usize
+ }
+
+ #[inline]
+ fn first_positive_bit_and_unset(n: &mut Block) -> usize {
+ /* Identify the first non zero bit */
+ let bit_idx = n.leading_zeros();
+
+ /* set that bit to zero */
+ let mask = !((1_usize) << (BITS as u32 - bit_idx - 1));
+ n.bitand_assign(mask);
+
+ bit_idx as usize
+ }
+}
+
+impl DoubleEndedIterator for IntoOnes {
+ fn next_back(&mut self) -> Option<Self::Item> {
+ while self.bitset_back == 0 {
+ match self.remaining_blocks.next_back() {
+ None => {
+ if self.bitset_front != 0 {
+ self.bitset_back = 0;
+ self.block_idx_back = self.block_idx_front;
+ return Some(
+ self.block_idx_front + BITS
+ - Self::first_positive_bit_and_unset(&mut self.bitset_front)
+ - 1,
+ );
+ } else {
+ return None;
+ }
+ }
+ Some(next_block) => {
+ self.bitset_back = next_block;
+ self.block_idx_back -= BITS;
+ }
+ };
+ }
+
+ Some(
+ self.block_idx_back - Self::first_positive_bit_and_unset(&mut self.bitset_back) + BITS
+ - 1,
+ )
+ }
+}
+
+impl Iterator for IntoOnes {
+ type Item = usize; // the bit position of the '1'
+
+ #[inline]
+ fn next(&mut self) -> Option<Self::Item> {
+ while self.bitset_front == 0 {
+ match self.remaining_blocks.next() {
+ Some(next_block) => {
+ self.bitset_front = next_block;
+ self.block_idx_front += BITS;
+ }
+ None => {
+ if self.bitset_back != 0 {
+ // not needed for iteration, but for size_hint
+ self.block_idx_front = self.block_idx_back;
+ self.bitset_front = 0;
+
+ return Some(
+ self.block_idx_back
+ + Self::last_positive_bit_and_unset(&mut self.bitset_back),
+ );
+ } else {
+ return None;
+ }
+ }
+ };
+ }
+
+ Some(self.block_idx_front + Self::last_positive_bit_and_unset(&mut self.bitset_front))
+ }
+
+ #[inline]
+ fn size_hint(&self) -> (usize, Option<usize>) {
+ (
+ 0,
+ (Some(self.block_idx_back - self.block_idx_front + 2 * BITS)),
+ )
+ }
+}
+
+// Ones will continue to return None once it first returns None.
+impl FusedIterator for IntoOnes {}
+
+impl<'a> BitAnd for &'a FixedBitSet {
+ type Output = FixedBitSet;
+ fn bitand(self, other: &FixedBitSet) -> FixedBitSet {
+ let (short, long) = {
+ if self.len() <= other.len() {
+ (self.as_simd_slice(), other.as_simd_slice())
+ } else {
+ (other.as_simd_slice(), self.as_simd_slice())
+ }
+ };
+ let mut data = Vec::from(short);
+ for (data, block) in data.iter_mut().zip(long.iter()) {
+ *data &= *block;
+ }
+ let len = core::cmp::min(self.len(), other.len());
+ FixedBitSet::from_blocks_and_len(data, len)
+ }
+}
+
+impl BitAndAssign for FixedBitSet {
+ fn bitand_assign(&mut self, other: Self) {
+ self.intersect_with(&other);
+ }
+}
+
+impl BitAndAssign<&Self> for FixedBitSet {
+ fn bitand_assign(&mut self, other: &Self) {
+ self.intersect_with(other);
+ }
+}
+
+impl<'a> BitOr for &'a FixedBitSet {
+ type Output = FixedBitSet;
+ fn bitor(self, other: &FixedBitSet) -> FixedBitSet {
+ let (short, long) = {
+ if self.len() <= other.len() {
+ (self.as_simd_slice(), other.as_simd_slice())
+ } else {
+ (other.as_simd_slice(), self.as_simd_slice())
+ }
+ };
+ let mut data = Vec::from(long);
+ for (data, block) in data.iter_mut().zip(short.iter()) {
+ *data |= *block;
+ }
+ let len = core::cmp::max(self.len(), other.len());
+ FixedBitSet::from_blocks_and_len(data, len)
+ }
+}
+
+impl BitOrAssign for FixedBitSet {
+ fn bitor_assign(&mut self, other: Self) {
+ self.union_with(&other);
+ }
+}
+
+impl BitOrAssign<&Self> for FixedBitSet {
+ fn bitor_assign(&mut self, other: &Self) {
+ self.union_with(other);
+ }
+}
+
+impl<'a> BitXor for &'a FixedBitSet {
+ type Output = FixedBitSet;
+ fn bitxor(self, other: &FixedBitSet) -> FixedBitSet {
+ let (short, long) = {
+ if self.len() <= other.len() {
+ (self.as_simd_slice(), other.as_simd_slice())
+ } else {
+ (other.as_simd_slice(), self.as_simd_slice())
+ }
+ };
+ let mut data = Vec::from(long);
+ for (data, block) in data.iter_mut().zip(short.iter()) {
+ *data ^= *block;
+ }
+ let len = core::cmp::max(self.len(), other.len());
+ FixedBitSet::from_blocks_and_len(data, len)
+ }
+}
+
+impl BitXorAssign for FixedBitSet {
+ fn bitxor_assign(&mut self, other: Self) {
+ self.symmetric_difference_with(&other);
+ }
+}
+
+impl BitXorAssign<&Self> for FixedBitSet {
+ fn bitxor_assign(&mut self, other: &Self) {
+ self.symmetric_difference_with(other);
+ }
+}
diff --git a/vendor/fixedbitset/src/range.rs b/vendor/fixedbitset/src/range.rs
new file mode 100644
index 00000000..9b385c42
--- /dev/null
+++ b/vendor/fixedbitset/src/range.rs
@@ -0,0 +1,45 @@
+use core::ops::{Range, RangeFrom, RangeFull, RangeTo};
+
+// Taken from https://github.com/bluss/odds/blob/master/src/range.rs.
+
+/// **IndexRange** is implemented by Rust's built-in range types, produced
+/// by range syntax like `..`, `a..`, `..b` or `c..d`.
+pub trait IndexRange<T = usize> {
+ #[inline]
+ /// Start index (inclusive)
+ fn start(&self) -> Option<T> {
+ None
+ }
+ #[inline]
+ /// End index (exclusive)
+ fn end(&self) -> Option<T> {
+ None
+ }
+}
+
+impl<T> IndexRange<T> for RangeFull {}
+
+impl<T: Copy> IndexRange<T> for RangeFrom<T> {
+ #[inline]
+ fn start(&self) -> Option<T> {
+ Some(self.start)
+ }
+}
+
+impl<T: Copy> IndexRange<T> for RangeTo<T> {
+ #[inline]
+ fn end(&self) -> Option<T> {
+ Some(self.end)
+ }
+}
+
+impl<T: Copy> IndexRange<T> for Range<T> {
+ #[inline]
+ fn start(&self) -> Option<T> {
+ Some(self.start)
+ }
+ #[inline]
+ fn end(&self) -> Option<T> {
+ Some(self.end)
+ }
+}
diff --git a/vendor/fixedbitset/src/serde_impl.rs b/vendor/fixedbitset/src/serde_impl.rs
new file mode 100644
index 00000000..98231599
--- /dev/null
+++ b/vendor/fixedbitset/src/serde_impl.rs
@@ -0,0 +1,150 @@
+#[cfg(not(feature = "std"))]
+use core as std;
+
+use crate::{Block, FixedBitSet, BYTES};
+use alloc::vec::Vec;
+use core::{convert::TryFrom, fmt};
+use serde::de::{self, Deserialize, Deserializer, MapAccess, SeqAccess, Visitor};
+use serde::ser::{Serialize, SerializeStruct, Serializer};
+
+struct BitSetByteSerializer<'a>(&'a FixedBitSet);
+
+impl Serialize for FixedBitSet {
+ fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
+ where
+ S: Serializer,
+ {
+ let mut struct_serializer = serializer.serialize_struct("FixedBitset", 2)?;
+ struct_serializer.serialize_field("length", &(self.length as u64))?;
+ struct_serializer.serialize_field("data", &BitSetByteSerializer(self))?;
+ struct_serializer.end()
+ }
+}
+
+impl<'a> Serialize for BitSetByteSerializer<'a> {
+ fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
+ where
+ S: Serializer,
+ {
+ let len = self.0.as_slice().len() * BYTES;
+ // PERF: Figure out a way to do this without allocating.
+ let mut temp = Vec::with_capacity(len);
+ for block in self.0.as_slice() {
+ temp.extend(&block.to_le_bytes());
+ }
+ serializer.serialize_bytes(&temp)
+ }
+}
+
+impl<'de> Deserialize<'de> for FixedBitSet {
+ fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
+ where
+ D: Deserializer<'de>,
+ {
+ enum Field {
+ Length,
+ Data,
+ }
+
+ fn bytes_to_data(length: usize, input: &[u8]) -> Vec<Block> {
+ let block_len = length / BYTES + 1;
+ let mut data = Vec::with_capacity(block_len);
+ for chunk in input.chunks(BYTES) {
+ match <&[u8; BYTES]>::try_from(chunk) {
+ Ok(bytes) => data.push(usize::from_le_bytes(*bytes)),
+ Err(_) => {
+ let mut bytes = [0u8; BYTES];
+ bytes[0..BYTES].copy_from_slice(chunk);
+ data.push(usize::from_le_bytes(bytes));
+ }
+ }
+ }
+ data
+ }
+
+ impl<'de> Deserialize<'de> for Field {
+ fn deserialize<D>(deserializer: D) -> Result<Field, D::Error>
+ where
+ D: Deserializer<'de>,
+ {
+ struct FieldVisitor;
+
+ impl<'de> Visitor<'de> for FieldVisitor {
+ type Value = Field;
+
+ fn expecting(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
+ formatter.write_str("`length` or `data`")
+ }
+
+ fn visit_str<E>(self, value: &str) -> Result<Field, E>
+ where
+ E: de::Error,
+ {
+ match value {
+ "length" => Ok(Field::Length),
+ "data" => Ok(Field::Data),
+ _ => Err(de::Error::unknown_field(value, FIELDS)),
+ }
+ }
+ }
+
+ deserializer.deserialize_identifier(FieldVisitor)
+ }
+ }
+
+ struct FixedBitSetVisitor;
+
+ impl<'de> Visitor<'de> for FixedBitSetVisitor {
+ type Value = FixedBitSet;
+
+ fn expecting(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
+ formatter.write_str("struct Duration")
+ }
+
+ fn visit_seq<V>(self, mut seq: V) -> Result<FixedBitSet, V::Error>
+ where
+ V: SeqAccess<'de>,
+ {
+ let length = seq
+ .next_element()?
+ .ok_or_else(|| de::Error::invalid_length(0, &self))?;
+ let data: &[u8] = seq
+ .next_element()?
+ .ok_or_else(|| de::Error::invalid_length(1, &self))?;
+ let data = bytes_to_data(length, data);
+ Ok(FixedBitSet::with_capacity_and_blocks(length, data))
+ }
+
+ fn visit_map<V>(self, mut map: V) -> Result<FixedBitSet, V::Error>
+ where
+ V: MapAccess<'de>,
+ {
+ let mut length = None;
+ let mut temp: Option<&[u8]> = None;
+ while let Some(key) = map.next_key()? {
+ match key {
+ Field::Length => {
+ if length.is_some() {
+ return Err(de::Error::duplicate_field("length"));
+ }
+ length = Some(map.next_value()?);
+ }
+ Field::Data => {
+ if temp.is_some() {
+ return Err(de::Error::duplicate_field("data"));
+ }
+ temp = Some(map.next_value()?);
+ }
+ }
+ }
+ let length = length.ok_or_else(|| de::Error::missing_field("length"))?;
+ let data = temp.ok_or_else(|| de::Error::missing_field("data"))?;
+ let data = bytes_to_data(length, data);
+ Ok(FixedBitSet::with_capacity_and_blocks(length, data))
+ }
+ }
+
+ const FIELDS: &'static [&'static str] = &["length", "data"];
+ deserializer.deserialize_struct("Duration", FIELDS, FixedBitSetVisitor)
+ }
+}