From 1b6a04ca5504955c571d1c97504fb45ea0befee4 Mon Sep 17 00:00:00 2001 From: Valentin Popov Date: Mon, 8 Jan 2024 01:21:28 +0400 Subject: Initial vendor packages Signed-off-by: Valentin Popov --- vendor/adler/src/algo.rs | 146 ++++++++++++++++++++++++ vendor/adler/src/lib.rs | 287 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 433 insertions(+) create mode 100644 vendor/adler/src/algo.rs create mode 100644 vendor/adler/src/lib.rs (limited to 'vendor/adler/src') diff --git a/vendor/adler/src/algo.rs b/vendor/adler/src/algo.rs new file mode 100644 index 0000000..650cffa --- /dev/null +++ b/vendor/adler/src/algo.rs @@ -0,0 +1,146 @@ +use crate::Adler32; +use std::ops::{AddAssign, MulAssign, RemAssign}; + +impl Adler32 { + pub(crate) fn compute(&mut self, bytes: &[u8]) { + // The basic algorithm is, for every byte: + // a = (a + byte) % MOD + // b = (b + a) % MOD + // where MOD = 65521. + // + // For efficiency, we can defer the `% MOD` operations as long as neither a nor b overflows: + // - Between calls to `write`, we ensure that a and b are always in range 0..MOD. + // - We use 32-bit arithmetic in this function. + // - Therefore, a and b must not increase by more than 2^32-MOD without performing a `% MOD` + // operation. + // + // According to Wikipedia, b is calculated as follows for non-incremental checksumming: + // b = n×D1 + (n−1)×D2 + (n−2)×D3 + ... + Dn + n*1 (mod 65521) + // Where n is the number of bytes and Di is the i-th Byte. We need to change this to account + // for the previous values of a and b, as well as treat every input Byte as being 255: + // b_inc = n×255 + (n-1)×255 + ... + 255 + n*65520 + // Or in other words: + // b_inc = n*65520 + n(n+1)/2*255 + // The max chunk size is thus the largest value of n so that b_inc <= 2^32-65521. + // 2^32-65521 = n*65520 + n(n+1)/2*255 + // Plugging this into an equation solver since I can't math gives n = 5552.18..., so 5552. + // + // On top of the optimization outlined above, the algorithm can also be parallelized with a + // bit more work: + // + // Note that b is a linear combination of a vector of input bytes (D1, ..., Dn). + // + // If we fix some value k Self { + U32X4([ + u32::from(bytes[0]), + u32::from(bytes[1]), + u32::from(bytes[2]), + u32::from(bytes[3]), + ]) + } +} + +impl AddAssign for U32X4 { + fn add_assign(&mut self, other: Self) { + for (s, o) in self.0.iter_mut().zip(other.0.iter()) { + *s += o; + } + } +} + +impl RemAssign for U32X4 { + fn rem_assign(&mut self, quotient: u32) { + for s in self.0.iter_mut() { + *s %= quotient; + } + } +} + +impl MulAssign for U32X4 { + fn mul_assign(&mut self, rhs: u32) { + for s in self.0.iter_mut() { + *s *= rhs; + } + } +} diff --git a/vendor/adler/src/lib.rs b/vendor/adler/src/lib.rs new file mode 100644 index 0000000..c7aa380 --- /dev/null +++ b/vendor/adler/src/lib.rs @@ -0,0 +1,287 @@ +//! Adler-32 checksum implementation. +//! +//! This implementation features: +//! +//! - Permissively licensed (0BSD) clean-room implementation. +//! - Zero dependencies. +//! - Zero `unsafe`. +//! - Decent performance (3-4 GB/s). +//! - `#![no_std]` support (with `default-features = false`). + +#![doc(html_root_url = "https://docs.rs/adler/1.0.2")] +// Deny a few warnings in doctests, since rustdoc `allow`s many warnings by default +#![doc(test(attr(deny(unused_imports, unused_must_use))))] +#![cfg_attr(docsrs, feature(doc_cfg))] +#![warn(missing_debug_implementations)] +#![forbid(unsafe_code)] +#![cfg_attr(not(feature = "std"), no_std)] + +#[cfg(not(feature = "std"))] +extern crate core as std; + +mod algo; + +use std::hash::Hasher; + +#[cfg(feature = "std")] +use std::io::{self, BufRead}; + +/// Adler-32 checksum calculator. +/// +/// An instance of this type is equivalent to an Adler-32 checksum: It can be created in the default +/// state via [`new`] (or the provided `Default` impl), or from a precalculated checksum via +/// [`from_checksum`], and the currently stored checksum can be fetched via [`checksum`]. +/// +/// This type also implements `Hasher`, which makes it easy to calculate Adler-32 checksums of any +/// type that implements or derives `Hash`. This also allows using Adler-32 in a `HashMap`, although +/// that is not recommended (while every checksum is a hash function, they are not necessarily a +/// good one). +/// +/// # Examples +/// +/// Basic, piecewise checksum calculation: +/// +/// ``` +/// use adler::Adler32; +/// +/// let mut adler = Adler32::new(); +/// +/// adler.write_slice(&[0, 1, 2]); +/// adler.write_slice(&[3, 4, 5]); +/// +/// assert_eq!(adler.checksum(), 0x00290010); +/// ``` +/// +/// Using `Hash` to process structures: +/// +/// ``` +/// use std::hash::Hash; +/// use adler::Adler32; +/// +/// #[derive(Hash)] +/// struct Data { +/// byte: u8, +/// word: u16, +/// big: u64, +/// } +/// +/// let mut adler = Adler32::new(); +/// +/// let data = Data { byte: 0x1F, word: 0xABCD, big: !0 }; +/// data.hash(&mut adler); +/// +/// // hash value depends on architecture endianness +/// if cfg!(target_endian = "little") { +/// assert_eq!(adler.checksum(), 0x33410990); +/// } +/// if cfg!(target_endian = "big") { +/// assert_eq!(adler.checksum(), 0x331F0990); +/// } +/// +/// ``` +/// +/// [`new`]: #method.new +/// [`from_checksum`]: #method.from_checksum +/// [`checksum`]: #method.checksum +#[derive(Debug, Copy, Clone)] +pub struct Adler32 { + a: u16, + b: u16, +} + +impl Adler32 { + /// Creates a new Adler-32 instance with default state. + #[inline] + pub fn new() -> Self { + Self::default() + } + + /// Creates an `Adler32` instance from a precomputed Adler-32 checksum. + /// + /// This allows resuming checksum calculation without having to keep the `Adler32` instance + /// around. + /// + /// # Example + /// + /// ``` + /// # use adler::Adler32; + /// let parts = [ + /// "rust", + /// "acean", + /// ]; + /// let whole = adler::adler32_slice(b"rustacean"); + /// + /// let mut sum = Adler32::new(); + /// sum.write_slice(parts[0].as_bytes()); + /// let partial = sum.checksum(); + /// + /// // ...later + /// + /// let mut sum = Adler32::from_checksum(partial); + /// sum.write_slice(parts[1].as_bytes()); + /// assert_eq!(sum.checksum(), whole); + /// ``` + #[inline] + pub fn from_checksum(sum: u32) -> Self { + Adler32 { + a: sum as u16, + b: (sum >> 16) as u16, + } + } + + /// Returns the calculated checksum at this point in time. + #[inline] + pub fn checksum(&self) -> u32 { + (u32::from(self.b) << 16) | u32::from(self.a) + } + + /// Adds `bytes` to the checksum calculation. + /// + /// If efficiency matters, this should be called with Byte slices that contain at least a few + /// thousand Bytes. + pub fn write_slice(&mut self, bytes: &[u8]) { + self.compute(bytes); + } +} + +impl Default for Adler32 { + #[inline] + fn default() -> Self { + Adler32 { a: 1, b: 0 } + } +} + +impl Hasher for Adler32 { + #[inline] + fn finish(&self) -> u64 { + u64::from(self.checksum()) + } + + fn write(&mut self, bytes: &[u8]) { + self.write_slice(bytes); + } +} + +/// Calculates the Adler-32 checksum of a byte slice. +/// +/// This is a convenience function around the [`Adler32`] type. +/// +/// [`Adler32`]: struct.Adler32.html +pub fn adler32_slice(data: &[u8]) -> u32 { + let mut h = Adler32::new(); + h.write_slice(data); + h.checksum() +} + +/// Calculates the Adler-32 checksum of a `BufRead`'s contents. +/// +/// The passed `BufRead` implementor will be read until it reaches EOF (or until it reports an +/// error). +/// +/// If you only have a `Read` implementor, you can wrap it in `std::io::BufReader` before calling +/// this function. +/// +/// # Errors +/// +/// Any error returned by the reader are bubbled up by this function. +/// +/// # Examples +/// +/// ```no_run +/// # fn run() -> Result<(), Box> { +/// use adler::adler32; +/// +/// use std::fs::File; +/// use std::io::BufReader; +/// +/// let file = File::open("input.txt")?; +/// let mut file = BufReader::new(file); +/// +/// adler32(&mut file)?; +/// # Ok(()) } +/// # fn main() { run().unwrap() } +/// ``` +#[cfg(feature = "std")] +#[cfg_attr(docsrs, doc(cfg(feature = "std")))] +pub fn adler32(mut reader: R) -> io::Result { + let mut h = Adler32::new(); + loop { + let len = { + let buf = reader.fill_buf()?; + if buf.is_empty() { + return Ok(h.checksum()); + } + + h.write_slice(buf); + buf.len() + }; + reader.consume(len); + } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn zeroes() { + assert_eq!(adler32_slice(&[]), 1); + assert_eq!(adler32_slice(&[0]), 1 | 1 << 16); + assert_eq!(adler32_slice(&[0, 0]), 1 | 2 << 16); + assert_eq!(adler32_slice(&[0; 100]), 0x00640001); + assert_eq!(adler32_slice(&[0; 1024]), 0x04000001); + assert_eq!(adler32_slice(&[0; 1024 * 1024]), 0x00f00001); + } + + #[test] + fn ones() { + assert_eq!(adler32_slice(&[0xff; 1024]), 0x79a6fc2e); + assert_eq!(adler32_slice(&[0xff; 1024 * 1024]), 0x8e88ef11); + } + + #[test] + fn mixed() { + assert_eq!(adler32_slice(&[1]), 2 | 2 << 16); + assert_eq!(adler32_slice(&[40]), 41 | 41 << 16); + + assert_eq!(adler32_slice(&[0xA5; 1024 * 1024]), 0xd5009ab1); + } + + /// Example calculation from https://en.wikipedia.org/wiki/Adler-32. + #[test] + fn wiki() { + assert_eq!(adler32_slice(b"Wikipedia"), 0x11E60398); + } + + #[test] + fn resume() { + let mut adler = Adler32::new(); + adler.write_slice(&[0xff; 1024]); + let partial = adler.checksum(); + assert_eq!(partial, 0x79a6fc2e); // from above + adler.write_slice(&[0xff; 1024 * 1024 - 1024]); + assert_eq!(adler.checksum(), 0x8e88ef11); // from above + + // Make sure that we can resume computing from the partial checksum via `from_checksum`. + let mut adler = Adler32::from_checksum(partial); + adler.write_slice(&[0xff; 1024 * 1024 - 1024]); + assert_eq!(adler.checksum(), 0x8e88ef11); // from above + } + + #[cfg(feature = "std")] + #[test] + fn bufread() { + use std::io::BufReader; + fn test(data: &[u8], checksum: u32) { + // `BufReader` uses an 8 KB buffer, so this will test buffer refilling. + let mut buf = BufReader::new(data); + let real_sum = adler32(&mut buf).unwrap(); + assert_eq!(checksum, real_sum); + } + + test(&[], 1); + test(&[0; 1024], 0x04000001); + test(&[0; 1024 * 1024], 0x00f00001); + test(&[0xA5; 1024 * 1024], 0xd5009ab1); + } +} -- cgit v1.2.3