diff options
author | Ellie Huxtable <e@elm.sh> | 2021-05-17 19:51:09 +0100 |
---|---|---|
committer | Ellie Huxtable <e@elm.sh> | 2021-05-17 19:51:09 +0100 |
commit | d0215a937a7889a97e11778ee4b0f9a12de01278 (patch) | |
tree | 5a8fd5ad62e6b5a4c218746ff2d4bd97373a48de /vendor/memchr-2.3.4 | |
parent | 802a2258cbd839c5b82d24f74d7aebe4a27d8dc5 (diff) |
Vendor dependenciesvendor
Just testing how CI works with this. I tend to prefer vendoring, as it
means that if you have a copy of the code *you can always build it*.
Even if you're 20 years in the future
This is the output of
```
cargo vendor --versioned-dirs
```
Diffstat (limited to 'vendor/memchr-2.3.4')
22 files changed, 3715 insertions, 0 deletions
diff --git a/vendor/memchr-2.3.4/.cargo-checksum.json b/vendor/memchr-2.3.4/.cargo-checksum.json new file mode 100644 index 00000000..cffbb8dc --- /dev/null +++ b/vendor/memchr-2.3.4/.cargo-checksum.json @@ -0,0 +1 @@ +{"files":{"COPYING":"01c266bced4a434da0051174d6bee16a4c82cf634e2679b6155d40d75012390f","Cargo.toml":"97791f3b1dcca115ef7340a312aad0320a79ff91b93cc8fc50b7dcd652721aa3","LICENSE-MIT":"0f96a83840e146e43c0ec96a22ec1f392e0680e6c1226e6f3ba87e0740af850f","README.md":"d2ab7c9c77235b68d1cc856ab5ef7b5115312098469edcac9d5611c5b74d3cd1","UNLICENSE":"7e12e5df4bae12cb21581ba157ced20e1986a0508dd10d0e8a4ab9a4cf94e85c","build.rs":"740225b5280e53e8d73971c6ccc55152d6af32e94132bc6980fdc0eb1fb3ab48","rustfmt.toml":"1ca600239a27401c4a43f363cf3f38183a212affc1f31bff3ae93234bbaec228","src/c.rs":"86fe35cbb46c8bece9927fbde20f1ca3af526defdde05ac969ad2f4bc9bb25e9","src/fallback.rs":"79519255d480a9c2667c06f9287931cfc2b85f6af6fcf92d453a11ee161dcb74","src/iter.rs":"23b1066d6b40159fe944388db7743c89422b1110ddb44667fde6d722f178ed4e","src/lib.rs":"8278d5f65db8081fa5ad3a0d17dac54d30de3d7a01b0dc4479927156e40b225c","src/naive.rs":"c7453bc99cc4e58eb37cf5a50c88688833e50a270ee1849baefddb8acc0ccd94","src/tests/iter.rs":"8d5999a2a5b8a3228c76cd82cd3ee86dfa6f7b4022405b1f06eedf7d74e4c704","src/tests/memchr.rs":"f30074eeab99a16ce5ca8a30f1890f86c43c0422523a7195cbb3ca5f3e465b67","src/tests/miri.rs":"27859a7ac1d0a9305b2d114e6dd8876f3d5e47fc46a81be96239c793ac6edb1f","src/tests/mod.rs":"2ad0c82d33b32562087254522641ea7bfa2a283130152be5e927a33f1978ebc7","src/tests/x86_64-soft_float.json":"c0e416487fe9b4809534edb7db2a9eff3453dc40d9f1e23362c37f45a77ec717","src/x86/avx.rs":"b19987410e49a079f33162424c42494626c91303c41824961e478be3b537c9c9","src/x86/mod.rs":"0b13becaabc150a0099f7528226c82e288136cc7ebcdb8e96cf5ee9aae0f05b6","src/x86/sse2.rs":"7c1b8248a8cd48396cb70a862d77f9a972f1e16324d65c260f39d13af73bf638","src/x86/sse42.rs":"f671ae9dd2b518a823e499a09ce32d4957bc5ae043db90d61c027e32f688f2b2"},"package":"0ee1c47aaa256ecabcaea351eae4a9b01ef39ed810004e298d2511ed284b1525"}
\ No newline at end of file diff --git a/vendor/memchr-2.3.4/COPYING b/vendor/memchr-2.3.4/COPYING new file mode 100644 index 00000000..bb9c20a0 --- /dev/null +++ b/vendor/memchr-2.3.4/COPYING @@ -0,0 +1,3 @@ +This project is dual-licensed under the Unlicense and MIT licenses. + +You may use this code under the terms of either license. diff --git a/vendor/memchr-2.3.4/Cargo.toml b/vendor/memchr-2.3.4/Cargo.toml new file mode 100644 index 00000000..1fc0b64f --- /dev/null +++ b/vendor/memchr-2.3.4/Cargo.toml @@ -0,0 +1,42 @@ +# THIS FILE IS AUTOMATICALLY GENERATED BY CARGO +# +# When uploading crates to the registry Cargo will automatically +# "normalize" Cargo.toml files for maximal compatibility +# with all versions of Cargo and also rewrite `path` dependencies +# to registry (e.g., crates.io) dependencies +# +# If you believe there's an error in this file please file an +# issue against the rust-lang/cargo repository. If you're +# editing this file be aware that the upstream Cargo.toml +# will likely look very different (and much more reasonable) + +[package] +name = "memchr" +version = "2.3.4" +authors = ["Andrew Gallant <jamslam@gmail.com>", "bluss"] +exclude = ["/ci/*", "/.travis.yml", "/Makefile", "/appveyor.yml"] +description = "Safe interface to memchr." +homepage = "https://github.com/BurntSushi/rust-memchr" +documentation = "https://docs.rs/memchr/" +readme = "README.md" +keywords = ["memchr", "char", "scan", "strchr", "string"] +license = "Unlicense/MIT" +repository = "https://github.com/BurntSushi/rust-memchr" +[profile.test] +opt-level = 3 + +[lib] +name = "memchr" +bench = false +[dependencies.libc] +version = "0.2.18" +optional = true +default-features = false +[dev-dependencies.quickcheck] +version = "0.9" +default-features = false + +[features] +default = ["std"] +std = [] +use_std = ["std"] diff --git a/vendor/memchr-2.3.4/LICENSE-MIT b/vendor/memchr-2.3.4/LICENSE-MIT new file mode 100644 index 00000000..3b0a5dc0 --- /dev/null +++ b/vendor/memchr-2.3.4/LICENSE-MIT @@ -0,0 +1,21 @@ +The MIT License (MIT) + +Copyright (c) 2015 Andrew Gallant + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in +all copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +THE SOFTWARE. diff --git a/vendor/memchr-2.3.4/README.md b/vendor/memchr-2.3.4/README.md new file mode 100644 index 00000000..f78a5a53 --- /dev/null +++ b/vendor/memchr-2.3.4/README.md @@ -0,0 +1,79 @@ +memchr +====== +The `memchr` crate provides heavily optimized routines for searching bytes. + +[![Build status](https://github.com/BurntSushi/rust-memchr/workflows/ci/badge.svg)](https://github.com/BurntSushi/rust-memchr/actions) +[![](http://meritbadge.herokuapp.com/memchr)](https://crates.io/crates/memchr) + +Dual-licensed under MIT or the [UNLICENSE](http://unlicense.org). + + +### Documentation + +[https://docs.rs/memchr](https://docs.rs/memchr) + + +### Overview + +The `memchr` function is traditionally provided by libc, but its +performance can vary significantly depending on the specific +implementation of libc that is used. They can range from manually tuned +Assembly implementations (like that found in GNU's libc) all the way to +non-vectorized C implementations (like that found in MUSL). + +To smooth out the differences between implementations of libc, at least +on `x86_64` for Rust 1.27+, this crate provides its own implementation of +`memchr` that should perform competitively with the one found in GNU's libc. +The implementation is in pure Rust and has no dependency on a C compiler or an +Assembler. + +Additionally, GNU libc also provides an extension, `memrchr`. This crate +provides its own implementation of `memrchr` as well, on top of `memchr2`, +`memchr3`, `memrchr2` and `memrchr3`. The difference between `memchr` and +`memchr2` is that `memchr2` permits finding all occurrences of two bytes +instead of one. Similarly for `memchr3`. + +### Compiling without the standard library + +memchr links to the standard library by default, but you can disable the +`std` feature if you want to use it in a `#![no_std]` crate: + +```toml +[dependencies] +memchr = { version = "2", default-features = false } +``` + +On x86 platforms, when the `std` feature is disabled, the SSE2 +implementation of memchr will be used in compilers that support it. When +`std` is enabled, the AVX implementation of memchr will be used if the CPU +is determined to support it at runtime. + +### Using libc + +`memchr` is a routine that is part of libc, although this crate does not use +libc by default. Instead, it uses its own routines, which are either vectorized +or generic fallback routines. In general, these should be competitive with +what's in libc, although this has not been tested for all architectures. If +using `memchr` from libc is desirable and a vectorized routine is not otherwise +available in this crate, then enabling the `libc` feature will use libc's +version of `memchr`. + +The rest of the functions in this crate, e.g., `memchr2` or `memrchr3`, are not +a standard part of libc, so they will always use the implementations in this +crate. One exception to this is `memrchr`, which is an extension commonly found +on Linux. On Linux, `memrchr` is used in precisely the same scenario as +`memchr`, as described above. + + +### Minimum Rust version policy + +This crate's minimum supported `rustc` version is `1.28.0`. + +The current policy is that the minimum Rust version required to use this crate +can be increased in minor version updates. For example, if `crate 1.0` requires +Rust 1.20.0, then `crate 1.0.z` for all values of `z` will also require Rust +1.20.0 or newer. However, `crate 1.y` for `y > 0` may require a newer minimum +version of Rust. + +In general, this crate will be conservative with respect to the minimum +supported version of Rust. diff --git a/vendor/memchr-2.3.4/UNLICENSE b/vendor/memchr-2.3.4/UNLICENSE new file mode 100644 index 00000000..68a49daa --- /dev/null +++ b/vendor/memchr-2.3.4/UNLICENSE @@ -0,0 +1,24 @@ +This is free and unencumbered software released into the public domain. + +Anyone is free to copy, modify, publish, use, compile, sell, or +distribute this software, either in source code form or as a compiled +binary, for any purpose, commercial or non-commercial, and by any +means. + +In jurisdictions that recognize copyright laws, the author or authors +of this software dedicate any and all copyright interest in the +software to the public domain. We make this dedication for the benefit +of the public at large and to the detriment of our heirs and +successors. We intend this dedication to be an overt act of +relinquishment in perpetuity of all present and future rights to this +software under copyright law. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, +EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF +MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. +IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR +OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, +ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +OTHER DEALINGS IN THE SOFTWARE. + +For more information, please refer to <http://unlicense.org/> diff --git a/vendor/memchr-2.3.4/build.rs b/vendor/memchr-2.3.4/build.rs new file mode 100644 index 00000000..e07ad6f1 --- /dev/null +++ b/vendor/memchr-2.3.4/build.rs @@ -0,0 +1,74 @@ +use std::env; + +fn main() { + enable_simd_optimizations(); + enable_libc(); +} + +// This adds various simd cfgs if this compiler and target support it. +// +// This can be disabled with RUSTFLAGS="--cfg memchr_disable_auto_simd", but +// this is generally only intended for testing. +// +// On targets which don't feature SSE2, this is disabled, as LLVM wouln't know +// how to work with SSE2 operands. Enabling SSE4.2 and AVX on SSE2-only targets +// is not a problem. In that case, the fastest option will be chosen at +// runtime. +fn enable_simd_optimizations() { + if is_env_set("CARGO_CFG_MEMCHR_DISABLE_AUTO_SIMD") + || !target_has_feature("sse2") + { + return; + } + println!("cargo:rustc-cfg=memchr_runtime_simd"); + println!("cargo:rustc-cfg=memchr_runtime_sse2"); + println!("cargo:rustc-cfg=memchr_runtime_sse42"); + println!("cargo:rustc-cfg=memchr_runtime_avx"); +} + +// This adds a `memchr_libc` cfg if and only if libc can be used, if no other +// better option is available. +// +// This could be performed in the source code, but it's simpler to do it once +// here and consolidate it into one cfg knob. +// +// Basically, we use libc only if its enabled and if we aren't targeting a +// known bad platform. For example, wasm32 doesn't have a libc and the +// performance of memchr on Windows is seemingly worse than the fallback +// implementation. +fn enable_libc() { + const NO_ARCH: &'static [&'static str] = &["wasm32", "windows"]; + const NO_ENV: &'static [&'static str] = &["sgx"]; + + if !is_feature_set("LIBC") { + return; + } + + let arch = match env::var("CARGO_CFG_TARGET_ARCH") { + Err(_) => return, + Ok(arch) => arch, + }; + let env = match env::var("CARGO_CFG_TARGET_ENV") { + Err(_) => return, + Ok(env) => env, + }; + if NO_ARCH.contains(&&*arch) || NO_ENV.contains(&&*env) { + return; + } + + println!("cargo:rustc-cfg=memchr_libc"); +} + +fn is_feature_set(name: &str) -> bool { + is_env_set(&format!("CARGO_FEATURE_{}", name)) +} + +fn is_env_set(name: &str) -> bool { + env::var_os(name).is_some() +} + +fn target_has_feature(feature: &str) -> bool { + env::var("CARGO_CFG_TARGET_FEATURE") + .map(|features| features.contains(feature)) + .unwrap_or(false) +} diff --git a/vendor/memchr-2.3.4/rustfmt.toml b/vendor/memchr-2.3.4/rustfmt.toml new file mode 100644 index 00000000..aa37a218 --- /dev/null +++ b/vendor/memchr-2.3.4/rustfmt.toml @@ -0,0 +1,2 @@ +max_width = 79 +use_small_heuristics = "max" diff --git a/vendor/memchr-2.3.4/src/c.rs b/vendor/memchr-2.3.4/src/c.rs new file mode 100644 index 00000000..63feca97 --- /dev/null +++ b/vendor/memchr-2.3.4/src/c.rs @@ -0,0 +1,44 @@ +// This module defines safe wrappers around memchr (POSIX) and memrchr (GNU +// extension). + +#![allow(dead_code)] + +extern crate libc; + +use self::libc::{c_int, c_void, size_t}; + +pub fn memchr(needle: u8, haystack: &[u8]) -> Option<usize> { + let p = unsafe { + libc::memchr( + haystack.as_ptr() as *const c_void, + needle as c_int, + haystack.len() as size_t, + ) + }; + if p.is_null() { + None + } else { + Some(p as usize - (haystack.as_ptr() as usize)) + } +} + +// memrchr is a GNU extension. We know it's available on Linux, so start there. +#[cfg(target_os = "linux")] +pub fn memrchr(needle: u8, haystack: &[u8]) -> Option<usize> { + // GNU's memrchr() will - unlike memchr() - error if haystack is empty. + if haystack.is_empty() { + return None; + } + let p = unsafe { + libc::memrchr( + haystack.as_ptr() as *const c_void, + needle as c_int, + haystack.len() as size_t, + ) + }; + if p.is_null() { + None + } else { + Some(p as usize - (haystack.as_ptr() as usize)) + } +} diff --git a/vendor/memchr-2.3.4/src/fallback.rs b/vendor/memchr-2.3.4/src/fallback.rs new file mode 100644 index 00000000..8bc32b27 --- /dev/null +++ b/vendor/memchr-2.3.4/src/fallback.rs @@ -0,0 +1,330 @@ +// This module defines pure Rust platform independent implementations of all +// the memchr routines. We do our best to make them fast. Some of them may even +// get auto-vectorized. + +use core::cmp; +use core::usize; + +#[cfg(target_pointer_width = "16")] +const USIZE_BYTES: usize = 2; + +#[cfg(target_pointer_width = "32")] +const USIZE_BYTES: usize = 4; + +#[cfg(target_pointer_width = "64")] +const USIZE_BYTES: usize = 8; + +// The number of bytes to loop at in one iteration of memchr/memrchr. +const LOOP_SIZE: usize = 2 * USIZE_BYTES; + +/// Return `true` if `x` contains any zero byte. +/// +/// From *Matters Computational*, J. Arndt +/// +/// "The idea is to subtract one from each of the bytes and then look for +/// bytes where the borrow propagated all the way to the most significant +/// bit." +#[inline(always)] +fn contains_zero_byte(x: usize) -> bool { + const LO_U64: u64 = 0x0101010101010101; + const HI_U64: u64 = 0x8080808080808080; + + const LO_USIZE: usize = LO_U64 as usize; + const HI_USIZE: usize = HI_U64 as usize; + + x.wrapping_sub(LO_USIZE) & !x & HI_USIZE != 0 +} + +/// Repeat the given byte into a word size number. That is, every 8 bits +/// is equivalent to the given byte. For example, if `b` is `\x4E` or +/// `01001110` in binary, then the returned value on a 32-bit system would be: +/// `01001110_01001110_01001110_01001110`. +#[inline(always)] +fn repeat_byte(b: u8) -> usize { + (b as usize) * (usize::MAX / 255) +} + +pub fn memchr(n1: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let confirm = |byte| byte == n1; + let loop_size = cmp::min(LOOP_SIZE, haystack.len()); + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = start_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr as *const usize).read_unaligned(); + if contains_zero_byte(chunk ^ vn1) { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = ptr.add(USIZE_BYTES - (start_ptr as usize & align)); + debug_assert!(ptr > start_ptr); + debug_assert!(end_ptr.sub(USIZE_BYTES) >= start_ptr); + while loop_size == LOOP_SIZE && ptr <= end_ptr.sub(loop_size) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let a = *(ptr as *const usize); + let b = *(ptr.add(USIZE_BYTES) as *const usize); + let eqa = contains_zero_byte(a ^ vn1); + let eqb = contains_zero_byte(b ^ vn1); + if eqa || eqb { + break; + } + ptr = ptr.add(LOOP_SIZE); + } + forward_search(start_ptr, end_ptr, ptr, confirm) + } +} + +/// Like `memchr`, but searches for two bytes instead of one. +pub fn memchr2(n1: u8, n2: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let vn2 = repeat_byte(n2); + let confirm = |byte| byte == n1 || byte == n2; + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = start_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr as *const usize).read_unaligned(); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + if eq1 || eq2 { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = ptr.add(USIZE_BYTES - (start_ptr as usize & align)); + debug_assert!(ptr > start_ptr); + debug_assert!(end_ptr.sub(USIZE_BYTES) >= start_ptr); + while ptr <= end_ptr.sub(USIZE_BYTES) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let chunk = *(ptr as *const usize); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + if eq1 || eq2 { + break; + } + ptr = ptr.add(USIZE_BYTES); + } + forward_search(start_ptr, end_ptr, ptr, confirm) + } +} + +/// Like `memchr`, but searches for three bytes instead of one. +pub fn memchr3(n1: u8, n2: u8, n3: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let vn2 = repeat_byte(n2); + let vn3 = repeat_byte(n3); + let confirm = |byte| byte == n1 || byte == n2 || byte == n3; + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = start_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr as *const usize).read_unaligned(); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + let eq3 = contains_zero_byte(chunk ^ vn3); + if eq1 || eq2 || eq3 { + return forward_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = ptr.add(USIZE_BYTES - (start_ptr as usize & align)); + debug_assert!(ptr > start_ptr); + debug_assert!(end_ptr.sub(USIZE_BYTES) >= start_ptr); + while ptr <= end_ptr.sub(USIZE_BYTES) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let chunk = *(ptr as *const usize); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + let eq3 = contains_zero_byte(chunk ^ vn3); + if eq1 || eq2 || eq3 { + break; + } + ptr = ptr.add(USIZE_BYTES); + } + forward_search(start_ptr, end_ptr, ptr, confirm) + } +} + +/// Return the last index matching the byte `x` in `text`. +pub fn memrchr(n1: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let confirm = |byte| byte == n1; + let loop_size = cmp::min(LOOP_SIZE, haystack.len()); + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = end_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr.sub(USIZE_BYTES) as *const usize).read_unaligned(); + if contains_zero_byte(chunk ^ vn1) { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = (end_ptr as usize & !align) as *const u8; + debug_assert!(start_ptr <= ptr && ptr <= end_ptr); + while loop_size == LOOP_SIZE && ptr >= start_ptr.add(loop_size) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let a = *(ptr.sub(2 * USIZE_BYTES) as *const usize); + let b = *(ptr.sub(1 * USIZE_BYTES) as *const usize); + let eqa = contains_zero_byte(a ^ vn1); + let eqb = contains_zero_byte(b ^ vn1); + if eqa || eqb { + break; + } + ptr = ptr.sub(loop_size); + } + reverse_search(start_ptr, end_ptr, ptr, confirm) + } +} + +/// Like `memrchr`, but searches for two bytes instead of one. +pub fn memrchr2(n1: u8, n2: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let vn2 = repeat_byte(n2); + let confirm = |byte| byte == n1 || byte == n2; + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = end_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr.sub(USIZE_BYTES) as *const usize).read_unaligned(); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + if eq1 || eq2 { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = (end_ptr as usize & !align) as *const u8; + debug_assert!(start_ptr <= ptr && ptr <= end_ptr); + while ptr >= start_ptr.add(USIZE_BYTES) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let chunk = *(ptr.sub(USIZE_BYTES) as *const usize); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + if eq1 || eq2 { + break; + } + ptr = ptr.sub(USIZE_BYTES); + } + reverse_search(start_ptr, end_ptr, ptr, confirm) + } +} + +/// Like `memrchr`, but searches for three bytes instead of one. +pub fn memrchr3(n1: u8, n2: u8, n3: u8, haystack: &[u8]) -> Option<usize> { + let vn1 = repeat_byte(n1); + let vn2 = repeat_byte(n2); + let vn3 = repeat_byte(n3); + let confirm = |byte| byte == n1 || byte == n2 || byte == n3; + let align = USIZE_BYTES - 1; + let start_ptr = haystack.as_ptr(); + let end_ptr = haystack[haystack.len()..].as_ptr(); + let mut ptr = end_ptr; + + unsafe { + if haystack.len() < USIZE_BYTES { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + let chunk = (ptr.sub(USIZE_BYTES) as *const usize).read_unaligned(); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + let eq3 = contains_zero_byte(chunk ^ vn3); + if eq1 || eq2 || eq3 { + return reverse_search(start_ptr, end_ptr, ptr, confirm); + } + + ptr = (end_ptr as usize & !align) as *const u8; + debug_assert!(start_ptr <= ptr && ptr <= end_ptr); + while ptr >= start_ptr.add(USIZE_BYTES) { + debug_assert_eq!(0, (ptr as usize) % USIZE_BYTES); + + let chunk = *(ptr.sub(USIZE_BYTES) as *const usize); + let eq1 = contains_zero_byte(chunk ^ vn1); + let eq2 = contains_zero_byte(chunk ^ vn2); + let eq3 = contains_zero_byte(chunk ^ vn3); + if eq1 || eq2 || eq3 { + break; + } + ptr = ptr.sub(USIZE_BYTES); + } + reverse_search(start_ptr, end_ptr, ptr, confirm) + } +} + +#[inline(always)] +unsafe fn forward_search<F: Fn(u8) -> bool>( + start_ptr: *const u8, + end_ptr: *const u8, + mut ptr: *const u8, + confirm: F, +) -> Option<usize> { + debug_assert!(start_ptr <= ptr); + debug_assert!(ptr <= end_ptr); + + while ptr < end_ptr { + if confirm(*ptr) { + return Some(sub(ptr, start_ptr)); + } + ptr = ptr.offset(1); + } + None +} + +#[inline(always)] +unsafe fn reverse_search<F: Fn(u8) -> bool>( + start_ptr: *const u8, + end_ptr: *const u8, + mut ptr: *const u8, + confirm: F, +) -> Option<usize> { + debug_assert!(start_ptr <= ptr); + debug_assert!(ptr <= end_ptr); + + while ptr > start_ptr { + ptr = ptr.offset(-1); + if confirm(*ptr) { + return Some(sub(ptr, start_ptr)); + } + } + None +} + +/// Subtract `b` from `a` and return the difference. `a` should be greater than +/// or equal to `b`. +fn sub(a: *const u8, b: *const u8) -> usize { + debug_assert!(a >= b); + (a as usize) - (b as usize) +} diff --git a/vendor/memchr-2.3.4/src/iter.rs b/vendor/memchr-2.3.4/src/iter.rs new file mode 100644 index 00000000..6217ae4a --- /dev/null +++ b/vendor/memchr-2.3.4/src/iter.rs @@ -0,0 +1,173 @@ +use {memchr, memchr2, memchr3, memrchr, memrchr2, memrchr3}; + +macro_rules! iter_next { + // Common code for the memchr iterators: + // update haystack and position and produce the index + // + // self: &mut Self where Self is the iterator + // search_result: Option<usize> which is the result of the corresponding + // memchr function. + // + // Returns Option<usize> (the next iterator element) + ($self_:expr, $search_result:expr) => { + $search_result.map(move |index| { + // split and take the remaining back half + $self_.haystack = $self_.haystack.split_at(index + 1).1; + let found_position = $self_.position + index; + $self_.position = found_position + 1; + found_position + }) + }; +} + +macro_rules! iter_next_back { + ($self_:expr, $search_result:expr) => { + $search_result.map(move |index| { + // split and take the remaining front half + $self_.haystack = $self_.haystack.split_at(index).0; + $self_.position + index + }) + }; +} + +/// An iterator for `memchr`. +pub struct Memchr<'a> { + needle: u8, + // The haystack to iterate over + haystack: &'a [u8], + // The index + position: usize, +} + +impl<'a> Memchr<'a> { + /// Creates a new iterator that yields all positions of needle in haystack. + #[inline] + pub fn new(needle: u8, haystack: &[u8]) -> Memchr { + Memchr { needle: needle, haystack: haystack, position: 0 } + } +} + +impl<'a> Iterator for Memchr<'a> { + type Item = usize; + + #[inline] + fn next(&mut self) -> Option<usize> { + iter_next!(self, memchr(self.needle, self.haystack)) + } + + #[inline] + fn size_hint(&self) -> (usize, Option<usize>) { + (0, Some(self.haystack.len())) + } +} + +impl<'a> DoubleEndedIterator for Memchr<'a> { + #[inline] + fn next_back(&mut self) -> Option<Self::Item> { + iter_next_back!(self, memrchr(self.needle, self.haystack)) + } +} + +/// An iterator for `memchr2`. +pub struct Memchr2<'a> { + needle1: u8, + needle2: u8, + // The haystack to iterate over + haystack: &'a [u8], + // The index + position: usize, +} + +impl<'a> Memchr2<'a> { + /// Creates a new iterator that yields all positions of needle in haystack. + #[inline] + pub fn new(needle1: u8, needle2: u8, haystack: &[u8]) -> Memchr2 { + Memchr2 { + needle1: needle1, + needle2: needle2, + haystack: haystack, + position: 0, + } + } +} + +impl<'a> Iterator for Memchr2<'a> { |