commit 353e30dbfbd1bc5c3ee32cf4da0a44e8e946dd42
parent 2858a84ec2ea7391d8ec73ca86aa47ecedb514ca
Author: Zack Newman <zack@philomathiclife.com>
Date: Sat, 7 Sep 2024 10:37:28 -0600
update deps. lint reasons
Diffstat:
7 files changed, 173 insertions(+), 172 deletions(-)
diff --git a/Cargo.toml b/Cargo.toml
@@ -9,33 +9,25 @@ license = "MIT OR Apache-2.0"
name = "ascii_domain"
readme = "README.md"
repository = "https://git.philomathiclife.com/repos/ascii_domain/"
-version = "0.6.1"
+rust-version = "1.81.0"
+version = "0.6.2"
-[lib]
-name = "ascii_domain"
-path = "src/lib.rs"
+[badges]
+maintenance = { status = "actively-developed" }
+
+[package.metadata.docs.rs]
+all-features = true
+rustdoc-args = ["--cfg", "docsrs"]
[dependencies]
-serde = { version = "1.0.197", default-features = false, features = ["alloc"], optional = true }
+serde = { version = "1.0.210", default-features = false, features = ["alloc"], optional = true }
[dev-dependencies]
-serde_json = { version = "1.0.115", default-features = false, features = ["alloc"] }
+serde_json = { version = "1.0.128", default-features = false, features = ["alloc"] }
+
-[build-dependencies]
-rustc_version = "0.4.0"
+### FEATURES #################################################################
[features]
+# Provides serde support.
serde = ["dep:serde"]
-std = []
-
-[package.metadata.docs.rs]
-all-features = true
-rustdoc-args = ["--cfg", "docsrs"]
-
-[badges]
-maintenance = { status = "actively-developed" }
-
-[profile.release]
-lto = true
-panic = 'abort'
-strip = true
diff --git a/README.md b/README.md
@@ -1,23 +1,33 @@
-# ascii_domain
-
-`ascii_domain` is a [library](https://docs.rs/ascii_domain/latest/ascii_domain) for efficiently parsing domains
-based on a supplied ASCII character set one wants to enforce each
-[`Label`](https://docs.rs/domain_parse/latest/domain_parse/dom/struct.Label.html) to conform to. The primary type
-in the library is [`Domain`](https://docs.rs/domain_parse/latest/domain_parse/dom/struct.Domain.html) which can be
-thought of as domains in _representation_ format. Technically since any ASCII `u8` except `b'.'` is allowed in a
-`Label`, it is more general than an actual representation format that doesn’t include some form of escape
-characters. For a full-fledged DNS library look elsewhere (e.g., [`domain`](https://docs.rs/domain/latest/domain/)).
-
+# ascii_domain
+
+`ascii_domain` is a library for efficiently parsing domains based on a supplied ASCII character set one wants to
+enforce each `Label` to conform to. The primary type in the library is `Domain` which can be thought of as a domain
+in _representation_ format. Technically since any ASCII `u8` except `b'.'` is allowed in a `Label`, it is more
+general than an actual representation format that doesn't include some form of escape characters. For a
+full-fledged DNS library look elsewhere (e.g., [`domain`](https://docs.rs/domain/latest/domain/)).
+
The purpose of this library is to allow efficient customization of domain name parsing while still retaining
the hierarchical structure of a domain. Depending on one’s use case, allowed formats and characters can
differ. If one wants to conform to the [Domain Name System (DNS)](https://www.rfc-editor.org/rfc/rfc2181),
all octets are allowed; but conforming to [RFC 1123](https://www.rfc-editor.org/rfc/rfc1123) or
[RFC 5891](https://datatracker.ietf.org/doc/html/rfc5891) requires stricter formats and a reduced character
-set.
-
-### Status
-
-This package is actively maintained.
-
+set.
+
+## License
+
+Licensed under either of
+
+* Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE) or http://www.apache.org/licenses/LICENSE-2.0).
+* MIT license ([LICENSE-MIT](LICENSE-MIT) or http://opensource.org/licenses/MIT).
+
+at your option.
+
+## Contribution
+
+Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you,
+as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.
+
+### Status
+
The crate is only tested on the `x86_64-unknown-linux-gnu` and `x86_64-unknown-openbsd` targets, but
it should work on any platform.
diff --git a/build.rs b/build.rs
@@ -1,12 +0,0 @@
-use rustc_version::{version_meta, Channel};
-
-fn main() {
- // Set cfg flags depending on release channel
- let channel = match version_meta().unwrap().channel {
- Channel::Stable => "CHANNEL_STABLE",
- Channel::Beta => "CHANNEL_BETA",
- Channel::Nightly => "CHANNEL_NIGHTLY",
- Channel::Dev => "CHANNEL_DEV",
- };
- println!("cargo:rustc-cfg={}", channel)
-}
diff --git a/src/char_set.rs b/src/char_set.rs
@@ -1,13 +1,9 @@
use core::{
+ error::Error,
fmt::{self, Display, Formatter},
str,
};
-#[cfg(feature = "std")]
-extern crate std;
-#[cfg(feature = "std")]
-use std::error::Error;
/// Error returned from [`AllowedAscii::try_from_unique_ascii`].
-#[allow(clippy::exhaustive_enums)]
#[derive(Clone, Copy, Debug, Eq, Hash, Ord, PartialEq, PartialOrd)]
pub enum AsciiErr {
/// Since `AllowedAscii` only allows unique ASCII characters and doesn't allow `b'.'`, the maximum count is
@@ -22,7 +18,6 @@ pub enum AsciiErr {
Duplicate(u8),
}
impl Display for AsciiErr {
- #[allow(unsafe_code)]
#[inline]
fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
match *self {
@@ -35,16 +30,15 @@ impl Display for AsciiErr {
Self::Contains46 => f.write_str("allowed ASCII contains '.'"),
Self::Duplicate(byt) => {
let input = [byt];
- // SAFETY:
- // This variant is only used _after_ verifying the passed `u8`s
- // are valid ASCII.
- let val = unsafe { str::from_utf8_unchecked(input.as_slice()) };
- write!(f, "allowed ASCII has the duplicate value '{val}'")
+ if let Ok(val) = str::from_utf8(input.as_slice()) {
+ write!(f, "allowed ASCII has the duplicate value '{val}'")
+ } else {
+ write!(f, "allowed ASCII has the invalid value '{byt}'")
+ }
}
}
}
}
-#[cfg(feature = "std")]
impl Error for AsciiErr {}
/// Container of the ASCII `u8`s that are allowed to appear in a [`crate::dom::Label`]. Note that while
/// [`crate::dom::Domain`] treats ASCII uppercase letters as lowercase, it still depends on such `u8`s being
@@ -105,7 +99,11 @@ impl<T: AsRef<[u8]>> AllowedAscii<T> {
/// use ascii_domain::char_set;
/// assert!(char_set::ASCII_LETTERS.len() == 52);
/// ```
- #[allow(clippy::as_conversions, clippy::cast_possible_truncation)]
+ #[expect(
+ clippy::as_conversions,
+ clippy::cast_possible_truncation,
+ reason = "comment justifies its correctness"
+ )]
#[inline]
#[must_use]
pub fn len(&self) -> u8 {
diff --git a/src/dom.rs b/src/dom.rs
@@ -1,12 +1,11 @@
extern crate alloc;
-#[cfg(feature = "std")]
-extern crate std;
use crate::char_set::{AllowedAscii, ASCII_HYPHEN_DIGITS_LETTERS};
use alloc::{string::String, vec::Vec};
use core::{
borrow::Borrow,
cmp::Ordering,
convert::{self, AsRef},
+ error::Error,
fmt::{self, Display, Formatter},
hash::{Hash, Hasher},
iter::FusedIterator,
@@ -14,8 +13,6 @@ use core::{
ops::Deref,
str,
};
-#[cfg(feature = "std")]
-use std::error::Error;
/// The `AllowedAscii` used by `Rfc1123Domain`.
static RFC_CHARS: &AllowedAscii<[u8; 63]> = &ASCII_HYPHEN_DIGITS_LETTERS;
/// Returned by [`Domain::cmp_by_domain_ordering`]. It is more informative than [`Ordering`] in that it
@@ -24,7 +21,6 @@ static RFC_CHARS: &AllowedAscii<[u8; 63]> = &ASCII_HYPHEN_DIGITS_LETTERS;
///
/// Another way to view this is that [`Self::Shorter`] is "closer" to being [`Self::Equal`] than [`Self::Less`]
/// since the `Domain`s are still part of the same branch in the DNS hierarchy. Ditto for [`Self::Longer`].
-#[allow(clippy::exhaustive_enums)]
#[derive(Clone, Copy, Debug, Eq, Hash, Ord, PartialEq, PartialOrd)]
pub enum DomainOrdering {
/// The `Domain` is less than another since a `Label` was less.
@@ -74,10 +70,10 @@ impl From<DomainOrdering> for Ordering {
/// case.
///
/// [^note]: It is a common misconception that the max length of a domain is 255, but that is only true for
-/// domains in _wire_ format. In representation format, which `Domain` can be thought of when only visible
-/// ASCII bytes are used, the max length is 253 when the last byte is not `b'.'`; otherwise the max length is
-/// 254. This is due to the fact that there is no way to explicitly represent the root label which in wire format
-/// contributes one byte due to each label being preceded by the octet that represents its length.
+/// domains in _wire_ format. In representation format, which `Domain` can be thought of when only visible
+/// ASCII bytes are used, the max length is 253 when the last byte is not `b'.'`; otherwise the max length is
+/// 254. This is due to the fact that there is no way to explicitly represent the root label which in wire format
+/// contributes one byte due to each label being preceded by the octet that represents its length.
///
/// Note this only contains `T`, so this is allocation-free and the same size as `T`.
#[derive(Clone, Copy, Debug)]
@@ -122,7 +118,11 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com.", &ASCII_LOWERCASE).unwrap().contains_trailing_dot());
/// ```
- #[allow(clippy::arithmetic_side_effects, clippy::indexing_slicing)]
+ #[expect(
+ clippy::arithmetic_side_effects,
+ clippy::indexing_slicing,
+ reason = "comments explain their correctness"
+ )]
#[inline]
pub fn contains_trailing_dot(&self) -> bool {
let bytes = self.value.as_ref();
@@ -161,21 +161,24 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com.", &ASCII_LOWERCASE).unwrap().len().get() == 11);
/// ```
- #[inline]
- #[allow(
+ #[expect(
unsafe_code,
+ reason = "we enforce nonzero lengths, so NonZeroU8::new_unchecked is fine"
+ )]
+ #[expect(
clippy::arithmetic_side_effects,
clippy::as_conversions,
- clippy::cast_lossless,
- clippy::cast_possible_truncation
+ clippy::cast_possible_truncation,
+ reason = "comments justify their correctness"
)]
+ #[inline]
pub fn len(&self) -> NonZeroU8 {
// No fear of underflow since the length of `value` is at least 1 _not including_ the
// trailing `b'.'` if there was one.
// `true as usize` is guaranteed to be 1 and `false as usize` is guaranteed to be 0.
// No fear of truncation either since the length is guaranteed to be less than 255.
// `Domain` is immutable ensuring such invariants are kept.
- let len = (self.value.as_ref().len() - self.contains_trailing_dot() as usize) as u8;
+ let len = (self.value.as_ref().len() - usize::from(self.contains_trailing_dot())) as u8;
// SAFETY:
// The only way to construct a `Domain` is via `try_from_bytes` which ensures `len` is
// is at least 1.
@@ -199,10 +202,9 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// # Errors
///
/// Returns [`DomainErr`] iff `v.as_ref()` is an invalid `Domain`.
- #[allow(
+ #[expect(
clippy::arithmetic_side_effects,
- clippy::indexing_slicing,
- clippy::into_iter_on_ref
+ reason = "comment justifies its correctness"
)]
#[inline]
pub fn try_from_bytes<T2: AsRef<[u8]>>(
@@ -210,29 +212,29 @@ impl<T: AsRef<[u8]>> Domain<T> {
allowed_ascii: &AllowedAscii<T2>,
) -> Result<Self, DomainErr> {
let val = v.as_ref();
- let value = match val.last() {
- None => return Err(DomainErr::Empty),
- Some(byt) => {
- if *byt == b'.' {
- if val.len() == 1 {
- return Err(DomainErr::RootDomain);
- }
- // We know `val.len` is at least 2.
- let len = val.len() - 1;
- if val[len - 1] == b'.' {
- return Err(DomainErr::EmptyLabel);
- }
- &val[..len]
+ let value = val
+ .split_last()
+ .ok_or(DomainErr::Empty)
+ .and_then(|(lst, rem)| {
+ if *lst == b'.' {
+ rem.split_last()
+ .ok_or(DomainErr::RootDomain)
+ .and_then(|(lst_2, _)| {
+ if *lst_2 == b'.' {
+ Err(DomainErr::EmptyLabel)
+ } else {
+ Ok(rem)
+ }
+ })
} else {
- val
+ Ok(val)
}
- }
- };
+ })?;
if value.len() > 253 {
Err(DomainErr::LenExceeds253(value.len()))
} else {
value
- .into_iter()
+ .iter()
.try_fold(0, |label_len, byt| {
let b = *byt;
if b == b'.' {
@@ -285,7 +287,7 @@ impl<T: AsRef<[u8]>> Domain<T> {
if self == right {
true
} else {
- self.into_iter()
+ self.iter()
.zip(right)
.try_fold(
(),
@@ -319,15 +321,14 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// assert!(matches!(dom1.cmp_by_domain_ordering(&dom3), DomainOrdering::Less));
/// assert!(matches!(dom3.cmp_by_domain_ordering(&dom1), DomainOrdering::Greater));
/// ```
- #[allow(clippy::unreachable)]
#[inline]
pub fn cmp_by_domain_ordering<T2: AsRef<[u8]>>(&self, right: &Domain<T2>) -> DomainOrdering {
// Faster to compare the entire value when we can instead each `Label`.
if self == right {
DomainOrdering::Equal
} else {
- let mut right_iter = right.into_iter();
- self.into_iter()
+ let mut right_iter = right.iter();
+ self.iter()
.try_fold(false, |_, label| {
right_iter
.next()
@@ -388,10 +389,10 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com", &ASCII_LOWERCASE).unwrap().first_label().as_str() == "example");
/// ```
- #[allow(clippy::unreachable)]
+ #[expect(clippy::unreachable, reason = "bug in code, so we want to crash")]
#[inline]
pub fn first_label(&self) -> Label<'_> {
- self.into_iter()
+ self.iter()
.next_back()
.unwrap_or_else(|| unreachable!("there is a bug in Domain::try_from_bytes"))
}
@@ -403,10 +404,10 @@ impl<T: AsRef<[u8]>> Domain<T> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com", &ASCII_LOWERCASE).unwrap().tld().as_str() == "com");
/// ```
- #[allow(clippy::unreachable)]
+ #[expect(clippy::unreachable, reason = "bug in code, so we want to crash")]
#[inline]
pub fn tld(&self) -> Label<'_> {
- self.into_iter()
+ self.iter()
.next()
.unwrap_or_else(|| unreachable!("there is a bug in Domain::try_from_bytes"))
}
@@ -485,7 +486,10 @@ impl<T: AsRef<[u8]>> Deref for Domain<T> {
}
}
impl From<Domain<Vec<u8>>> for Domain<String> {
- #[allow(unsafe_code)]
+ #[expect(
+ unsafe_code,
+ reason = "we enforce ASCII, so String::from_utf8_unchecked is fine"
+ )]
#[inline]
fn from(value: Domain<Vec<u8>>) -> Self {
// SAFETY:
@@ -519,7 +523,10 @@ impl From<Domain<String>> for Domain<Vec<u8>> {
}
}
impl<'a: 'b, 'b> From<Domain<&'a [u8]>> for Domain<&'b str> {
- #[allow(unsafe_code)]
+ #[expect(
+ unsafe_code,
+ reason = "we enforce ASCII, so str::from_utf8_unchecked is fine"
+ )]
#[inline]
fn from(value: Domain<&'a [u8]>) -> Self {
// SAFETY:
@@ -565,12 +572,15 @@ impl<'a: 'b, 'b> From<Domain<&'a str>> for &'b str {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LETTERS};
/// assert!(<&str>::from(Domain::try_from_bytes("Example.com.", &ASCII_LETTERS).unwrap()) == "Example.com");
/// ```
- #[allow(unsafe_code, clippy::as_conversions, clippy::indexing_slicing)]
+ #[expect(
+ unsafe_code,
+ reason = "we enforce ASCII, so str::from_utf8_unchecked is fine"
+ )]
+ #[expect(clippy::indexing_slicing, reason = "comment justifies its correctness")]
#[inline]
fn from(value: Domain<&'a str>) -> Self {
- // `value.len().get() as usize` is fine since it's a positive `u8`.
// Indexing won't `panic` since `value.len()` is at most as long as `value.value`.
- let utf8 = &value.value.as_bytes()[..value.len().get() as usize];
+ let utf8 = &value.value.as_bytes()[..usize::from(value.len().get())];
// SAFETY:
// Only ASCII is allowed, so this is fine.
unsafe { str::from_utf8_unchecked(utf8) }
@@ -605,16 +615,14 @@ impl<'a: 'b, 'b> From<Domain<&'a [u8]>> for &'b [u8] {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LETTERS};
/// assert!(<&[u8]>::from(Domain::try_from_bytes(b"Example.com.".as_slice(), &ASCII_LETTERS).unwrap()) == b"Example.com");
/// ```
- #[allow(clippy::as_conversions, clippy::indexing_slicing)]
+ #[expect(clippy::indexing_slicing, reason = "comment justifies its correctness")]
#[inline]
fn from(value: Domain<&'a [u8]>) -> Self {
- // `value.len().get() as usize` is fine since it's a positive `u8`.
// Indexing won't `panic` since `value.len()` is at most as long as `value.value`.
- &value.value[..value.len().get() as usize]
+ &value.value[..usize::from(value.len().get())]
}
}
/// Error returned from [`Domain::try_from_bytes`].
-#[allow(clippy::exhaustive_enums)]
#[derive(Clone, Copy, Debug, Eq, Hash, Ord, PartialEq, PartialOrd)]
pub enum DomainErr {
/// The domain was empty.
@@ -651,7 +659,6 @@ impl Display for DomainErr {
}
}
}
-#[cfg(feature = "std")]
impl Error for DomainErr {}
/// A label of a [`Domain`]. The total length of a `Label` is inclusively between 1 and 63.
#[derive(Clone, Copy, Debug)]
@@ -681,13 +688,12 @@ impl<'a> Label<'a> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com", &ASCII_LOWERCASE).unwrap().into_iter().next().map_or(false, |label| label.is_alphabetic()));
/// ```
- #[allow(clippy::into_iter_on_ref)]
#[inline]
#[must_use]
pub fn is_alphabetic(self) -> bool {
self.value
.as_bytes()
- .into_iter()
+ .iter()
.try_fold((), |(), byt| {
if byt.is_ascii_alphabetic() {
Ok(())
@@ -705,13 +711,12 @@ impl<'a> Label<'a> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_DIGITS_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.123", &ASCII_DIGITS_LOWERCASE).unwrap().into_iter().next().map_or(false, |label| label.is_digits()));
/// ```
- #[allow(clippy::into_iter_on_ref)]
#[inline]
#[must_use]
pub fn is_digits(self) -> bool {
self.value
.as_bytes()
- .into_iter()
+ .iter()
.try_fold((), |(), byt| {
if byt.is_ascii_digit() {
Ok(())
@@ -729,13 +734,12 @@ impl<'a> Label<'a> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_DIGITS_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.1com", &ASCII_DIGITS_LOWERCASE).unwrap().into_iter().next().map_or(false, |label| label.is_alphanumeric()));
/// ```
- #[allow(clippy::into_iter_on_ref)]
#[inline]
#[must_use]
pub fn is_alphanumeric(self) -> bool {
self.value
.as_bytes()
- .into_iter()
+ .iter()
.try_fold((), |(), byt| {
if byt.is_ascii_alphanumeric() {
Ok(())
@@ -753,13 +757,12 @@ impl<'a> Label<'a> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_HYPHEN_DIGITS_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.1-com", &ASCII_HYPHEN_DIGITS_LOWERCASE).unwrap().into_iter().next().map_or(false, |label| label.is_hyphen_or_alphanumeric()));
/// ```
- #[allow(clippy::into_iter_on_ref)]
#[inline]
#[must_use]
pub fn is_hyphen_or_alphanumeric(self) -> bool {
self.value
.as_bytes()
- .into_iter()
+ .iter()
.try_fold((), |(), byt| {
if *byt == b'-' || byt.is_ascii_alphanumeric() {
Ok(())
@@ -777,7 +780,15 @@ impl<'a> Label<'a> {
/// use ascii_domain::{dom::Domain, char_set::ASCII_LOWERCASE};
/// assert!(Domain::try_from_bytes("example.com.", &ASCII_LOWERCASE).unwrap().into_iter().next().map_or(false, |label| label.len().get() == 3));
/// ```
- #[allow(unsafe_code, clippy::as_conversions, clippy::cast_possible_truncation)]
+ #[expect(
+ unsafe_code,
+ reason = "we enforce label lengths, so NonZeroU8::new_unchecked is fine"
+ )]
+ #[expect(
+ clippy::as_conversions,
+ clippy::cast_possible_truncation,
+ reason = "comments justify their correctness"
+ )]
#[inline]
#[must_use]
pub const fn len(self) -> NonZeroU8 {
@@ -864,17 +875,20 @@ pub struct LabelIter<'a> {
}
impl<'a> Iterator for LabelIter<'a> {
type Item = Label<'a>;
- #[allow(
+ #[expect(
unsafe_code,
+ reason = "we only allow ASCII, so str::from_utf8_unchecked is fine"
+ )]
+ #[expect(
clippy::arithmetic_side_effects,
clippy::indexing_slicing,
- clippy::into_iter_on_ref
+ reason = "comments justify their correctness"
)]
#[inline]
fn next(&mut self) -> Option<Self::Item> {
(!self.domain.is_empty()).then(|| {
self.domain
- .into_iter()
+ .iter()
.rev()
.try_fold(1, |count, byt| {
if *byt == b'.' {
@@ -908,7 +922,6 @@ impl<'a> Iterator for LabelIter<'a> {
{
self.next_back()
}
- #[allow(clippy::integer_division)]
#[inline]
fn size_hint(&self) -> (usize, Option<usize>) {
if self.domain.is_empty() {
@@ -920,7 +933,7 @@ impl<'a> Iterator for LabelIter<'a> {
// The min size of a `Label` is 1; and all but the last have a `b'.'` that follow it.
// This means the max number of `Label`s is the ceiling of the length divided by 2.
(
- (self.domain.len() / 64).max(1),
+ (self.domain.len() >> 6).max(1),
Some(self.domain.len().div_ceil(2)),
)
}
@@ -928,17 +941,20 @@ impl<'a> Iterator for LabelIter<'a> {
}
impl FusedIterator for LabelIter<'_> {}
impl DoubleEndedIterator for LabelIter<'_> {
- #[allow(
+ #[expect(
unsafe_code,
+ reason = "we only allow ASCII, so str::from_utf8_unchecked is fine"
+ )]
+ #[expect(
clippy::arithmetic_side_effects,
clippy::indexing_slicing,
- clippy::into_iter_on_ref
+ reason = "comments justify their correctness"
)]
#[inline]
fn next_back(&mut self) -> Option<Self::Item> {
(!self.domain.is_empty()).then(|| {
self.domain
- .into_iter()
+ .iter()
.try_fold(0, |count, byt| {
if *byt == b'.' {
// `count + 1` < `self.domain.len()` since there is at least one more `Label` and `Label`s
@@ -996,7 +1012,6 @@ impl<'a> IntoIterator for Domain<&'a [u8]> {
}
}
/// Error returned from [`Rfc1123Domain::try_from`] and [`Rfc1123Domain::try_from_bytes`].
-#[allow(clippy::exhaustive_enums)]
#[derive(Clone, Copy, Debug, Eq, Hash, Ord, PartialEq, PartialOrd)]
pub enum Rfc1123Err {
/// The inputs was not a valid [`Domain`].
@@ -1022,7 +1037,6 @@ impl Display for Rfc1123Err {
}
}
}
-#[cfg(feature = "std")]
impl Error for Rfc1123Err {}
/// **TL;DR** Wrapper type around a [`Domain`] that enforces conformance to
/// [RFC 1123](https://www.rfc-editor.org/rfc/rfc1123#page-13).
@@ -1030,7 +1044,7 @@ impl Error for Rfc1123Err {}
/// * Each [`Label`] must only contain ASCII digits, letters, or hyphen.
/// * Each `Label` must not begin or end with a hyphen.
/// * The last `Label` (i.e., TLD) must either contain only ASCII letters or have length of at least five and
-/// begin with `xn--`.
+/// begin with `xn--`.
/// ---
/// Unsurprisingly, RFC 1123 is not super precise as it uses "host name" to mean label and also domain:
/// "Host software MUST handle host names \[labels\] of up to 63 characters and SHOULD handle host
@@ -1053,7 +1067,7 @@ impl Error for Rfc1123Err {}
/// * More relaxed: enforce the "spirit" that the TLD must have the same format of a valid TLD.
/// * Much more relaxed: enforce the "spirit" that the domain cannot have the form of an IPv4 address.
/// * Most relaxed: treat TLDs no differently than other labels (i.e., don't make assumptions about what will be
-/// a valid TLD in the future).
+/// a valid TLD in the future).
///
/// RFC 1123 is not obsolete, and it is clear from more recent RFCs like
/// [RFC 5891](https://www.rfc-editor.org/rfc/rfc5891) that it is designed to be a foundation (i.e., domains that
@@ -1134,13 +1148,12 @@ impl<T: AsRef<[u8]>> Rfc1123Domain<T> {
/// # Errors
///
/// Returns [`Rfc1123Err`] iff `v.as_ref()` is an invalid `Rfc1123Domain`.
- #[allow(
+ #[expect(
clippy::arithmetic_side_effects,
- clippy::as_conversions,
clippy::indexing_slicing,
- clippy::into_iter_on_ref,
- clippy::redundant_else
+ reason = "comments justify their correctness"
)]
+ #[expect(clippy::redundant_else, reason = "prefer else with else-if")]
#[inline]
pub fn try_from_bytes(v: T) -> Result<Self, Rfc1123Err> {
// The easiest implementation would be redirecting to `Domain::try_from_bytes`; and upon success,
@@ -1177,7 +1190,7 @@ impl<T: AsRef<[u8]>> Rfc1123Domain<T> {
} else {
let mut count = 0;
value
- .into_iter()
+ .iter()
.try_fold(0, |label_len, byt| {
let b = *byt;
if b == b'.' {
@@ -1209,10 +1222,12 @@ impl<T: AsRef<[u8]>> Rfc1123Domain<T> {
})
.and_then(|tld_len| {
// `tld_len <= value.len()`.
- let tld = &value[value.len() - tld_len as usize..];
- if (tld_len > 4 && tld[..4] == *b"xn--".as_slice())
+ let tld = &value[value.len() - usize::from(tld_len)..];
+ if (tld
+ .split_at_checked(4)
+ .map_or(false, |(fst, rem)| !rem.is_empty() && fst == b"xn--"))
|| tld
- .into_iter()
+ .iter()
.try_fold((), |(), byt| {
if byt.is_ascii_alphabetic() {
Ok(())
@@ -1410,20 +1425,22 @@ impl<'a: 'b, 'b> From<Rfc1123Domain<&'a str>> for Rfc1123Domain<&'b [u8]> {
}
impl<T: AsRef<[u8]>> TryFrom<Domain<T>> for Rfc1123Domain<T> {
type Error = Rfc1123Err;
- #[allow(
+ #[expect(
clippy::arithmetic_side_effects,
clippy::indexing_slicing,
- clippy::into_iter_on_ref,
- clippy::unreachable
+ clippy::unreachable,
+ reason = "comments explain their correctness"
)]
#[inline]
fn try_from(value: Domain<T>) -> Result<Self, Self::Error> {
- let mut labels = value.into_iter();
+ let mut labels = value.iter();
let tld = labels
.next()
.unwrap_or_else(|| unreachable!("there is a bug in Domain::try_from_bytes"));
if tld.is_alphabetic()
- || (tld.len().get() > 4 && tld.as_bytes()[..4] == *b"xn--".as_slice())
+ || tld
+ .split_at_checked(4)
+ .map_or(false, |(fst, rem)| !rem.is_empty() && fst == "xn--")
{
labels
.try_fold((), |(), label| {
@@ -1435,7 +1452,7 @@ impl<T: AsRef<[u8]>> TryFrom<Domain<T>> for Rfc1123Domain<T> {
} else if bytes[bytes.len() - 1] == b'-' {
Err(Rfc1123Err::LabelEndsWithAHyphen)
} else {
- bytes.into_iter().try_fold((), |(), byt| match *byt {
+ bytes.iter().try_fold((), |(), byt| match *byt {
b'-' | b'0'..=b'9' | b'A'..=b'Z' | b'a'..=b'z' => Ok(()),
val => Err(Rfc1123Err::DomainErr(DomainErr::InvalidByte(val))),
})
@@ -1522,8 +1539,8 @@ mod tests {
assert!(Domain::try_from_bytes("aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa", &allowed_ascii).map_or(false, |d| d.len().get() == 253 ));
// Test max labels.
assert!(Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a", &allowed_ascii).map_or_else(|e| e == DomainErr::LenExceeds253(255), |_| false));
- assert!(Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a", &allowed_ascii).map_or(false, |d| d.into_iter().count() == 127 && d.len().get() == 253));
- assert!(Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.", &allowed_ascii).map_or(false, |d| d.into_iter().count() == 127 && d.len().get() == 253));
+ assert!(Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a", &allowed_ascii).map_or(false, |d| d.iter().count() == 127 && d.len().get() == 253));
+ assert!(Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.", &allowed_ascii).map_or(false, |d| d.iter().count() == 127 && d.len().get() == 253));
// Test removal of trailing '.'.
assert!(
Domain::try_from_bytes("com.", &allowed_ascii).map_or(false, |d| d.as_str() == "com")
@@ -1589,7 +1606,7 @@ mod tests {
let allowed_ascii = ASCII_FIREFOX;
assert!(
Domain::try_from_bytes("www.example.com", &allowed_ascii).map_or(false, |d| {
- let mut iter = d.into_iter();
+ let mut iter = d.iter();
let Some(l) = iter.next() else {
return false;
};
@@ -1611,7 +1628,7 @@ mod tests {
);
assert!(
Domain::try_from_bytes("www.example.com", &allowed_ascii).map_or(false, |d| {
- let mut iter = d.into_iter();
+ let mut iter = d.iter();
let Some(l) = iter.next_back() else {
return false;
};
@@ -1635,7 +1652,7 @@ mod tests {
);
assert!(
Domain::try_from_bytes("www.example.com", &allowed_ascii).map_or(false, |d| {
- let mut iter = d.into_iter();
+ let mut iter = d.iter();
let Some(l) = iter.next_back() else {
return false;
};
@@ -1772,8 +1789,8 @@ mod tests {
assert!(Rfc1123Domain::try_from_bytes("aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa").map_or(false, |d| d.len().get() == 253 ));
// Test max labels.
assert!(Rfc1123Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a").map_or_else(|e| e == Rfc1123Err::DomainErr(DomainErr::LenExceeds253(255)), |_| false));
- assert!(Rfc1123Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a").map_or(false, |d| d.into_iter().count() == 127 && d.len().get() == 253));
- assert!(Rfc1123Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.").map_or(false, |d| d.into_iter().count() == 127 && d.len().get() == 253));
+ assert!(Rfc1123Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a").map_or(false, |d| d.iter().count() == 127 && d.len().get() == 253));
+ assert!(Rfc1123Domain::try_from_bytes("a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.a.").map_or(false, |d| d.iter().count() == 127 && d.len().get() == 253));
// Test removal of trailing '.'.
assert!(Rfc1123Domain::try_from_bytes("com.").map_or(false, |d| d.as_str() == "com"));
// Test single label.
diff --git a/src/lib.rs b/src/lib.rs
@@ -13,8 +13,8 @@
//! all octets are allowed; but conforming to [RFC 1123](https://www.rfc-editor.org/rfc/rfc1123) or
//! [RFC 5891](https://datatracker.ietf.org/doc/html/rfc5891) requires stricter formats and a reduced character
//! set.
+#![cfg_attr(docsrs, feature(doc_cfg))]
#![no_std]
-#![cfg_attr(all(doc, CHANNEL_NIGHTLY), feature(doc_auto_cfg))]
#![deny(
future_incompatible,
let_underscore,
@@ -38,14 +38,15 @@
clippy::style,
clippy::suspicious
)]
-#![allow(
+#![expect(
clippy::blanket_clippy_restriction_lints,
+ clippy::exhaustive_enums,
clippy::implicit_return,
clippy::min_ident_chars,
clippy::missing_trait_methods,
- clippy::single_call_fn,
+ clippy::question_mark_used,
clippy::single_char_lifetime_names,
- clippy::unseparated_literal_suffix
+ reason = "noisy, opinionated, and likely doesn't prevent bugs or improve APIs"
)]
/// Contains [`char_set::AllowedAscii`] which is how one dictates the character set [`dom::Domain::try_from_bytes`]
/// uses.
@@ -56,5 +57,6 @@ pub mod char_set;
pub mod dom;
/// Contains a Serde [`Visitor`](https://docs.rs/serde/latest/serde/de/trait.Visitor.html) that can be used to help
/// deserialize [`dom::Domain`] wrappers.
+#[cfg_attr(docsrs, doc(cfg(feature = "serde")))]
#[cfg(feature = "serde")]
pub mod serde;
diff --git a/src/serde.rs b/src/serde.rs
@@ -1,4 +1,3 @@
-#![cfg(feature = "serde")]
extern crate alloc;
use crate::{
char_set::{AllowedAscii, PRINTABLE_ASCII},
@@ -56,7 +55,10 @@ impl<T: AsRef<[u8]>> Serialize for Rfc1123Domain<T> {
/// }
/// }
/// ```
-#[allow(clippy::partial_pub_fields)]
+#[expect(
+ clippy::partial_pub_fields,
+ reason = "we don't expost PhantomData for obvious reasons, so this is fine"
+)]
#[derive(Clone, Copy)]
pub struct DomainVisitor<'a, T, T2> {
/// Phantom.
@@ -65,8 +67,6 @@ pub struct DomainVisitor<'a, T, T2> {
pub allowed_ascii: &'a AllowedAscii<T>,
}
/// Converts `DomainErr` to a Serde `de::Error`.
-#[allow(clippy::as_conversions, clippy::cast_lossless)]
-#[inline]
fn dom_err_to_serde<E: de::Error>(value: DomainErr) -> E {
match value {
DomainErr::Empty => E::invalid_length(
@@ -89,7 +89,7 @@ fn dom_err_to_serde<E: de::Error>(value: DomainErr) -> E {
&"a valid domain containing labels of length inclusively between 1 and 63",
),
DomainErr::InvalidByte(byt) => E::invalid_value(
- Unexpected::Unsigned(byt as u64),
+ Unexpected::Unsigned(u64::from(byt)),
&"a valid domain containing only the supplied ASCII subset",
),
}
@@ -167,7 +167,6 @@ impl<'de: 'a, 'a> Deserialize<'de> for Domain<&'a str> {
}
}
/// Converts `Rfc1123Err` to a Serde `de::Error`.
-#[inline]
fn rfc_err_to_serde<E: de::Error>(value: Rfc1123Err) -> E {
match value {
Rfc1123Err::DomainErr(err) => dom_err_to_serde(err),
@@ -181,11 +180,9 @@ fn rfc_err_to_serde<E: de::Error>(value: Rfc1123Err) -> E {
struct Rfc1123Visitor<T>(PhantomData<fn() -> T>);
impl<'de: 'a, 'a> Visitor<'de> for Rfc1123Visitor<&'a str> {
type Value = Rfc1123Domain<&'a str>;
- #[inline]
fn expecting(&self, formatter: &mut Formatter<'_>) -> fmt::Result {
formatter.write_str("Rfc1123Domain")
}
- #[inline]
fn visit_borrowed_str<E>(self, v: &'de str) -> Result<Self::Value, E>
where
E: de::Error,
@@ -195,18 +192,15 @@ impl<'de: 'a, 'a> Visitor<'de> for Rfc1123Visitor<&'a str> {
}
impl<'de> Visitor<'de> for Rfc1123Visitor<String> {
type Value = Rfc1123Domain<String>;
- #[inline]
fn expecting(&self, formatter: &mut Formatter<'_>) -> fmt::Result {
formatter.write_str("Rfc1123Domain")
}
- #[inline]
fn visit_string<E>(self, v: String) -> Result<Self::Value, E>
where
E: de::Error,
{
Self::Value::try_from_bytes(v).map_err(|err| rfc_err_to_serde(err))
}
- #[inline]
fn visit_str<E>(self, v: &str) -> Result<Self::Value, E>
where
E: de::Error,
