1// Copyright 2016 The rust-url developers.
2//
3// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
4// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
5// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
6// option. This file may not be copied, modified, or distributed
7// except according to those terms.
8
9//! This Rust crate implements IDNA
10//! [per the WHATWG URL Standard](https://url.spec.whatwg.org/#idna).
11//!
12//! It also exposes the underlying algorithms from [*Unicode IDNA Compatibility Processing*
13//! (Unicode Technical Standard #46)](http://www.unicode.org/reports/tr46/)
14//! and [Punycode (RFC 3492)](https://tools.ietf.org/html/rfc3492).
15//!
16//! Quoting from [UTS #46’s introduction](http://www.unicode.org/reports/tr46/#Introduction):
17//!
18//! > Initially, domain names were restricted to ASCII characters.
19//! > A system was introduced in 2003 for internationalized domain names (IDN).
20//! > This system is called Internationalizing Domain Names for Applications,
21//! > or IDNA2003 for short.
22//! > This mechanism supports IDNs by means of a client software transformation
23//! > into a format known as Punycode.
24//! > A revision of IDNA was approved in 2010 (IDNA2008).
25//! > This revision has a number of incompatibilities with IDNA2003.
26//! >
27//! > The incompatibilities force implementers of client software,
28//! > such as browsers and emailers,
29//! > to face difficult choices during the transition period
30//! > as registries shift from IDNA2003 to IDNA2008.
31//! > This document specifies a mechanism
32//! > that minimizes the impact of this transition for client software,
33//! > allowing client software to access domains that are valid under either system.
34#![no_std]
35
36// For forwards compatibility
37#[cfg(feature = "std")]
38extern crate std;
39
40extern crate alloc;
41
42#[cfg(not(feature = "alloc"))]
43compile_error!("the `alloc` feature must be enabled");
44
45// Avoid a breaking change if in the future there's a use case for
46// having a Bring-Your-Own-ICU4X-Data constructor for `Uts46` and
47// not also having compiled data in the binary.
48#[cfg(not(feature = "compiled_data"))]
49compile_error!("the `compiled_data` feature must be enabled");
50
51use alloc::borrow::Cow;
52use alloc::string::String;
53pub use uts46::AsciiDenyList;
54use uts46::Uts46;
55
56mod deprecated;
57pub mod punycode;
58pub mod uts46;
59
60#[allow(deprecated)]
61pub use crate::deprecated::{Config, Idna};
62
63/// Type indicating that there were errors during UTS #46 processing.
64#[derive(Default, Debug)]
65#[non_exhaustive]
66pub struct Errors {}
67
68impl From<Errors> for Result<(), Errors> {
69 fn from(e: Errors) -> Result<(), Errors> {
70 Err(e)
71 }
72}
73
74#[cfg(feature = "std")]
75impl std::error::Error for Errors {}
76
77#[cfg(not(feature = "std"))]
78impl core::error::Error for Errors {}
79
80impl core::fmt::Display for Errors {
81 fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result {
82 core::fmt::Debug::fmt(self, f)
83 }
84}
85
86/// The [domain to ASCII](https://url.spec.whatwg.org/#concept-domain-to-ascii) algorithm;
87/// version returning a `Cow`.
88///
89/// Most applications should be using this function rather than the sibling functions,
90/// and most applications should pass [`AsciiDenyList::URL`] as the second argument.
91/// Passing [`AsciiDenyList::URL`] as the second argument makes this function also
92/// perform the [forbidden domain code point](https://url.spec.whatwg.org/#forbidden-domain-code-point)
93/// check in addition to the [domain to ASCII](https://url.spec.whatwg.org/#concept-domain-to-ascii)
94/// algorithm.
95///
96/// Returns the ASCII representation a domain name,
97/// normalizing characters (upper-case to lower-case and other kinds of equivalence)
98/// and using Punycode as necessary.
99///
100/// This process may fail.
101///
102/// If you have a `&str` instead of `&[u8]`, just call `.to_bytes()` on it before
103/// passing it to this function. It's still preferable to use this function over
104/// the sibling functions that take `&str`.
105pub fn domain_to_ascii_cow(
106 domain: &[u8],
107 ascii_deny_list: AsciiDenyList,
108) -> Result<Cow<'_, str>, Errors> {
109 Uts46::new().to_ascii(
110 domain,
111 ascii_deny_list,
112 uts46::Hyphens::Allow,
113 uts46::DnsLength::Ignore,
114 )
115}
116
117/// The [domain to ASCII](https://url.spec.whatwg.org/#concept-domain-to-ascii) algorithm;
118/// version returning `String` and no ASCII deny list (i.e. _UseSTD3ASCIIRules=false_).
119///
120/// This function exists for backward-compatibility. Consider using [`domain_to_ascii_cow`]
121/// instead.
122///
123/// Return the ASCII representation a domain name,
124/// normalizing characters (upper-case to lower-case and other kinds of equivalence)
125/// and using Punycode as necessary.
126///
127/// This process may fail.
128pub fn domain_to_ascii(domain: &str) -> Result<String, Errors> {
129 domain_to_ascii_cow(domain.as_bytes(), AsciiDenyList::EMPTY).map(|cow: Cow<'_, str>| cow.into_owned())
130}
131
132/// The [domain to ASCII](https://url.spec.whatwg.org/#concept-domain-to-ascii) algorithm,
133/// with the `beStrict` flag set.
134///
135/// Note that this rejects various real-world names including:
136/// * YouTube CDN nodes
137/// * Some GitHub user pages
138/// * Pseudo-hosts used by various TXT record-based protocols.
139pub fn domain_to_ascii_strict(domain: &str) -> Result<String, Errors> {
140 Uts46::new()
141 .to_ascii(
142 domain.as_bytes(),
143 uts46::AsciiDenyList::STD3,
144 uts46::Hyphens::Check,
145 uts46::DnsLength::Verify,
146 )
147 .map(|cow: Cow<'_, str>| cow.into_owned())
148}
149
150/// The [domain to Unicode](https://url.spec.whatwg.org/#concept-domain-to-unicode) algorithm;
151/// version returning `String` and no ASCII deny list (i.e. _UseSTD3ASCIIRules=false_).
152///
153/// This function exists for backward-compatibility. Consider using [`Uts46::to_user_interface`]
154/// or [`Uts46::to_unicode`].
155///
156/// Return the Unicode representation of a domain name,
157/// normalizing characters (upper-case to lower-case and other kinds of equivalence)
158/// and decoding Punycode as necessary.
159///
160/// If the second item of the tuple indicates an error, the first item of the tuple
161/// denotes errors using the REPLACEMENT CHARACTERs in order to be able to illustrate
162/// errors to the user. When the second item of the return tuple signals an error,
163/// the first item of the tuple must not be used in a network protocol.
164pub fn domain_to_unicode(domain: &str) -> (String, Result<(), Errors>) {
165 let (cow: Cow<'_, str>, result: Result<(), Errors>) = Uts46::new().to_unicode(
166 domain_name:domain.as_bytes(),
167 ascii_deny_list:uts46::AsciiDenyList::EMPTY,
168 uts46::Hyphens::Allow,
169 );
170 (cow.into_owned(), result)
171}
172