diff options
Diffstat (limited to 'vendor/icu_locid/src/extensions/unicode')
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/attribute.rs | 34 | ||||
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/attributes.rs | 115 | ||||
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/key.rs | 31 | ||||
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/keywords.rs | 404 | ||||
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/mod.rs | 233 | ||||
| -rw-r--r-- | vendor/icu_locid/src/extensions/unicode/value.rs | 199 |
6 files changed, 1016 insertions, 0 deletions
diff --git a/vendor/icu_locid/src/extensions/unicode/attribute.rs b/vendor/icu_locid/src/extensions/unicode/attribute.rs new file mode 100644 index 000000000..ba4b70924 --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/attribute.rs @@ -0,0 +1,34 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +impl_tinystr_subtag!( + /// An attribute used in a set of [`Attributes`](super::Attributes). + /// + /// An attribute has to be a sequence of alphanumerical characters no + /// shorter than three and no longer than eight characters. + /// + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Attribute; + /// use icu::locid::extensions_unicode_attribute as attribute; + /// + /// let attr: Attribute = + /// "buddhist".parse().expect("Failed to parse an Attribute."); + /// + /// assert_eq!(attr, attribute!("buddhist")); + /// ``` + Attribute, + extensions::unicode::Attribute, + extensions_unicode_attribute, + 3..=8, + s, + s.is_ascii_alphanumeric(), + s.to_ascii_lowercase(), + s.is_ascii_alphanumeric() && s.is_ascii_lowercase(), + InvalidExtension, + ["foo12"], + ["no", "toolooong"], +); diff --git a/vendor/icu_locid/src/extensions/unicode/attributes.rs b/vendor/icu_locid/src/extensions/unicode/attributes.rs new file mode 100644 index 000000000..1f9536bfa --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/attributes.rs @@ -0,0 +1,115 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +use super::Attribute; + +use alloc::vec::Vec; +use core::ops::Deref; + +/// A set of [`Attribute`] elements as defined in [`Unicode Extension Attributes`]. +/// +/// [`Unicode Extension Attributes`]: https://unicode.org/reports/tr35/tr35.html#u_Extension +/// +/// # Examples +/// +/// ``` +/// use icu::locid::extensions::unicode::{Attribute, Attributes}; +/// +/// let attribute1: Attribute = +/// "foobar".parse().expect("Failed to parse a variant subtag."); +/// +/// let attribute2: Attribute = "testing" +/// .parse() +/// .expect("Failed to parse a variant subtag."); +/// let mut v = vec![attribute1, attribute2]; +/// v.sort(); +/// v.dedup(); +/// +/// let attributes: Attributes = Attributes::from_vec_unchecked(v); +/// assert_eq!(attributes.to_string(), "foobar-testing"); +/// ``` +#[derive(Default, Debug, PartialEq, Eq, Clone, Hash, PartialOrd, Ord)] +pub struct Attributes(Vec<Attribute>); + +impl Attributes { + /// Returns a new empty set of attributes. Same as [`default()`](Default::default()), but is `const`. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Attributes; + /// + /// assert_eq!(Attributes::new(), Attributes::default()); + /// ``` + #[inline] + pub const fn new() -> Self { + Self(Vec::new()) + } + + /// A constructor which takes a pre-sorted list of [`Attribute`] elements. + /// + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::{Attribute, Attributes}; + /// + /// let attribute1: Attribute = "foobar".parse().expect("Parsing failed."); + /// let attribute2: Attribute = "testing".parse().expect("Parsing failed."); + /// let mut v = vec![attribute1, attribute2]; + /// v.sort(); + /// v.dedup(); + /// + /// let attributes = Attributes::from_vec_unchecked(v); + /// ``` + /// + /// Notice: For performance- and memory-constrained environments, it is recommended + /// for the caller to use [`binary_search`](slice::binary_search) instead of [`sort`](slice::sort) + /// and [`dedup`](Vec::dedup()). + pub fn from_vec_unchecked(input: Vec<Attribute>) -> Self { + Self(input) + } + + /// Empties the [`Attributes`] list. + /// + /// Returns the old list. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::{Attribute, Attributes}; + /// + /// let attribute1: Attribute = "foobar".parse().expect("Parsing failed."); + /// let attribute2: Attribute = "testing".parse().expect("Parsing failed."); + /// let mut v = vec![attribute1, attribute2]; + /// + /// let mut attributes: Attributes = Attributes::from_vec_unchecked(v); + /// + /// assert_eq!(attributes.to_string(), "foobar-testing"); + /// + /// attributes.clear(); + /// + /// assert_eq!(attributes.to_string(), ""); + /// ``` + pub fn clear(&mut self) -> Self { + core::mem::take(self) + } + + pub(crate) fn for_each_subtag_str<E, F>(&self, f: &mut F) -> Result<(), E> + where + F: FnMut(&str) -> Result<(), E>, + { + self.deref().iter().map(|t| t.as_str()).try_for_each(f) + } +} + +impl_writeable_for_subtag_list!(Attributes, "foobar", "testing"); + +impl Deref for Attributes { + type Target = [Attribute]; + + fn deref(&self) -> &[Attribute] { + self.0.deref() + } +} diff --git a/vendor/icu_locid/src/extensions/unicode/key.rs b/vendor/icu_locid/src/extensions/unicode/key.rs new file mode 100644 index 000000000..bdfdd4e5c --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/key.rs @@ -0,0 +1,31 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +impl_tinystr_subtag!( + /// A key used in a list of [`Keywords`](super::Keywords). + /// + /// The key has to be a two ASCII alphanumerical characters long, with the first + /// character being alphanumeric, and the second being alphabetic. + /// + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Key; + /// + /// assert!("ca".parse::<Key>().is_ok()); + /// ``` + Key, + extensions::unicode::Key, + extensions_unicode_key, + 2..=2, + s, + s.all_bytes()[0].is_ascii_alphanumeric() && s.all_bytes()[1].is_ascii_alphabetic(), + s.to_ascii_lowercase(), + (s.all_bytes()[0].is_ascii_lowercase() || s.all_bytes()[0].is_ascii_digit()) + && s.all_bytes()[1].is_ascii_lowercase(), + InvalidExtension, + ["ca", "8a"], + ["a", "a8", "abc"], +); diff --git a/vendor/icu_locid/src/extensions/unicode/keywords.rs b/vendor/icu_locid/src/extensions/unicode/keywords.rs new file mode 100644 index 000000000..dc9a15921 --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/keywords.rs @@ -0,0 +1,404 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +use core::borrow::Borrow; +use core::cmp::Ordering; +use core::iter::FromIterator; +use litemap::LiteMap; + +use super::Key; +use super::Value; +use crate::helpers::ShortVec; +use crate::ordering::SubtagOrderingResult; + +/// A list of [`Key`]-[`Value`] pairs representing functional information +/// about locale's internationnalization preferences. +/// +/// Here are examples of fields used in Unicode: +/// - `hc` - Hour Cycle (`h11`, `h12`, `h23`, `h24`) +/// - `ca` - Calendar (`buddhist`, `gregory`, ...) +/// - `fw` - First Day Of the Week (`sun`, `mon`, `sat`, ...) +/// +/// You can find the full list in [`Unicode BCP 47 U Extension`] section of LDML. +/// +/// [`Unicode BCP 47 U Extension`]: https://unicode.org/reports/tr35/tr35.html#Key_And_Type_Definitions_ +/// +/// # Examples +/// +/// Manually build up a [`Keywords`] object: +/// +/// ``` +/// use icu::locid::extensions::unicode::{Key, Keywords, Value}; +/// +/// let key: Key = "hc".parse().expect("Failed to parse a Key."); +/// let value: Value = "h23".parse().expect("Failed to parse a Value."); +/// let keywords: Keywords = vec![(key, value)].into_iter().collect(); +/// +/// assert_eq!(&keywords.to_string(), "hc-h23"); +/// ``` +/// +/// Access a [`Keywords`] object from a [`Locale`]: +/// +/// ``` +/// use icu::locid::{ +/// extensions_unicode_key as key, extensions_unicode_value as value, +/// Locale, +/// }; +/// +/// let loc: Locale = "und-u-hc-h23-kc-true".parse().expect("Valid BCP-47"); +/// +/// assert_eq!(loc.extensions.unicode.keywords.get(&key!("ca")), None); +/// assert_eq!( +/// loc.extensions.unicode.keywords.get(&key!("hc")), +/// Some(&value!("h23")) +/// ); +/// assert_eq!( +/// loc.extensions.unicode.keywords.get(&key!("kc")), +/// Some(&value!("true")) +/// ); +/// +/// assert_eq!(loc.extensions.unicode.keywords.to_string(), "hc-h23-kc"); +/// ``` +/// +/// [`Locale`]: crate::Locale +#[derive(Clone, PartialEq, Eq, Debug, Default, Hash, PartialOrd, Ord)] +pub struct Keywords(LiteMap<Key, Value, ShortVec<(Key, Value)>>); + +impl Keywords { + /// Returns a new empty list of key-value pairs. Same as [`default()`](Default::default()), but is `const`. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Keywords; + /// + /// assert_eq!(Keywords::new(), Keywords::default()); + /// ``` + #[inline] + pub const fn new() -> Self { + Self(LiteMap::new()) + } + + /// Create a new list of key-value pairs having exactly one pair, callable in a `const` context. + #[inline] + pub const fn new_single(key: Key, value: Value) -> Self { + Self(LiteMap::from_sorted_store_unchecked(ShortVec::new_single( + (key, value), + ))) + } + + /// Returns `true` if there are no keywords. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Keywords; + /// use icu::locid::locale; + /// use icu::locid::Locale; + /// + /// let loc1 = Locale::try_from_bytes(b"und-t-h0-hybrid").unwrap(); + /// let loc2 = locale!("und-u-ca-buddhist"); + /// + /// assert!(loc1.extensions.unicode.keywords.is_empty()); + /// assert!(!loc2.extensions.unicode.keywords.is_empty()); + /// ``` + pub fn is_empty(&self) -> bool { + self.0.is_empty() + } + + /// Returns `true` if the list contains a [`Value`] for the specified [`Key`]. + /// + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::{Key, Keywords, Value}; + /// use litemap::LiteMap; + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// let value: Value = "gregory".parse().expect("Failed to parse a Value."); + /// let keywords: Keywords = vec![(key, value)].into_iter().collect(); + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// assert!(&keywords.contains_key(&key)); + /// ``` + pub fn contains_key<Q>(&self, key: &Q) -> bool + where + Key: Borrow<Q>, + Q: Ord, + { + self.0.contains_key(key) + } + + /// Returns a reference to the [`Value`] corresponding to the [`Key`]. + /// + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::{Key, Keywords, Value}; + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// let value: Value = "buddhist".parse().expect("Failed to parse a Value."); + /// let keywords: Keywords = vec![(key, value)].into_iter().collect(); + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// assert_eq!( + /// keywords.get(&key).map(|v| v.to_string()), + /// Some("buddhist".to_string()) + /// ); + /// ``` + pub fn get<Q>(&self, key: &Q) -> Option<&Value> + where + Key: Borrow<Q>, + Q: Ord, + { + self.0.get(key) + } + + /// Returns a mutable reference to the [`Value`] corresponding to the [`Key`]. + /// + /// Returns `None` if the key doesn't exist or if the key has no value. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::{Key, Keywords, Value}; + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// let value: Value = "buddhist".parse().expect("Failed to parse a Value."); + /// let mut keywords: Keywords = vec![(key, value)].into_iter().collect(); + /// + /// let key: Key = "ca".parse().expect("Failed to parse a Key."); + /// if let Some(value) = keywords.get_mut(&key) { + /// *value = "gregory".parse().expect("Failed to parse a Value."); + /// } + /// assert_eq!( + /// keywords.get(&key).map(|v| v.to_string()), + /// Some("gregory".to_string()) + /// ); + /// ``` + pub fn get_mut<Q>(&mut self, key: &Q) -> Option<&mut Value> + where + Key: Borrow<Q>, + Q: Ord, + { + self.0.get_mut(key) + } + + /// Sets the specified keyword, returning the old value if it already existed. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Key; + /// use icu::locid::extensions::unicode::Value; + /// use icu::locid::Locale; + /// use icu::locid::{ + /// extensions_unicode_key as key, extensions_unicode_value as value, + /// }; + /// + /// let mut loc: Locale = "und-u-hello-ca-buddhist-hc-h12" + /// .parse() + /// .expect("valid BCP-47 identifier"); + /// let old_value = loc + /// .extensions + /// .unicode + /// .keywords + /// .set(key!("ca"), value!("japanese")); + /// + /// assert_eq!(old_value, Some(value!("buddhist"))); + /// assert_eq!(loc, "und-u-hello-ca-japanese-hc-h12".parse().unwrap()); + /// ``` + pub fn set(&mut self, key: Key, value: Value) -> Option<Value> { + self.0.insert(key, value) + } + + /// Removes the specified keyword, returning the old value if it existed. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Key; + /// use icu::locid::extensions_unicode_key as key; + /// use icu::locid::Locale; + /// + /// let mut loc: Locale = "und-u-hello-ca-buddhist-hc-h12" + /// .parse() + /// .expect("valid BCP-47 identifier"); + /// loc.extensions.unicode.keywords.remove(key!("ca")); + /// assert_eq!(loc, "und-u-hello-hc-h12".parse().unwrap()); + /// ``` + pub fn remove<Q: Borrow<Key>>(&mut self, key: Q) -> Option<Value> { + self.0.remove(key.borrow()) + } + + /// Clears all Unicode extension keywords, leaving Unicode attributes. + /// + /// Returns the old Unicode extension keywords. + /// + /// # Example + /// + /// ``` + /// use icu::locid::Locale; + /// + /// let mut loc: Locale = "und-u-hello-ca-buddhist-hc-h12".parse().unwrap(); + /// loc.extensions.unicode.keywords.clear(); + /// assert_eq!(loc, "und-u-hello".parse().unwrap()); + /// ``` + pub fn clear(&mut self) -> Self { + core::mem::take(self) + } + + /// Retains a subset of keywords as specified by the predicate function. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions_unicode_key as key; + /// use icu::locid::Locale; + /// + /// let mut loc: Locale = "und-u-ca-buddhist-hc-h12-ms-metric".parse().unwrap(); + /// + /// loc.extensions + /// .unicode + /// .keywords + /// .retain_by_key(|&k| k == key!("hc")); + /// assert_eq!(loc, "und-u-hc-h12".parse().unwrap()); + /// + /// loc.extensions + /// .unicode + /// .keywords + /// .retain_by_key(|&k| k == key!("ms")); + /// assert_eq!(loc, Locale::UND); + /// ``` + pub fn retain_by_key<F>(&mut self, mut predicate: F) + where + F: FnMut(&Key) -> bool, + { + self.0.retain(|k, _| predicate(k)) + } + + /// Compare this [`Keywords`] with BCP-47 bytes. + /// + /// The return value is equivalent to what would happen if you first converted this + /// [`Keywords`] to a BCP-47 string and then performed a byte comparison. + /// + /// This function is case-sensitive and results in a *total order*, so it is appropriate for + /// binary search. The only argument producing [`Ordering::Equal`] is `self.to_string()`. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Keywords; + /// use icu::locid::Locale; + /// use std::cmp::Ordering; + /// + /// let bcp47_strings: &[&str] = + /// &["ca-hebrew", "ca-japanese", "ca-japanese-nu-latn", "nu-latn"]; + /// + /// for ab in bcp47_strings.windows(2) { + /// let a = ab[0]; + /// let b = ab[1]; + /// assert!(a.cmp(b) == Ordering::Less); + /// let a_kwds = format!("und-u-{}", a) + /// .parse::<Locale>() + /// .unwrap() + /// .extensions + /// .unicode + /// .keywords; + /// assert_eq!(a, a_kwds.to_string()); + /// assert!(a_kwds.strict_cmp(a.as_bytes()) == Ordering::Equal); + /// assert!(a_kwds.strict_cmp(b.as_bytes()) == Ordering::Less); + /// } + /// ``` + pub fn strict_cmp(&self, other: &[u8]) -> Ordering { + self.strict_cmp_iter(other.split(|b| *b == b'-')).end() + } + + /// Compare this [`Keywords`] with an iterator of BCP-47 subtags. + /// + /// This function has the same equality semantics as [`Keywords::strict_cmp`]. It is intended as + /// a more modular version that allows multiple subtag iterators to be chained together. + /// + /// For an additional example, see [`SubtagOrderingResult`]. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Keywords; + /// use icu::locid::locale; + /// use std::cmp::Ordering; + /// + /// let subtags: &[&[u8]] = &[b"ca", b"buddhist"]; + /// + /// let kwds = locale!("und-u-ca-buddhist").extensions.unicode.keywords; + /// assert_eq!( + /// Ordering::Equal, + /// kwds.strict_cmp_iter(subtags.iter().copied()).end() + /// ); + /// + /// let kwds = locale!("und").extensions.unicode.keywords; + /// assert_eq!( + /// Ordering::Less, + /// kwds.strict_cmp_iter(subtags.iter().copied()).end() + /// ); + /// + /// let kwds = locale!("und-u-nu-latn").extensions.unicode.keywords; + /// assert_eq!( + /// Ordering::Greater, + /// kwds.strict_cmp_iter(subtags.iter().copied()).end() + /// ); + /// ``` + pub fn strict_cmp_iter<'l, I>(&self, mut subtags: I) -> SubtagOrderingResult<I> + where + I: Iterator<Item = &'l [u8]>, + { + let r = self.for_each_subtag_str(&mut |subtag| { + if let Some(other) = subtags.next() { + match subtag.as_bytes().cmp(other) { + Ordering::Equal => Ok(()), + not_equal => Err(not_equal), + } + } else { + Err(Ordering::Greater) + } + }); + match r { + Ok(_) => SubtagOrderingResult::Subtags(subtags), + Err(o) => SubtagOrderingResult::Ordering(o), + } + } + + pub(crate) fn for_each_subtag_str<E, F>(&self, f: &mut F) -> Result<(), E> + where + F: FnMut(&str) -> Result<(), E>, + { + for (k, v) in self.0.iter() { + f(k.as_str())?; + v.for_each_subtag_str(f)?; + } + Ok(()) + } + + /// This needs to be its own method to help with type inference in helpers.rs + #[cfg(test)] + pub(crate) fn from_tuple_vec(v: Vec<(Key, Value)>) -> Self { + v.into_iter().collect() + } +} + +impl From<LiteMap<Key, Value, ShortVec<(Key, Value)>>> for Keywords { + fn from(map: LiteMap<Key, Value, ShortVec<(Key, Value)>>) -> Self { + Self(map) + } +} + +impl FromIterator<(Key, Value)> for Keywords { + fn from_iter<I: IntoIterator<Item = (Key, Value)>>(iter: I) -> Self { + LiteMap::from_iter(iter).into() + } +} + +impl_writeable_for_key_value!(Keywords, "ca", "islamic-civil", "mm", "mm"); diff --git a/vendor/icu_locid/src/extensions/unicode/mod.rs b/vendor/icu_locid/src/extensions/unicode/mod.rs new file mode 100644 index 000000000..fabf1036c --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/mod.rs @@ -0,0 +1,233 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +//! Unicode Extensions provide information about user preferences in a given locale. +//! +//! The main struct for this extension is [`Unicode`] which contains [`Keywords`] and +//! [`Attributes`]. +//! +//! +//! # Examples +//! +//! ``` +//! use icu::locid::extensions::unicode::{Attribute, Key, Unicode, Value}; +//! use icu::locid::{LanguageIdentifier, Locale}; +//! +//! let mut loc: Locale = +//! "en-US-u-foobar-hc-h12".parse().expect("Parsing failed."); +//! +//! let key: Key = "hc".parse().expect("Parsing key failed."); +//! let value: Value = "h12".parse().expect("Parsing value failed."); +//! let attribute: Attribute = +//! "foobar".parse().expect("Parsing attribute failed."); +//! +//! assert_eq!(loc.extensions.unicode.keywords.get(&key), Some(&value)); +//! assert!(loc.extensions.unicode.attributes.contains(&attribute)); +//! +//! assert_eq!(&loc.extensions.unicode.to_string(), "-u-foobar-hc-h12"); +//! ``` +mod attribute; +mod attributes; +mod key; +mod keywords; +mod value; + +use alloc::vec; +pub use attribute::Attribute; +pub use attributes::Attributes; +pub use key::Key; +pub use keywords::Keywords; +pub use value::Value; + +use crate::parser::ParserError; +use crate::parser::SubtagIterator; +use litemap::LiteMap; + +/// Unicode Extensions provide information about user preferences in a given locale. +/// +/// A list of [`Unicode BCP47 U Extensions`] as defined in [`Unicode Locale +/// Identifier`] specification. +/// +/// Unicode extensions provide subtags that specify language and/or locale-based behavior +/// or refinements to language tags, according to work done by the Unicode Consortium. +/// (See [`RFC 6067`] for details). +/// +/// [`Unicode BCP47 U Extensions`]: https://unicode.org/reports/tr35/#u_Extension +/// [`RFC 6067`]: https://www.ietf.org/rfc/rfc6067.txt +/// [`Unicode Locale Identifier`]: https://unicode.org/reports/tr35/#Unicode_locale_identifier +/// +/// # Examples +/// +/// ``` +/// use icu::locid::extensions::unicode::{Key, Value}; +/// use icu::locid::Locale; +/// +/// let mut loc: Locale = +/// "de-u-hc-h12-ca-buddhist".parse().expect("Parsing failed."); +/// +/// let key: Key = "ca".parse().expect("Parsing key failed."); +/// let value: Value = "buddhist".parse().expect("Parsing value failed."); +/// assert_eq!(loc.extensions.unicode.keywords.get(&key), Some(&value)); +/// ``` +#[derive(Clone, PartialEq, Eq, Debug, Default, Hash, PartialOrd, Ord)] +#[allow(clippy::exhaustive_structs)] // spec-backed stable datastructure +pub struct Unicode { + /// The key-value pairs present in this locale extension, with each extension key subtag + /// associated to its provided value subtag. + pub keywords: Keywords, + /// A canonically ordered sequence of single standalone subtags for this locale extension. + pub attributes: Attributes, +} + +impl Unicode { + /// Returns a new empty map of Unicode extensions. Same as [`default()`](Default::default()), but is `const`. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Unicode; + /// + /// assert_eq!(Unicode::new(), Unicode::default()); + /// ``` + #[inline] + pub const fn new() -> Self { + Self { + keywords: Keywords::new(), + attributes: Attributes::new(), + } + } + + /// Returns [`true`] if there list of keywords and attributes is empty. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::Locale; + /// + /// let loc: Locale = "en-US-u-foo".parse().expect("Parsing failed."); + /// + /// assert!(!loc.extensions.unicode.is_empty()); + /// ``` + pub fn is_empty(&self) -> bool { + self.keywords.is_empty() && self.attributes.is_empty() + } + + /// Clears all Unicode extension keywords and attributes, effectively removing + /// the Unicode extension. + /// + /// # Example + /// + /// ``` + /// use icu::locid::Locale; + /// + /// let mut loc: Locale = + /// "und-t-mul-u-hello-ca-buddhist-hc-h12".parse().unwrap(); + /// loc.extensions.unicode.clear(); + /// assert_eq!(loc, "und-t-mul".parse().unwrap()); + /// ``` + pub fn clear(&mut self) { + self.keywords.clear(); + self.attributes.clear(); + } + + pub(crate) fn try_from_iter(iter: &mut SubtagIterator) -> Result<Self, ParserError> { + let mut attributes = vec![]; + let mut keywords = LiteMap::new(); + + let mut current_keyword = None; + let mut current_type = vec![]; + + while let Some(subtag) = iter.peek() { + if let Ok(attr) = Attribute::try_from_bytes(subtag) { + if let Err(idx) = attributes.binary_search(&attr) { + attributes.insert(idx, attr); + } + } else { + break; + } + iter.next(); + } + + while let Some(subtag) = iter.peek() { + let slen = subtag.len(); + if slen == 2 { + if let Some(kw) = current_keyword.take() { + keywords.try_insert(kw, Value::from_vec_unchecked(current_type)); + current_type = vec![]; + } + current_keyword = Some(Key::try_from_bytes(subtag)?); + } else if current_keyword.is_some() { + match Value::parse_subtag(subtag) { + Ok(Some(t)) => current_type.push(t), + Ok(None) => {} + Err(_) => break, + } + } else { + break; + } + iter.next(); + } + + if let Some(kw) = current_keyword.take() { + keywords.try_insert(kw, Value::from_vec_unchecked(current_type)); + } + + // Ensure we've defined at least one attribute or keyword + if attributes.is_empty() && keywords.is_empty() { + return Err(ParserError::InvalidExtension); + } + + Ok(Self { + keywords: keywords.into(), + attributes: Attributes::from_vec_unchecked(attributes), + }) + } + + pub(crate) fn for_each_subtag_str<E, F>(&self, f: &mut F) -> Result<(), E> + where + F: FnMut(&str) -> Result<(), E>, + { + if self.is_empty() { + return Ok(()); + } + f("u")?; + self.attributes.for_each_subtag_str(f)?; + self.keywords.for_each_subtag_str(f)?; + Ok(()) + } +} + +writeable::impl_display_with_writeable!(Unicode); + +impl writeable::Writeable for Unicode { + fn write_to<W: core::fmt::Write + ?Sized>(&self, sink: &mut W) -> core::fmt::Result { + if self.is_empty() { + return Ok(()); + } + sink.write_str("-u")?; + if !self.attributes.is_empty() { + sink.write_char('-')?; + writeable::Writeable::write_to(&self.attributes, sink)?; + } + if !self.keywords.is_empty() { + sink.write_char('-')?; + writeable::Writeable::write_to(&self.keywords, sink)?; + } + Ok(()) + } + + fn writeable_length_hint(&self) -> writeable::LengthHint { + if self.is_empty() { + return writeable::LengthHint::exact(0); + } + let mut result = writeable::LengthHint::exact(2); + if !self.attributes.is_empty() { + result += writeable::Writeable::writeable_length_hint(&self.attributes) + 1; + } + if !self.keywords.is_empty() { + result += writeable::Writeable::writeable_length_hint(&self.keywords) + 1; + } + result + } +} diff --git a/vendor/icu_locid/src/extensions/unicode/value.rs b/vendor/icu_locid/src/extensions/unicode/value.rs new file mode 100644 index 000000000..ce9982a4c --- /dev/null +++ b/vendor/icu_locid/src/extensions/unicode/value.rs @@ -0,0 +1,199 @@ +// This file is part of ICU4X. For terms of use, please see the file +// called LICENSE at the top level of the ICU4X source tree +// (online at: https://github.com/unicode-org/icu4x/blob/main/LICENSE ). + +use crate::helpers::ShortVec; +use crate::parser::{get_subtag_iterator, ParserError}; +use alloc::vec::Vec; +use core::ops::RangeInclusive; +use core::str::FromStr; +use tinystr::TinyAsciiStr; + +/// A value used in a list of [`Keywords`](super::Keywords). +/// +/// The value has to be a sequence of one or more alphanumerical strings +/// separated by `-`. +/// Each part of the sequence has to be no shorter than three characters and no +/// longer than 8. +/// +/// +/// # Examples +/// +/// ``` +/// use icu::locid::extensions::unicode::Value; +/// +/// let value1: Value = "gregory".parse().expect("Failed to parse a Value."); +/// let value2: Value = +/// "islamic-civil".parse().expect("Failed to parse a Value."); +/// let value3: Value = "true".parse().expect("Failed to parse a Value."); +/// +/// assert_eq!(&value1.to_string(), "gregory"); +/// assert_eq!(&value2.to_string(), "islamic-civil"); +/// +/// // The value "true" is special-cased to an empty value +/// assert_eq!(&value3.to_string(), ""); +/// ``` +#[derive(Debug, PartialEq, Eq, Clone, Hash, PartialOrd, Ord)] +pub struct Value(ShortVec<TinyAsciiStr<{ *VALUE_LENGTH.end() }>>); + +const VALUE_LENGTH: RangeInclusive<usize> = 3..=8; +const TRUE_VALUE: TinyAsciiStr<8> = tinystr::tinystr!(8, "true"); + +impl Value { + /// A constructor which takes a utf8 slice, parses it and + /// produces a well-formed [`Value`]. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Value; + /// + /// let value = Value::try_from_bytes(b"buddhist").expect("Parsing failed."); + /// + /// assert_eq!(&value.to_string(), "buddhist"); + /// ``` + pub fn try_from_bytes(input: &[u8]) -> Result<Self, ParserError> { + let mut v = ShortVec::new(); + + if !input.is_empty() { + for subtag in get_subtag_iterator(input) { + let val = Self::subtag_from_bytes(subtag)?; + if let Some(val) = val { + v.push(val); + } + } + } + Ok(Self(v)) + } + + /// Const constructor for when the value contains only a single subtag. + /// + /// # Examples + /// + /// ``` + /// use icu::locid::extensions::unicode::Value; + /// + /// Value::try_from_single_subtag(b"buddhist").expect("valid subtag"); + /// Value::try_from_single_subtag(b"#####").expect_err("invalid subtag"); + /// Value::try_from_single_subtag(b"foo-bar").expect_err("not a single subtag"); + /// ``` + pub const fn try_from_single_subtag(subtag: &[u8]) -> Result<Self, ParserError> { + match Self::subtag_from_bytes(subtag) { + Err(_) => Err(ParserError::InvalidExtension), + Ok(option) => Ok(Self::from_tinystr(option)), + } + } + + #[doc(hidden)] + pub fn as_tinystr_slice(&self) -> &[TinyAsciiStr<8>] { + self.0.as_slice() + } + + #[doc(hidden)] + pub const fn as_single_subtag(&self) -> Option<&TinyAsciiStr<8>> { + self.0.single() + } + + #[doc(hidden)] + pub const fn from_tinystr(subtag: Option<TinyAsciiStr<8>>) -> Self { + match subtag { + None => Self(ShortVec::new()), + Some(val) => { + debug_assert!(val.is_ascii_alphanumeric()); + debug_assert!(!matches!(val, TRUE_VALUE)); + Self(ShortVec::new_single(val)) + } + } + } + + pub(crate) fn from_vec_unchecked(input: Vec<TinyAsciiStr<8>>) -> Self { + Self(input.into()) + } + + #[doc(hidden)] + pub const fn subtag_from_bytes(bytes: &[u8]) -> Result<Option<TinyAsciiStr<8>>, ParserError> { + Self::parse_subtag_from_bytes_manual_slice(bytes, 0, bytes.len()) + } + + pub(crate) fn parse_subtag(t: &[u8]) -> Result<Option<TinyAsciiStr<8>>, ParserError> { + Self::parse_subtag_from_bytes_manual_slice(t, 0, t.len()) + } + + pub(crate) const fn parse_subtag_from_bytes_manual_slice( + bytes: &[u8], + start: usize, + end: usize, + ) -> Result<Option<TinyAsciiStr<8>>, ParserError> { + let slice_len = end - start; + if slice_len > *VALUE_LENGTH.end() || slice_len < *VALUE_LENGTH.start() { + return Err(ParserError::InvalidExtension); + } + + match TinyAsciiStr::from_bytes_manual_slice(bytes, start, end) { + Ok(TRUE_VALUE) => Ok(None), + Ok(s) if s.is_ascii_alphanumeric() => Ok(Some(s.to_ascii_lowercase())), + Ok(_) => Err(ParserError::InvalidExtension), + Err(_) => Err(ParserError::InvalidSubtag), + } + } + + pub(crate) fn for_each_subtag_str<E, F>(&self, f: &mut F) -> Result<(), E> + where + F: FnMut(&str) -> Result<(), E>, + { + self.0.as_slice().iter().map(|t| t.as_str()).try_for_each(f) + } +} + +impl FromStr for Value { + type Err = ParserError; + + fn from_str(source: &str) -> Result<Self, Self::Err> { + Self::try_from_bytes(source.as_bytes()) + } +} + +impl_writeable_for_tinystr_list!(Value, "", "islamic", "civil"); + +/// A macro allowing for compile-time construction of valid Unicode [`Value`] subtag. +/// +/// The macro only supports single-subtag values. +/// +/// # Examples +/// +/// ``` +/// use icu::locid::Locale; +/// use icu::locid::{ +/// extensions_unicode_key as key, extensions_unicode_value as value, +/// }; +/// +/// let loc: Locale = "de-u-ca-buddhist".parse().unwrap(); +/// +/// assert_eq!( +/// loc.extensions.unicode.keywords.get(&key!("ca")), +/// Some(&value!("buddhist")) +/// ); +/// ``` +/// +/// [`Value`]: crate::extensions::unicode::Value +#[macro_export] +macro_rules! extensions_unicode_value { + ($value:literal) => {{ + // What we want: + // const R: $crate::extensions::unicode::Value = + // match $crate::extensions::unicode::Value::try_from_single_subtag($value.as_bytes()) { + // Ok(r) => r, + // #[allow(clippy::panic)] // const context + // _ => panic!(concat!("Invalid Unicode extension value: ", $value)), + // }; + // Workaround until https://github.com/rust-lang/rust/issues/73255 lands: + const R: $crate::extensions::unicode::Value = + $crate::extensions::unicode::Value::from_tinystr( + match $crate::extensions::unicode::Value::subtag_from_bytes($value.as_bytes()) { + Ok(r) => r, + _ => panic!(concat!("Invalid Unicode extension value: ", $value)), + }, + ); + R + }}; +} |