Struct icu::experimental::transliterate::Transliterator

pub struct Transliterator { /* private fields */ }

A Transliterator allows transliteration based on UTS #35 transform rules, including overrides with custom implementations.

Examples

A transliterator with a custom alias referenced by another:

use icu::experimental::transliterate::{Transliterator, CustomTransliterator, RuleCollection};
use icu::locale::Locale;

// Set up a transliterator with 3 custom rules.
// Note: These rules are for demonstration purposes only! Do not use.

// 1. Main entrypoint: a chain of several transliterators
let mut collection = RuleCollection::default();
collection.register_source(
    &"und-t-und-x0-custom".parse().unwrap(),
    "::NFD; ::FlattenLowerUmlaut; ::[:Nonspacing_Mark:] Remove; ::AsciiUpper; ::NFC;".to_string(),
    [],
    false,
    true,
);

// 2. A custom ruleset that expands lowercase umlauts
collection.register_source(
    &"und-t-und-x0-dep1".parse().unwrap(),
    r#"
      [ä {a \u0308}] → ae;
      [ö {o \u0308}] → oe;
      [ü {u \u0308}] → ue;
    "#.to_string(),
    ["Any-FlattenLowerUmlaut"],
    false,
    true,
);

// 3. A custom transliterator that uppercases all ASCII characters
#[derive(Debug)]
struct AsciiUpperTransliterator;
impl CustomTransliterator for AsciiUpperTransliterator {
    fn transliterate(&self, input: &str, range: std::ops::Range<usize>) -> String {
        input.to_ascii_uppercase()
    }
}
collection.register_aliases(
    &"und-t-und-x0-dep2".parse().unwrap(),
    ["Any-AsciiUpper"],
);

// Create a transliterator from the main entrypoint:
let provider = collection.as_provider();
let t = Transliterator::try_new_with_override_unstable(
    &provider,
    &provider,
    &"und-t-und-x0-custom".parse().unwrap(),
    |locale| locale.normalizing_eq("und-t-und-x0-dep2").then_some(Ok(Box::new(AsciiUpperTransliterator))),
)
.unwrap();

// Test the behavior:
// - The uppercase 'Ü' is stripped of its umlaut
// - The lowercase 'ä' is expanded to "ae"
// - All ASCII characters are uppercased: not 'ß', which is not ASCII
let r = t.transliterate("Übermäßig".to_string());
assert_eq!(r, "UBERMAEßIG");

Implementations

impl Transliterator

pub fn try_new(locale: &Locale) -> Result<Transliterator, DataError>

Construct a Transliterator from the given Locale.

Examples

use icu::experimental::transliterate::Transliterator;
// BCP-47-T ID for Bengali to Arabic transliteration
let locale = "und-Arab-t-und-beng".parse().unwrap();
let t = Transliterator::try_new(&locale).unwrap();
let output = t.transliterate("অকার্যতানাযা".to_string());

assert_eq!(output, "اكاريتانايا");

pub fn try_new_with_any_provider( provider: &(impl AnyProvider + ?Sized), locale: &Locale, ) -> Result<Transliterator, DataError>

A version of Self::try_new that uses custom data provided by an AnyProvider.

📚 Help choosing a constructor

pub fn try_new_with_buffer_provider( provider: &(impl BufferProvider + ?Sized), locale: &Locale, ) -> Result<Transliterator, DataError>

A version of Self::try_new that uses custom data provided by a BufferProvider.

✨ Enabled with the serde feature.

📚 Help choosing a constructor

pub fn try_new_unstable<PT, PN>( transliterator_provider: &PT, normalizer_provider: &PN, locale: &Locale, ) -> Result<Transliterator, DataError>

where PT: DataProvider<TransliteratorRulesV1Marker> + ?Sized, PN: DataProvider<CanonicalDecompositionDataV2Marker> + DataProvider<CompatibilityDecompositionDataV2Marker> + DataProvider<CanonicalDecompositionTablesV1Marker> + DataProvider<CompatibilityDecompositionTablesV1Marker> + DataProvider<CanonicalCompositionsV1Marker> + ?Sized,

A version of Self::try_new that uses custom data provided by a DataProvider.

📚 Help choosing a constructor

⚠️ The bounds on provider may change over time, including in SemVer minor releases.

pub fn try_new_with_override<F>( locale: &Locale, lookup: F, ) -> Result<Transliterator, DataError>

where F: Fn(&Locale) -> Option<Result<Box<dyn CustomTransliterator>, DataError>>,

Construct a Transliterator from the given Locale using overrides provided by lookup.

This allows clients to override the nested transliterators used by this transliterator. Any nested transliterator will first try to be loaded with lookup, and only fall back to the nested transliterator defined by the data if it returns None. See CustomTransliterator.

Example

Overriding "de-t-de-d0-ascii"’s dependency on "und-t-und-Latn-d0-ascii":

use core::ops::Range;
use icu::experimental::transliterate::{
    CustomTransliterator, Transliterator,
};
use icu::locale::Locale;

#[derive(Debug)]
struct FunkyGermanToAscii;
impl CustomTransliterator for FunkyGermanToAscii {
    fn transliterate(
        &self,
        input: &str,
        allowed_range: Range<usize>,
    ) -> String {
        input[allowed_range].replace("oeverride", "overridden")
    }
}

let override_locale: Locale = "und-t-und-Latn-d0-ascii".parse().unwrap();
let locale = "de-t-de-d0-ascii".parse().unwrap();
let t = Transliterator::try_new_with_override(&locale, |locale| {
    override_locale
        .eq(locale)
        .then_some(Ok(Box::new(FunkyGermanToAscii)))
})
.unwrap();
let output = t.transliterate("This is an överride example".to_string());

assert_eq!(output, "This is an overridden example");

pub fn try_new_with_override_with_any_provider<F>( provider: &(impl AnyProvider + ?Sized), locale: &Locale, lookup: F, ) -> Result<Transliterator, DataError>

where F: Fn(&Locale) -> Option<Result<Box<dyn CustomTransliterator>, DataError>>,

A version of Self::try_new_with_override that uses custom data provided by an AnyProvider.

📚 Help choosing a constructor

pub fn try_new_with_override_with_buffer_provider<F>( provider: &(impl BufferProvider + ?Sized), locale: &Locale, lookup: F, ) -> Result<Transliterator, DataError>

where F: Fn(&Locale) -> Option<Result<Box<dyn CustomTransliterator>, DataError>>,

A version of Self::try_new_with_override that uses custom data provided by a BufferProvider.

✨ Enabled with the serde feature.

📚 Help choosing a constructor

pub fn try_new_with_override_unstable<PT, PN, F>( transliterator_provider: &PT, normalizer_provider: &PN, locale: &Locale, lookup: F, ) -> Result<Transliterator, DataError>

where PT: DataProvider<TransliteratorRulesV1Marker> + ?Sized, PN: DataProvider<CanonicalDecompositionDataV2Marker> + DataProvider<CompatibilityDecompositionDataV2Marker> + DataProvider<CanonicalDecompositionTablesV1Marker> + DataProvider<CompatibilityDecompositionTablesV1Marker> + DataProvider<CanonicalCompositionsV1Marker> + ?Sized, F: Fn(&Locale) -> Option<Result<Box<dyn CustomTransliterator>, DataError>>,

A version of Self::try_new_with_override that uses custom data provided by a DataProvider.

📚 Help choosing a constructor

⚠️ The bounds on provider may change over time, including in SemVer minor releases.

pub fn transliterate(&self, input: String) -> String

Transliterates input and returns its transliteration.

Trait Implementations

impl Debug for Transliterator

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.