unicode_intervals/

lib.rs

//! [![github]](https://github.com/Stranger6667/unicode-intervals) [![crates-io]](https://crates.io/crates/unicode-intervals) [![docs-rs]](https://docs.rs/unicode-intervals)
//!
//! [github]: https://img.shields.io/badge/github-8da0cb?style=flat-square&labelColor=555555&logo=github
//! [crates-io]: https://img.shields.io/badge/crates.io-fc8d62?style=flat-square&labelColor=555555&logo=rust
//! [docs-rs]: https://img.shields.io/badge/docs.rs-66c2a5?style=flat-square&labelColor=555555&logo=docs.rs
//!
//! <br>
//!
//! This library provides a way to search for Unicode code point intervals by categories, ranges,
//! and custom character sets.
//!
//! The main purpose of `unicode-intervals` is to simplify generating strings that matching
//! specific criteria.
//!
//! # Examples
//!
//! Raw Unicode codepoint intervals from the latest Unicode version:
//!
//! ```rust
//! use unicode_intervals::UnicodeCategory;
//!
//! let intervals = unicode_intervals::query()
//!     .include_categories(
//!         UnicodeCategory::UPPERCASE_LETTER |
//!         UnicodeCategory::LOWERCASE_LETTER
//!     )
//!     .max_codepoint(128)
//!     .include_characters("☃")
//!     .intervals()
//!     .expect("Invalid query input");
//! assert_eq!(intervals, &[(65, 90), (97, 122), (9731, 9731)]);
//! ```
//!
//! `IntervalSet` for index-like access to the underlying codepoints:
//!
//! ```rust
//! use unicode_intervals::UnicodeCategory;
//!
//! let interval_set = unicode_intervals::query()
//!     .include_categories(UnicodeCategory::UPPERCASE_LETTER)
//!     .interval_set()
//!     .expect("Invalid query input");
//! // Get 10th codepoint in this interval set
//! assert_eq!(interval_set.codepoint_at(10), Some('K' as u32));
//! assert_eq!(interval_set.index_of('K'), Some(10));
//! ```
//!
//! Query specific Unicode version:
//!
//! ```rust
//! use unicode_intervals::UnicodeVersion;
//!
//! let intervals = UnicodeVersion::V11_0_0.query()
//!     .max_codepoint(128)
//!     .include_characters("☃")
//!     .intervals()
//!     .expect("Invalid query input");
//! assert_eq!(intervals, &[(0, 128), (9731, 9731)]);
//! ```
//!
//! Restrict the output to code points within a certain range:
//!
//! ```rust
//! let intervals = unicode_intervals::query()
//!     .min_codepoint(65)
//!     .max_codepoint(128)
//!     .intervals()
//!     .expect("Invalid query input");
//! assert_eq!(intervals, &[(65, 128)])
//! ```
//!
//! Include or exclude specific characters:
//!
//! ```rust
//! # use unicode_intervals::UnicodeCategory;
//! let intervals = unicode_intervals::query()
//!     .include_categories(UnicodeCategory::PARAGRAPH_SEPARATOR)
//!     .include_characters("-123")
//!     .intervals()
//!     .expect("Invalid query input");
//! assert_eq!(intervals, &[(45, 45), (49, 51), (8233, 8233)])
//! ```
//!
//! # Unicode version support
//!
//! `unicode-intervals` supports Unicode 9.0.0 - 17.0.0.
#![warn(
    clippy::cast_possible_truncation,
    clippy::doc_markdown,
    clippy::explicit_iter_loop,
    clippy::map_unwrap_or,
    clippy::match_same_arms,
    clippy::needless_borrow,
    clippy::needless_pass_by_value,
    clippy::print_stdout,
    clippy::redundant_closure,
    clippy::trivially_copy_pass_by_ref,
    missing_debug_implementations,
    missing_docs,
    trivial_casts,
    trivial_numeric_casts,
    unused_extern_crates,
    unused_import_braces,
    variant_size_differences,
    clippy::arithmetic_side_effects,
    clippy::unwrap_used,
    clippy::semicolon_if_nothing_returned,
    clippy::cargo
)]
#![allow(clippy::redundant_static_lifetimes)]
use crate::constants::MAX_CODEPOINT;
use core::fmt;
use std::str::FromStr;

mod categories;
mod constants;
mod error;
mod intervals;
mod intervalset;
mod query;
mod tables;
pub use crate::{
    categories::{as_general_categories, UnicodeCategory, UnicodeCategorySet},
    error::Error,
    intervalset::{Codepoints, IntervalSet},
};

#[cfg(feature = "__benchmark_internals")]
/// Internals used for benchmarking.
pub mod internals {
    /// Unicode categories.
    pub mod categories {
        pub use crate::categories::merge;
    }

    /// Intervals manipulation.
    pub mod intervals {
        pub use crate::intervals::{from_str, merge, subtract};
    }

    /// Querying Unicode intervals.
    pub mod query {
        pub use crate::query::{intervals_for_set, query};
    }
}

/// Interval between two Unicode codepoints.
pub type Interval = (u32, u32);

/// Supported Unicode versions.
#[derive(Debug, Copy, Clone, Eq, PartialEq, Hash)]
#[non_exhaustive]
pub enum UnicodeVersion {
    /// Unicode 9.0.0
    V9_0_0,
    /// Unicode 10.0.0
    V10_0_0,
    /// Unicode 11.0.0
    V11_0_0,
    /// Unicode 12.0.0
    V12_0_0,
    /// Unicode 12.1.0
    V12_1_0,
    /// Unicode 13.0.0
    V13_0_0,
    /// Unicode 14.0.0
    V14_0_0,
    /// Unicode 15.0.0
    V15_0_0,
    /// Unicode 15.1.0
    V15_1_0,
    /// Unicode 16.0.0
    V16_0_0,
    /// Unicode 17.0.0
    V17_0_0,
}

impl fmt::Display for UnicodeVersion {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        f.write_str(self.as_str())
    }
}

impl FromStr for UnicodeVersion {
    type Err = Error;

    fn from_str(s: &str) -> Result<Self, Self::Err> {
        match s {
            "9.0.0" => Ok(UnicodeVersion::V9_0_0),
            "10.0.0" => Ok(UnicodeVersion::V10_0_0),
            "11.0.0" => Ok(UnicodeVersion::V11_0_0),
            "12.0.0" => Ok(UnicodeVersion::V12_0_0),
            "12.1.0" => Ok(UnicodeVersion::V12_1_0),
            "13.0.0" => Ok(UnicodeVersion::V13_0_0),
            "14.0.0" => Ok(UnicodeVersion::V14_0_0),
            "15.0.0" => Ok(UnicodeVersion::V15_0_0),
            "15.1.0" => Ok(UnicodeVersion::V15_1_0),
            "16.0.0" => Ok(UnicodeVersion::V16_0_0),
            "17.0.0" => Ok(UnicodeVersion::V17_0_0),
            _ => Err(Error::InvalidVersion(s.to_string().into_boxed_str())),
        }
    }
}

impl UnicodeVersion {
    /// Every bundled Unicode version, oldest to newest.
    pub const ALL: [UnicodeVersion; 11] = [
        UnicodeVersion::V9_0_0,
        UnicodeVersion::V10_0_0,
        UnicodeVersion::V11_0_0,
        UnicodeVersion::V12_0_0,
        UnicodeVersion::V12_1_0,
        UnicodeVersion::V13_0_0,
        UnicodeVersion::V14_0_0,
        UnicodeVersion::V15_0_0,
        UnicodeVersion::V15_1_0,
        UnicodeVersion::V16_0_0,
        UnicodeVersion::V17_0_0,
    ];
    /// Unicode version as a string.
    #[must_use]
    pub const fn as_str(self) -> &'static str {
        match self {
            UnicodeVersion::V9_0_0 => "9.0.0",
            UnicodeVersion::V10_0_0 => "10.0.0",
            UnicodeVersion::V11_0_0 => "11.0.0",
            UnicodeVersion::V12_0_0 => "12.0.0",
            UnicodeVersion::V12_1_0 => "12.1.0",
            UnicodeVersion::V13_0_0 => "13.0.0",
            UnicodeVersion::V14_0_0 => "14.0.0",
            UnicodeVersion::V15_0_0 => "15.0.0",
            UnicodeVersion::V15_1_0 => "15.1.0",
            UnicodeVersion::V16_0_0 => "16.0.0",
            UnicodeVersion::V17_0_0 => "17.0.0",
        }
    }
    /// Get the latest Unicode version.
    #[must_use]
    pub const fn latest() -> UnicodeVersion {
        UnicodeVersion::V17_0_0
    }
    /// A sorted slice of slices where each item is a slice of intervals for every Unicode category.
    /// They are sorted alphabetically by their full name.
    #[inline]
    #[must_use]
    pub const fn table(self) -> &'static [&'static [Interval]] {
        match self {
            UnicodeVersion::V9_0_0 => tables::v9_0_0::BY_NAME,
            UnicodeVersion::V10_0_0 => tables::v10_0_0::BY_NAME,
            UnicodeVersion::V11_0_0 => tables::v11_0_0::BY_NAME,
            UnicodeVersion::V12_0_0 => tables::v12_0_0::BY_NAME,
            UnicodeVersion::V12_1_0 => tables::v12_1_0::BY_NAME,
            UnicodeVersion::V13_0_0 => tables::v13_0_0::BY_NAME,
            UnicodeVersion::V14_0_0 => tables::v14_0_0::BY_NAME,
            UnicodeVersion::V15_0_0 => tables::v15_0_0::BY_NAME,
            UnicodeVersion::V15_1_0 => tables::v15_1_0::BY_NAME,
            UnicodeVersion::V16_0_0 => tables::v16_0_0::BY_NAME,
            UnicodeVersion::V17_0_0 => tables::v17_0_0::BY_NAME,
        }
    }

    /// Get a slice of intervals for the provided Unicode category.
    #[inline]
    #[must_use]
    pub const fn intervals_for(self, category: UnicodeCategory) -> &'static [Interval] {
        self.table()[category as usize]
    }

    /// Unicode categories sorted by the number of intervals inside.
    #[inline]
    #[must_use]
    pub const fn normalized_categories(self) -> [UnicodeCategory; 30] {
        // Pair each category with a sort key: interval count, ties broken by abbreviation
        // (rank < 30, so `count * 30 + rank` orders by count then abbreviation).
        let mut keyed: [(UnicodeCategory, usize); 30] = [(UnicodeCategory::Cc, 0); 30];
        let table = self.table();
        let mut idx = 0;
        // `idx` stays below 30, so the cast and increment can't overflow.
        #[allow(clippy::arithmetic_side_effects, clippy::cast_possible_truncation)]
        while idx < table.len() {
            if let Some(category) = UnicodeCategory::from_index(idx as u8) {
                keyed[idx] = (
                    category,
                    table[idx].len() * 30 + category.abbrev_rank() as usize,
                );
            }
            idx += 1;
        }
        // Bubble sort by key (stable, and works in a `const` context).
        loop {
            let mut swapped = false;
            let mut idx = 1;
            // Arithmetic here will not overflow as it is always less than 30 and more than 1
            #[allow(clippy::arithmetic_side_effects)]
            while idx < keyed.len() {
                let (lcat, lkey) = keyed[idx - 1];
                let (rcat, rkey) = keyed[idx];
                if lkey > rkey {
                    keyed[idx - 1] = (rcat, rkey);
                    keyed[idx] = (lcat, lkey);
                    swapped = true;
                }
                idx += 1;
            }
            if !swapped {
                break;
            }
        }

        // Collect the sorted categories, forcing Cc (control) and Cs (surrogate) to the very
        // end regardless of size. `output[28]` keeps its `Cc` default; `output[29]` is `Cs`.
        let mut output = [UnicodeCategory::Cc; 30];
        output[29] = UnicodeCategory::Cs;
        let mut idx = 0;
        let mut ptr = 0;
        // `idx` and `ptr` stay below 30, so the increments can't overflow.
        #[allow(clippy::arithmetic_side_effects)]
        while idx < keyed.len() {
            let (category, _) = keyed[idx];
            if category as u8 != UnicodeCategory::Cc as u8
                && category as u8 != UnicodeCategory::Cs as u8
            {
                output[ptr] = category;
                ptr += 1;
            }
            idx += 1;
        }
        output
    }

    /// A Query builder for specifying the input parameters to `intervals()` / `interval_set` methods.
    #[must_use]
    #[inline]
    pub fn query<'a>(self) -> IntervalQuery<'a> {
        IntervalQuery::new(self)
    }

    fn intervals_impl(
        self,
        include_categories: Option<UnicodeCategorySet>,
        exclude_categories: UnicodeCategorySet,
        include_characters: Option<&str>,
        exclude_characters: Option<&str>,
        min_codepoint: u32,
        max_codepoint: u32,
    ) -> Result<Vec<Interval>, Error> {
        if min_codepoint > MAX_CODEPOINT || max_codepoint > MAX_CODEPOINT {
            return Err(Error::CodepointNotInRange(min_codepoint, max_codepoint));
        }
        if min_codepoint > max_codepoint {
            return Err(Error::InvalidCodepoints(min_codepoint, max_codepoint));
        }
        Ok(query::query(
            self,
            include_categories,
            exclude_categories,
            include_characters.unwrap_or(""),
            exclude_characters.unwrap_or(""),
            min_codepoint,
            max_codepoint,
        ))
    }
}

/// A builder for querying Unicode intervals.
///
/// # Examples
///
/// ```rust
/// use unicode_intervals::{UnicodeVersion, UnicodeCategory};
///
/// let intervals = UnicodeVersion::V15_0_0.query()
///     .include_categories(UnicodeCategory::UPPERCASE_LETTER | UnicodeCategory::LOWERCASE_LETTER)
///     .max_codepoint(128)
///     .include_characters("☃")
///     .intervals()
///     .expect("Invalid query input");
/// assert_eq!(intervals, &[(65, 90), (97, 122), (9731, 9731)]);
/// ```
#[derive(Debug, Clone, PartialEq)]
pub struct IntervalQuery<'a> {
    version: UnicodeVersion,
    include_categories: Option<UnicodeCategorySet>,
    exclude_categories: Option<UnicodeCategorySet>,
    include_characters: Option<&'a str>,
    exclude_characters: Option<&'a str>,
    min_codepoint: u32,
    max_codepoint: u32,
}

impl<'a> IntervalQuery<'a> {
    fn new(version: UnicodeVersion) -> IntervalQuery<'a> {
        IntervalQuery {
            version,
            include_categories: None,
            exclude_categories: None,
            include_characters: None,
            exclude_characters: None,
            min_codepoint: 0,
            max_codepoint: MAX_CODEPOINT,
        }
    }
    /// Set `include_categories`.
    #[must_use]
    pub fn include_categories(
        mut self,
        include_categories: impl Into<Option<UnicodeCategorySet>>,
    ) -> IntervalQuery<'a> {
        self.include_categories = include_categories.into();
        self
    }
    /// Set `exclude_categories`.
    #[must_use]
    pub fn exclude_categories(
        mut self,
        exclude_categories: impl Into<Option<UnicodeCategorySet>>,
    ) -> IntervalQuery<'a> {
        self.exclude_categories = exclude_categories.into();
        self
    }
    /// Set `include_characters`.
    #[must_use]
    pub fn include_characters(mut self, include_characters: &'a str) -> IntervalQuery<'a> {
        self.include_characters = Some(include_characters);
        self
    }
    /// Set `exclude_characters`.
    #[must_use]
    pub fn exclude_characters(mut self, exclude_characters: &'a str) -> IntervalQuery<'a> {
        self.exclude_characters = Some(exclude_characters);
        self
    }
    /// Set `min_codepoint`.
    #[must_use]
    pub fn min_codepoint(mut self, min_codepoint: u32) -> IntervalQuery<'a> {
        self.min_codepoint = min_codepoint;
        self
    }
    /// Set `max_codepoint`.
    #[must_use]
    pub fn max_codepoint(mut self, max_codepoint: u32) -> IntervalQuery<'a> {
        self.max_codepoint = max_codepoint;
        self
    }
    /// Find intervals matching the query.
    ///
    /// # Errors
    ///
    ///   - `min_codepoint > max_codepoint`
    ///   - `min_codepoint > 1114111` or `max_codepoint > 1114111`
    pub fn intervals(&self) -> Result<Vec<Interval>, Error> {
        let exclude_categories = self.exclude_categories.unwrap_or_default();
        self.version.intervals_impl(
            self.include_categories,
            exclude_categories,
            self.include_characters,
            self.exclude_characters,
            self.min_codepoint,
            self.max_codepoint,
        )
    }
    /// Build an `IndexSet` for the intervals matching the query.
    ///
    /// # Errors
    ///
    ///   - `min_codepoint > max_codepoint`
    ///   - `min_codepoint > 1114111` or `max_codepoint > 1114111`
    pub fn interval_set(&self) -> Result<IntervalSet, Error> {
        Ok(IntervalSet::new(self.intervals()?))
    }
}

/// Build a query that finds Unicode intervals matching the query criteria.
///
/// Uses the latest available Unicode version.
pub fn query<'a>() -> IntervalQuery<'a> {
    UnicodeVersion::latest().query()
}

#[cfg(test)]
mod tests {
    use super::*;
    use std::{
        collections::hash_map::DefaultHasher,
        hash::{Hash, Hasher},
    };
    use test_case::test_case;

    #[test]
    fn test_all_versions() {
        assert_eq!(UnicodeVersion::ALL.len(), 11);
        assert_eq!(UnicodeVersion::ALL[0], UnicodeVersion::V9_0_0);
        assert_eq!(
            *UnicodeVersion::ALL.last().expect("non-empty"),
            UnicodeVersion::latest()
        );
        for v in UnicodeVersion::ALL {
            assert_eq!(v.as_str().parse::<UnicodeVersion>().expect("round-trip"), v);
        }
    }

    #[test_case(None, None, &[(95, 95), (8255, 8256), (8276, 8276), (65075, 65076), (65101, 65103), (65343, 65343)])]
    #[test_case(None, Some(128), &[(95, 95)])]
    #[test_case(Some(65077), None, &[(65101, 65103), (65343, 65343)])]
    #[test_case(Some(65076), Some(65102), &[(65076, 65076), (65101, 65102)])]
    fn test_intervals(
        min_codepoint: Option<u32>,
        max_codepoint: Option<u32>,
        expected: &[Interval],
    ) {
        let mut query = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::Pc);
        if let Some(min) = min_codepoint {
            query = query.min_codepoint(min);
        }
        if let Some(max) = max_codepoint {
            query = query.max_codepoint(max);
        }
        let intervals = query.intervals().expect("Invalid query");
        assert_eq!(intervals, expected);
    }

    #[test]
    fn test_interval_set() {
        let interval_set = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::Lu)
            .max_codepoint(128)
            .interval_set()
            .expect("Invalid query");
        assert_eq!(interval_set.index_of('A'), Some(0));
    }

    #[test]
    fn test_top_level_query() {
        assert_eq!(
            query().intervals().expect("Invalid query"),
            vec![(0, MAX_CODEPOINT)]
        );
    }

    #[test]
    fn test_query_include_only_characters() {
        let intervals = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::Pc)
            .min_codepoint(0)
            .max_codepoint(50)
            .include_characters("abc")
            .intervals()
            .expect("Invalid query");
        assert_eq!(intervals, &[(97, 99)]);
    }

    #[test]
    fn test_query_exclude_only_characters() {
        let intervals = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::UPPERCASE_LETTER)
            .max_codepoint(90)
            .exclude_characters("ABC")
            .intervals()
            .expect("Invalid query");
        assert_eq!(intervals, &[(68, 90)]);
    }

    #[test]
    fn test_query_exclude_categories() {
        let intervals = UnicodeVersion::V15_0_0
            .query()
            .exclude_categories(UnicodeCategory::UPPERCASE_LETTER)
            .max_codepoint(90)
            .intervals()
            .expect("Invalid query");
        assert_eq!(intervals, &[(0, 64)]);
    }

    #[test]
    fn test_multi_category_codepoint_range_is_sorted() {
        // `Ll` (a-z) and `Lu` (A-Z) concatenate out of iteration order before merging,
        // so the result must still come back sorted and non-overlapping.
        let intervals = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::Lu | UnicodeCategory::Ll)
            .max_codepoint(128)
            .intervals()
            .expect("Invalid query");
        assert_eq!(intervals, &[(65, 90), (97, 122)]);
    }

    #[test]
    fn test_query_include_category_and_characters() {
        let intervals = UnicodeVersion::V15_0_0
            .query()
            .include_categories(UnicodeCategory::Pc)
            .include_characters("abc")
            .intervals()
            .expect("Invalid query");
        assert_eq!(
            intervals,
            &[
                (95, 95),
                (97, 99),
                (8255, 8256),
                (8276, 8276),
                (65075, 65076),
                (65101, 65103),
                (65343, 65343)
            ]
        );
    }

    #[test_case(
        1073741824,
        2147483648,
        "Codepoints should be in [0; 1114111] range. Got: [1073741824; 2147483648]"
    )]
    #[test_case(
        0,
        2147483648,
        "Codepoints should be in [0; 1114111] range. Got: [0; 2147483648]"
    )]
    #[test_case(
        5,
        1,
        "Minimum codepoint should be less or equal than maximum codepoint. Got 5 < 1"
    )]
    fn test_query_invalid_codepoints(min_codepoint: u32, max_codepoint: u32, expected: &str) {
        let error = UnicodeVersion::V15_0_0
            .query()
            .min_codepoint(min_codepoint)
            .max_codepoint(max_codepoint)
            .intervals()
            .expect_err("Should error");
        assert_eq!(error.to_string(), expected);
        let error = UnicodeVersion::V15_0_0
            .query()
            .min_codepoint(min_codepoint)
            .max_codepoint(max_codepoint)
            .interval_set()
            .expect_err("Should error");
        assert_eq!(error.to_string(), expected);
    }

    #[test]
    fn test_intervals_for() {
        assert_eq!(
            UnicodeVersion::V15_0_0.intervals_for(UnicodeCategory::Pc),
            &[
                (95, 95),
                (8255, 8256),
                (8276, 8276),
                (65075, 65076),
                (65101, 65103),
                (65343, 65343),
            ]
        );
    }

    #[test]
    fn test_normalized_categories() {
        assert_eq!(
            UnicodeVersion::V15_0_0.normalized_categories(),
            [
                UnicodeCategory::Zl,
                UnicodeCategory::Zp,
                UnicodeCategory::Co,
                UnicodeCategory::Me,
                UnicodeCategory::Pc,
                UnicodeCategory::Zs,
                UnicodeCategory::Lt,
                UnicodeCategory::Pf,
                UnicodeCategory::Pi,
                UnicodeCategory::Nl,
                UnicodeCategory::Pd,
                UnicodeCategory::Cf,
                UnicodeCategory::Sc,
                UnicodeCategory::Sk,
                UnicodeCategory::Nd,
                UnicodeCategory::Sm,
                UnicodeCategory::Lm,
                UnicodeCategory::No,
                UnicodeCategory::Pe,
                UnicodeCategory::Ps,
                UnicodeCategory::Mc,
                UnicodeCategory::So,
                UnicodeCategory::Po,
                UnicodeCategory::Mn,
                UnicodeCategory::Lo,
                UnicodeCategory::Lu,
                UnicodeCategory::Ll,
                UnicodeCategory::Cn,
                UnicodeCategory::Cc,
                UnicodeCategory::Cs,
            ]
        );
    }

    #[test_case(UnicodeVersion::V9_0_0)]
    #[test_case(UnicodeVersion::V10_0_0)]
    #[test_case(UnicodeVersion::V11_0_0)]
    #[test_case(UnicodeVersion::V12_0_0)]
    #[test_case(UnicodeVersion::V12_1_0)]
    #[test_case(UnicodeVersion::V13_0_0)]
    #[test_case(UnicodeVersion::V14_0_0)]
    #[test_case(UnicodeVersion::V15_0_0)]
    #[test_case(UnicodeVersion::V15_1_0)]
    #[test_case(UnicodeVersion::V16_0_0)]
    #[test_case(UnicodeVersion::V17_0_0)]
    fn test_successive_union(version: UnicodeVersion) {
        let mut x = vec![];
        for v in version.table() {
            x.extend_from_slice(v);
        }
        intervals::merge(&mut x);
        assert_eq!(x, vec![(0, MAX_CODEPOINT)]);
    }

    #[test_case(UnicodeVersion::V9_0_0, "9.0.0")]
    #[test_case(UnicodeVersion::V10_0_0, "10.0.0")]
    #[test_case(UnicodeVersion::V11_0_0, "11.0.0")]
    #[test_case(UnicodeVersion::V12_0_0, "12.0.0")]
    #[test_case(UnicodeVersion::V12_1_0, "12.1.0")]
    #[test_case(UnicodeVersion::V13_0_0, "13.0.0")]
    #[test_case(UnicodeVersion::V14_0_0, "14.0.0")]
    #[test_case(UnicodeVersion::V15_0_0, "15.0.0")]
    #[test_case(UnicodeVersion::V15_1_0, "15.1.0")]
    #[test_case(UnicodeVersion::V16_0_0, "16.0.0")]
    #[test_case(UnicodeVersion::V17_0_0, "17.0.0")]
    fn test_display(version: UnicodeVersion, expected: &str) {
        let string = version.to_string();
        assert_eq!(string, expected);
        assert_eq!(
            UnicodeVersion::from_str(&string).expect("Invalid version"),
            version
        );
    }

    #[test_case("9.0.0", UnicodeVersion::V9_0_0)]
    #[test_case("10.0.0", UnicodeVersion::V10_0_0)]
    #[test_case("11.0.0", UnicodeVersion::V11_0_0)]
    #[test_case("12.0.0", UnicodeVersion::V12_0_0)]
    #[test_case("12.1.0", UnicodeVersion::V12_1_0)]
    #[test_case("13.0.0", UnicodeVersion::V13_0_0)]
    #[test_case("14.0.0", UnicodeVersion::V14_0_0)]
    #[test_case("15.0.0", UnicodeVersion::V15_0_0)]
    #[test_case("15.1.0", UnicodeVersion::V15_1_0)]
    #[test_case("16.0.0", UnicodeVersion::V16_0_0)]
    #[test_case("17.0.0", UnicodeVersion::V17_0_0)]
    fn test_version_from_str(version: &str, expected: UnicodeVersion) {
        assert_eq!(
            UnicodeVersion::from_str(version).expect("Invalid version"),
            expected
        );
    }

    #[test]
    fn test_version_from_str_error() {
        assert_eq!(
            UnicodeVersion::from_str("invalid")
                .expect_err("Should fail")
                .to_string(),
            "'invalid' is not a valid Unicode version"
        );
    }

    #[test]
    #[allow(clippy::clone_on_copy)]
    fn test_unicode_version_traits() {
        let version = UnicodeVersion::V15_0_0;
        let mut hasher = DefaultHasher::new();
        version.hash(&mut hasher);
        let _ = hasher.finish();
        let _ = version.clone();
        assert_eq!(format!("{version:?}"), "V15_0_0");
    }

    #[test]
    fn test_interval_query_traits() {
        let query = UnicodeVersion::V15_0_0.query();
        let _ = query.clone();
        assert_eq!(
            format!("{query:?}"), 
            "IntervalQuery { version: V15_0_0, include_categories: None, exclude_categories: None, include_characters: None, exclude_characters: None, min_codepoint: 0, max_codepoint: 1114111 }"
        );
        assert_eq!(query, query);
    }
}