enhnace(frontend): 文字列比較のためのローマナイズを強化(設定の検索) (#15632)

* enhnace(frontend): 文字列比較のためのローマナイズを強化

* docs

* fix

* fix

* fix

* comment

* wanakanaの初回ロードをコンポーネント内に移動

* comment

* fix

* add tests

---------

Co-authored-by: syuilo <4439005+syuilo@users.noreply.github.com>
This commit is contained in:
かっこかり 2025-03-09 14:21:23 +09:00 committed by GitHub
parent bdb74539d4
commit f35eb0f6d9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 279 additions and 11 deletions

View file

@ -0,0 +1,97 @@
/*
* SPDX-FileCopyrightText: syuilo and misskey-project
* SPDX-License-Identifier: AGPL-3.0-only
*/
import { versatileLang } from '@@/js/intl-const.js';
import type { toHiragana as toHiraganaType } from 'wanakana';
let toHiragana: typeof toHiraganaType = (str?: string) => str ?? '';
let isWanakanaLoaded = false;
/**
* lazy-loading
*
* 使
*/
export async function initIntlString(forceWanakana = false) {
if ((!versatileLang.includes('ja') && !forceWanakana) || isWanakanaLoaded) return;
const { toHiragana: _toHiragana } = await import('wanakana');
toHiragana = _toHiragana;
isWanakanaLoaded = true;
}
/**
* -
* -
* - : `` ``
* -
* -
* -
*/
export function normalizeString(str: string) {
const segmenter = new Intl.Segmenter(versatileLang, { granularity: 'grapheme' });
return [...segmenter.segment(str)].map(({ segment }) => segment.normalize('NFKC')).join('').toLowerCase().trim();
}
// https://qiita.com/non-caffeine/items/77360dda05c8ce510084
const hyphens = [
0x002d, // hyphen-minus
0x02d7, // modifier letter minus sign
0x1173, // hangul jongseong eu
0x1680, // ogham space mark
0x1b78, // balinese musical symbol left-hand open pang
0x2010, // hyphen
0x2011, // non-breaking hyphen
0x2012, // figure dash
0x2013, // en dash
0x2014, // em dash
0x2015, // horizontal bar
0x2043, // hyphen bullet
0x207b, // superscript minus
0x2212, // minus sign
0x25ac, // black rectangle
0x2500, // box drawings light horizontal
0x2501, // box drawings heavy horizontal
0x2796, // heavy minus sign
0x30fc, // katakana-hiragana prolonged sound mark
0x3161, // hangul letter eu
0xfe58, // small em dash
0xfe63, // small hyphen-minus
0xff0d, // fullwidth hyphen-minus
0xff70, // halfwidth katakana-hiragana prolonged sound mark
0x10110, // aegean number ten
0x10191, // roman uncia sign
];
const hyphensCodePoints = hyphens.map(code => `\\u{${code.toString(16).padStart(4, '0')}}`);
/** ハイフンを統一(ローマ字半角入力時に`ー`と`-`が判定できない問題の調整) */
export function normalizeHyphens(str: string) {
return str.replace(new RegExp(`[${hyphensCodePoints.join('')}]`, 'ug'), '\u002d');
}
/**
* `normalizeString`
*
* `normalizeString`
*/
export function normalizeStringWithHiragana(str: string) {
return normalizeHyphens(toHiragana(normalizeString(str), { convertLongVowelMark: false }));
}
/** aとbが同じかどうか */
export function compareStringEquals(a: string, b: string) {
return (
normalizeString(a) === normalizeString(b) ||
normalizeStringWithHiragana(a) === normalizeStringWithHiragana(b)
);
}
/** baseにqueryが含まれているかどうか */
export function compareStringIncludes(base: string, query: string) {
return (
normalizeString(base).includes(normalizeString(query)) ||
normalizeStringWithHiragana(base).includes(normalizeStringWithHiragana(query))
);
}