yomichan/ext/bg/js/japanese.js

459 lines
13 KiB
JavaScript
Raw Normal View History

2017-07-19 06:07:46 +00:00
/*
2020-01-01 17:00:00 +00:00
* Copyright (C) 2016-2020 Alex Yatskov <alex@foosoft.net>
2017-07-19 06:07:46 +00:00
* Author: Alex Yatskov <alex@foosoft.net>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
2017-08-26 19:21:28 +00:00
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
2017-07-19 06:07:46 +00:00
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
2020-01-01 17:00:31 +00:00
* along with this program. If not, see <https://www.gnu.org/licenses/>.
2017-07-19 06:07:46 +00:00
*/
2020-03-11 02:30:36 +00:00
/* global
* wanakana
*/
2017-07-19 06:07:46 +00:00
2020-01-23 01:54:06 +00:00
const JP_HALFWIDTH_KATAKANA_MAPPING = new Map([
2019-12-22 19:07:30 +00:00
['ヲ', 'ヲヺ-'],
['ァ', 'ァ--'],
['ィ', 'ィ--'],
['ゥ', 'ゥ--'],
['ェ', 'ェ--'],
['ォ', 'ォ--'],
['ャ', 'ャ--'],
['ュ', 'ュ--'],
['ョ', 'ョ--'],
['ッ', 'ッ--'],
['ー', 'ー--'],
['ア', 'ア--'],
['イ', 'イ--'],
['ウ', 'ウヴ-'],
['エ', 'エ--'],
['オ', 'オ--'],
['カ', 'カガ-'],
['キ', 'キギ-'],
['ク', 'クグ-'],
['ケ', 'ケゲ-'],
['コ', 'コゴ-'],
['サ', 'サザ-'],
['シ', 'シジ-'],
['ス', 'スズ-'],
['セ', 'セゼ-'],
['ソ', 'ソゾ-'],
['タ', 'タダ-'],
['チ', 'チヂ-'],
['ツ', 'ツヅ-'],
['テ', 'テデ-'],
['ト', 'トド-'],
['ナ', 'ナ--'],
['ニ', 'ニ--'],
['ヌ', 'ヌ--'],
['ネ', 'ネ--'],
['ノ', '--'],
['ハ', 'ハバパ'],
['ヒ', 'ヒビピ'],
['フ', 'フブプ'],
['ヘ', 'ヘベペ'],
['ホ', 'ホボポ'],
['マ', 'マ--'],
['ミ', 'ミ--'],
['ム', 'ム--'],
['メ', 'メ--'],
['モ', 'モ--'],
['ヤ', 'ヤ--'],
['ユ', 'ユ--'],
['ヨ', 'ヨ--'],
['ラ', 'ラ--'],
['リ', 'リ--'],
['ル', 'ル--'],
['レ', 'レ--'],
['ロ', 'ロ--'],
['ワ', 'ワ--'],
['ン', 'ン--']
]);
2020-01-23 02:41:32 +00:00
const JP_HIRAGANA_RANGE = [0x3040, 0x309f];
const JP_KATAKANA_RANGE = [0x30a0, 0x30ff];
const JP_KANA_RANGES = [JP_HIRAGANA_RANGE, JP_KATAKANA_RANGE];
const JP_CJK_COMMON_RANGE = [0x4e00, 0x9fff];
const JP_CJK_RARE_RANGE = [0x3400, 0x4dbf];
const JP_CJK_RANGES = [JP_CJK_COMMON_RANGE, JP_CJK_RARE_RANGE];
const JP_ITERATION_MARK_CHAR_CODE = 0x3005;
// Japanese character ranges, roughly ordered in order of expected frequency
const JP_JAPANESE_RANGES = [
JP_HIRAGANA_RANGE,
JP_KATAKANA_RANGE,
JP_CJK_COMMON_RANGE,
JP_CJK_RARE_RANGE,
[0xff66, 0xff9f], // Halfwidth katakana
[0x30fb, 0x30fc], // Katakana punctuation
[0xff61, 0xff65], // Kana punctuation
[0x3000, 0x303f], // CJK punctuation
[0xff10, 0xff19], // Fullwidth numbers
[0xff21, 0xff3a], // Fullwidth upper case Latin letters
[0xff41, 0xff5a], // Fullwidth lower case Latin letters
[0xff01, 0xff0f], // Fullwidth punctuation 1
[0xff1a, 0xff1f], // Fullwidth punctuation 2
[0xff3b, 0xff3f], // Fullwidth punctuation 3
[0xff5b, 0xff60], // Fullwidth punctuation 4
2020-02-09 01:48:47 +00:00
[0xffe0, 0xffee] // Currency markers
2020-01-23 02:41:32 +00:00
];
2020-01-23 01:54:06 +00:00
// Helper functions
function _jpIsCodePointInRanges(codePoint, ranges) {
for (const [min, max] of ranges) {
if (codePoint >= min && codePoint <= max) {
return true;
}
}
return false;
}
// Character code testing functions
function jpIsCodePointKanji(codePoint) {
return _jpIsCodePointInRanges(codePoint, JP_CJK_RANGES);
}
function jpIsCodePointKana(codePoint) {
return _jpIsCodePointInRanges(codePoint, JP_KANA_RANGES);
}
function jpIsCodePointJapanese(codePoint) {
return _jpIsCodePointInRanges(codePoint, JP_JAPANESE_RANGES);
}
// String testing functions
function jpIsStringEntirelyKana(str) {
if (str.length === 0) { return false; }
for (const c of str) {
if (!jpIsCodePointKana(c.codePointAt(0))) {
return false;
}
}
return true;
}
function jpIsStringPartiallyJapanese(str) {
if (str.length === 0) { return false; }
for (const c of str) {
if (jpIsCodePointJapanese(c.codePointAt(0))) {
return true;
}
}
return false;
}
// Conversion functions
2017-07-19 06:07:46 +00:00
function jpKatakanaToHiragana(text) {
2017-08-15 06:10:59 +00:00
let result = '';
for (const c of text) {
if (wanakana.isKatakana(c)) {
result += wanakana.toHiragana(c);
} else {
result += c;
}
}
return result;
2017-07-19 06:07:46 +00:00
}
2019-11-13 18:24:11 +00:00
function jpHiraganaToKatakana(text) {
let result = '';
for (const c of text) {
if (wanakana.isHiragana(c)) {
result += wanakana.toKatakana(c);
} else {
result += c;
}
}
return result;
}
function jpToRomaji(text) {
return wanakana.toRomaji(text);
}
function jpConvertReading(expressionFragment, readingFragment, readingMode) {
switch (readingMode) {
case 'hiragana':
return jpKatakanaToHiragana(readingFragment || '');
case 'katakana':
return jpHiraganaToKatakana(readingFragment || '');
case 'romaji':
if (readingFragment) {
return jpToRomaji(readingFragment);
} else {
if (jpIsStringEntirelyKana(expressionFragment)) {
2019-11-13 18:24:11 +00:00
return jpToRomaji(expressionFragment);
}
}
return readingFragment;
2020-01-25 00:56:14 +00:00
case 'none':
return null;
2019-11-13 18:24:11 +00:00
default:
return readingFragment;
}
}
2017-08-26 18:57:34 +00:00
function jpDistributeFurigana(expression, reading) {
const fallback = [{furigana: reading, text: expression}];
if (!reading) {
return fallback;
2017-08-24 03:56:52 +00:00
}
2017-08-26 18:57:34 +00:00
let isAmbiguous = false;
2020-02-17 20:21:30 +00:00
const segmentize = (reading2, groups) => {
if (groups.length === 0 || isAmbiguous) {
2017-08-26 18:57:34 +00:00
return [];
2017-08-24 03:56:52 +00:00
}
2017-08-26 18:57:34 +00:00
const group = groups[0];
if (group.mode === 'kana') {
2020-02-17 20:21:30 +00:00
if (jpKatakanaToHiragana(reading2).startsWith(jpKatakanaToHiragana(group.text))) {
const readingLeft = reading2.substring(group.text.length);
2017-08-26 18:57:34 +00:00
const segs = segmentize(readingLeft, groups.splice(1));
if (segs) {
2019-11-03 03:08:57 +00:00
return [{text: group.text}].concat(segs);
2017-08-26 18:57:34 +00:00
}
}
} else {
let foundSegments = null;
2020-02-17 20:21:30 +00:00
for (let i = reading2.length; i >= group.text.length; --i) {
const readingUsed = reading2.substring(0, i);
const readingLeft = reading2.substring(i);
2017-08-26 18:57:34 +00:00
const segs = segmentize(readingLeft, groups.slice(1));
if (segs) {
if (foundSegments !== null) {
// more than one way to segmentize the tail, mark as ambiguous
isAmbiguous = true;
return null;
}
foundSegments = [{text: group.text, furigana: readingUsed}].concat(segs);
}
// there is only one way to segmentize the last non-kana group
if (groups.length === 1) {
break;
2017-08-24 03:56:52 +00:00
}
2017-08-26 18:57:34 +00:00
}
return foundSegments;
2017-08-26 18:57:34 +00:00
}
};
const groups = [];
let modePrev = null;
for (const c of expression) {
const codePoint = c.codePointAt(0);
const modeCurr = jpIsCodePointKanji(codePoint) || codePoint === JP_ITERATION_MARK_CHAR_CODE ? 'kanji' : 'kana';
2017-08-26 18:57:34 +00:00
if (modeCurr === modePrev) {
groups[groups.length - 1].text += c;
} else {
groups.push({mode: modeCurr, text: c});
modePrev = modeCurr;
2017-08-24 03:56:52 +00:00
}
}
2017-08-26 18:57:34 +00:00
const segments = segmentize(reading, groups);
if (segments && !isAmbiguous) {
return segments;
}
return fallback;
2017-08-24 03:56:52 +00:00
}
2019-11-03 03:08:57 +00:00
function jpDistributeFuriganaInflected(expression, reading, source) {
const output = [];
let stemLength = 0;
const shortest = Math.min(source.length, expression.length);
const sourceHiragana = jpKatakanaToHiragana(source);
const expressionHiragana = jpKatakanaToHiragana(expression);
while (stemLength < shortest && sourceHiragana[stemLength] === expressionHiragana[stemLength]) {
2019-11-03 03:08:57 +00:00
++stemLength;
}
const offset = source.length - stemLength;
2019-12-08 20:13:22 +00:00
const stemExpression = source.substring(0, source.length - offset);
const stemReading = reading.substring(
0,
offset === 0 ? reading.length : reading.length - expression.length + stemLength
2019-11-11 19:58:04 +00:00
);
for (const segment of jpDistributeFurigana(stemExpression, stemReading)) {
2019-11-03 03:08:57 +00:00
output.push(segment);
}
if (stemLength !== source.length) {
2019-12-08 20:13:22 +00:00
output.push({text: source.substring(stemLength)});
2019-11-03 03:08:57 +00:00
}
return output;
}
2019-12-22 19:07:30 +00:00
function jpConvertHalfWidthKanaToFullWidth(text, sourceMapping) {
let result = '';
const hasSourceMapping = Array.isArray(sourceMapping);
// This function is safe to use charCodeAt instead of codePointAt, since all
// the relevant characters are represented with a single UTF-16 character code.
for (let i = 0, ii = text.length; i < ii; ++i) {
2019-12-22 19:07:30 +00:00
const c = text[i];
2020-01-23 01:54:06 +00:00
const mapping = JP_HALFWIDTH_KATAKANA_MAPPING.get(c);
2019-12-22 19:07:30 +00:00
if (typeof mapping !== 'string') {
result += c;
continue;
}
let index = 0;
switch (text.charCodeAt(i + 1)) {
case 0xff9e: // dakuten
index = 1;
break;
case 0xff9f: // handakuten
index = 2;
break;
}
let c2 = mapping[index];
if (index > 0) {
if (c2 === '-') { // invalid
index = 0;
c2 = mapping[0];
} else {
++i;
}
}
if (hasSourceMapping && index > 0) {
index = result.length;
const v = sourceMapping.splice(index + 1, 1)[0];
sourceMapping[index] += v;
}
result += c2;
}
return result;
}
function jpConvertNumericTofullWidth(text) {
let result = '';
for (const char of text) {
let c = char.codePointAt(0);
2019-12-22 19:07:30 +00:00
if (c >= 0x30 && c <= 0x39) { // ['0', '9']
c += 0xff10 - 0x30; // 0xff10 = '0' full width
result += String.fromCodePoint(c);
2019-12-22 19:07:30 +00:00
} else {
result += char;
2019-12-22 19:07:30 +00:00
}
}
return result;
}
function jpConvertAlphabeticToKana(text, sourceMapping) {
let part = '';
let result = '';
const ii = text.length;
if (sourceMapping.length === ii) {
sourceMapping.length = ii;
sourceMapping.fill(1);
}
for (const char of text) {
// Note: 0x61 is the character code for 'a'
let c = char.codePointAt(0);
2019-12-22 19:07:30 +00:00
if (c >= 0x41 && c <= 0x5a) { // ['A', 'Z']
c += (0x61 - 0x41);
2019-12-22 19:07:30 +00:00
} else if (c >= 0x61 && c <= 0x7a) { // ['a', 'z']
// NOP; c += (0x61 - 0x61);
} else if (c >= 0xff21 && c <= 0xff3a) { // ['A', 'Z'] fullwidth
c += (0x61 - 0xff21);
} else if (c >= 0xff41 && c <= 0xff5a) { // ['a', 'z'] fullwidth
c += (0x61 - 0xff41);
2020-01-24 22:58:05 +00:00
} else if (c === 0x2d || c === 0xff0d) { // '-' or fullwidth dash
c = 0x2d; // '-'
2019-12-22 19:07:30 +00:00
} else {
if (part.length > 0) {
result += jpToHiragana(part, sourceMapping, result.length);
part = '';
}
result += char;
2019-12-22 19:07:30 +00:00
continue;
}
part += String.fromCodePoint(c);
2019-12-22 19:07:30 +00:00
}
if (part.length > 0) {
result += jpToHiragana(part, sourceMapping, result.length);
}
return result;
}
function jpToHiragana(text, sourceMapping, sourceMappingStart) {
const result = wanakana.toHiragana(text);
// Generate source mapping
if (Array.isArray(sourceMapping)) {
if (typeof sourceMappingStart !== 'number') { sourceMappingStart = 0; }
let i = 0;
let resultPos = 0;
const ii = text.length;
while (i < ii) {
// Find smallest matching substring
let iNext = i + 1;
let resultPosNext = result.length;
while (iNext < ii) {
const t = wanakana.toHiragana(text.substring(0, iNext));
if (t === result.substring(0, t.length)) {
resultPosNext = t.length;
break;
}
++iNext;
}
// Merge characters
const removals = iNext - i - 1;
if (removals > 0) {
let sum = 0;
const vs = sourceMapping.splice(sourceMappingStart + 1, removals);
for (const v of vs) { sum += v; }
sourceMapping[sourceMappingStart] += sum;
}
++sourceMappingStart;
// Empty elements
const additions = resultPosNext - resultPos - 1;
for (let j = 0; j < additions; ++j) {
sourceMapping.splice(sourceMappingStart, 0, 0);
++sourceMappingStart;
}
i = iNext;
resultPos = resultPosNext;
}
}
return result;
}