parseChineseNumber / parseRomanNumber / parseEnglishNumber

Parse numbers written for humans: 第二十三章, Chapter XIV, Part Three.

All three share one contract: return null rather than a guess when the input cannot be parsed in full. These parsers usually feed a "is this line a heading?" decision, and one wrong number poisons the whole sequence check.

API

Function	Accepts
parseChineseNumber(value)	Digits (half- and full-width), 一二三…, units 十百千万/萬, simplified + traditional
parseRomanNumber(value)	IVXLCDM, any case, subtractive notation (IV, IX)
parseEnglishNumber(value)	Digits, English number words one–twenty, then Roman numerals

Related string helpers: toHalfWidth(value) / toFullWidth(value) normalize full-width characters, which parseChineseNumber applies for you.

Example

import { parseChineseNumber, parseRomanNumber, parseEnglishNumber, toHalfWidth } from 'ranuts';

parseChineseNumber('二十三'); // 23
parseChineseNumber('一百零三'); // 103
parseChineseNumber('三萬'); // 30000
parseChineseNumber('第三章'); // null — extract the number segment first

parseRomanNumber('MCMXCIV'); // 1994
parseEnglishNumber('Three'); // 3
toHalfWidth('（１）'); // '(1)'

Notes

1. Pass only the number segment. 第三章 returns null — pull 三 out with your own pattern first, then parse it.
2. 十 with nothing in front is 1, so 十五 is 15, not 5.
3. parseEnglishNumber tries digits, then words, then Roman. twenty-one and above are not covered; extend the word table if you need them.