中文数字转阿拉伯数字 Convert Chinese number count into Arabic integers.

numzh2num-ver02.py

#!/usr/bin/python3
# vim:fileencoding=utf-8:sw=4:et

# 参考 <https://blog.csdn.net/grllery/article/details/89430363>
def numzh2int(txt):
    """Convert Chinese number count to Arabic integer"""
    cn_nums = {'〇': 0, '一': 1, '二': 2, '三': 3, '四': 4,
            '五': 5, '六': 6, '七': 7, '八': 8, '九': 9,
            '零': 0, '壹': 1, '贰': 2, '叁': 3, '肆': 4,
            '伍': 5, '陆': 6, '柒': 7, '捌': 8, '玖': 9,
            '貮': 2, '两': 2, }

    uint = {'十': 10, '拾': 10,
            '百': 100, '佰': 100,
            '千': 1000, '仟': 1000,
            '万': 10000, '萬': 10000,
            '亿': 100000000, '億': 100000000,
            '兆': 1000000000000, }

    units = set(uint.values())
    unit_of_unit = [10**4, 10**8, 10**12]  # 万亿兆: unit of unit

    nums = []
    for c in reversed(txt):
        if c in cn_nums:
            num = cn_nums[c]
            nums.append(num)
        else:
            u = uint[c]
            nums.append(u)

    res = 0
    unit_c = 1  # current unit
    unit_of_unit_c = 1  # current unit of unit
    # print(nums)
    for i, num in enumerate(nums):
        if num in units:
            if num in unit_of_unit and num > unit_of_unit_c:
                unit_of_unit_c = num
                unit_c = num
            else:
                unit_c = num * unit_of_unit_c
            if i == 1:  # Fix for 二百五 == [5, 100, 2]
                res = res * unit_c // 10                
        else:
            res = res + num * unit_c
            unit_c = 1

    # Fix for 十一 == [1, 10].
    # Extra unit_c not reset by the digit place.
    if unit_c > 1:
        res += unit_c
    return res

def main():
    test_nums = [
("八", 8),
("十", 10),
("十一", 11),
("五十", 50),
("五百", 500),
("三千", 3_000),
("一万", 10_000),
("十万", 100_000),
("百万", 1_000_000),
("三十二", 32),
("二百五", 250),
("两万五", 25_000),
("万八千", 18_000),
("一百万", 1_000_000),
("三千万", 30_000_000),
("四万亿", 4_000_000_000_000),
("五百零三", 503),
("五百一十", 510),
("一百二十三", 123),
("一千二百零三", 1_203),
("一万一千一百", 11_100),
("一万一千一百零一", 11_101),
("一万一千一百一十", 11_110),
("十万三千六百零九", 103_609),
("十万零三千六百零九", 103_609),
("一百二十三万四千五百六十七", 1_234_567),
("一千一百二十三万四千五百六十七", 11_234_567),
("一亿一千一百二十三万四千五百六十七", 111_234_567),
("一百零二亿五千零一万一千零三十八", 10_250_011_038),
("一万三千零三十二亿五千四百零一万一千零三十八", 1_303_254_011_038),
("一兆三千零三十二亿五千四百零一万一千零三十八", 1_303_254_011_038),
("一千兆三千零三十二亿五千四百零一万一千零三十八", 1_000_303_254_011_038),
("一万兆三千零三十二亿五千四百零一万一千零三十八", 10_000_303_254_011_038),
            ]
    #item = "一千兆三千零三十二亿五千四百零一万零一千零三十八"; print(f'{item}, {numzh2int(item):,d}'); return
    for item, val in test_nums:
        r = numzh2int(item)
        print(f'{r==val} {item}, {r:,d}')

if __name__ == '__main__':
    main()