rime-ice/lua/number_translator.lua

-- 来源 https://github.com/yanhuacuo/98wubi-tables > http://98wb.ysepan.com/
-- 数字、金额大写
-- 触发前缀默认为 recognizer/patterns/number 的第 2 个字符，即 R

local function splitNumPart(str)
    local part = {}
    part.int, part.dot, part.dec = string.match(str, "^(%d*)(%.?)(%d*)")
    return part
end

local function GetPreciseDecimal(nNum, n)
    if type(nNum) ~= "number" then nNum = tonumber(nNum) end
    n = n or 0;
    n = math.floor(n)
    if n < 0 then n = 0 end
    local nDecimal = 10 ^ n
    local nTemp = math.floor(nNum * nDecimal);
    local nRet = nTemp / nDecimal;
    return nRet;
end

local function decimal_func(str, posMap, valMap)
    local dec
    posMap = posMap or { [1] = "角", [2] = "分", [3] = "厘", [4] = "毫" }
    valMap = valMap or { [0] = "零", "壹", "贰", "叁", "肆", "伍", "陆", "柒", "捌", "玖" }
    if #str > 4 then dec = string.sub(tostring(str), 1, 4) else dec = tostring(str) end
    dec = string.gsub(dec, "0+$", "")

    if dec == "" then return "整" end

    local result = ""
    for pos = 1, #dec do
        local val = tonumber(string.sub(dec, pos, pos))
        if val ~= 0 then result = result .. valMap[val] .. posMap[pos] else result = result .. valMap[val] end
    end
    result = result:gsub(valMap[0] .. valMap[0], valMap[0])
    return result:gsub(valMap[0] .. valMap[0], valMap[0])
end

-- 把数字串按千分位四位数分割，进行转换为中文
local function formatNum(num, t)
    local digitUnit, wordFigure
    local result = ""
    num = tostring(num)
    if tonumber(t) < 1 then digitUnit = { "", "十", "百", "千" } else digitUnit = { "", "拾", "佰", "仟" } end
    if tonumber(t) < 1 then
        wordFigure = { "〇", "一", "二", "三", "四", "五", "六", "七", "八", "九" }
    else
        wordFigure = { "零", "壹", "贰", "叁", "肆", "伍", "陆", "柒", "捌", "玖" }
    end
    if string.len(num) > 4 or tonumber(num) == 0 then return wordFigure[1] end
    local lens = string.len(num)
    for i = 1, lens do
        local n = wordFigure[tonumber(string.sub(num, -i, -i)) + 1]
        if n ~= wordFigure[1] then result = n .. digitUnit[i] .. result else result = n .. result end
    end
    result = result:gsub(wordFigure[1] .. wordFigure[1], wordFigure[1])
    result = result:gsub(wordFigure[1] .. "$", "")
    result = result:gsub(wordFigure[1] .. "$", "")

    return result
end

-- 数值转换为中文
local function number2cnChar(num, flag, digitUnit, wordFigure) --flag=0中文小写反之为大写
    local result = ""

    if tonumber(flag) < 1 then
        digitUnit = digitUnit or { [1] = "万", [2] = "亿" }
        wordFigure = wordFigure or { [1] = "〇", [2] = "一", [3] = "十", [4] = "元" }
    else
        digitUnit = digitUnit or { [1] = "万", [2] = "亿" }
        wordFigure = wordFigure or { [1] = "零", [2] = "壹", [3] = "拾", [4] = "元" }
    end
    local lens = string.len(num)
    if lens < 5 then
        result = formatNum(num, flag)
    elseif lens < 9 then
        result = formatNum(string.sub(num, 1, -5), flag) .. digitUnit[1] .. formatNum(string.sub(num, -4, -1), flag)
    elseif lens < 13 then
        result = formatNum(string.sub(num, 1, -9), flag) ..
            digitUnit[2] ..
            formatNum(string.sub(num, -8, -5), flag) .. digitUnit[1] .. formatNum(string.sub(num, -4, -1), flag)
    else
        result = ""
    end

    result = result:gsub("^" .. wordFigure[1], "")
    result = result:gsub(wordFigure[1] .. digitUnit[1], "")
    result = result:gsub(wordFigure[1] .. digitUnit[2], "")
    result = result:gsub(wordFigure[1] .. wordFigure[1], wordFigure[1])
    result = result:gsub(wordFigure[1] .. "$", "")
    if lens > 4 then result = result:gsub("^" .. wordFigure[2] .. wordFigure[3], wordFigure[3]) end
    if result ~= "" then result = result .. wordFigure[4] else result = "数值超限！" end

    return result
end

local function number2zh(num, t)
    local result, wordFigure
    result = ""
    if tonumber(t) < 1 then
        wordFigure = { "〇", "一", "二", "三", "四", "五", "六", "七", "八", "九" }
    else
        wordFigure = { "零", "壹", "贰", "叁", "肆", "伍", "陆", "柒", "捌", "玖" }
    end
    if tostring(num) == nil then return "" end
    for pos = 1, string.len(num) do
        result = result .. wordFigure[tonumber(string.sub(num, pos, pos) + 1)]
    end
    result = result:gsub(wordFigure[1] .. wordFigure[1], wordFigure[1])
    return result:gsub(wordFigure[1] .. wordFigure[1], wordFigure[1])
end

local function number_translatorFunc(num)
    local numberPart = splitNumPart(num)
    local result = {}
    if numberPart.dot ~= "" then
        table.insert(result,
            { number2cnChar(numberPart.int, 0, { "万", "亿" }, { "〇", "一", "十", "点" }) .. number2zh(numberPart.dec, 0),
                "〔数字小写〕" })
        table.insert(result,
            { number2cnChar(numberPart.int, 1, { "萬", "億" }, { "〇", "一", "十", "点" }) .. number2zh(numberPart.dec, 1),
                "〔数字大写〕" })
    else
        table.insert(result, { number2cnChar(numberPart.int, 0, { "万", "亿" }, { "〇", "一", "十", "" }), "〔数字小写〕" })
        table.insert(result, { number2cnChar(numberPart.int, 1, { "萬", "億" }, { "零", "壹", "拾", "" }), "〔数字大写〕" })
    end
    table.insert(result,
        { number2cnChar(numberPart.int, 0) ..
        decimal_func(numberPart.dec, { [1] = "角", [2] = "分", [3] = "厘", [4] = "毫" },
            { [0] = "〇", "一", "二", "三", "四", "五", "六", "七", "八", "九" }), "〔金额小写〕" })

    local number2cnCharInt = number2cnChar(numberPart.int, 1)
    local number2cnCharDec = decimal_func(numberPart.dec, { [1] = "角", [2] = "分", [3] = "厘", [4] = "毫" }, { [0] = "零", "壹", "贰", "叁", "肆", "伍", "陆", "柒", "捌", "玖" })
    if string.len(numberPart.int) > 4 and number2cnCharInt:find('^拾[壹贰叁肆伍陆柒捌玖]?') and number2cnCharInt:find('[万亿]')  then -- 简易地规避 utf8 匹配问题
        local number2cnCharInt_var = number2cnCharInt:gsub('^拾', '壹拾')
        table.insert(result, { number2cnCharInt_var .. number2cnCharDec , "〔金额大写〕"})
        -- 会计书写要求 https://github.com/iDvel/rime-ice/issues/989
    else
        table.insert(result, { number2cnCharInt .. number2cnCharDec , "〔金额大写〕"})
    end
    return result
end

local function number_translator(input, seg, env)
    -- 获取 recognizer/patterns/number 的第 2 个字符作为触发前缀
    env.number_keyword = env.number_keyword or
        env.engine.schema.config:get_string('recognizer/patterns/number'):sub(2, 2)
    local str, num, numberPart
    if env.number_keyword ~= '' and input:sub(1, 1) == env.number_keyword then
        str = string.gsub(input, "^(%a+)", "")
        numberPart = number_translatorFunc(str)
        if str and #str > 0 and #numberPart > 0 then
            for i = 1, #numberPart do
                yield(Candidate(input, seg.start, seg._end, numberPart[i][1], numberPart[i][2]))
            end
        end
    end
end

-- print(#number_translatorFunc(3355.433))
return number_translator