Lua UTF-8 全角转半角

根据UTF-8的编码规则，可以知道：

1. 全角空格为12288，半角空格为32 

2. 其他字符半角(33-126)与全角(65281-65374)的对应关系是：均相差65248 

但是UTF-8不能字节转换位整型变量，因此需要一个函数，做一个转换：

function utf8_to_num(raw_string)    local result = {}    pos = 1    while pos <= #raw_string do        local count_1_of_byte = get_continuous_1_count_of_byte(string.byte(raw_string, pos))        local num = 0        if count_1_of_byte < 1 then             num = string.byte(raw_string, pos)            count_1_of_byte = 1        else             boundary = 8            i = count_1_of_byte + 1            while i < boundary * count_1_of_byte do                if 0 == i % boundary then                    i = i + 2                end                if (1 << (boundary - i % boundary - 1)) & string.byte(raw_string, pos + math.floor(i / boundary)) ~= 0 then                    --print(1)                    num = (num << 1) | 1                else                    --print(0)                    num = num << 1                end                i= i + 1            end        end        pos = pos + count_1_of_byte        table.insert(result, num)    end    return resultend

为了方便将一个utf8字符，转换成整型，还需要判断一个utf8字符占了多少个字节，为此需要一个函数来判断（具体参考：http://blog.csdn.net/SKY453589103/article/details/76337557）：

--获取一个字节中，从最高位开始连续的1的个数function  get_continuous_1_count_of_byte(num)    if nil == num then         return -1    end    local count = 0    while num & 0x80 ~= 0 do        count = count + 1        num = num << 1    end    return countend

接下来就只转换的函数：

function full_width_to_half_width(raw_string)    local new_string = {}    local pos = 1    while pos <= #raw_string do        local count_1_of_byte = get_continuous_1_count_of_byte(string.byte(raw_string, pos))        if 3 == count_1_of_byte then            char = string.sub(raw_string, pos, pos + 2)            num_of_char = utf8_to_num(char)[1]            if 12288 == num_of_char then                num_of_char = 32                table.insert(new_string, string.char(num_of_char))            elseif 65281 <= num_of_char and num_of_char <= 65374 then                 num_of_char = num_of_char - 65248                table.insert(new_string, string.char(num_of_char))            end            pos = pos + count_1_of_byte         else            table.insert(new_string, string.sub(raw_string, pos, pos))            pos = pos + 1        end    end    return table.concat(new_string)end

比较简单，就不做解释了。