大端 Unicode 字符串

首先，lua 内部使用 utf8 存储字符串。

-- 转换成大端字符串
local function utf8_to_unicode(convertStr)
    if type(convertStr)~="string" then
        return convertStr
    end

    local resultStr=""
    local i=1
    local num1=string.byte(convertStr,i)

    while num1~=nil do
        local tempVar1,tempVar2

        if num1 >= 0x00 and num1 <= 0x7f then
            tempVar1=num1
            tempVar2=0
        elseif bit.band(num1,0xe0)== 0xc0 then
            local t1 = 0
            local t2 = 0
            t1 = bit.band(num1,bit.rshift(0xff,3))
            i=i+1
            num1=string.byte(convertStr,i)
            t2 = bit.band(num1,bit.rshift(0xff,2))

            tempVar1=bit.bor(t2,bit.lshift(bit.band(t1,bit.rshift(0xff,6)),6))
            tempVar2=bit.rshift(t1,2)
        elseif bit.band(num1,0xf0)== 0xe0 then
            local t1 = 0
            local t2 = 0
            local t3 = 0
            t1 = bit.band(num1,bit.rshift(0xff,3))
            i=i+1
            num1=string.byte(convertStr,i)
            t2 = bit.band(num1,bit.rshift(0xff,2))
            i=i+1
            num1=string.byte(convertStr,i)
            t3 = bit.band(num1,bit.rshift(0xff,2))

            tempVar1=bit.bor(bit.lshift(bit.band(t2,bit.rshift(0xff,6)),6),t3)
            tempVar2=bit.bor(bit.lshift(t1,4),bit.rshift(t2,2))

        end

        resultStr=resultStr..string.format("\\u%02x%02x",tempVar2,tempVar1) -- tempVar1, tempVar2 就是小端结果

        i=i+1
        num1=string.byte(convertStr,i)
    end
    return resultStr
end

使用：

local ret = string.upper(string.gsub(utf8_to_unicode('吉米 abc'), '\\u', ''))
print(ret)

配合 java 验证：

public static void printbyte(byte[] bt)  
{  
      for (int i = 0; i < bt.length; i++){  
       int hex = (int)bt[i] & 0xff;  
       System.out.print(Integer.toHexString(hex) + " ");  
      }  
       System.out.println(" length = "+bt.length);  
    }  
  
   @Test  
public void test1() throws UnsupportedEncodingException {  
      
    String name = "杨彬abc";  
    // TODO Auto-generated method stub  
     byte[] defaultBytes = name.getBytes();  
     printbyte(defaultBytes);  
     byte[] utf_8 = name.getBytes("utf-8");  
     printbyte(utf_8);  
     byte[] utf_16be = name.getBytes("utf-16be");  
     printbyte(utf_16be);  
     byte[] utf_16le = name.getBytes("utf-16le");  
     printbyte(utf_16le);  
       
     byte[] gbk = name.getBytes("gbk");  
     printbyte(gbk);  
           /* 
            *  
            e6 9d a8 e5 bd ac 61 62 63  length = 9   默认 
            e6 9d a8 e5 bd ac 61 62 63  length = 9 utf-8             汉字占三个字节 
            而utf-8 采用  3 个字节存储汉字  1个字节存储英文字符 
            unicode 采用两个字节存储汉字及英文字符  
            67 68 5f 6c 0 61 0 62 0 63  length = 10 大端方式    汉字占两个字节 
            68 67 6c 5f 61 0 62 0 63 0  length = 10 小端方式 
            gbk采用两个字节存储汉字   
            d1 ee b1 f2 61 62 63  length = 7    
             */  
     //网络编程socket 通讯 针对字符串采用  utf-16le 这种方式存储  
}  

http://www.cocoachina.com/bbs/read.php?tid-312194-page-1.html
http://yangbinfx.iteye.com/blog/1768501
http://blog.csdn.net/operhero1990/article/details/47044697