523066680

回复 11# CrLf

   《Why didn't I think of that》

happy886rr

回复 12# wskwfkbdn
unicode和unicode big endian一般是靠bom来区分字节序的。如果没有bom，你可以直接在tl的C代码的编码检测群中添加一个函数，检测00 0A或0A 00。从而判断大小端。

/*判断大小端*/
int isUnicode(const char* Str)
{
	if(!Str){
		return false;
	}
	const unsigned char* bytes=(const unsigned char *)Str;
	while(*bytes){
		if(bytes[0]==0x00 && bytes[1]==0x0A){
			return 1;  //小端
		}else if(bytes[0]==0x0A && bytes[1]==0x00){
			return 2;  //大端
		}
		bytes+=2;
	}
	return 0;                  //除非这个文本一个换行符都没有
}
复制代码

wskwfkbdn

TL的bug ，unicode 无BOM头识别失败

CrLf

回复 10# happy886rr


    是哦，我怎么没想到

happy886rr

回复 9# CrLf
小写aa会被干掉吗？

CrLf

附顶楼压缩代码：

@echo off
REM 原始数据
set "LINE=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**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"
setlocal enabledelayedexpansion
set "Z=A"
for %%Z in (@,-,#,$,_,},{,],[) do (
	for %%S in (!Z!) do (
		set "LINE=!LINE:%%S%%S=%%Z!"
	)
	set "Z=%%Z"
)
echo !LINE!
pause
复制代码

CrLf

回复 6# happy886rr


    效果很好，尤其是 EXE 的 Base64，但是目前只适用于 A，若要扩展到其他字符，如何区分不同的层？

pcl_test

回复 5# yu2n

可以想象成俄罗斯套娃

happy886rr

回复 5# yu2n
加权，谁的权重大，就优先压缩谁。层层嵌套，这只是一层，只压了40%，如果是3层，可以实现体积减少80%，不过得携带通用解压字典。类似过去的电报译码表。经过几次测试，这种算法的有点经常超越7Z的压缩率。要想压得狠，就得带字典。

比如论坛代码遇到g c d这几个组合就会被吃掉，所以可以把**定义为一个字母，总之给一些常用的组合编码，制作成码表，解压的时候根据码表重组。就好比基因工程，基因决定了你的生物性状，但基因只是压缩的编码，生物是由基因调度氨基酸组合成蛋白质构成细胞结构体。这也是仿生学。

yu2n

::******Base64加权解密器*************
setlocal enabledelayedexpansion
set "Z=A"&(for %%Z in ([,],{,},_,$,#,-,@) do (set "Z=!Z!!Z!"&for %%S in (!Z!) do (set "LINE=!LINE:%%Z=%%S!")))&echo !LINE:.=!>base64_minitl
复制代码

压缩前 4780 个字符，压缩后 2859 个字符，压缩了 40.19% 。
这算法很好，可惜看不懂。

CrLf

卧槽，史上最快版本迭代

Spring

4096把长行拆成几行了，classpath超出你相信

wskwfkbdn

真牛           ~