<# :
cls
@echo off
cd /d "%~dp0"
powershell -NoProfile -ExecutionPolicy bypass "Invoke-Command -ScriptBlock ([ScriptBlock]::Create([IO.File]::ReadAllText('%~f0',[Text.Encoding]::Default))) -Args '%~f0'"
pause
exit
#>
$self=get-item -liter $args[0];
$path=$self.Directory.FullName
$len=15;
$newfolder=$path+'\#result';
if(-not (test-path -liter $newfolder)){[void](md $newfolder)};
$files=@(dir -liter $path|?{('.txt' -eq $_.Extension) -and ($_ -is [System.IO.FileInfo])});
for($i=0;$i -lt $files.length;$i++){
    write-host $files[$i].Name;
    [System.Collections.ArrayList]$s=@();
    $dic=New-Object 'System.Collections.Generic.Dictionary[string,string]';
    $enc=[Text.Encoding]::Default;
    $text=[IO.File]::ReadAllLines($files[$i].FullName, $enc);
    for($j=0;$j -lt $text.count;$j++){
        $tmpline=$text[$j] -replace '[^\u4e00-\u9fa5]','';
        $a=$tmpline.length;
        $line=$tmpline;
        if($len -lt $a){$line=$tmpline.substring(0, $len);}
        if(-not $dic.ContainsKey($line)){
            $dic.add($line,'');
            [void]$s.add($text[$j]);
        };
    };
    $newfile=$newfolder+'\'+$files[$i].Name;
    [IO.File]::WriteAllLines($newfile, $s, $enc);
};
复制代码

#! /usr/bin/env python3
# coding:utf-8

"""
取前15字符重复的行的第一行
"""

source_file = '2.txt'   # 源文件
new_file = '3.txt'      # 新文件路径
pre_len = 15
d = dict()
with open(source_file,'r',encoding='utf-8') as sou, open(new_file,'w',encoding='utf-8') as new:
    for i in sou:
        if i[:pre_len] not in d:
            d[i[:pre_len]] = i       # 前15个字符
    new.write(''.join(d.values()))
复制代码