China DOS Union

I have a text with the following content:
Zhang San
Li Si
Wang Er Ma Zi
Xiao Bai
Xiao Hei
Xiao Qiang
...... rem: there are N lines in the middle
Xiao Er
Zhao Da
QQ
DOS
WIN98
WINXP
Now I want to achieve swapping WINXP and Zhang San to get the following text: (that is, swap the first line and the last line of the text)
WINXP
Li Si
Wang Er Ma Zi
Xiao Bai
Xiao Hei
Xiao Qiang
...... rem: there are N lines in the middle
Xiao Er
Zhao Da
QQ
DOS
WIN98
Zhang San

@echo %dbg% off

setlocal enabledelayedexpansion

if == goto :error

if exist %4 del /q %4



for /F "delims=" %%. in (%1) do (set /a line+=1 && set "num!line!=%%.")

(set swapTemp=!num%3!) && (set num%3=!num%2!) && (set num%2=!swapTemp!)

for /L %%i in (1,1,!line!) do (echo !num%%i!>>%4)

start %4

goto :eof



:error

    echo.

    echo Usage: %0.Bat  File to process  Line to swap  Line to be swapped  Swapped file name

    echo.

    echo Example (A): %0.Bat swap.dat  5 10 a.txt

    echo             Process the swap.dat file, swap the data of line 5 and line 10, and generate a new file: a.txt

    echo.

Usage: If no parameters are added, the usage instructions will be prompted.

Usage: Swap.BAT  File to process  Line to swap  Line to be swapped  Swapped file name





    Example:  Swap.BAT    swap.dat    5    10    a.txt

    It means: Process the swap.dat file, swap the data of line 5 and line 10, and generate a new file: a.txt

              The Swap.BAT file is the batch file name for this processing to swap data, and the owner can name it arbitrarily.

Principle: The first wife and the second wife want to exchange their gifts with each other.
So the first wife first temporarily stores her gift with me. After her hand is empty, she takes the second wife's gift.
After the second wife's gift is handed over to the first wife, her hand is empty. At this time, I give the gift that the first wife temporarily stored with me to the second wife.
So they both exchanged gifts ~ : )


If the specified new file to be generated already exists, it will be automatically deleted without prompt.

[ Last edited by redtek on 2006-11-6 at 05:50 AM ]

Suppose the test.txt text you need to process has no sensitive characters under CMD (such as <, >, |, etc.), the first line is not empty, and there are no duplicate lines, then the following code can basically meet your needs:

@echo off

:: Idea:

::     Extract the content of the first line and the last line (if the last line is an empty line, then the extracted content is the non-empty last line content),

:: Then write the extracted last line at the first line of the new text, use the findstr statement to filter out the first line and the last line of the original text, and add the extracted first line after the last line.

:: Because there are conditional restrictions when findstr processes the last line, so first generate an empty line after the last line to ensure the accuracy of extraction

copy test.txt test.bak>nul

echo.>>test.txt

set num=0

setlocal enabledelayedexpansion

for /f "delims=" %%i in (test.txt) do (

    set str=%%i

    set /a num+=1

    if !num! equ 1 set var=%%i

)

>test_.txt echo.%str%

findstr /b /e /v /c:"%str%" /c:"%var%" test.txt>>test_.txt

>>test_.txt echo.%var%

move test.bak test.txt

start test_.txt

[ Last edited by namejm on 2006-11-6 at 02:03 PM ]

Originally, it should be for the convenience of the original poster's application.
For example: If you want to swap the 3rd line and the 3rd line from the end, then

　　　Swap.BAT swap.dat 3 -3 a.txt
　or　Swap.BAT swap.dat 3 a.txt

It would be more convenient for the original poster if it could be like this. I really can't do it, forgive me~

On the basis of 3F, the following code is obtained, which can handle lines with repeated content. The principle is similar to 3F, but instead of filtering line content, it filters line numbers:

@echo off

:: Note: Because when findstr /n counts the number of lines, the last empty line will be ignored

:: So, if there is an empty line at the end, please ensure that the number of empty lines is greater than 1

findstr /n .* test.txt>tmp1.txt

for /f "tokens=1* delims=:" %%i in (tmp1.txt) do (

    if %%i equ 1 set str=%%j

    set line=%%i&set var=%%j

)

>tmp2.txt echo %line%:%var%

findstr /b /v /c:"1:" /c:"%line%:" tmp1.txt>>tmp2.txt

echo 1:%str%>>tmp2.txt

echo.>>tmp2.txt

cd.>result.txt

for /f "tokens=1* delims=:" %%i in (tmp2.txt) do >>result.txt echo.%%j

del /q tmp1.txt tmp2.txt

start result.txt

sed -n "$p" >temp.txt
sed "1{d;p}" test.txt |sed "${d;p}" >>temp.txt
sed -n "1p" >>temp.txt
Not tested, theoretically feasible

Thanks, but namejm, the command on the 3rd floor didn't work!!!

The code of 3F has many restrictions. If I miss some situations, can you post part of your txt content?

I'm sorry. When I posted the message, I copied the code before modification and didn't post the corrected content, resulting in the code for 3F missing the statements for writing the first and last lines. Now it has been corrected. Please test again.

test.bat
@echo off
::%1 is life, swap %2 and %3
::For example, swap line 1 and line 9 in test.txt, use method test.bat test.txt,1,9
::Temporarily cannot change 3,-3 and so on, but you can use sed -n "$=" %1 to see how many lines there are in total~ Then set/a to calculate the actual + lines

if *%1==* (echo No file specified&pause>nul&goto :eof)
if *%2==* (set s=1) else (set s=%2)
if *%3==* (set e=$) else (set e=%3)
if not "%e%"=="$" if %e% lss %s% (set "s=%3"&set "e=%2")

for /f "tokens=*" %%? in ('sed -n "%e%p" %1') do (
sed -e "%s%h" -e "%e%x" %1|sed "%s% c\%%?" |more>temp.txt
)
pause
exit



Test test.txt
`-=\;',./~!@#o^&*()_+{}|:"<>?
ab
AB
12 34
One two three four five six seven eight nine ○
！·￥……—（）——：“？》《
http ://abc.abc.abc/

[ Last edited by vkill on 2006-11-7 at 09:55 AM ]

The systems in internet cafes are generally cut down. The commands with security issues may all be disabled.

Originally posted by namejm at 2006-11-7 09:23:
The systems in internet cafes are generally the ones that have been cut down, and commands with security issues may all have been disabled.

Brother namejm, test it. The principle is to first replace 9 with 1, and then copy 1 as the content of the original file 9~

11 F code update

[ Last edited by vkill on 2006-11-7 at 09:58 AM ]

Actually, the sed program wasn't saved. I know that Brother Wunaikai just posted it recently, but I can't find the download at the moment. Sweat~~ Can you give a download address?

I also add a sed one, the file should not be too huge.

sed -re :a -e "N;$!ba;s/(*)\n(.*)\n(.*)/\3\n\2\n\1/"

sed download: http://www.student.northpark.edu/pemente/sed/gsed407x.zip