How to handle file name with brackets in Windows CMD Batch for loop?

Question

I want to write a script that move files which dropped on it to somewhere. So this is how my batch code goes:

@echo off
for %%i in (%*) do move "%%~i" "somewhere\%%~nxi"
pause

Then I find that when I drop a file with the name that contains ')' and does not contain a space (eg. fig(1).jpg) it will report an error that says "There should be no .jpg)".

I know that it is fine if I write ./xxx.bat "fig(1).jpg" in terminal straightly, but I do need to drop some files on it. any help?

Answer 1

Use PowerShell instead. Just save this as move-to-somewhere.ps1

ls $args | mv -d "somewhere"

Then create a new shortcut and paste this

powershell.exe -noprofile -noexit -f path\to\move-to-somewhere.ps1

Now just drag the files you want to the newly created shortcut and they'll be moved as expected

The full command is Get-ChildItem -LiteralPath $args | Move-Item -Destination "somewhere" and you can also append -WhatIf/-wi to do a dry run before actually moving the files

In fact it should be easier but for some reason the methods to make drag-and-drop files on *.ps1 file directly don't work and the simplest temporary workaround is to use a shortcut. You can also create a batch file and forward all arguments to PowerShell but the file contents will be slightly different. For more information as well as other methods read

• Drag and Drop to a Powershell script
• Running powershell scripts by drag-n-drop file or folder on it
• Process files dragged to BAT file in PowerShell

Answer 2

The error is being emitted by FOR, which has even stricter escaping requirements for special characters than the rules for filenames.

You can't quote the %* because the interpreter doesn't consider %* as an list it can iterate over, but as a single string comprised of the numbered arguments, as they were passed to the script. The interpreter is substituting the string literally, and so FOR is seeing

for %%i in (fig(1).jpg) do move "%%~i" "somewhere\%%~nxi"
           ^     ^

The closing-bracket in the filename is actually closing what FOR thinks is the list, leaving an extra .jpg) at the end, which the interpreter doesn't understand.

Thankfully, the string-manipulation features that you are using in your FOR variables are also available to numbered args (%1, %2, etc.) so instead of looping over a badly-parsed list, you can instead use a real list (of arguments) and iterate using a :label and a GOTO:

@ECHO OFF

:loop_args
IF NOT "{%~1}"=="{}" (
  MOVE "%~1" "somewhere\%~nx1"
  SHIFT
  GOTO :loop_args
)

Essentially, all the time the first argument %~1 is not empty, we perform three actions:

1. move the file to its string-manipulated path
2. pop the first argument off the list (moving the second to the first, third to second, etc)
3. return to the label

When there are no more arguments, the first argument becomes empty, so the interpreter moves on to the next statement following the closing-bracket ), or quits if there are no more statements.

Note that you can write this short-hand, if it reads better for you, by separating the three commands with a &:

:loop_args
IF NOT "{%~1}"=="{}" ( MOVE "%~1" "somewhere\%~nx1" & SHIFT & GOTO :loop_args )

If your action is more complicated than a simple move, put your action in its own "function" using another label:

@ECHO OFF
SETLOCAL 

:::: MAIN LOGIC ::::

ECHO Starting with arguments: %*
ECHO.

:loop_args
IF NOT "{%~1}"=="{}" ( CALL :do_something "%~1" & SHIFT & GOTO :loop_args )

ECHO.
ECHO Finished processing %NUM_CALLS% argument(s)

ENDLOCAL
GOTO :EOF

:::: FUNCTIONS ::::

:do_something
  IF DEFINED NUM_CALLS (
    ECHO.
    ECHO.******* 
    ECHO.
    SET /a NUM_CALLS += 1
  ) ELSE (
    SET NUM_CALLS=1
  )

  ECHO You passed the value ^[ %1 ^]
  ECHO Unquoted, this reads ^[ %~1 ^]
  ECHO This resolves to the file: %~f1
  ECHO Here is just the filename: %~n1
  ECHO Here is it's extension: %~x1
  ECHO How about a size, too^?: %~z1

  IF EXIST "%~f1" (
    ECHO Here is where it lives... 
    DIR "%~dp1"
  )
GOTO :EOF

When you CALL a function, you can pass arguments to it, which the interpreter will store in its own set of numbered variables, and you can manipulate them as you can the FOR variables. When the function finishes, you get back the numbered variables that you had in the caller.

Unlike the single IF statement, you can manipulate environment variables inside a function. These changes are global to the script unless you call SETLOCAL at the beginning of the function and ENDLOCAL at the end.

Note the use of GOTO :EOF which is a special label which marks the end of the function or the end of the script, if you are not currently running a function. In particular, note the one between the main logic and the function -- this stops the interpreter running the code inside the function after it finishes the main routine.

For both approaches, also note that you can only iterate over an argument list once, as once each argument has been SHIFTed out of the list, it is gone. You would need to employ a different array technique if you wanted to retain them, but that is a problem for another question.