Command/script to copy files with certain ext, which may have the same name, from folder structure to singe directory and rename

Question

From a source directory structure (multiple folders), I need to copy all csv files to a single target directory. I found a command to do this, though files of the same name exist within different folders of the source structure, causing obvious issues when copied to one folder.

How can the files with duplicate names be renamed during the copy please (ideally: report.csv, reportcopy2.csv etc)? The job currently only copies a single instance of each file. Thanks for help.

Answer 1

A simple PowerShell-Function can also do this:

Function Start-FileCopyWDupliCheck(){
    # specify input-path, output-path, renaming-scheme, and file-extension here:
    param(
        [string]$path_in = 'D:\Temp\bla [ ] pfad\[ ]',
        [string]$path_out = 'D:\Temp\mirrbla [123] 123',
        [string]$appendix = "_copy",
        [string]$fileext = "*.*"
    )

    # get all the files (and the necessary attributes)
    [array]$file_input = @(Get-ChildItem -LiteralPath $path_in -Filter $fileext -Recurse)
    [array]$file_input_path = @($file_input | ForEach-Object {$_.FullName})
    [array]$file_input_name = @($file_input | ForEach-Object {$_.BaseName})
    [array]$file_input_ext = @($file_input | ForEach-Object {$_.Extension})

    # check file-names for duplicates:
    for($i = 0; $i -lt $file_input_path.Length; $i++){
        $inter = "$path_out\$($file_input_name[$i])$($file_input_ext[$i])" -replace '[[+]*?()\\.]','\$&'
        if((Test-Path -LiteralPath $inter) -eq $false){
            Write-Host "$inter not found in output-path -> leaving the name the same" -ForegroundColor Green
            $ready_for_copying = $inter
        }else{
            $j = 1
            Write-Host "$inter found in output-path -> appending `"$($appendix)XY`"..." -ForegroundColor Yellow
            while($true){
                $inter = "$path_out\$($file_input_name[$i])$appendix$j$($file_input_ext[$i])" -replace '[[+]*?()\\.]','\$&'
                if((Test-Path -LiteralPath $inter) -eq $true){
                    $j++
                    continue
                }else{
                    Write-Host "$appendix$j is working" -ForegroundColor Green
                    $ready_for_copying = $inter
                    break
                }
            }
        }
        # finally, copy the file:
        Copy-Item -LiteralPath $($file_input_path[$i]) -Destination $ready_for_copying
    }
}

Start-FileCopyWDupliCheck

Of course one could remove the "write-host"-commands - they simply provide some visual feedback. Also, it would be possible to first gather the file-names and then later copy all files in a separate loop, possibly speeding things up.

• simply make $ready_for_copying an array,
• add (+=) the $inter-values to it,
• add the check for already specified output-names in the if-conditions (e.g. $inter -in $ready_for_copying),
• then move the copy-item-command to a new for-block (e.g. for($i = 0; $i -lt $ready_for_copying.Length; $i++)

Answer 2

This is a slight work around - I know you wanted to name differently but I wondered if this will work for you.

I would suggest you rename on copy to something like Foldername.Filename.csv.

Use something like

echo f | xcopy /f /y srcfile destfile

EDIT

I have tried for a few hours, I don't think what you want is possible with the CMD prompt or bat files.

This is what I have

set sDir=C:\Documents and Settings\drook\Desktop
set dDir="C:\Documents and Settings\drook\Desktop\Folder\"

cd C:\Documents and Settings\drook\Desktop\
FOR /F %%a in ("*.txt") DO (     
    xcopy "%%a" "C:\Documents and Settings\drook\Desktop\Folder\"
    cd\ 
    cd C:\Documents and Settings\drook\Desktop\Folder\
    ren "%%a" "newName-%dir%.txt"
    cd\
    cd C:\Documents and Settings\drook\Desktop\
)

pause

It fails on the rename because it ignores the variable. So where I'm showing newName-%dir% (where dir is the variable) it would fail on newName-%%a as well...

Sorry, I don't think it is possible.

Having said that, this makes it look like it is possible: windows-batch-file-to-copy-and-keep-duplicates

Answer 3

This is similar to aflat's answer:

setlocal enableDelayedExpansion
set counter=1
for /r src %%F in (*.csv) do (
    copy %%F dest\%%~nF-!counter!%%~xF
    set /a counter=counter+1
)

Answer 4

This should do it:

@echo off
setlocal
set "sourcePath=c:\source"
set "targetPath=c:\temp\target"
set "pattern=*.csv"
set prev=""
set count=0

for /f  "tokens=1,2 delims=?" %%F in ('^(for /r "%sourcePath%" %%N in ^("%pattern%"^) do ^@echo %%~nxN?%%N?^)^|sort') do (
  call :copyFile "%%F" "%%G"
  set prev="%%F"
)

goto :eof

:copyFile
  if /I %prev%==%1 (set /a count+=1) else (set count=0)
  if %count%==0 (set sufix=) else (set sufix=Copy%count%)
  echo copy "%~2" "%targetPath%\%~n1%sufix%%~x1%" 
goto :eof

What it does:

• Lists all files matching pattern (most inner for) formatted as name?path
• Sorts the list (by file name)
• Parses sorted list line by line (outer for) and for each line (file)
  • if file name appears first time, copies it unmodified
  • if file name repeats, adds suffix and count_nr to a copy

As it is now it will just echo copy command instead of executing it, so you can safely run it and check it's output. When you're ready, just remove echo from
echo copy "%~2" "%targetPath%\%~n1%sufix%%~x1%"