Copy altered CSV Data to new CSV - powershell

Question

The whole point of this issue is going to be: How to copy data from one CSV to another without knowing/listing the headers of the original CSV.
The cmdlet I'm building is meant to convert a report from CSV to a spreadsheet eventually. And if I write the column headers to the code, each time somebody changes the report, the code will break and it would have to be updated.
The steps I would take right now:
# Import the Source CSV. Gonna pull data from this later.
$SourceCSV = Import-Csv -Path $reportSourceCSV -Delimiter ";"
# Remove NULL characters, white spaces and change comma separator to semicolon.
(Get-Content -Path $reportSourceCSV | Where-Object {-not [string]::IsNullOrWhiteSpace($PSItem)}).Replace('","',";") | Out-File -FilePath $TMP1
# Import the modified new temp CSV.
$Input = Import-Csv -Path $TMP1 -Delimiter ";"
# Take existing CSV file headers and append some new ones. Rename a long column name.
((($GetHeaders = foreach ($Header in $SourceCSV[0].PSObject.Properties.Name) {
"`"$Header`""
}) + '"column4"','"column5"','"column6"') -join ";").Replace("VerylongOldColumnName","ShortName") | Out-File -FilePath $TMP2
foreach ($Item in $Input) {
"`"$($Item.column1)`";`"$($Item.'column2')`";`"$($Item.column3)`"" | Out-File -FilePath $TMP2 -Append
}
$exportToXLSX = Import-Csv -Path $TMP2 -Delimiter ";" | Export-Excel -Path $Target -WorkSheetname "reportname" -TableName "tablename" -TableStyle Medium2 -FreezeTopRow -AutoSize -PassThru
$exportToXLSX.Save()
$exportToXLSX.Dispose()
Remove-Item -Path $TMP1, $TMP2
This works! But I don't want to create infinite amount of different reports and just as many different logic blocks to process all these reports.
So this is as far as I was able to get trying a more dynamic way of processing the report CSVs:
(Get-Content -Path $reportSourceCSV | Where-Object {-not [string]::IsNullOrWhiteSpace($PSItem)}).Replace('","',";") | Out-File -FilePath $TMP1
$import = Import-Csv -Path $TMP1 -Delimiter ";"
$headers = ($import[0].PSObject.Properties.Name).Replace("VerylongOldColumnName","ShortName")
$headers | Out-File -FilePath "C:\TEMP\test.csv"
foreach ($item in $import) {
for ($h = 0; $h -le ($headers).Count; $h++) {
$($item.$($headers[$h]))
}
}
Now, this works... kind of. If I run the script like this, it shows me the output I want, but I was NOT able to export this to CSV.
I added Export-Csv to this line: $($item.$($headers[$h])) so this particular line would look like this:
$($item.$($headers[$h])) | Export-Csv -Path $Output -Delimiter ";" -Append -NoTypeInformation
And this is the error I get:
Export-Csv : Cannot append CSV content to the following file: C:\TEMP\test.csv.
The appended object does not have a property that corresponds to the following
column: column1. To continue with mismatched properties, add the -Force parameter,
and then retry the command.
At line:11 char:36
+ ... ers[$h])) | Export-Csv -Path $Output -Delimiter ";" -Append -NoTypeIn ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : InvalidData: (column1:String) [Export-Csv], InvalidOperationException
+ FullyQualifiedErrorId : CannotAppendCsvWithMismatchedPropertyNames,Microsoft.PowerShell.Commands.ExportCsvCommand
If I add -Force parameter, the output will be the headers and a bunch of empty lines.
As little as I understand, is that the output is for some reason a string? To my knowledge everything should be an object in PS, unless converted to string (Write-Host cmdlet being an exception). And I don't really know how to force the output back to being objects.
Edit: Added sample source CSV
"Plugin","Plugin Name","Family","Severity","IP Address","Protocol","Port","Exploit?","Repository","DNS Name","NetBIOS Name","Plugin Text","Synopsis","Description","Solution","See Also","Vulnerability Priority Rating","CVSS V3 Base Score","CVSS V3 Temporal Score","CVSS V3 Vector","CPE","CVE","Cross References","First Discovered","Last Observed","Vuln Publication Date","Patch Publication Date","Exploit Ease","Exploit Frameworks"
"65057","Insecure Windows Service Permissions","Windows","High","127.0.0.1","TCP","445","No","Individual Scan","computer.domain.tld","NetBIOS Name","Plugin Output:
Path : c:\program files (x86)\application\folder\service.exe
Used by services : application
File write allowed for groups : Users, Authenticated Users
Full control of directory allowed for groups : Users, Authenticated Users","At least one improperly configured Windows service may have a privilege escalation vulnerability.","At least one Windows service executable with insecure permissions was detected on the remote host. Services configured to use an executable with weak permissions are vulnerable to privilege escalation attacks.
An unprivileged user could modify or overwrite the executable with arbitrary code, which would be executed the next time the service is started. Depending on the user that the service runs as, this could result in privilege escalation.
This plugin checks if any of the following groups have permissions to modify executable files that are started by Windows services :
- Everyone
- Users
- Domain Users
- Authenticated Users","Ensure the groups listed above do not have permissions to modify or write service executables. Additionally, ensure these groups do not have Full Control permission to any directories that contain service executables.","http://www.nessus.org/u?e4e766b2","","8.4","","AV:L/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","cpe:/o:microsoft:windows","","","Jul 11, 2029 06:48:20 CEST","Jul 11, 2029 06:48:20 CEST","N/A","N/A","",""
Edit: I think I found another way how to accomplish this and looking at it, it looks I tried to overdo it quite a bit.
# Doing cleanup, changing delimiters, renaming that one known column. All in one line.
$importCSV = 'C:\TEMP\sourceReport.csv'
(Get-Content -Path $importCSV | Where-Object {-not [string]::IsNullOrWhiteSpace($PSItem)}).Replace('","','";"').Replace"VerylongOldColumnName","ShortName") | Out-File -FilePath C:\TEMP\tmp1.csv
# Adding additional columns and exporting it all to result CSV.
Import-Csv -Path C:\TEMP\tmp1.csv -Delimiter ";" | Select-Object *, "Column1", "Column2" | Export-Csv -Path C:\TEMP\result.csv -NoTypeInformation -Delimiter ";"

Answer 1

You should not simply replace , with ; because the fields actually contain commas as in ..Additionally, ensure these groups .. By replacing just like that, the field will get separated from the rest of its content and you'll end up with a mis-aligned csv.
The below approach will do what you want, leaving the structure of the csv file intact:
$importCSV = 'C:\TEMP\sourceReport.csv'
$exportCSV = 'C:\TEMP\result.csv'
$columnsToAdd = "Column1", "Column2"
# read the file as string array, not including empty lines
$content = Get-Content -Path $importCSV | Where-Object { $_ -match '\S' }
# replace the column header in the top line only
$content[0] = $content[0].Replace("VerylongOldColumnName", "ShortName")
# join the string array with newlines and convert that to an object with ConvertFrom-Csv
# add the columns to the object and export it using the semi-colon as delimiter
($content -join [Environment]::NewLine) | ConvertFrom-Csv |
Select-Object *, $columnsToAdd |
Export-Csv -Path $exportCSV -NoTypeInformation -Delimiter ";"