Pergunta

I'm trying to export a complete CSV to Excel by using Powershell. I stuck at a point where static column names are used. But this doesn't work if my CSV has generic unknown header names.

Steps to reproduce

Open your PowerShell ISE and copy & paste the following standalone code. Run it with F5
"C:\Windows\system32\WindowsPowerShell\v1.0\powershell_ise.exe"

Get-Process | Export-Csv -Path $env:temp\process.csv -NoTypeInformation

$processes = Import-Csv -Path $env:temp\process.csv 
$Excel = New-Object -ComObject excel.application 
$workbook = $Excel.workbooks.add() 

$i = 1 
foreach($process in $processes) 
{ 
 $excel.cells.item($i,1) = $process.name
 $excel.cells.item($i,2) = $process.vm
 $i++ 
} 
Remove-Item $env:temp\process.csv
$Excel.visible = $true

What it does

  1. The script will export a list of all active processes as a CSV to your temp folder. This file is only for our example. It could be any CSV with any data
  2. It reads in the newly created CSV and saves it under the $processes variable
  3. It creates a new and empty Excel workbook where we can write data
  4. It iterates through all rows (?) and writes all values from the name and vm column to Excel

My questions

  • What if I don't know the column headers? (In our example name and vm). How do I address values where I don't know their header names?
  • How do I count how many columns a CSV has? (after reading it with Import-Csv)

I just want to write an entire CSV to Excel with Powershell

Foi útil?

Solução

Ups, I entirely forgot this question. In the meantime I got a solution.
This Powershell script converts a CSV to XLSX in the background

Gimmicks are

  • Preserves all CSV values as plain text like =B1+B2 or 0000001.
    You don't see #Name or anything like that. No autoformating is done.
  • Automatically chooses the right delimiter (comma or semicolon) according to your regional setting
  • Autofit columns

PowerShell Code

### Set input and output path
$inputCSV = "C:\somefolder\input.csv"
$outputXLSX = "C:\somefolder\output.xlsx"

### Create a new Excel Workbook with one empty sheet
$excel = New-Object -ComObject excel.application 
$workbook = $excel.Workbooks.Add(1)
$worksheet = $workbook.worksheets.Item(1)

### Build the QueryTables.Add command
### QueryTables does the same as when clicking "Data » From Text" in Excel
$TxtConnector = ("TEXT;" + $inputCSV)
$Connector = $worksheet.QueryTables.add($TxtConnector,$worksheet.Range("A1"))
$query = $worksheet.QueryTables.item($Connector.name)

### Set the delimiter (, or ;) according to your regional settings
$query.TextFileOtherDelimiter = $Excel.Application.International(5)

### Set the format to delimited and text for every column
### A trick to create an array of 2s is used with the preceding comma
$query.TextFileParseType  = 1
$query.TextFileColumnDataTypes = ,2 * $worksheet.Cells.Columns.Count
$query.AdjustColumnWidth = 1

### Execute & delete the import query
$query.Refresh()
$query.Delete()

### Save & close the Workbook as XLSX. Change the output extension for Excel 2003
$Workbook.SaveAs($outputXLSX,51)
$excel.Quit()

Outras dicas

I am using excelcnv.exe to convert csv into xlsx and that seemed to work properly. You will have to change the directory to where your excelcnv is. If 32 bit, it goes to Program Files (x86)

Start-Process -FilePath 'C:\Program Files\Microsoft Office\root\Office16\excelcnv.exe' -ArgumentList "-nme -oice ""$xlsFilePath"" ""$xlsToxlsxPath"""

Why would you bother? Load your CSV into Excel like this:

$csv = Join-Path $env:TEMP "process.csv"
$xls = Join-Path $env:TEMP "process.xlsx"

$xl = New-Object -COM "Excel.Application"
$xl.Visible = $true

$wb = $xl.Workbooks.OpenText($csv)

$wb.SaveAs($xls, 51)

You just need to make sure that the CSV export uses the delimiter defined in your regional settings. Override with -Delimiter if need be.


Edit: A more general solution that should preserve the values from the CSV as plain text. Code for iterating over the CSV columns taken from here.

$csv = Join-Path $env:TEMP "input.csv"
$xls = Join-Path $env:TEMP "output.xlsx"

$xl = New-Object -COM "Excel.Application"
$xl.Visible = $true

$wb = $xl.Workbooks.Add()
$ws = $wb.Sheets.Item(1)

$ws.Cells.NumberFormat = "@"

$i = 1
Import-Csv $csv | ForEach-Object {
  $j = 1
  foreach ($prop in $_.PSObject.Properties) {
    if ($i -eq 1) {
      $ws.Cells.Item($i, $j++).Value = $prop.Name
    } else {
      $ws.Cells.Item($i, $j++).Value = $prop.Value
    }
  }
  $i++
}

$wb.SaveAs($xls, 51)
$wb.Close()

$xl.Quit()
[System.Runtime.Interopservices.Marshal]::ReleaseComObject($xl)

Obviously this second approach won't perform too well, because it's processing each cell individually.

This topic really helped me, so I'd like to share my improvements. All credits go to the nixda, this is based on his answer.

For those who need to convert multiple csv's in a folder, just modify the directory. Outputfilenames will be identical to input, just with another extension.

Take care of the cleanup in the end, if you like to keep the original csv's you might not want to remove these.

Can be easily modifed to save the xlsx in another directory.

$workingdir = "C:\data\*.csv"
$csv = dir -path $workingdir
foreach($inputCSV in $csv){
$outputXLSX = $inputCSV.DirectoryName + "\" + $inputCSV.Basename + ".xlsx"
### Create a new Excel Workbook with one empty sheet
$excel = New-Object -ComObject excel.application 
$excel.DisplayAlerts = $False
$workbook = $excel.Workbooks.Add(1)
$worksheet = $workbook.worksheets.Item(1)

### Build the QueryTables.Add command
### QueryTables does the same as when clicking "Data » From Text" in Excel
$TxtConnector = ("TEXT;" + $inputCSV)
$Connector = $worksheet.QueryTables.add($TxtConnector,$worksheet.Range("A1"))
$query = $worksheet.QueryTables.item($Connector.name)

### Set the delimiter (, or ;) according to your regional settings
### $Excel.Application.International(3) = ,
### $Excel.Application.International(5) = ;
$query.TextFileOtherDelimiter = $Excel.Application.International(5)

### Set the format to delimited and text for every column
### A trick to create an array of 2s is used with the preceding comma
$query.TextFileParseType  = 1
$query.TextFileColumnDataTypes = ,2 * $worksheet.Cells.Columns.Count
$query.AdjustColumnWidth = 1

### Execute & delete the import query
$query.Refresh()
$query.Delete()

### Save & close the Workbook as XLSX. Change the output extension for Excel 2003
$Workbook.SaveAs($outputXLSX,51)
$excel.Quit()
}
## To exclude an item, use the '-exclude' parameter (wildcards if needed)
remove-item -path $workingdir -exclude *Crab4dq.csv

If you want to convert CSV to Excel without Excel being installed, you can use the great .NET library EPPlus (under LGPL license) to create and modify Excel Sheets and also convert CSV to Excel really fast!

Preparation

  1. Download the latest stable EPPlus version
  2. Extract EPPlus to your preferred location (e.g. to $HOME\Documents\WindowsPowerShell\Modules\EPPlus)
  3. Right Click EPPlus.dll, select Properties and at the bottom of the General Tab click "Unblock" to allow loading of this dll. If you don't have the rights to do this, try [Reflection.Assembly]::UnsafeLoadFrom($DLLPath) | Out-Null

Detailed Powershell Commands to import CSV to Excel

# Create temporary CSV and Excel file names
$FileNameCSV = "$HOME\Downloads\test.csv"
$FileNameExcel = "$HOME\Downloads\test.xlsx"

# Create CSV File (with first line containing type information and empty last line)
Get-Process | Export-Csv -Delimiter ';' -Encoding UTF8 -Path $FileNameCSV

# Load EPPlus
$DLLPath = "$HOME\Documents\WindowsPowerShell\Modules\EPPlus\EPPlus.dll"
[Reflection.Assembly]::LoadFile($DLLPath) | Out-Null

# Set CSV Format
$Format = New-object -TypeName OfficeOpenXml.ExcelTextFormat
$Format.Delimiter = ";"
# use Text Qualifier if your CSV entries are quoted, e.g. "Cell1","Cell2"
$Format.TextQualifier = '"'
$Format.Encoding = [System.Text.Encoding]::UTF8
$Format.SkipLinesBeginning = '1'
$Format.SkipLinesEnd = '1'

# Set Preferred Table Style
$TableStyle = [OfficeOpenXml.Table.TableStyles]::Medium1

# Create Excel File
$ExcelPackage = New-Object OfficeOpenXml.ExcelPackage 
$Worksheet = $ExcelPackage.Workbook.Worksheets.Add("FromCSV")

# Load CSV File with first row as heads using a table style
$null=$Worksheet.Cells.LoadFromText((Get-Item $FileNameCSV),$Format,$TableStyle,$true) 

# Load CSV File without table style
#$null=$Worksheet.Cells.LoadFromText($file,$format) 

# Fit Column Size to Size of Content
$Worksheet.Cells[$Worksheet.Dimension.Address].AutoFitColumns()

# Save Excel File
$ExcelPackage.SaveAs($FileNameExcel) 

Write-Host "CSV File $FileNameCSV converted to Excel file $FileNameExcel"

This is a slight variation that worked better for me.

$csv = Join-Path $env:TEMP "input.csv"
$xls = Join-Path $env:TEMP "output.xlsx"

$xl = new-object -comobject excel.application
$xl.visible = $false
$Workbook = $xl.workbooks.open($CSV)
$Worksheets = $Workbooks.worksheets

$Workbook.SaveAs($XLS,1)
$Workbook.Saved = $True

$xl.Quit()

I had some problem getting the other examples to work.

EPPlus and other libraries produces OpenDocument Xml format, which is not the same as you get when you save from Excel as xlsx.

macks example with open CSV and just re-saving didn't work, I never managed to get the ',' delimiter to be used correctly.

Ansgar Wiechers example has some slight error which I found the answer for in the commencts.

Anyway, this is a complete working example. Save this in a File CsvToExcel.ps1

param (
[Parameter(Mandatory=$true)][string]$inputfile,
[Parameter(Mandatory=$true)][string]$outputfile
)

$excel = New-Object -ComObject Excel.Application
$excel.Visible = $false

$wb = $excel.Workbooks.Add()
$ws = $wb.Sheets.Item(1)

$ws.Cells.NumberFormat = "@"

write-output "Opening $inputfile"

$i = 1
Import-Csv $inputfile | Foreach-Object { 
    $j = 1
    foreach ($prop in $_.PSObject.Properties)
    {
        if ($i -eq 1) {
            $ws.Cells.Item($i, $j) = $prop.Name
        } else {
            $ws.Cells.Item($i, $j) = $prop.Value
        }
        $j++
    }
    $i++
}

$wb.SaveAs($outputfile,51)
$wb.Close()
$excel.Quit()
write-output "Success"

Execute with:

.\CsvToExcel.ps1 -inputfile "C:\Temp\X\data.csv" -outputfile "C:\Temp\X\data.xlsx"

I found this while passing and looking for answers on how to compile a set of csvs into a single excel doc with the worksheets (tabs) named after the csv files. It is a nice function. Sadly, I cannot run them on my network :( so i do not know how well it works.

Function Release-Ref ($ref)
{
    ([System.Runtime.InteropServices.Marshal]::ReleaseComObject(
    [System.__ComObject]$ref) -gt 0)
    [System.GC]::Collect()
    [System.GC]::WaitForPendingFinalizers()
    }
    Function ConvertCSV-ToExcel
    {
    <#
    .SYNOPSIS
    Converts     one or more CSV files into an excel file.
    
    .DESCRIPTION
    Converts one or more CSV files into an excel file. Each CSV file is imported into its own worksheet with the name of the
    file being the name of the worksheet.
        
    .PARAMETER inputfile
    Name of the CSV file being converted
    
    .PARAMETER output
    Name of the converted excel file
    
    .EXAMPLE
    Get-ChildItem *.csv | ConvertCSV-ToExcel -output ‘report.xlsx’
    
    .EXAMPLE
    ConvertCSV-ToExcel -inputfile ‘file.csv’ -output ‘report.xlsx’
    
    .EXAMPLE
    ConvertCSV-ToExcel -inputfile @(“test1.csv”,”test2.csv”) -output ‘report.xlsx’
    
    .NOTES
    Author:     Boe Prox
    Date Created: 01SEPT210
    Last Modified:
    
    #>
    
    #Requires -version 2.0
    [CmdletBinding(
    SupportsShouldProcess = $True,
    ConfirmImpact = ‘low’,
    DefaultParameterSetName = ‘file’
    )]
    Param (
    [Parameter(
    ValueFromPipeline=$True,
    Position=0,
    Mandatory=$True,
    HelpMessage=”Name of CSV/s to import”)]
    [ValidateNotNullOrEmpty()]
    [array]$inputfile,
    [Parameter(
    ValueFromPipeline=$False,
    Position=1,
    Mandatory=$True,
    HelpMessage=”Name of excel file output”)]
    [ValidateNotNullOrEmpty()]
    [string]$output
    )
    
    Begin {
    #Configure regular expression to match full path of each file
    [regex]$regex = “^\w\:\\”
    
    #Find the number of CSVs being imported
    $count = ($inputfile.count -1)
    
    #Create Excel Com Object
    $excel = new-object -com excel.application
    
    #Disable alerts
    $excel.DisplayAlerts = $False
    
    #Show Excel application
    $excel.V    isible = $False
    
    #Add workbook
    $workbook = $excel.workbooks.Add()
    
    #Remove other worksheets
    $workbook.worksheets.Item(2).delete()
    #After the first worksheet is removed,the next one takes its place
    $workbook.worksheets.Item(2).delete()
    
    #Define initial worksheet number
    $i = 1
    }
    
    Process {
    ForEach ($input in $inputfile) {
    #If more than one file, create another worksheet for each file
    If ($i -gt 1) {
    $workbook.worksheets.Add() | Out-Null
    }
    #Use the first worksheet in the workbook (also the newest created worksheet is always 1)
    $worksheet = $workbook.worksheets.Item(1)
    #Add name of CSV as worksheet name
    $worksheet.name = “$((GCI $input).basename)”
    
    #Open the CSV file in Excel, must be converted into complete path if no already done
    If ($regex.ismatch($input)) {
    $tempcsv = $excel.Workbooks.Open($input)
    }
    ElseIf ($regex.ismatch(“$($input.fullname)”)) {
    $tempcsv = $excel.Workbooks.Open(“$($input.fullname)”)
    }
    Else {
    $tempcsv = $excel.Workbooks.Open(“$($pwd)\$input”)
    }
    $tempsheet = $tempcsv.Worksheets.Item(1)
    #Copy contents of the CSV file
    $tempSheet.UsedRange.Copy() | Out-Null
    #Paste contents of CSV into existing workbook
    $worksheet.Paste()
    
    #Close temp workbook
    $tempcsv.close()
    
    #Select all used cells
    $range = $worksheet.UsedRange
    
    #Autofit the columns
    $range.EntireColumn.Autofit() | out-null
    $i++
    }
    }
    
    End {
    #Save spreadsheet
    $workbook.saveas(“$pwd\$output”)
    
    Write-Host -Fore Green “File saved to $pwd\$output”
    
    #Close Excel
    $excel.quit()
    
    #Release processes for Excel
    $a = Release-Ref($range)
    }
}
Licenciado em: CC-BY-SA com atribuição
Não afiliado a StackOverflow
scroll top