I have data file
$data=gc c:\blabla.cfg
inside it i have all kind of lines, i want to extract the data, example, in the file I have :
SMTP_SERVER = smtp.bla.local
so if i want to create variable inside the script for example : $smtp which have this information, this is how i do it:
$tmp = $data | Select-string SMTP_SERVER
$smtp=($tmp -split ' ')[2]
It may look kind of dumb its just that i have lots of configuration to take from the file and id like to take it in one liners with pipe,
if i do this:
$smtp = $data | Select-String SMTP_SERVER | $_.Split(' ')[2]
I'm getting an error saying Expression are only allowed as the first element of pipeline
Use a ForEach-Object
Select-String -Path "c:\blabla.cfg" -SimpleMatch "SMTP_SERVER" | ForEach-Object { $_.ToString().Split(' ')[2] }
Related
I'm using PowerShell to try and get a specific field from a CSV, store it as a variable, and output it as another csv. This is mainly because I want to use it as part of a larger script, but I'm having problems...
Import-Csv C:\EmailsListNoBlanks.csv | ForEach-Object{
$Email = $_.Member -split ';'
}
$Email | Out-File C:\EmailListCOMP.csv
However in my CSV I'm only ever getting the last 4 values, whereas I'm expecting a few hundred...
Is there something I'm missing here?
Thanks
Matt
I tried something similar to what you are attempting to do and this is what I came up with to have each item on it's own line in the output:
Import-Csv C:\EmailsListNoBlanks.csv | ForEach-Object {
$Email += ($_.Member -split ';') + ("`n")
}
$Email | Out-File C:\EmailListCOMP.csv
The output is not in a csv format, but just a regular text file. The `n adds a newline to the text so that the output is one entry per line.
Try this:
Add-Type -AssemblyName System.Collections
$Email = [System.Collections.Generic.List[string]]::new()
Import-Csv C:\EmailsListNoBlanks.csv | ForEach-Object{
[void]$Email.Add( $_.Member -split ';' )
}
$Email | Out-File C:\EmailListCOMP.csv
First line of csv looks like this spaces are at after Path as well
author ,Revision ,Date ,SVNFolder ,Rev,Status,Path
I am trying to remove spaces only and rest of the content will be the same .
author,Revision,Date,SVNFolder,Rev,Status,Path
I tried below
Import-CSV .\script.csv | ForEach-Object {$_.Trimend()}
expanding on the comment with an example since it looks like you may be new:
$text = get-content .\script.csv
$text[0] = $text[0] -replace " ", ""
$csv = $text | ConvertFrom-CSV
Note: The solutions below avoid loading the entire CSV file into memory.
First, get the header row and fix it by removing all whitespace from it:
$header = (Get-Content -TotalCount 1 .\script.csv) -replace '\s+'
If you want to rewrite the CSV file to fix its header problem:
# Write the corrected header and the remaining lines to the output file.
# Note: I'm outputting to a *new* file, to be safe.
# If the file fits into memory as a whole, you can enclose
# Get-Content ... | Select-Object ... in (...) and write back to the
# input file, but note that there's a small risk of data loss, if
# writing back gets interrupted.
& { $header; Get-Content .\script.csv | Select-Object -Skip 1 } |
Set-content -Encoding utf8 .\fixed.csv
Note: I've chosen -Encoding utf8 as the example output character encoding; adjust as needed; note that the default is ASCII(!), which can result in data loss.
If you just want to import the CSV using the fixed headers:
& { $header; Get-Content .\script.csv | Select-Object -Skip 1 } | ConvertFrom-Csv
As for what you tried:
Import-Csv uses the column names in the header as property names of the custom objects it constructs from the input rows.
This property names are locked in at the time of reading the file, and cannot be changed later - unless you explicitly construct new custom objects from the old ones with the property names trimmed.
Import-Csv ... | ForEach-Object {$_.Trimend()}
Since Import-Csv outputs [pscustomobject] instances, reflected one by one in $_ in the ForEach-Object block, your code tries call .TrimEnd() directly on them, which will fail (because it is only [string] instances that have such a method).
Aside from that, as stated, your goal is to trim the property names of these objects, and that cannot be done without constructing new objects.
Read the whole file into an array:
$a = Get-Content test.txt
Replace the spaces in the first array element ([0]) with empty strings:
$a[0] = $a[0] -replace " ", ""
Write over the original file: (Don't forget backups!)
$a | Set-Content test.txt
$inFilePath = "C:\temp\headerwithspaces.csv"
$content = Get-Content $inFilePath
$csvColumnNames = ($content | Select-Object -First 1) -Replace '\s',''
$csvColumnNames = $csvColumnNames -Replace '\s',''
$remainingFile = ($content | Select-Object -Skip 1)
I have a big file which content somme blocks of text data like this.
[SERVER1]
LIBELLE=DATA, SOMME DATA
VARIABLES=A,B,C,D,E
PHYSICAL NAME=E:\SOMME\PATH\FILE.INI
ARTICLE SIZE=50
MACHINE=SOME SERVER
PARAMETER =
OPTION =
[SERVER2]
LIBELLE=DATA2, SOMME DATA2
VARIABLES=A,B,C,D,E
PHYSICAL NAME=Z:\SOMME\PATH\FILE2.INI
ARTICLE SIZE=150
MACHINE=SOME SERVER XY
PARAMETER =
OPTION 1 = VOID
OPTION 2 =
OPTION 3 =
OPTION 4 =
OPTION 5 =
What i would like to do is retrieving every block [SERVERX] and put it into a CSV file in this format (horizontally)
ColumnA__ | ColumnB (LIBELLE)___ | ColumnC (VARIABLES) | ColumnD ect...
[SERVER1] | DATA1, SOMME DATA1 | A,B,C,D,E___________ | Ect...
[SERVER2] | DATA2, SOMME DATA2 | A,B,C,D,E___________ | Ect...
I've tried this, the output work as i want but it need to be automated and exported to scv, which doesn't work for me.
$mydata = Get-Content my_file.txt
write-Host $mydata[0] $mydata[1] $mydata[2] $mydata[3] $mydata[4] $mydata[5] | Export-Csv -Path $rep\results.csv -Force -UseCulture -NoTypeInformation
Tried also somthing with select-string but i dont know if this is the right way to do my job..
select-String -path $my_file -Pattern '\[*\]', 'IDENTIFIANTS=','LIBELLE=','VARIABLES=' | Select-Object -Property LineNumber, Line | Export-Csv -Path $rep\results.csv -Force -UseCulture -NoTypeInformation
Thanks for your advices.
So, I would read the whole file in as a multi-line string using the -Raw switch for Get-Content. Then split the file up based on the [ character to denote records. The get the properties from the ConvertFrom-StringData cmdlet (have to prepend "SERVER=" to each record), and make an object from it. Then we find out what all properties any given record can have, make sure to add them all to the first record if it doesn't have it (this is done because when you export to CSV it bases the columns off of the first entries' property list). Then you can export a CSV.
$Data = (Get-Content my_file.txt -Raw) -split "(\[[^[]+)" | ?{![string]::IsNullOrWhiteSpace($_)}
$Records = $Data -replace '\\','\\'|%{$Record="SERVER="+$_.trim()|ConvertFrom-StringData;New-Object PSObject -Prop $Record}
$Props = $Records|%{$_.psobject.properties.name}|select -Unique
$Props | Where{$_ -notin $Records[0].PSObject.Properties.Name}|%{Add-Member -InputObject $Records[0] -NotepropertyName $_ -NotepropertyValue $Null}
$Records|Export-CSV .\my_file.csv -notype
Edit: For those of you out there running PowerShell 2.0 (3 versions out of date at this point in time), you can't use the -Raw parameter. Here's the alternative:
$Data = (Get-Content my_file.txt) -Join "`r`n" -split "(\[[^[]+)" | ?{![string]::IsNullOrWhiteSpace($_)}
Alternative: Thanks #Matt for the suggestion, it is always good to have a different point of view on these things. As Matt suggested, you can use Out-String to combine the array of strings that Get-Content generates, and end up with a single multi-line string. Here's the usage!
$Data = (GC my_file.txt | Out-String) -split "(\[[^[]+)" | ?{![string]::IsNullOrWhiteSpace($_)}
I'm using powershell to locate an email address in a htm file and write it out to a text file.
I'm using select-string which finds the string OK, but writes the line number as well as the email address to the file.
All I want is the email address! It seems simple enough, but I can't crack it.
Here's my code:
$List_htm = Get-ChildItem -Filter *.htm
# Loop:
foreach ($htm in $List_htm)
{
# Locate recipient email address to send to:
# Regex pattern to match:
$pattern = '(^\W*.*#.*\.{1,}\w*$)'
$sel = select-string -list $htm -pattern $pattern | Select-Object Line
If ($Sel -eq $null)
{
write-host "FAILS - $htm does NOT contain $pattern"
}
Else
{
write-host "WORKS! $pattern `n$sel"
}
Write-host "end"
$EmailAddressee = $PDFFolder + "EmailAddressee.txt"
$sel | Out-File $EmailAddressee
}
However emailaddressee.txt looks like this:
Line
----
fred.bloggs#helpmeplease.com
All I want is a single line with the email address in:
fred.bloggs#helpmeplease.com
I could obviously further process this results file in powershell to get this, but I'm hoping someone can come up with a simple one stage result.
Thanks
Ian
Change the following line:
$sel = select-string -list $htm -pattern $pattern | Select-Object Line
To:
$sel = select-string -list $htm -pattern $pattern | Select-Object -ExpandProperty Line
That will ensure you write the property of the object rather than the textual representation of the object itself
I'm using PowerShell To import a TAB separated file with headers. The generated file has a few empty strings "" on the end of first line of headers. PowerShell fails with an error:
"Cannot process argument because the
value of argument "name" is invalid.
Change the value of the "name"
argument and run the operation again"
because the header's require a name.
I'm wondering if anyone has any ideas on how to manipulate the file to either remove the double quotes or enumerate them with a "1" "2" "3" ... "10" etc.
Ideally I would like to not modify my original file. I was thinking something like this
$fileContents = Get-Content -Path = $tsvFileName
$firstLine = $fileContents[0].ToString().Replace('`t""',"")
$fileContents[0] = $firstLine
Import-Csv $fileContents -Delimiter "`t"
But Import-Csv is expecting $fileContents to be a path. Can I get it to use Content as a source?
You can either provide your own headers and ignore the first line of the csv, or you can use convertfrom-csv on the end like Keith says.
ps> import-csv -Header a,b,c,d,e foo.csv
Now the invalid headers in the file is just a row that you can skip.
-Oisin
If you want to work with strings instead use ConvertFrom-Csv e.g.:
PS> 'FName,LName','John,Doe','Jane,Doe' | ConvertFrom-Csv | Format-Table -Auto
FName LName
----- -----
John Doe
Jane Doe
I ended up needing to handle multiple instances of this issue. Rather than use the -Header and manually setting up each import instance I wrote a more generic method to apply to all of them. I cull out all of the `t"" instances of the first line and save the file to open as a $filename + _bak and import that one.
$fileContents = Get-Content -Path $tsvFileName
if( ([string]$fileContents[0]).ToString().Contains('""') )
{
[string]$fixedFirstLine = $fileContents[0].ToString().Replace('`t""',"")
$fileContents[0] = $fixedFirstLine
$tsvFileName = [string]::Format("{0}_bak",$tsvFileName
$fileContents | Out-File -FilePath $tsvFileName
}
Import-Csv $tsvFileName -Delimiter "`t"
My Solution if you have much columns :
$index=0
$ColumnsName=(Get-Content "C:\temp\yourCSCFile.csv" | select -First 1) -split ";" | %{
$index++
"Header_{0:d5}" -f $index
}
import-csv "C:\temp\yourCSCFile.csvv" -Header $ColumnsName