Search code examples
loopspowershellstreamreaderstreamwriterimport-csv

Find strings in one file in another and output certain columns


I have a file that contains CampaignNames and IDs. The two fields are separated by a pipe |. The IDs are separated by a space. I want to find all rows in a file (thorpe þ delimited) that contain the IDs, and output those rows into separate files per name. This file is usually 4-7 GB, sometimes larger.

campaigns.txt:

Name|NameID
FirstName|123 212 445 39
SecondName|313 939
ThirdName|219

Data ID File:

DateþIDþCode
10-22-14þ123þAbc
10-24-16þ212þPow
09-18-15þ219

So I would want 3 files created. FirstName.txt contains 2 rows. SecondName.txt contains 0 rows. ThirdName.txt contains 1 row.

I cobbled together some code from various sources and came up with this. However, I'm wondering if there's a better way than having to read through the data file multiple times. Any thoughts out there?

$campaigns = Import-Csv "campaigns.txt" -Delimiter "|"
$datafile = "5282_10-19-2016"
$encoding = [Text.Encoding]::GetEncoding('iso-8859-1')

echo "Starting.."
Get-Date -Format g

foreach ($campaign in $campaigns) {
    $campaignname = $campaign.CampaignName
    $campaignids = $campaign.CampaignID.split(" ")
    echo "Looking for $campaignname - $campaignids"
    $writer = New-Object System.IO.StreamWriter($campaignname + "_filtered.txt")
    foreach ($campaignid in $campaignids) {
        $datareader = New-Object System.IO.StreamReader($datafile, $encoding)
        while ($dataline = $datareader.ReadLine()) {
            if ($dataline -match $campaignid) {
                $data = $dataline.Split("þ")
                $writer.WriteLine('{0}|{1}|{2}|{3}|{4}|{5}|{6}|{7}', $data[0], $data[3], $data[5], $data[8], $data[12], $data[14], $data[19], $data[20])
            }
        }
    }
    $writer.Close()
}

echo "Done!"
Get-Date -Format g

Solution

  • Process the huge datafile just once.
    Pick the campaign names from a hashtable built from campaign.txt.
    Assuming there are not many campaigns (say, less than 1000) write to as many StreamWriters.

    $campaignByID = @{}
    foreach ($c in (Import-Csv 'campaigns.txt' -Delimiter '|')) {
        foreach ($id in ($c.CampaignID -split ' ')) {
            $campaignByID[$id] = $c.CampaignName
        }
    }
    
    $campaignWriters = @{}
    $datareader = New-Object IO.StreamReader($datafile, $encoding)
    while (!$datareader.EndOfStream) {
        $data = $datareader.ReadLine().Split('þ')
        $campaignName = $campaignByID[$data[1]]
        if ($campaignName) {
            $writer = $campaignWriters[$campaignName]
            if (!$writer) {
                $writer = $campaignWriters[$campaignName] =
                    New-Object IO.StreamWriter($campaignName + '_filtered.txt')
            }
            $writer.WriteLine(($data[0,3,5,8,12,14,19,20] -join '|'))
        }
    }
    
    $datareader.Close()
    foreach ($writer in $campaignWriters.Values) {
        $writer.Close()
    }
    

    To display progress use Write-Progress based on $datareader.BaseStream.Position / $datareader.BaseStream.Length * 100 but don't do it for every datafile line because it'll slow down the processing, do it every 1 second, for example, using a datetime variable: update it when a second has elapsed and display the progress.