PowerShell 提高阵列更新性能

Question

I have two arrays with user details.

Array 1

Let us say there are around 2000 different users, and they are present in the "Distinct Users" array. Distinct users is an array of user objects, with properties like:

$userDetailsObject=@{
    UserName = "UserName1"
    UserEmail = "UserEmail1"
    Category = "Category1"

} 

Array 2

The second array is a list of all users that are part of multiple groups - named "All Group Users". So there can be repetition of users, as the same user can be part of mutliple groups. The "All Group Users" array doesnt have the category detail of the users. There are around 12000 objects in this array.

$groupUserDetailsObject=@{
    UserName = "UserName1"
    UserEmail = "UserEmail1"
    GroupName = "Group1"
    Category = ""

}

I am trying to get group wise user details with category, and need to compare "All Group Users" array and "Distinct Users" array ( based on the email address which will be unique), and update the Category property for the users in "All Group Users".

What will be the fastest way to do this? I tried different combinatons like nested foreach loops and regex matches, but haven't been able to speed this up considerably, takes around 20 seconds to go through each user in the "Distinct Users" array and update all matching objects in the "All Group Users" array.

Adding the mock code that I was using to test this out:

$Users=[System.Collections.ArrayList]@()
$a = @(1..12000)
$i=0

#build 12000 sample users with 2000 unique users by adding sample values
foreach ($item in $a)
{   
    if($i -eq 2000) #reset pattern after every 200 users
    {
        $i=0
    }
    $userDetailsObject=@{
        UserName = "UserName$i"
        UserEmail = "UserEmail$i"
        Category  = ""

    } 
    $obj = New-Object -Type PSObject -Prop $userDetailsObject
    $Users.Add($obj)
    $i++
}

#loop through 2000 unique users and compare by email address against 12000 users
for($i=0 ;$i -lt 2000 ;$i++)
{
    [regex] $regex = '(?i)^('+[regex]::escape("UserEmail$i") + ')$'

    Measure-Command {foreach($item in ($Users.where{$_.UserEmail -match $regex}))
    {
        $item.Category  = "Category$i"
    }
}| Select-Object -Property TotalSeconds

}

Thanks.

Answer 1

If you need to compare fast, use a Hashtable to store the user objects in. This would mean every item in the hash must have a unique key , and for that you can use the users EmailAddress property (which in a network should be unique for each user).

Something like:

Write-Host 'Creating dummy users'
# build a hashtable with 12000 sample users with unique email addresses (as in real life they would have)
$userHash = [ordered]@{}
foreach ($item in @(1..12000)) {
    $userHash["UserEmail$item"] = [PsCustomObject]@{
        UserName  = "UserName$item"
        UserEmail = "UserEmail$item"
        Category  = ""
    } 
}

Write-Host 'Setting Category on 2000 users'
Measure-Command {
    for($i = 0; $i -lt 2000; $i++) {
        if ($userHash.Contains("UserEmail$i")) {
            $userHash["UserEmail$i"].Category = "Category$i"
        }
    }
} | Select-Object -Property TotalSeconds

If you want to see the results

$userHash.Values | Select-Object * | Format-Table

---

UPDATE

Thanks for adding extra explanation. If I understand correctly, this is what you want to do:

$distinctUsers = 2000
$allGroupUsers = 12000

Write-Host "Creating $distinctUsers DISTINCT users"
# build a hashtable with $distinctUsers sample users with unique email addresses
# this doesn't need to be an ordered list, it is just for lookup
$distinctHash = @{}
foreach ($i in @(1..$distinctUsers)) {
    $distinctHash["UserEmail$i"] = [PsCustomObject]@{
        UserName  = "UserName$i"
        UserEmail = "UserEmail$i"
        Category  = "Category$i"
    } 
}

Write-Host "Creating $allGroupUsers GROUP users"
# build a hashtable with $allGroupUsers sample users
# to have the output easier on the eyes, make this hash ordered
$groupHash = [ordered]@{}
for ($i = 1; $i -le $allGroupUsers; $i++) {
    # uniqify the key using the email addresses combined with the value of $i
    # reset the user counter on every $distinctUsers users
    $currentUser = (($i - 1) % $distinctUsers) + 1
    $uniqueEmail = '{0}-{1}' -f $i, "UserEmail$currentUser"
    $groupHash[$uniqueEmail] = [PsCustomObject]@{
        UserName  = "UserName$currentUser"
        UserEmail = "UserEmail$currentUser"
        GroupName = "Group$i"
        Category  = ""
    } 
}


Write-Host "Setting Category property on the $($groupHash.Count) GROUP users"
Measure-Command {
    foreach ($uniqueEmail in $groupHash.Keys) {
        $realEmail = ($uniqueEmail -split '-', 2)[-1]
        if ($distinctHash.ContainsKey($realEmail)) {
            $groupHash[$uniqueEmail].Category = $distinctHash[$realEmail].Category
        }
    }
} | Select-Object -Property TotalSeconds

After this, the $groupHash will be updated.
To see the result:

$groupHash.Values | Select-Object * | Format-Table -AutoSize