I have been using code from this answer to check for additions/changes to class rosters from MS Teams:

$set = [System.Collections.Generic.HashSet[string]]::new(
    [string[]] (Import-CSV -Path stundent.csv).UserPrincipalName,
    [System.StringComparer]::InvariantCultureIgnoreCase
)
Import-Csv ad.csv | Where-Object { $set.Add($_.UserPrincipalName) } |
    Export-Csv path\to\output.csv -NoTypeInformation

Ideally, I want to be able to check if there have been removals when compared to a new file, swap the import file positions, and check for additions. If my files look like Source1 and Source2 (below), the check for removals would return Export1, and the check for additions would return Export2. Since there will be multiple instances of students across multiple classes, I want to include TeamDesc in the filter query to make sure only the specific instance of that student with that class is returned.

Source1.csv

Source2.csv

Export1.csv

Export2.csv

CodePudding user response：

Try the following, which uses Compare-Object to compare the CSV files by two fields, simply by passing the properties of interest to -Property; the resulting output is split into two collections based on which input side a differing property combination is unique to, using the intrinsic .Where() method:

$removed, $added = (
  Compare-Object (Import-Csv Source1.csv) (Import-Csv Source2.csv) -PassThru `
                 -Property TeamDesc, UserPrincipalName
).Where({ $_.SideIndicator -eq '=>'}, 'Split')

$removed | 
  Select-Object -ExcludeProperty SideIndicator |
  Export-Csv -NoTypeInformation Export1.csv

$added |
  Select-Object -ExcludeProperty SideIndicator |
  Export-Csv -NoTypeInformation Export2.csv

CodePudding user response：

Assuming both Csvs are stored in memory, Source1.csv is $csv1 and Source2.csv is $csv2, you already have the logic for Export2.csv using the HashSet<T>:

$set = [System.Collections.Generic.HashSet[string]]::new(
    [string[]] $csv1.UserPrincipalName,
    [System.StringComparer]::InvariantCultureIgnoreCase
)
$csv2 | Where-Object { $set.Add($_.UserPrincipalName) }

Outputs:

TeamDesc UserPrincipalName   Name
-------- -----------------   ----
Team 2   [email protected] tim tams

For the first requirement, Export1.csv, the reference object would be $csv2 and instead of a HashSet<T> you could use a hash table, Group-Object -AsHashTable makes it really easy in this case:

$map = $csv2 | Group-Object UserPrincipalName -AsHashTable -AsString

# if Csv2 has unique values for `UserPrincipalName`
$csv1 | Where-Object { $map[$_.UserPrincipalName].TeamDesc -ne $_.TeamDesc }

# if Csv2 has duplicated values for `UserPrincipalName`
$csv1 | Where-Object { $_.TeamDesc -notin $map[$_.UserPrincipalName].TeamDesc }

Outputs:

TeamDesc UserPrincipalName   Name
-------- -----------------   ----
Team 1   [email protected] john smith