calculate from values in a text file's first column, ":" delimited, bat, VBs, PS

[kdb2012]

Yes sidewinder's script throws errors, but I am concentrating more on VB as I have some prior experience. But as I mentioned in my first post that the files I receive are dynamic and the contents changes every time. But the overall pattern remains the same.
1: server name (may or may not have \instance)
2: -----------------------------
3:domain\AD_Group
...
n:domain\AD_Group
(n+1):--------------------------
(n+2):AD_User that belongs to AD_Group(s) above
...
m:AD_User that belongs to AD_Group(s) above


That's it.
Thanks!

[Salmon Trout]

The example text file you provided does not throw any errors with the script I posted, so something must be different. I wonder if there are extra blank lines at the end of the files that cause the error? Could you post an example of a file that gave the error messages?

[Sidewinder]

Yes sidewinder's script throws errors, but I am concentrating more on VB as I have some prior experience.

I too would be interested to know what those errors are. I can't get it to fail with your original test file, or my souped up test file or even a test file with embedded blank records.

Posting an example of a test file that produced any errors would be most helpful.

 

[Salmon Trout]

See the CR/LF?

Code: [Select]

C:\Batch\Test>Readfile-CH.vbs Input2.txt
Kalvin Rodger = KALDB\AREADER
Kalvin Rodger = KALDB\PREADER
Kalvin Rodger = KALDB\READER
Kalvin Rodger = KALDB\SWRITER
C:\Batch\Test\Readfile-CH.vbs(51, 4) Microsoft VBScript runtime error: Subscript out of range: '[number: 0]'



[Sidewinder]

See the CR/LF?

Indeed I do. However the regular expression filtered out any records that did not match the pattern. Powershell was used to have access to a sort technique in case the AD_Group was not sequenced. I still can't get the Powershell version to fail even with the hanging CR/LF.

I know the OP prefers the VBS, just looking for why the PS version throws errors.

 

[kdb2012]

This is the files I worked on today:

ServerName: KADB
------------------------------------
2:KALDB\AD Development
15:KALDB\DB-BDATAREADER)
20:KALDB\ETL_USERS
24:KALDB\API2
30:KALDB\prod_development
------------------------------------

Note: In this the user is not found (this line is not part of the file)

Yesterday it was this file
ServerName: KADB
------------------------------------
2:KALDB\AD QA
15:KALDB\BIREADER)
20:KALDB\ETL_USERS
24:KALDB\BIWriter
30:KALDB\dbowners
------------------------------------
3:Sue Ling
25:Sue Ling
31:Sue Heimer

[kdb2012]

With Sidewinder's script nothing is returned for this file:

ServerName: KADB
------------------------------------
2:KALDB\AD Development
15:KALDB\DB-BDATAREADER)
20:KALDB\ETL_USERS
24:KALDB\API2
30:KALDB\prod_development
------------------------------------
3:Sue Ling
25:Sue Ling
31:Sue Hemmer

[oldun]

Try this:

Code: [Select]

Option Explicit
Dim file, domain(), users(), i, j, temp, tmparr
If WScript.Arguments.Count=0 Then
WScript.Echo "Required filename not specified....Quitting"
wscript.quit()
End If
file=WScript.Arguments(0)
UpdateArrays file, domain, users
on error resume next
i=ubound(users)
if err.number then
wscript.Echo "No users...Quitting"
wscript.quit
end if
i=ubound(domain)
if err.number then
wscript.Echo "No domain...Quitting"
wscript.quit
end if
on error goto 0
for i=ubound(domain)-1 to 0 Step -1
for j=0 To i
if domain(j)>domain(j+1) then
temp=domain(j+1)
domain(j+1)=domain(j)
domain(j)=temp
end if
next
next

for i=0 To ubound(users)
for j=1 To ubound(domain)
if (domain(j)>users(i)) Then
tmparr=Split(Users(i),":")
temp=tmparr(1) & " = "
tmparr=split(domain(j-1),":")
temp=temp & tmparr(1)
wscript.echo temp
exit for
elseif users(i)>domain(ubound(domain)) then
tmparr=Split(Users(i),":")
temp=tmparr(1) & " = "
tmparr=split(domain(ubound(domain)),":")
temp=temp & tmparr(1)
wscript.echo temp
exit for
end if
next
next

Sub UpdateArrays(file,ByRef ar1,ByRef ar2)
Const ForReading=1
Dim fso,infile,i1,i2,line
Dim regex1,regex2,match,matches1,matches2
Set regex1=New RegExp
Set regex2=New RegExp
regex1.Pattern="^\d+:\w+\\\w+$"
regex2.Pattern="^\d+:\w+\s\w+$"
regex1.Global=True
regex2.Global=True
Set fso=CreateObject("Scripting.FileSystemObject")
if (Not fso.FileExists(file)) Then
WScript.Echo file & " does not exist...Quitting"
WScript.Quit()
End If
Set inFile=fso.OpenTextFile(file,ForReading)
i1=0
i2=0
Do While Not infile.AtEndOfStream
line=infile.ReadLine
Set matches1=regex1.Execute(line)
Set matches2=regex2.Execute(line)
for each match in matches1
redim preserve ar1(i1)
line=Split(line,":")
ar1(i1)=Right(String(4,"0") & line(0),4) & ":" & line(1)
i1=i1+1
next
for each match in matches2
redim preserve ar2(i2)
line=Split(line,":")
ar2(i2)=Right(String(4,"0") & line(0),4) & ":" & line(1)
i2=i2+1
next
loop
infile.close()
End Sub


[Sidewinder]

I dislike unfinished business. Regular Expressions (RegEx) are difficult to read, difficult to write and difficult to maintain. When the data pattern changes, as it did in the second sample, the RegEx must change also. So with a little help from a RegEx tool, a tweak here and there and much headbanging, Version 3.0 has arrived:

Code: [Select]

#Requires -Version 2.0

$file = Join-Path -Path $pwd -ChildPath myFile.txt
$arrGroups = @()
$arrUsers = @()

# Load the Groups into Array
#
(Get-Content -Path $file) -match "^[0-9]{1,3}:[A-Za-z0-9]{1,}\\(\s|\S){1,}$" |
  ForEach-Object {
    $arrGroups += , @( [int] $_.split(":")[0], $_.split(":")[1] )
  }
 
$arrGroups = $arrGroups | Sort-Object @{Expression={$_[0]}; Ascending=$true}

# Load the Users into Array
#
(Get-Content -Path $file) -match "^[0-9]{1,3}:[A-Z]{1,}\s[A-Z]{1,}$" |
  ForEach-Object {
    $arrUsers += , @( [int] $_.split(":")[0], $_.split(":")[1] )
  }
   
# Do the Lookups
#
for($j=0; $j -lt $arrUsers.Count; $j++) {
  for($i=0; $i -lt $arrGroups.Count - 1; $i++) {

    if( ($arrUsers[$j][0] -eq $arrGroups[$i+1][0]) ) `
      { Write-Host $arrUsers[$j][1], "=", $arrGroups[$i+1][1]; break }

    if( ($arrUsers[$j][0] -gt $arrGroups[$i][0]) -and ($arrUsers[$j][0] -lt $arrGroups[$i+1][0]) ) `
        { Write-Host $arrUsers[$j][1], "=", $arrGroups[$i][1]; break }
       
    if( ($arrUsers[$j][0] -ge $arrGroups[$arrGroups.Count - 1][0]) ) `
      { Write-Host $arrUsers[$j][1], "=", $arrGroups[$arrGroups.Count-1][1]; break }
       
  }
}

The $file variable holds the name of the input file. The default is to have both the script and the file in the current directory, so you can run the script as: .\scriptname.ps1
This can be changed by the user.

If the input data pattern changes again, you have an opportunity to make the changes. Good luck 

[kdb2012]

@Oldun,
remarkably fast and accurate.

Thank you a lot.

@SideWinder
this is perfect version, cheers!

All of you deserve Kudos