Splunk Search

Multivalue field extraction

Contributor

Hi, I'm struggling to get this extracted correctly so it's usable.
The raw data is presented as:

    Privileges: SeSecurityPrivilege
            SeBackupPrivilege
            SeRestorePrivilege
            SeTakeOwnershipPrivilege
            SeDebugPrivilege
            SeSystemEnvironmentPrivilege
            SeLoadDriverPrivilege
            SeImpersonatePrivilege
            SeAssignPrimaryTokenPrivilege

By default, only the first entry is assigned to the field eg:

"SeSecurityPrivilege"

I can extract the remaining lines into the field using the regex "Privileges:\t(?P(\w+\s+|\t\t\t\w+\s+)+)", which returns this:

"SeSecurityPrivilege SeBackupPrivilege SeRestorePrivilege SeTakeOwnershipPrivilege SeDebugPrivilege SeSystemEnvironmentPrivilege SeLoadDriverPrivilege SeImpersonatePrivilege"

Now what I need is to split each of these into the Privileges field as separate values.

These values are not a static number of entries, so Privileges could contain anything between 1 and 10 lines.

Any advice would be great as the docs and answers I've read don't seem to help this situation.

TIA
Steve

1 Solution

Revered Legend

Give this a try

index=main EventCode="576" |rex field=_raw "Privileges:\t(?P<Privileges>(\w+\s+|\t\t\t\w+\s+)+)" | rex field=Privileges max_match=0 "(?<Privilege>\S+)\s+" | mvexpand Privilege

View solution in original post

Contributor

Thanks to everyone that has offered help. Much appreciated.

0 Karma

Motivator

Hi cdstealer

Try to add this, after your field extraction.

main search | rex field=_raw ... | makemv delim=" " Privileges | mvexpand Privileges 

Hope it helps.

0 Karma

Revered Legend

Give this a try

index=main EventCode="576" |rex field=_raw "Privileges:\t(?P<Privileges>(\w+\s+|\t\t\t\w+\s+)+)" | rex field=Privileges max_match=0 "(?<Privilege>\S+)\s+" | mvexpand Privilege

View solution in original post

Contributor

Perfect! I think you've solved a couple of my past questions. So thank you very much again 🙂

0 Karma

Contributor

Hi again, Now I'm struggling transferring the search into props/transforms. I'm probably missing something simple.

props.conf:

EXTRACT-Privileges = Privileges:\t(?P<Privileges>(\w+\s+|\t\t\t\w+\s+)+)
REPORT-fields = privilege

transforms.conf:

[privilege]
SOURCE_KEY = Privileges
REGEX = "(?<privilege>\S+)\s+)
REPEAT_MATCH = true
MV_ADD = true

So the new field that contains the extracted values is not happening.

0 Karma

Super Champion

Hello, did you manage to solve this ?

0 Karma

Splunk Employee
Splunk Employee

You have an extra " in front of your REGEX in transforms.conf.

0 Karma

Motivator

try :

  your_base_search|rex field=_raw "Privileges:\t(?P<Privileges>(\w+\s+|\t\t\t\w+\s+)+)"| eval Privileges=split(field_name_extraction, " ")| mvexpand Privileges|...
0 Karma

Contributor

Hi fdi01, Thanks. Unfortunately this does not change the output.

Steve

0 Karma

Motivator

your rex |rex field=_raw "Privileges:\t(?P<Privileges>(\w+\s+|\t\t\t\w+\s+)+)"| capture all results in Privileges field ???

0 Karma

Contributor

This is what I'm getting.
alt text

0 Karma
State of Splunk Careers

Access the Splunk Careers Report to see real data that shows how Splunk mastery increases your value and job satisfaction.

Find out what your skills are worth!