Using regex on Extracted field

[373782073]

Hi,
I have incoming syslog events for which I've used the Field Extraction wizard in SPLUNK to separate a the filename of the events which results in the field values look like this:

abc-servicen.City-backup-1986-01-08-16:00:43-level1.tar

I would like to eventually extract and tabulate the filename (I've called it bkp_filename when I extracted this field earlier) as well as the date YYYY-MM-DD for each filename in a separate field (call it bkp_file_date).

Can anyone suggest how to use regex or an in-built function to achieve the date extraction into a separate field or column?

Thanks

[MuS · Accepted solution]

Hi 373782073,

Give this a try:

| makeresults 
| eval example="abc-servicen.City-backup-1986-01-08-16:00:43-level1.tar" 
| rex field=example "(?<bkp_filename>[a-zA-Z-\.]+)-(?<bkp_file_date>\d{4}-\d{2}-\d{2})"

Hope this helps ...

cheers, MuS

[373782073]

Great. I was able to add the rex statement to my existing search and strip the info I wanted into separate fields.
Thanks a lot!

[niketn]

@373782073 is the above example the backup file name bkp_filename from which you need to extract date?

[373782073]

Yes this is the field extracted which has the date embedded