topic Re: Extract the User-Agent from HTTP request in Splunk Search

Extract the User-Agent from HTTP request

ashishmgupta — Tue, 27 Jul 2021 17:09:57 GMT

Below the excerpt from my HTTP request and I'm trying to get the User-Agent value from it and so far not successful. Will appreciate any help.

This Splunk editor is removing the carriage return and line feed characters so below is the regex101 link.https://regex101.com/r/rdu8yE/1

Also attached is the screenshot of the HTTP request.

Re: Extract the User-Agent from HTTP request

richgalloway — Tue, 27 Jul 2021 19:32:33 GMT

Your regex was very close. This worked for me using the one example:

User-Agent: (?<UserAgent>[^\\]*)

Re: Extract the User-Agent from HTTP request

ashishmgupta — Tue, 27 Jul 2021 19:39:25 GMT

Error in 'rex' command: Encountered the following error while compiling the regex 'User-Agent: (?<UserAgent>[^\]*)': Regex: missing terminating ] for character class.

Re: Extract the User-Agent from HTTP request

richgalloway — Tue, 27 Jul 2021 19:53:53 GMT

To use backslashes in a regex in SPL you have to escape them.

User-Agent: (?<UserAgent>[^\\\\\\]*)

Yes, that's 6 backslashes to get two.

Re: Extract the User-Agent from HTTP request

ashishmgupta — Tue, 27 Jul 2021 20:13:08 GMT

looks like 4 backslashes to get two? Below worked. Thank you for your help.

User-Agent: (?<UserAgent>[^\\\\]*)