I am trying to set a regex that works when i use say regexr.com but doesn't apply in my transforms/props file.
I am wanting to not ingest any apache logs that contain: assets/js, assets/css, assets/img
I can set one up singular, and it works fine, but the two commented out lines, even though they work in a regex case, don't seem to apply in my transforms file. Any insight if I may be doing something wrong?
Thank you for any assistance.
[drop_assets]
REGEX = .*assets\/js.*
#REGEX = .*(assets\/js|assets\/css|assets\/img).*
#REGEX = .*assets/js.*|.*assets/css.*|.*assets/img.*
DEST_KEY = queue
FORMAT = nullQueue
[apache]
TRANSFORMS-drop = drop_assets
in the end i think i made a simple mistake of naming my stanza in my props file "apache" when my sourcetype was "ApacheAccess". That helped me, but also, the regex above gave me the results. Splunk seems to be able to read the escaping / "\/" also.
Thank you all.
in the end i think i made a simple mistake of naming my stanza in my props file "apache" when my sourcetype was "ApacheAccess". That helped me, but also, the regex above gave me the results. Splunk seems to be able to read the escaping / "\/" also.
Thank you all.
Here's another regex to try. Note that the leading and trailing ".*" is implied unless the "^" or "$" anchor is used.
REGEX = assets\/(?:js|css|img)
Also, slash ("/") doesn't need escaping.
regex101.com says otherwise.
regex101 highlits a lone slash as an eror because by default it considers the given regex as delimited by slashes (it's described as PHP-compatible). If you switch to Python mode, then slash will not be highlited as error, but double quotes will, since Python regexes are delimited by double quotes.
Splunk doesn't use either of those as delimiters so they don't need to be escaped.