Splunk Search

Regex with conditional statement

krusovice
Path Finder

Hi there,

I need some help to form regex command. My requirement is to first search for code=SEND then stats count the CountryCode, secondly search for code=RECEIVE then stats count the CountryCode.

This is my XML log:

                    <Cust>
                    <Code>SEND</Code>
                           <CountryCode>CN</CountryCode>
                           <Lty>
                              <CtyNm>BEIJING</CtyNm>
                              <Zip>100176</Zip>
                           </Lty>
                  </Cust>
                  <Cust>
                     <Code>RECEIVE</Code>
                           <CountryCode>JP</CountryCode>
                           <Lty>
                              <CtyNm>TOKYO</CtyNm>
                              <Zip>1000001</Zip>
                           </Lty>
                  </Cust>

I'm having this query formed but not meeting the above requirement, it only matches the first code=SEND and perform the count.

index=*
| rex "(?P[\w+\s+]+)"
| stats count by country

Appreciate for your help.

Tags (2)
0 Karma
1 Solution

493669
Super Champion

If you do the ingestion in such a way that each is a separate event in Splunk with valid xml syntax, the field extraction can be done by adding KV_MODE = xml in props.conf on Search Head(s). So, main focus should be getting the data ingested correctly. if your data doesn't have timestamp then here current time is considered as the _time value for the event.

Try this for props.conf on your Indexer/Heavy Forwarder.

 [SourceType_name]
 SHOULD_LINEMERGE = false
 LINE_BREAKER = ([\r\n]+)(?=\s*\<Cust\>)
 DATETIME_CONFIG = CURRENT

Search head props.conf

 [SourceType_name]
 KV_MODE =xml

If you separate events on the basis of tag then it will be easier to perform evaluation

View solution in original post

0 Karma

493669
Super Champion

If you do the ingestion in such a way that each is a separate event in Splunk with valid xml syntax, the field extraction can be done by adding KV_MODE = xml in props.conf on Search Head(s). So, main focus should be getting the data ingested correctly. if your data doesn't have timestamp then here current time is considered as the _time value for the event.

Try this for props.conf on your Indexer/Heavy Forwarder.

 [SourceType_name]
 SHOULD_LINEMERGE = false
 LINE_BREAKER = ([\r\n]+)(?=\s*\<Cust\>)
 DATETIME_CONFIG = CURRENT

Search head props.conf

 [SourceType_name]
 KV_MODE =xml

If you separate events on the basis of tag then it will be easier to perform evaluation

0 Karma

krusovice
Path Finder

I'm not familiar with props.conf, the ... is nested inside a long output of XML, and there are many events with these tags. Could you further explain by editing props.conf, how can I achieve what I needed to query?

0 Karma

493669
Super Champion

have you indexed your xml log ?
if not then you can separate events on the basis of start/end of tag
like here in your case I supposed your starting tag is <Cust> so on the the basis of this I have separated events .
refer LINE_BREAKER = part in docs
http://docs.splunk.com/Documentation/Splunk/latest/Admin/Propsconf?utm_source=answers&utm_medium=in-...

0 Karma

krusovice
Path Finder

Thanks for explaining and I got your point now. The log already indexed and I can't un-do it since I'm not admin. Furthermore, the log is not purely just XML but with other text and SOAP messages as well.

0 Karma

493669
Super Champion

try this:

   ... |rex field=_raw max_match=0 "(?m)<code>(?<code>[^<]+).*(?[^<]+)"|eval merged=mvzip(Code,CountryCode,";;")|mvexpand merged| eval merged=split(merged,";;") | eval Code=mvindex(merged,0)| eval CountryCode=mvindex(merged,1)

try this run anywhere search:

|makeresults|eval raw="<Cust> <Code>SEND</Code> <CountryCode>CN</CountryCode> <Lty> <CtyNm>BEIJING</CtyNm> <Zip>100176</Zip> </Lty> </Cust>
<Cust> <Code>RECEIVE</Code> <CountryCode>JP</CountryCode> <Lty> <CtyNm>TOKYO</CtyNm> <Zip>1000001</Zip> </Lty> </Cust>"
|rex field=raw max_match=0 "(?m)<Code>(?<Code>[^<]+).*<CountryCode>(?<CountryCode>[^<]+)"|eval merged=mvzip(Code,CountryCode,";;")|mvexpand merged| eval merged=split(merged,";;") | eval Code=mvindex(merged,0)| eval CountryCode=mvindex(merged,1)
0 Karma

krusovice
Path Finder

Thank you. I get some ideas with your answer.

0 Karma

493669
Super Champion

is this only one event? why you have not separated event on basis of tag

0 Karma
Get Updates on the Splunk Community!

Stay Connected: Your Guide to May Tech Talks, Office Hours, and Webinars!

Take a look below to explore our upcoming Community Office Hours, Tech Talks, and Webinars this month. This ...

They're back! Join the SplunkTrust and MVP at .conf24

With our highly anticipated annual conference, .conf, comes the fez-wearers you can trust! The SplunkTrust, as ...

Enterprise Security Content Update (ESCU) | New Releases

Last month, the Splunk Threat Research Team had two releases of new security content via the Enterprise ...