parse XML embedded in a field

[moneybox]

Hi there,

I have a CSV file with 2 fields : time,xml_data.
Is there anyway I can parse the xml_data field as XML ?

This is a nested XML inside a CSV field and I would prefer to parse it on index time (if not, parsing on search time is also acceptable).

Thanks

[nickhills]

I have a CSV file with 2 fields : time,xml_data

As long as you know the name of the field in the CSV which contains the XML spath will work at seach time

<your search which includes your csv events>| spath xml_data

[kamlesh_vaghela]

Hi @moneybox,

You can use spath for same. Check below search.

|inputlookup mylookup
| eval _raw=XML_DATA_FIELD 
| spath output=.. path=..
| table ..

Please check my sample search from below doc

https://docs.splunk.com/Documentation/SplunkCloud/6.6.3/SearchReference/Spath

| makeresults 
| eval A="<?xml version=\"1.0\">
<purchases>
   <book>
         <author>Martin, George R.R.</author>
         <title yearPublished=1996>A Game of Thrones</title>
         <title yearPublished=1998>A Clash of Kings</title>
  </book>
   <book>
         <author>Clarke, Susanna</author>
         <title yearPublished=2004>Jonathan Strange and Mr. Norrell</title>
   </book>
   <book>
         <author>Kay, Guy Gavriel</author>
         <title yearPublished=1990>Tigana</title>
   </book>
   <book>
         <author>Bujold, Lois McMasters</author>
         <title yearPublished=1986>The Warrior's Apprentice</title>
   </book>
</purchases>" 
| eval _raw=A 
| spath output=dates path=purchases.book.title{@yearPublished} 
| table dates

Thanks

[moneybox]

Thank you, but that works for cases I know what fields should I expect.
Is there anything that could automatically convert the nested XML to searchable fields in Index Time ?
Meaning, I want to extract all fields from the nested XML without knowing them.

Thanks again