Splunk Search

Extract JSON objects

vishaltaneja070
Motivator

How can i extract this:
"properties": {"nextLink": null,
"columns": [
{"name": "Cost", "type": "Number"},
{"name": "Date", "type": "Number"},
{"name": "Charge", "type": "String"},
{"name": "Publisher", "type": "String"},
{"name": "Resource", "type": "String"},
{"name": "Resource", "type": "String"},
{"name": "Service", "type": "String"},
{"name": "Standard", "type": "String"},
"rows": [
[2.06, 20210807, "usage", "uuuu", "hhh", "gd", "bandwidth", "azy", "HHH"],
[2.206, 20210807, "usage", "uuuhhh", "ggg", "gd", "bandwidth", "new", "YYY"] ]

No of columns can be increased.

 

 

Labels (1)
Tags (1)
0 Karma
1 Solution

ITWhisperer
SplunkTrust
SplunkTrust

Assuming columns is supposed to be an array of name/type pairs (added closing ]) and that there are supposed to be 9 of these pairs (added Comment), and that you have a properly formatted JSON string (added surrounding and closing braces), then you could do something like this

 

| makeresults 
| eval _raw="{\"properties\": {\"nextLink\": null,
\"columns\": [
{\"name\": \"Cost\", \"type\": \"Number\"},
{\"name\": \"Date\", \"type\": \"Number\"},
{\"name\": \"Charge\", \"type\": \"String\"},
{\"name\": \"Publisher\", \"type\": \"String\"},
{\"name\": \"Resource\", \"type\": \"String\"},
{\"name\": \"Resource\", \"type\": \"String\"},
{\"name\": \"Service\", \"type\": \"String\"},
{\"name\": \"Standard\", \"type\": \"String\"},
{\"name\": \"Comment\", \"type\": \"String\"}],
\"rows\": [
[2.06, 20210807, \"usage\", \"uuuu\", \"hhh\", \"gd\", \"bandwidth\", \"azy\", \"HHH\"],
[2.206, 20210807, \"usage\", \"uuuhhh\", \"ggg\", \"gd\", \"bandwidth\", \"new\", \"YYY\"]] }}"



| spath path="properties.columns{}.name" output=columnnames
| spath path="properties.rows{}{}" output=rows
| streamstats count as event 
| mvexpand rows
| streamstats count as row by event
| eval index=(row-1)%mvcount(columnnames)
| eval name=mvindex(columnnames,index)
| eval {name}=rows
| eval row=floor((row-1)/mvcount(columnnames))
| fields - columnnames name index rows
| stats values(*) as * by row event

 

 

View solution in original post

0 Karma

ITWhisperer
SplunkTrust
SplunkTrust

Assuming columns is supposed to be an array of name/type pairs (added closing ]) and that there are supposed to be 9 of these pairs (added Comment), and that you have a properly formatted JSON string (added surrounding and closing braces), then you could do something like this

 

| makeresults 
| eval _raw="{\"properties\": {\"nextLink\": null,
\"columns\": [
{\"name\": \"Cost\", \"type\": \"Number\"},
{\"name\": \"Date\", \"type\": \"Number\"},
{\"name\": \"Charge\", \"type\": \"String\"},
{\"name\": \"Publisher\", \"type\": \"String\"},
{\"name\": \"Resource\", \"type\": \"String\"},
{\"name\": \"Resource\", \"type\": \"String\"},
{\"name\": \"Service\", \"type\": \"String\"},
{\"name\": \"Standard\", \"type\": \"String\"},
{\"name\": \"Comment\", \"type\": \"String\"}],
\"rows\": [
[2.06, 20210807, \"usage\", \"uuuu\", \"hhh\", \"gd\", \"bandwidth\", \"azy\", \"HHH\"],
[2.206, 20210807, \"usage\", \"uuuhhh\", \"ggg\", \"gd\", \"bandwidth\", \"new\", \"YYY\"]] }}"



| spath path="properties.columns{}.name" output=columnnames
| spath path="properties.rows{}{}" output=rows
| streamstats count as event 
| mvexpand rows
| streamstats count as row by event
| eval index=(row-1)%mvcount(columnnames)
| eval name=mvindex(columnnames,index)
| eval {name}=rows
| eval row=floor((row-1)/mvcount(columnnames))
| fields - columnnames name index rows
| stats values(*) as * by row event

 

 

0 Karma
Get Updates on the Splunk Community!

Say goodbye to manually analyzing phishing and malware threats with Splunk Attack ...

In today’s evolving threat landscape, we understand you’re constantly bombarded with phishing and malware ...

AppDynamics is now part of Splunk Ideas

Hello Splunkers, We have exciting news for you! AppDynamics has been added to the Splunk Ideas Portal. Which ...

Advanced Splunk Data Management Strategies

Join us on Wednesday, May 14, 2025, at 11 AM PDT / 2 PM EDT for an exclusive Tech Talk that delves into ...