Jolt transformation based on unstable json format

Question

Hi, I am looking how to handle cases when json format may vary between different client responses. On response

{
"data": [
{
"ID": "123",
"Name": "Queen"
},
{
"ID": "456",
"Name": "King"
}
]
}

we apply following jolt transform:

[
{
"operation": "shift",
"spec": {
"data": {
"*": {
"ID": {
"456": {
"@(2,Name)": "Name"
}
}
}
}
}
}
]

but some times response is coming without leading "{"data":
and our transformation fails. So, generally, how can we query json with JOLT if we know for instance that Name and ID are next to each other somewhere in the structure?

PeterKa · Accepted Answer

Generally, I agree with the comments and JOLT is not so flexible for this purpose. In my case I did it with multiple matches for various responses, but still format is needed. I will close the ticket, thank you Andrés @vafs

vafs · Answer

Hello @PeterKa,

Thanks for reaching our Community.

For this, JOLT always need to have the values, it's not very much flexible if an expected value is missing.

But, something that JOLT offers is the default operation, that may help you to add "{"data": if it is already included.
You can see it here: https://lucabiscotti.github.io/jolt-guide.github.io/#default

That option may help you to have the needed format always before the shift.

Sign up

Please use your Cloudera account.

Login to the community

Please use your Cloudera account.