24. How to Copy Complex multi-array JSON into tabular format using mapping dataflow

In this video, we learnt how to Copy Complex multi-array JSON into tabular format using mapping dataflow

INPUT JSON:
{
"id": "XXXX",
"name": "XYZ",
"type": "cost",
"location": null,
"sku": null,
"eTag": null,
"properties": {
"nextLink": null,
"columns": [
{
"name": "Amount",
"type": "Number"
},
{
"name": "Date",
"type": "Number"
},
{
"name": "Group",
"type": "String"
},
{
"name": "Type",
"type": "String"
},
{
"name": "Location",
"type": "String"
},
{
"name": "Currency",
"type": "String"
}
],
"rows": [
[
0. 0012,
20220901,
"RAM",
"storage",
"EU",
"GBP"
],
[
0. 0015,
20220902,
"VM",
"Compute",
"EU",
"GBP"
],
[
0. 0017,
20220903,
"Disk",
"storage",
"EU",
"GBP"
]
]
}
}

#adf #azuredatafactory #datafactory #azuresynapse #azuresynapseanalytics

[andrespi]

Hi, At minute 21:40 with the iff you explain to us about the number of records, how can this number of records be obtained automatically based on what the json file contains?
Thanks for your help

[MohammedKhan-npdn]

Hi, What about in case the input has n of rows since you mentioned example for 3 array set.

[KentuckyJIm]

Thank you for this post! At the step where you create the group by using the surrogate (around 21 minutes), you first sort by column name. How do we know the second maintains the integrity of the original rows? That is, how do we know that Amount 1 and Currency 1 both came from the original row 1? Seems possible that ADF could jumble this up.