Merge two schemas into one in Apache nifi

Question

I'm trying to merge two csv files into a json using Apache nifi. Two csv's are persons.csv containing information about people:

Id|Name|Surname
ABC-123|John|Smith
ABC-111|Allan|Wood
ABC-001|Grace|Kelly

And the second csv contains list of events these people have attended:

EId|PId|Date|Desc
1|ABC-123|2017-05-01|"Groove party"
2|ABC-111|2017-06-01|"Snack No. One"
3|ABC-123|2017-06-01|"The night out"

I'm using a flow of (Nifi flow on git hub):

• GetFile
• UpdateAttribute (schema.name)
• Split Records
• ExtractText
• UpdateAttribute (correlation.id, newschema)
• Funnel
• MergeRecords / Merge Content
• PutFile

Trying to achieve final json:

{
"Person": {
    "Id": "ABC-123",
    "Name": "John",
    "Surname": "Smith",
    "Events": [{
        "Date": "2017-05-01",
        "Name": "Groove party"
    }, {
        "Date": "2017-06-01",
        "Name": "The night out"
    }]
}
}

But I'm not sure how to set up Merge Record, or how to join multiple csv lines after Merge Content into a single json. Is there a way how to do it?

Answer 1

You can actually achieve this using

1- ConvertRecord(CSV to JSON) - using Avro Schema in your case

CSVReader

JsonRecordSetWriter

AvroSchemaRegistry

   {
 "name": "person",
 "namespace": "nifi",
 "type": "record",
 "fields": [
 {"name": "Id" , "type" : "string"},
 {"name": "Name" , "type" : "string"},
 {"name": "Surname" , "type" : "string"}
 ]
}

• 2 - SplitJson
• 3 - EvaluateJson
• 4 - LookupAttribute (Get Desc SimpleCsvFileLookupService)
• 5 - LookupAttribute (Get Date SimpleCsvFileLookupService)
  • 6 - AttriutetoJson