Hello, I think it is because the types of "Double" need to be "double".
If the provided schema is not a URI, then it calls Schema.Parser().parse(literal). -Bryan On Wed, Oct 14, 2015 at 7:08 AM, Karl Schulmeisters <[email protected]> wrote: > I am building my first NiFi flow and am testing it by feeding it a series > of > CSV files that are really one large Excell table broken up. I want to use > the CSV to Avro converter, but I cannot seem to get past the validation of > the schema in the processor. > > I'm providing this as the input schema, into the processor and it gives me > a > validation error > > > > {"namespace": "ERDF.avro", > > "type": "record", > > "name": "Meters", > > "fields": [ > > {"name": "Meterid", "type": "string"}, > > {"name": "City", "type": "string"}, > > {"name": "Lat", "type": "Double"}, > > {"name": "Lon", "type": "Double"}, > > {"name": "Readings", "type": ["int", "null"]} > > > > ] > > } > > > > Any documentation on the syntax for the Avro converter would be appreciated > >
