Hi,   I think you were close with the last option but did not use the correct syntax with null values where instead of '<> null',  use 'is not null' as follows:   select * from FLOWFILE where RPATH(fields, '/field') is not null   However when I run that I got the correct result but for some reason a MapRecord Syntax is added to each array element :   [ { "fields" : [ "MapRecord[{field=field1}]", "MapRecord[{field=field1}]" ] }, { "fields" : [ "MapRecord[{field=field2}]", "MapRecord[{field=field2}]" ] }, { "fields" : [ "MapRecord[{field=field3}]", "MapRecord[{field=field3}]" ] } ] Not sure why this happens but that should get you close to what you are trying to accomplish. Hope that helps. If it does, please accept solution. Thanks   ... View more

Hi, I dont think there is a way to do that in one processor, check out this post around the problem with possible solution: https://community.cloudera.com/t5/Support-Questions/Unable-to-return-a-scalar-value-for-the-expression-from-NIFI/td-p/213677   If you find this helpful please accept solution. ... View more

Hi, Based on the link below , it doesnt seem like you can access the variable registry directly from an ExecuteScript processor. You can either use the API - as the referenced article in the link below- to retrieve the value or just use an UpdateAttribute processor to get the value of the variable using Expression Language and store into flowfile attribute before passing it to the groovy script where its easy to retrieve the incoming flowfile attribute. https://community.cloudera.com/t5/Support-Questions/NiFi-Is-it-possible-to-access-Processor-Group-variables/m-p/204283 If that helps please accept solution. Thanks ... View more

Hi,   Is it always the case that when you have one address you will have a map and if its multiple addresses then its an array? If that is true, then can use the EvaluateJsonPath processor to get the size of child elements of the "addresses" key with the following json path expression: $.addresses.length() If you have one address then the result is 1 otherwise >1. Then you can use RouteOnAttribute when the length is 1 to Jolt Transformation processor with the following spec to convert to Array: [ { "operation": "shift", "spec": { "addresses": { "*": "addresses.[].&" }, "*": "&" } } ]   If that helps please accept solution. Thanks     ... View more

Hi,   You dont have to use the ExtractText processor for this. Use the EvaluateJsonPath processor with the following configuration:     If you find this helpful please accept solution. Thanks ... View more

I dont think there is an out of the box processor where you can utilize such thing. However you can do some workaround where you can use the ExecuteSQL processor instead since this processor allows you to return the stored proc output in Avro format in new flowfile based on whatever your select statement is in the ExecuteSQL SQL Select Query property. Since this will generate new flowfile, the assumption here is that you dont care about the original flowfile.  before going farther and give you an example how to do it, do you want to preserve the original flow file and you were thinking of adding the stored proc output as attribute?   ... View more