Hi @SAMSAL , I'm using the CSV readers and writers in Query Record processor and CSVtoJSON Record Converter processor. I am expecting around 10 types of CSV files(I mean to say, files with different columns). But this number might increase in the future upto 50. Even though I know what will these 50 files exactly contain, I'll still have to write all the 50 schemas in a schema registry. Which is complex to maintain and work with. So I thought of this method in which I'll append all the values in a csv file with double quotes such that the reader and writer will consider it as a string. But it still continues to read it as an integer Input File: col1,col2,col3,col4,col5 999,C10,100,010,0 999,C06,10,010,0 This is what the CSV writer is giving: avro.schema {"type":"record","name":"nifiRecord","namespace":"org.apache.nifi","fields":[{"name":"col1","type":["int","null"]},{"name":"col2","type":["int","null"]},{"name":"col3","type":["string","null"]},{"name":"col4","type":["int","null"]},{"name":"col5","type":["int","null"]}]}
... View more
