Brendan Buhr created NIFI-13387:
-----------------------------------
Summary: ExcelReader - Avro Schema Registry Schema Implementation
Key: NIFI-13387
URL: https://issues.apache.org/jira/browse/NIFI-13387
Project: Apache NiFi
Issue Type: Bug
Affects Versions: 1.26.0, 1.25.0
Reporter: Brendan Buhr
Attachments: excel_record_reader_testcase.zip
As a user I need the ability to define a schema and apply it on the excel
record reader to ensure consistency regardless of column ordering
There is an issue when using the Avro Schema / Avro Schema Registry with the
excel record reader, since there is no implementation currently where you can
define the first row as a header row (Ticket Logged Here:
https://issues.apache.org/jira/browse/NIFI-12491) this translates that when you
apply a schema while the outcome appears correct since the schema column order
and the file column order align and the data is applied correctly the result
appears to be correct.
In the event that the order changes it will throw errors due to mismatches on
the data type for the column.
I believe that if the schema mapping with aliasing is correctly applied by
being able to set first row as header where the labels will align with the
schema like any other it should resolve this issue so if the ticket linked is
addressed it should resolve this as well.
I attached a little flow where I setup the reader with a few sample files to
run this and just test a basic file with an identifier, label, decimal, int and
then randomised the columns but kept the schema intact and it would errors on
the randomised file
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
