Also, I think there will be a high likelihood of format-specific processors like ValidateCSV going away in NiFi 2.0. I wouldn't start any new data engineering work using these processors.
On Wed, Jan 18, 2023 at 4:02 PM Pierre Villard <pierre.villard...@gmail.com> wrote: > > Hi, > > ValidateCSV was released before and is using a specific library that is > allowing you to define specific rules on what the CSV should contain (like > for a given column that should contain numerical values, confirm the values > are in a specific range, etc) [1]. ValidateRecord would only check things > against a schema (type of fields, column names). I do believe that > ValidateRecord + QueryRecord can achieve all of the features provided by > ValidateCSV. > > [1] > https://nifi.apache.org/docs/nifi-docs/components/org.apache.nifi/nifi-standard-nar/1.19.1/org.apache.nifi.processors.standard.ValidateCsv/additionalDetails.html > > HTH, > Pierre > > Le mer. 18 janv. 2023 à 21:50, Dan S <dsti...@gmail.com> a écrit : > > > I am trying to wrap my head around why there is a ValidateCsv processor > > when there is a ValidateRecord processor which can do I believe everything > > ValidateCsv can do and more. What is the community stance on this? Should > > only ValidateRecord be used or is there still a use case for using > > ValidateCsv? > >
