search

confluentinc / kafka-connect-datagen

Connector that generates data for demos
Apache License 2.0
13 stars 84 forks source link

"price_per_unit" field of the "purchases" quickstart not deserializing correctly #100

Open chuck-confluent opened 2 years ago

chuck-confluent commented 2 years ago

I was just running datagen from Confluent Cloud today and noticed something odd. The price_per_unit field of the purchases data was showing values like

\t\u00f5

and

\u0007\u00B1

I looked up the unicode characters and they're just weird characters. Not sure what's going on.

chuck-confluent commented 2 years ago

This was with AVRO value format and confluent CLI version 2.5.0

ybyzek commented 2 years ago

Please see https://github.com/confluentinc/kafka-connect-datagen/blob/master/config/connector_purchases.config#L10, a config parameter that should display the price_per_unit field properly and was validated locally. PR where this was discussed: https://github.com/confluentinc/kafka-connect-datagen/pull/89 . I don't know if this config parameter is accepted for Datagen in CCloud.

chuck-confluent commented 2 years ago

It is not accepted in Confluent Cloud for avro value format. There is a parameter like this for json value format, but not for avro. Screenshot:

image