Why does this PR show evidence of support for JSON from an S3 source when Airbyte Cloud can't support JSON?

I raised your question to the connector team, Erik. I’m not fully aware of all the changes in the S3 connector. However, since that contribution, the connector has undergone a significant refactor and has been migrated to the Airbyte CDK. It now uses the same structure as the other file connectors. Why it doesn’t support JSON and JSONL only the connector team can provide a better answer, let’s wait for their input here.

Hi @Erik Webb, you can select the jsonl format from the format dropdown when adding a stream to your configuration

Thanks for the help here, @[DEPRECATED] Marcos Marx and @Alexandre Girard (Airbyte). Json and Jsonl are not interchangeable, correct? We're getting an error trying to use the jsonl format to ingest json files.

each line should represent a json object

I'm also curious about your use case! JSONL would be one file with an array, with each entry in that array being moved as an Airbyte record. I interpret the ask of moving JSON files as, perhaps, you've got a directory full of objects, 1-per-file, that you want to move.

We actually got this up and running with the S3 source using the Jsonl format, as suggested. Some processing was needed once Airbyte loaded the data into Snowflake, including a lateral flatten to unnest individual transactions from the Plaid payload.

Great! Would you be able to share an example of what the plaid payload looks like?