What is the process to add support of Basic Normalization to Java Custom Destination connector?

Hello. I think you and I have the same situation. I found this thread and it seems they're talking about the same matter. Reading that, apparently we need to add the normalization in here. So it's like normalization is not available yet for custom destination connectors. (not 100% sure but it seems that way) For now what I'm doing is not following the standard. Like, usually you would follow the raw standard (

_airbyte_ab_id

,

_airbyte_emitted_at

and

_airbyte_data

) and then add basic normalization. But in this case, I just set my destination connector with the structure the streams define

Hi @Marlon Gómez Thank You for your reply. I modified normalization python files for supporting normalization for teradata but still it is failing. here the code. Oh, you defined your own streams configuration instead of default raw load. I am creating PR request and will take help from airbyte team.

Hi @Marlon Gómez, How did you set the destination connector with structure the streams define and how are you avoind raw standard?

@Satish Chinthanippu Well, I'm not very familiar with these templates, to be honest. I know they exist and kind of how they work. My approach could be brittle and rough, since it moves away from the standard, which it could be kind of bad, depending on how well you manage to handle everything. @Rahul Borse let's see for example this destination-firestore. So, you want to write your data using a writer implementation. If you see here, this method only has a

stream

and the

data

you want to store. From the

configuredCatalog

you can have access to the properties of each stream. If I'm not mistaken, that's what they do here I want to be clear that doing something like this breaks the pattern and contract that Airbyte is proposing. How bad can it get? I suppose it depends on the scenario. In a perfect world, you should try to understand those transformation templates and create your own