I am working on a tap that uses different paging and response object names for each endpoint. Is the intent of

client.py

to handle each streams parsing and post-processing conditionally within

client.py

or to break it out by stream in

streams.py

?

Ideally apis should be consistent and thus only require the base stream class in

client.py

to handle pagination, etc. but you can always override

get_next_page_token

,

parse_response

and

post_process

on every stream subclass

ok, thanks.

Exactly as @edgar_ramirez_mondragon. 👍 For most implementations, the “hard work” of dealing with the api can be addressed in

client.py

and then the schema definitions and special behavior of individual streams can be handled in

streams.py

.