typed entity registries #2236
typed entity registries #2236
Conversation
…ommon destination files
✅ Deploy Preview for dlt-hub-docs canceled.
|
| @@ -942,6 +942,98 @@ async def r_async(): | |||
| assert async_list in allowed_results | |||
|
|
|||
|
|
|||
| def test_limit_yield_cleanup() -> None: | |||
There was a problem hiding this comment.
this is a good test! i had briefly thought about wether the limit influences the incremental state but I think I forgot to add a test. There are a couple of print statements remaining further down here.
There was a problem hiding this comment.
I wrote this test because I realized that we cannot limit pg_replication this way which is doing yield from iterator but full iterators MUST be consumed. So we'd need to add some kind of groupings.
| AnySourceFactory = SourceFactory[Any, DltSource] | ||
|
|
||
|
|
||
| class SourceReference: |
There was a problem hiding this comment.
Maybe these classes are to different, but isn't there a common mechanism between the source and the destination reference that would warrant a mutual baseclass?
|
Very cool PR! I read all of it but as always with these huge PRs can't promise to understand every part of it, but some semi related or even not related observations are:
|
|
@sh-rp I've got all tests fixed I think |
Description
replaces #2183
also
sql_clientfor filesystempipeline.runon top of passing source instances