confluentinc / confluentinc/confluent-sql
Inventory the `ddl` aspect referrents of ExecutionMode, propose simplification plan
- Dominant language
- Python
- Stars
- 6
- Forks
- 1
- Avg merge
- 21h 47m
- Merged PRs (30d)
- 26
Description
From @fdolce :
> So, I think we already discussed this at some point, but maybe it's worth thinking about it again after all the knowledge we gained. I'd love if the execution mode could simply be snapshot or streaming, and whether results can be fetched or not could be inferred from the info we get after we submit the statement. If we can infer what state we'll have to wait for from statement response, and whether it's possible to fetch results from either the statement response, or a failed call to the results endpoint, it would simplify the interface a lot. You don't have to think whether the query you'll send is DDL or not beforehand. You either want snapshot queries, or streaming queries, but whether you'll be able to fetch results or not only depends on the content of the query as seen by the server. Also, I don't remember if we noticed this (or even if it's there due to a request from james), but there's an execution_mode field in spec in the statement response in early_access
> It's also the experience you get in the SQL workspace. You can toggle snapshot or streaming, you don't have to specify whether it's ddl anywhere
Then somewhat misnamed things like `execute_streaming_ddl()` might could both be simplified or removed.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.