Make DataFrame API consuming (#4621) #4624
Conversation
| @@ -265,20 +265,18 @@ impl SessionContext { | |||
| if_not_exists, | |||
| or_replace, | |||
| }) => { | |||
| let input = Arc::try_unwrap(input).unwrap_or_else(|e| e.as_ref().clone()); | |||
There was a problem hiding this comment.
This is kind of unfortunate, I hope to rework this all as part of #4617
| @@ -431,7 +431,7 @@ mod tests { | |||
| ) | |||
| .await?; | |||
|
|
|||
| ctx.register_table("t1", ctx.table("test")?)?; | |||
| ctx.register_table("t1", Arc::new(ctx.table("test")?))?; | |||
There was a problem hiding this comment.
This API is still problematic as it leads to a reference cycle (#2659), I intend to address this as part of reworking these interfaces
|
FYI @andygrove I'm not sure if this has some implication for the python bindings? |
That's a good point .. thanks, I'll take a look (cc @isidentical) |
|
I think there a couple of places where we need to make some adjustments (and another interesting idea that comes with this is whether we want tests with nightly datafusion revisions on the datafusion-python) but (at least by a very light look) I don't think this should be a problem. |
|
Thank you for taking a look 👍 |
github-actions
bot
added
logical-expr
tustvold
force-pushed
the
make-dataframe-consuming
branch
from
ca93d5b
to
23f5e40
Compare
github-actions
bot
removed
logical-expr
|
I'm going to merge this in as I don't think it all that controversial and there are changes backed up on it, happy to address feedback in follow up PRs |
|
Benchmark runs are scheduled for baseline = 3611d91 and contender = 5c558e9. 5c558e9 is a master commit associated with this PR. Results will be available as each benchmark for each run completes. |
Which issue does this PR close?
Closes #4621
Part of #4349
Relates to #4617
Rationale for this change
What changes are included in this PR?
Are these changes tested?
Are there any user-facing changes?