alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2685457670
The tests are green again on main!
https://github.com/apache/datafusion/actions/runs/13545248421/job/37855153112
--
This is an automated message from the Apache Git Service.
To res
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2685167196
> Change in 46: `count_all()` expr_fn function now displayed as `count(1)`
rather than `count(*)` #14894
Thansk! Note I did file
- https://github.com/apache/datafusion/issues/
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2685102083
Thanks @alamb. I will file related issue as follow-up
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
alamb commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971612657
##
datafusion/core/tests/dataframe/mod.rs:
##
@@ -2455,7 +2455,7 @@ async fn test_count_wildcard_on_sort() -> Result<()> {
let ctx = create_join_context()?;
alamb commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971606770
##
datafusion/optimizer/tests/optimizer_integration.rs:
##
@@ -198,7 +198,7 @@ fn between_date32_plus_interval() -> Result<()> {
WHERE col_date32 between '1998-
alamb merged PR #14824:
URL: https://github.com/apache/datafusion/pull/14824
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@datafusi
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2684989262
Let's get the tests clean
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific c
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2684827953
NM I think things are clean now
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the spec
jayzhan211 commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971494414
##
datafusion/core/tests/dataframe/mod.rs:
##
@@ -2455,7 +2455,7 @@ async fn test_count_wildcard_on_sort() -> Result<()> {
let ctx = create_join_context()?
jayzhan211 commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971494414
##
datafusion/core/tests/dataframe/mod.rs:
##
@@ -2455,7 +2455,7 @@ async fn test_count_wildcard_on_sort() -> Result<()> {
let ctx = create_join_context()?
jayzhan211 commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971489661
##
datafusion/core/tests/dataframe/mod.rs:
##
@@ -2455,7 +2455,7 @@ async fn test_count_wildcard_on_sort() -> Result<()> {
let ctx = create_join_context()?
jayzhan211 commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971489661
##
datafusion/core/tests/dataframe/mod.rs:
##
@@ -2455,7 +2455,7 @@ async fn test_count_wildcard_on_sort() -> Result<()> {
let ctx = create_join_context()?
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2684776705
I think we need to update the datafusion-testing pin -- closing/reopening
this PR to rerun the tests to make sure
--
This is an automated message from the Apache Git Service.
To resp
alamb closed pull request #14824: fix duplicated schema name error from count
wildcard
URL: https://github.com/apache/datafusion/pull/14824
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specif
alamb commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1971470732
##
datafusion/expr/src/expr_rewriter/mod.rs:
##
@@ -286,6 +286,7 @@ pub struct NamePreserver {
/// If the qualified name of an expression is remembered, it will b
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2684762035
I just merged https://github.com/apache/datafusion-testing/pull/7
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and u
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2684734611
After merging https://github.com/apache/datafusion-testing/pull/7 and update
commit, I guess is good to go
--
This is an automated message from the Apache Git Service.
To respon
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2683457657
> > I think the issue is that the runner in
https://github.com/Omega359/sqllogictest-rs is based on an older version of the
sqllogictests than we use in datafusion.
> > I have an id
Omega359 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2683289068
> I think the issue is that the runner in
https://github.com/Omega359/sqllogictest-rs is based on an older version of the
sqllogictests than we use in datafusion.
>
> I have
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2683234475
I think the issue is that the runner in
https://github.com/Omega359/sqllogictest-rs is based on an older version of the
sqllogictests than we use in datafusion.
I have an idea
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2683198971
I am working on helping here as I think getting the tests back green is
quite importatn
--
This is an automated message from the Apache Git Service.
To respond to the message, please
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682380265
> error[E0407]: method `shutdown` is not a member of trait
`sqllogictest::AsyncDB`
This is likely related to a change made in the most recent version of
sqllogictests . Here
Omega359 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682252789
> Run with
`PG_URI="postgresql://postgres:postgres@$POSTGRES_HOST:$POSTGRES_PORT/db_test"
bash datafusion/sqllogictest/regenerate_sqlite_files.sh`
> error[E0407]: method `shu
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682233579
```
Running the sqllogictests with sqlite completion. This will take
approximately an hour to run
Updating git repository `https://github.com/Omega359/sqllogictest-rs.git
jonahgao commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682031390
> But execute it again comes out error
SQLite tests use a value-wise format. Their file headers have a control
directive: `control resultmode valuewise`
--
This is an autom
Omega359 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682099360
> We should probably document that somewhere in the sqllite readme file
https://github.com/apache/datafusion/blob/1fedb4e000293e3997b477d87d575f3a5453171e/datafusion/sqllogict
Omega359 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682111560
> Is it possible to run with `INCLUDE_SQLITE=true cargo test --profile
release-nonlto --test sqllogictests -- --complete` like what we have in
sqllogictest. Why do we have `control
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682110043
We convert count wildcard in planner which is before plan built unlike
previously we convert after plan is created. There are many changes required
--
This is an automated messa
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682104096
Is it possible to run with `INCLUDE_SQLITE=true cargo test --profile
release-nonlto --test sqllogictests -- --complete` like what we have in
sqllogictest. Why do we have `control
Omega359 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682103367
> anyone familiar with sqlite test?
>
> I update test with `--complete`, ends up the format like what we have in
datafusion `0 Null` and not
>
> ```
> 0
> Null
alamb commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2682063992
> > But execute it again comes out error
>
> SQLite tests use a value-wise format. Their file headers have a control
directive: `control resultmode valuewise`
I think you
jayzhan211 commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2681422325
anyone familiar with sqlite test?
I update test with `--complete`, ends up the format like what we have in
datafusion `0 Null`.
But execute it again comes out error
findepi commented on PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#issuecomment-2681390655
> I think we can fix this with the generated projections (and I think it is
what @jonahgao is implemented)
@alamb I am not sure what is the "generated projections"?
then, w
jayzhan211 commented on code in PR #14824:
URL: https://github.com/apache/datafusion/pull/14824#discussion_r1966519181
##
datafusion/physical-plan/src/aggregates/mod.rs:
##
@@ -1231,6 +1233,13 @@ fn evaluate(
expr: &[Arc],
batch: &RecordBatch,
) -> Result> {
+// h
34 matches
Mail list logo
