Flyangz opened a new issue, #1845:
URL: https://github.com/apache/auron/issues/1845
**Is your feature request related to a problem? Please describe.**
Common Sub-expression Elimination (CSE) is currently failing for queries
that access multiple indices of the same array transformation. This leads to
performance degradation, occasionally making Auron slower than vanilla Spark.
For example, in the following query:
```sql
select
split(content, ',')[0],
split(content, ',')[1]
from test_table
```
Spark identifies split(content, ',') as a common sub-expression. However,
Auron fails to handle the cached native result during the GetArrayItem
conversion.
**Describe the solution you'd like**
Update the `GetArrayItem` pattern matching in `NativeConverters`. It should
support scenarios where the ordinal index has already been converted into a
`NativeExprWrapper`.
**Additional context**
Current failure signature:
```
GetArrayItem(
child: NativeExprWrapper(ArrayType...),
ordinal: NativeExprWrapper(IntegerType...)
)
```
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
