panbingkun
panbingkun
> We **shall not break** existing SQL user's code only because it is inconsistent to Scala/Python APIs. As I said let's document the existing behaviour of SQL func first of...
> @panbingkun I made slight modifications to your PR description for better readability: > > ``` > - ``` -> ```scala > - Added ` to escape $"str" > ```...
cc @cloud-fan @yaooqinn
- Yes, in the syntax `SHOW COLUMNS ... {FROM | IN} ... {FROM | IN} ...`, the second clause `{FROM | IN} ..` is actually unnecessary, but due to historical...
> @panbingkun can you update `Does this PR introduce any user-facing change`? I think it does affect v2 tables. Okay
> @panbingkun can you update `Does this PR introduce any user-facing change`? I think it does affect v2 tables. Updated.
> since it introduces a new error, shall we add a legacy config? Okay, let me to add it.
> since it introduces a new error, shall we add a legacy config? Updated. (`spark.sql.legacy.showColumnsCheckNamespace`)
> @yaooqinn you are right! We don't need a config then :) Okay, updated!
> Merged into master. Thanks @panbingkun @cloud-fan and @yaooqinn Thanks all @LuciferYang @cloud-fan @yaooqinn