Github user dongjoon-hyun commented on the issue:
https://github.com/apache/spark/pull/20208
Of course! Those command are an illustration of use case of schema
evolution in Hive-side in general.
This PR aims to provide *schema evolution* test coverage in Spark-side. As
you see, this is for only `sql/core`, not for `sql/hive` module at all.
Apache Spark 2.3 already supports the same schema evolution in terms of
**userGivenSchema**. (In Hive case, it's HMS schema instead of file schema.)
```scala
val df = spark.read
.schema(userGivenSchema)
.format(format)
.options(options)
.load(path)
```
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
