Github user HyukjinKwon commented on the issue: https://github.com/apache/spark/pull/22775 Actually, that usecase can more easily accomplished by simply inferring schema by JSON datasource. Yea, I indeed suggested that as workaround for this issue before. Let's say, `spark.read.json(df.select("json").as[String]).schema`. I know it's not super clear about the usecase of `schema_of_json` and actually that's also partly why I want to allow what we need for this expression now. @rxin, WDYT? This PR tries to allow what we only need for now. Let's say disallow: ```scala schema_of_json(column) ``` and only allow ```scala schema_of_json(literal) ``` because the main usecase is: ```scala from_json(schema_of_json(literal)) ``` and ```scala from_json(schema_of_json(column)) ``` is already not being supported. My judgement was `schema_of_json(column)` doesn't have to be exposed for now and actually want to have a talk with @MaxGekk about this when he comes back from his vacation.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org