Github user MaxGekk commented on a diff in the pull request:
https://github.com/apache/spark/pull/22237#discussion_r216444450
--- Diff:
sql/core/src/test/scala/org/apache/spark/sql/JsonFunctionsSuite.scala ---
@@ -469,4 +470,26 @@ class JsonFunctionsSuite extends QueryTest with
SharedSQLContext {
checkAnswer(sql("""select json[0] from jsonTable"""), Seq(Row(null)))
}
+
+ test("from_json invalid json - check modes") {
+ val df = Seq("""{"a" 1}""", """{"a": 2}""").toDS()
+ val schema = new StructType().add("a", IntegerType)
+
+ checkAnswer(
+ df.select(from_json($"value", schema, Map("mode" -> "PERMISSIVE"))),
+ Row(Row(null)) :: Row(Row(2)) :: Nil)
+
+ val exception1 = intercept[SparkException] {
+ df.select(from_json($"value", schema, Map("mode" ->
"FAILFAST"))).collect()
+ }.getMessage
+ assert(exception1.contains(
+ "Malformed records are detected in record parsing. Parse Mode:
FAILFAST."))
+
+ val exception2 = intercept[AnalysisException] {
+ df.select(from_json($"value", schema, Map("mode" ->
"DROPMALFORMED"))).collect()
--- End diff --
I replaced it by `AnalysisException` but I think it is wrong decision.
Throwing of `AnalysisException` at run-time looks ugly:
```
Caused by: org.apache.spark.sql.AnalysisException: from_json() doesn't
support the DROPMALFORMED mode. Acceptable modes are PERMISSIVE and FAILFAST.;
at
org.apache.spark.sql.catalyst.expressions.JsonToStructs.parser$lzycompute(jsonExpressions.scala:568)
at
org.apache.spark.sql.catalyst.expressions.JsonToStructs.parser(jsonExpressions.scala:564)
...
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
```
I am going to replace it by something else or revert back to
`IllegalArgumentException`.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
