[ https://issues.apache.org/jira/browse/SPARK-18891?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15756842#comment-15756842 ]
Michal Šenkýř commented on SPARK-18891: --------------------------------------- [PR #16240|https://github.com/apache/spark/pull/16240] should be relevant. It adds deserialization support for arbitrary sequence types (serialization was already supported). The only thing not working yet is the {{Seq.toDS}} method due to implicits resolution ({{SparkSession.createDataset}} works fine). > Support for specific collection types > ------------------------------------- > > Key: SPARK-18891 > URL: https://issues.apache.org/jira/browse/SPARK-18891 > Project: Spark > Issue Type: Improvement > Components: SQL > Affects Versions: 1.6.3, 2.1.0 > Reporter: Michael Armbrust > Priority: Critical > > Encoders treat all collections the same (i.e. {{Seq}} vs {{List}}) which > force users to only define classes with the most generic type. > An [example > error|https://databricks-prod-cloudfront.cloud.databricks.com/public/4027ec902e239c93eaaa8714f173bcfc/1023043053387187/2398463439880241/2840265927289860/latest.html]: > {code} > case class SpecificCollection(aList: List[Int]) > Seq(SpecificCollection(1 :: Nil)).toDS().collect() > {code} > {code} > java.lang.RuntimeException: Error while decoding: > java.util.concurrent.ExecutionException: java.lang.Exception: failed to > compile: org.codehaus.commons.compiler.CompileException: File > 'generated.java', Line 98, Column 120: No applicable constructor/method found > for actual parameters "scala.collection.Seq"; candidates are: > "line29e7e4b1e36445baa3505b2e102aa86b29.$read$$iw$$iw$$iw$$iw$SpecificCollection(scala.collection.immutable.List)" > {code} -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org