stefankandic commented on code in PR #45383:
URL: https://github.com/apache/spark/pull/45383#discussion_r1519324890


##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/TypeCoercion.scala:
##########
@@ -764,6 +773,94 @@ abstract class TypeCoercionBase {
     }
   }
 
+  object CollationTypeCasts extends TypeCoercionRule {
+    override def transform: PartialFunction[Expression, Expression] = {
+      case e if !e.childrenResolved => e
+
+      case b @ BinaryComparison(left, right) if shouldCast(Seq(left.dataType, 
right.dataType)) =>
+        val newChildren = collateToSingleType(Seq(left, right))
+        b.withNewChildren(newChildren)
+    }
+
+    def shouldCast(types: Seq[DataType]): Boolean = {
+      types.forall(_.isInstanceOf[StringType]) && types.distinct.length > 1
+    }
+
+    /**
+     *  Collates the input expressions to a single collation.
+     */
+    def collateToSingleType(exprs: Seq[Expression]): Seq[Expression] = {
+      val collationId = getOutputCollation(exprs)
+
+      exprs.map { expression =>
+        expression.dataType match {
+          case st: StringType if st.collationId == collationId =>
+            expression
+          case _: StringType =>
+            Cast(expression, StringType(collationId))
+        }
+      }
+    }
+
+    /**
+     * Based on the data types of the input expressions this method determines
+     * a collation type which the output will have.
+     */
+    def getOutputCollation(exprs: Seq[Expression], failOnIndeterminate: 
Boolean = true): Int = {
+      val explicitTypes = 
exprs.filter(hasExplicitCollation).map(_.dataType).distinct
+
+      explicitTypes.size match {
+        case 1 => explicitTypes.head.asInstanceOf[StringType].collationId
+        case size if size > 1 =>
+          throw QueryCompilationErrors
+            .explicitCollationMismatchError(
+              explicitTypes.map(t => t.asInstanceOf[StringType].typeName)
+            )
+        case _ =>
+          val dataTypes = exprs.map(_.dataType.asInstanceOf[StringType])
+
+          if (hasIndeterminate(dataTypes)) {
+            if (failOnIndeterminate) {

Review Comment:
   also a sidenote about failOnIndeterminate; currently if will result 
sometimes in analysis and sometimes in runtime. Is this okay and should we 
maybe only fail in one place?



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

Reply via email to