[GitHub] [spark] LuciferYang commented on a diff in pull request #38867: [WIP] [SPARK-41234][SQL][PYTHON] Add functionality for array_insert

[GitBox <gi...@apache.org>]

LuciferYang commented on code in PR #38867:
URL: https://github.com/apache/spark/pull/38867#discussion_r1041915384


##########
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/collectionOperations.scala:
##########
@@ -4600,3 +4600,155 @@ case class ArrayExcept(left: Expression, right: Expression) extends ArrayBinaryL
   override protected def withNewChildrenInternal(
     newLeft: Expression, newRight: Expression): ArrayExcept = copy(left = newLeft, right = newRight)
 }
+
+@ExpressionDescription(
+  usage = "_FUNC_(x, pos, val) - Places val into index pos of array x (array indices start at 1)",
+  examples = """
+    Examples:
+      > SELECT _FUNC_(array(1, 2, 3, 4), 5, 5);
+       [1,2,3,4,5]
+      > SELECT _FUNC_(array(5, 3, 2, 1), 2, 4);
+       [5,4,3,2,1]
+  """,
+  group = "array_funcs",
+  since = "3.4.0")
+case class ArrayInsert(srcArrayExpr: Expression, posExpr: Expression, itemExpr: Expression)
+  extends TernaryExpression with ImplicitCastInputTypes with SupportQueryContext
+    with QueryErrorsBase {
+
+  @transient private lazy val elementType: DataType =
+    srcArrayExpr.dataType.asInstanceOf[ArrayType].elementType
+
+  override def dataType: DataType = srcArrayExpr.dataType
+  override def inputTypes: Seq[AbstractDataType] = {
+    (srcArrayExpr.dataType, posExpr.dataType, itemExpr.dataType) match {
+      case (ArrayType(e1, hasNull), e2: IntegralType, e3) if (e2 != LongType) =>
+        TypeCoercion.findTightestCommonType(e1, e3) match {
+          case Some(dt) => Seq(ArrayType(dt, hasNull), IntegerType, dt)
+          case _ => Seq.empty
+        }
+      case (e1, e2, e3) => Seq.empty
+    }
+    Seq.empty
+  }
+
+  override def checkInputDataTypes(): TypeCheckResult = {
+    (first.dataType, second.dataType, third.dataType) match {
+      case (_: ArrayType, e2, e3) if e2 != IntegerType =>
+        DataTypeMismatch(
+          errorSubClass = "UNEXPECTED_INPUT_TYPE",
+          messageParameters = Map(
+            "paramIndex" -> "2",
+            "requiredType" -> toSQLType(IntegerType),
+            "inputSql" -> toSQLExpr(second),
+            "inputType" -> toSQLType(second.dataType))
+        )
+      case _ => TypeCheckResult.TypeCheckSuccess
+    }
+  }
+
+  override def eval(input: InternalRow): Any = {
+    val arr = first.eval(input)
+    val pos = second.eval(input)
+    val item = third.eval(input)
+
+    if (arr == null || pos == null) {
+      null
+    } else {
+      nullSafeEval(arr, pos, item)
+    }
+  }
+
+  override def nullSafeEval(arr: Any, pos: Any, item: Any): Any = {
+    val baseArr = arr.asInstanceOf[ArrayData]
+    val posInt = pos.asInstanceOf[Int]
+
+    if (baseArr.numElements() + 1  > ByteArrayMethods.MAX_ROUNDED_ARRAY_LENGTH) {
+      throw QueryExecutionErrors.concatArraysWithElementsExceedLimitError(baseArr.numElements() + 1)
+    }
+    val validatedPosInt = this.getValidPosIndexOrThrow(posInt, baseArr.numElements())
+
+    if (validatedPosInt < 0 || validatedPosInt > baseArr.numElements()) {
+      null
+    } else {
+      val newArray = new Array[Any](arr.asInstanceOf[ArrayData].numElements() + 1)
+      arr.asInstanceOf[ArrayData].foreach(elementType, (i, v) =>
+        if (i >= validatedPosInt) {
+          newArray(i + 1) = v
+        } else {
+          newArray(i) = v
+        }
+      )
+      if (item != null) {
+        newArray(validatedPosInt) = item
+      }
+      new GenericArrayData(newArray)
+    }
+  }
+
+  private def getValidPosIndexOrThrow(pos: Int, numArrayElements: Int): Int = {
+    val validIdx = if (pos == 0) {

Review Comment:
   https://docs.snowflake.com/en/developer-guide/snowpark/reference/python/api/snowflake.snowpark.functions.array_insert.html
   
   For `pos` is 0, Is the behavior ``inserting at position 0 has the same effect as using array_prepend`` easy to implement?



-- 
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-unsubscribe@spark.apache.org

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


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