[GitHub] spark pull request #16476: [SPARK-19084][SQL] Implement expression field

[gczsjdy <gi...@git.apache.org>]

Github user gczsjdy commented on a diff in the pull request:

    https://github.com/apache/spark/pull/16476#discussion_r106157856
  
    --- Diff: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/conditionalExpressions.scala ---
    @@ -340,3 +343,105 @@ object CaseKeyWhen {
         CaseWhen(cases, elseValue)
       }
     }
    +
    +/**
    + * A function that returns the index of expr in (expr1, expr2, ...) list or 0 if not found.
    + * It takes at least 2 parameters, and all parameters should be subtype of AtomicType or NullType.
    + * It's also acceptable to give parameters of different types. When the parameters have different
    + * types, comparing will be done based on type firstly. For example, ''999'' 's type is StringType,
    + * while 999's type is IntegerType, so that no further comparison need to be done since they have
    + * different types.
    + * If the search expression is NULL, the return value is 0 because NULL fails equality comparison
    + * with any value.
    + * To also point out, no implicit cast will be done in this expression.
    + */
    +// scalastyle:off line.size.limit
    +@ExpressionDescription(
    +  usage = "_FUNC_(expr, expr1, expr2, ...) - Returns the index of expr in the expr1, expr2, ... or 0 if not found.",
    +  extended = """
    +    Examples:
    +      > SELECT _FUNC_(10, 9, 3, 10, 4);
    +       3
    +      > SELECT _FUNC_('a', 'b', 'c', 'd', 'a');
    +       4
    +      > SELECT _FUNC_('999', 'a', 999, 9.99, '999');
    +       4
    +  """)
    +// scalastyle:on line.size.limit
    +case class Field(children: Seq[Expression]) extends Expression {
    +
    +  /** Even if expr is not found in (expr1, expr2, ...) list, the value will be 0, not null */
    +  override def nullable: Boolean = false
    +  override def foldable: Boolean = children.forall(_.foldable)
    +
    +  private lazy val ordering = TypeUtils.getInterpretedOrdering(children(0).dataType)
    +
    +  private val dataTypeMatchIndex: Array[Int] = children.zipWithIndex.tail.filter(
    +    _._1.dataType.sameType(children.head.dataType)).map(_._2).toArray
    +
    +  override def checkInputDataTypes(): TypeCheckResult = {
    --- End diff --
    
    I think I should add an interface(and`Field` mixin that) like `ImplicitCastInputTypes`, the difference is that, the new interface casts all parameter to the same type(at least, they can be cast to string type)\u3002
    In the new interface, I fist try to cast all parameters to NumericType(if don't do this, s"2.3" won't be determined equal to s"2.30"), if that fails, I cast all parameters to StringType.
    Do you think this is a good idea? @cloud-fan 


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---

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