spark-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From rxin <...@git.apache.org>
Subject [GitHub] spark pull request: [SPARK-9157][SQL] codegen substring
Date Mon, 20 Jul 2015 23:08:48 GMT
Github user rxin commented on a diff in the pull request:

    https://github.com/apache/spark/pull/7534#discussion_r35054871
  
    --- Diff: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/stringOperations.scala
---
    @@ -628,24 +623,59 @@ case class Substring(str: Expression, pos: Expression, len: Expression)
     
       override def eval(input: InternalRow): Any = {
         val string = str.eval(input)
    -    val po = pos.eval(input)
    -    val ln = len.eval(input)
    -
    -    if ((string == null) || (po == null) || (ln == null)) {
    -      null
    -    } else {
    -      val start = po.asInstanceOf[Int]
    -      val length = ln.asInstanceOf[Int]
    -      string match {
    -        case ba: Array[Byte] =>
    -          val (st, end) = slicePos(start, length, () => ba.length)
    -          ba.slice(st, end)
    -        case s: UTF8String =>
    +    if (string != null) {
    +      val po = pos.eval(input)
    +      if (po != null) {
    +        val ln = len.eval(input)
    +        if (ln != null) {
    +          val start = po.asInstanceOf[Int]
    +          val length = ln.asInstanceOf[Int]
    +          val s = string.asInstanceOf[UTF8String]
    --- End diff --
    
    slicepos is going to be really bad for performance, since it uses pattern matching and
creates tuples. can we remove the function?


---
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


Mime
View raw message