From 8d8725208771a8815a60160a5a30dc6ea87a7e6a Mon Sep 17 00:00:00 2001 From: Sean Owen Date: Sun, 7 Aug 2016 12:20:07 +0100 Subject: [SPARK-16409][SQL] regexp_extract with optional groups causes NPE ## What changes were proposed in this pull request? regexp_extract actually returns null when it shouldn't when a regex matches but the requested optional group did not. This makes it return an empty string, as apparently designed. ## How was this patch tested? Additional unit test Author: Sean Owen Closes #14504 from srowen/SPARK-16409. --- .../spark/sql/catalyst/expressions/regexpExpressions.scala | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) (limited to 'sql/catalyst/src/main/scala/org/apache') diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/regexpExpressions.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/regexpExpressions.scala index be82b3b8f4..d25da3fd58 100644 --- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/regexpExpressions.scala +++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/regexpExpressions.scala @@ -329,7 +329,12 @@ case class RegExpExtract(subject: Expression, regexp: Expression, idx: Expressio val m = pattern.matcher(s.toString) if (m.find) { val mr: MatchResult = m.toMatchResult - UTF8String.fromString(mr.group(r.asInstanceOf[Int])) + val group = mr.group(r.asInstanceOf[Int]) + if (group == null) { // Pattern matched, but not optional group + UTF8String.EMPTY_UTF8 + } else { + UTF8String.fromString(group) + } } else { UTF8String.EMPTY_UTF8 } @@ -367,7 +372,11 @@ case class RegExpExtract(subject: Expression, regexp: Expression, idx: Expressio ${termPattern}.matcher($subject.toString()); if (${matcher}.find()) { java.util.regex.MatchResult ${matchResult} = ${matcher}.toMatchResult(); - ${ev.value} = UTF8String.fromString(${matchResult}.group($idx)); + if (${matchResult}.group($idx) == null) { + ${ev.value} = UTF8String.EMPTY_UTF8; + } else { + ${ev.value} = UTF8String.fromString(${matchResult}.group($idx)); + } $setEvNotNull } else { ${ev.value} = UTF8String.EMPTY_UTF8; -- cgit v1.2.3