Fall to cpu when && found in character class (#5658)

Signed-off-by: Anthony Chang <antchang@nvidia.com>
NVIDIA · May 26, 2022 · 4144c3d · 4144c3d
1 parent e2b3e00
commit 4144c3d
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 0 deletions.
diff --git a/sql-plugin/src/main/scala/com/nvidia/spark/rapids/RegexParser.scala b/sql-plugin/src/main/scala/com/nvidia/spark/rapids/RegexParser.scala
@@ -234,6 +234,14 @@ class RegexParser(pattern: String) {
                   RegexEscaped(ch)
                 case other => other
               }
+            case '&' => 
+              peek() match {
+                case Some('&') => 
+                  throw new RegexUnsupportedException("" +
+                    "cuDF does not support class intersection operator &&", Some(pos))
+                case _ => // ignore
+              }
+              RegexChar('&')
             case ch =>
               RegexChar(ch)
           }

diff --git a/tests/src/test/scala/com/nvidia/spark/rapids/RegularExpressionTranspilerSuite.scala b/tests/src/test/scala/com/nvidia/spark/rapids/RegularExpressionTranspilerSuite.scala
@@ -150,6 +150,15 @@ class RegularExpressionTranspilerSuite extends FunSuite with Arm {
         "cuDF does not support null characters in regular expressions"))
   }
 
+  test("cuDF does not support class intersection &&") {
+    val patterns = Seq("[a&&b]", "[&&1]")
+    patterns.foreach(pattern =>
+      assertUnsupported(pattern, RegexFindMode, 
+        "cuDF does not support class intersection operator &&"))
+
+    assertCpuGpuMatchesRegexpReplace(Seq("a&&b", "[a&b&c]"), Seq("a&&b", "b&c"))
+  }
+
   test("octal digits - find") {
     val patterns = Seq(raw"\07", raw"\077", raw"\0177", raw"\01772", raw"\0200", 
       raw"\0376", raw"\0377", raw"\02002")