Try to stabilize sealed class completion tests

[mbien]

the new code completion tests added in #7966 have a high failure rate in CI. E.g https://github.com/apache/netbeans/actions/runs/12197542394?pr=8019 is still failing after 5 attempts.

I can't reproduce it locally unfortunately.

diff --git a/harness/nbjunit/src/org/netbeans/junit/NbTestCase.java b/harness/nbjunit/src/org/netbeans/junit/NbTestCase.java
index 2d9f96007c80..b70f164df48a 100644
--- a/harness/nbjunit/src/org/netbeans/junit/NbTestCase.java
+++ b/harness/nbjunit/src/org/netbeans/junit/NbTestCase.java
@@ -752,6 +752,8 @@ public static void assertFile(String message, File test, File pass, File diff, D
         } else {
             try {
                 if (diffImpl.diff(test, pass, diffFile)) {
+                    System.out.println("test: "+test+"\n"+Files.readString(test.toPath()));
+                    System.out.println("pass: "+pass+"\n"+Files.readString(pass.toPath()));
                     throw new AssertionFileFailedError(message+"\n diff: "+diffFile, null == diffFile ? "" : diffFile.getAbsolutePath());
                 }
             } catch (IOException e) {

showed

    [junit] test: /home/runner/work/netbeans/netbeans/java/java.completion/build/test/unit/work/o.n.m.j.c.J/ststfg/testSealedTypeSwitchTypeFilteringGuard.out
    [junit] I
    [junit] null
    [junit] com
    [junit] java
    [junit] javax
    [junit] org
    [junit] sun
    [junit] 
    [junit] pass: /home/runner/work/netbeans/netbeans/java/java.completion/build/test/unit/data/goldenfiles/org/netbeans/modules/java/completion/JavaCompletionTaskTest/17/sealedTypeSwitch.pass
    [junit] public static final EJ EJ.A
    [junit] public static final EJ EJ.B
    [junit] public static final EJ EJ.C
    [junit] CI
    [junit] CJ1
    [junit] CJ2
    [junit] EJ
    [junit] I
    [junit] J
    [junit] null
    [junit] com
    [junit] java
    [junit] javax
    [junit] org
    [junit] sun
    [junit] test
    [junit] 

which might indicate missed code scanning updates or other indexer race conditions

[mbien]

closing in favor of #8066