RE: Unicode test noise

["Jeroen Frijters" <jeroen@sumatra.nl>]

[-- Attachment #1: Type: text/plain, Size: 581 bytes --]

Michael Koch wrote:
> Am Montag, 31. Januar 2005 12:34 schrieb Jeroen Frijters:
> > Hi,
> >
> > Does anyone have any objections against this patch to eliminate the
> > noise that the unicode test generates?
> 
> Why don't you just skip the tests ? You change would make the tests 
> don't show up in summaries when they succeed.

I don't want to skip the tests, it's just that I don't think it is
meaningful to have that single test count as 3.5 million tests, and in
addition it makes the -verbose option incredibly slow.

How about this patch?

Regards,
Jeroen

[-- Attachment #2: unicode.txt --]
[-- Type: text/plain, Size: 1235 bytes --]

Index: gnu/testlet/java/lang/Character/UnicodeBase.java
===================================================================
RCS file: /cvs/mauve/mauve/gnu/testlet/java/lang/Character/UnicodeBase.java,v
retrieving revision 1.1
diff -u -r1.1 UnicodeBase.java
--- gnu/testlet/java/lang/Character/UnicodeBase.java	8 Apr 2004 16:05:36 -0000	1.1
+++ gnu/testlet/java/lang/Character/UnicodeBase.java	31 Jan 2005 12:13:11 -0000
@@ -43,7 +43,7 @@
   public int failures;
   public int tests;
   TestHarness harness;
-
+  private boolean failed;
 
   public UnicodeBase()
   {
@@ -266,17 +266,18 @@
 	
   protected void reportError( String what)
   {
-    harness.check(false, what);
+    failed = true;
+    harness.debug(what);
   }
 	
   protected void reportError( int ch, String what)
   {
-    harness.check(false, stringChar(ch) +" incorrectly reported as " + what);
+    failed = true;
+    harness.debug(stringChar(ch) +" incorrectly reported as " + what);
   }
 	
   protected void checkPassed()
   {
-    harness.check(true);
   }
 
   public boolean range(int mid, int low, int high)
@@ -675,6 +676,8 @@
 	    else checkPassed();
 	  }
       }
+
+    harness.check(!failed);
   }
 
 }