* build.xml:

Changed the way the Clover report is generated.
  Used the existing targets, instead of having to maintain two different sets
    of targets to build with or without Clover.
  Changed order of target dependencies for clover.tests so that instrumented
    classes are built before the test classes.
  Removed the separate target folder for instrumented classes.  If a separate
    target folder is used, separate classpaths are required when testing with
    Clover than when not using Clover.

* project.properties:
  Removed the target.clover.dir property definition.
  Updated clover.dbdir to put the DB somewhere safe.
  Set clover.excluded so as not instrument the test classes themselves.

project.properties

@@ -75,9 +75,8 @@ target.otherclasses.dir=${target.dir}/other-classes
 # Where to find code formatting rules
 jalopy.xml=jalopy.xml
 
-# Where we generate instrumented (Cloverised) classes
-target.clover.dir=${target.dir}/clover
-clover.dbdir=${target.clover.dir}
+# Clover properties
+clover.dbdir=${target.dir}/clover
 clover.dbfile=acegi_security_coverage.db
 clover.initstring=${clover.dbdir}/${clover.dbfile}
 target.clover.html.dir=${target.dir}/clover-reports
@@ -85,7 +84,7 @@ target.clover.html.dir=${target.dir}/clover-reports
 # Global property to exclude selected classes from Clover instrumentation.
 # Typically we'll use this to exclude experimental packages from coverage
 # analysis, or classes that are intended to support development.
-clover.excludes=
+clover.excludes=**/*Exception.java:**/*Tests.java
 
 # ------------------------------------------------------------------------ 
 #     docbook reference documentation