QuickCheck

QuickCheck is an implementation of the Haskell QuickCheck specification based test tool. Each specification is tested with generated values.

Basically quickcheck is about generators of data. (The QuickCheck runner method is just a fancy for loop implementation.) Quickcheck can help in scenarios where whole classes of test cases have to be tested and it is not feasible to write tests for all distinct test scenarios. (E.g.: "This algorithm has to work for unicode string values of unlimited size.")

Generators support different:

• types (primitive: int, byte, long, String, char, null, boolean; collections: arrays, lists; POJOS)
• value ranges (integer, chars)
• distinct values (uniqueValues, ensureValues, fixedValues, enumValues, clonedValues)
• sizes (String, Lists, Arrays)
• distributions
• generator strategies (composition of generator values (oneOf, frequency, list, array, nullsAnd), transformation, mutation)
• value frequencies (frequency, oneOf)
• determinism (random, deterministic )

Javadoc

Example

The following example asserts that a sorted list implementation is actual sorted. The values inserted into the list are arbitrary integer values.

public class SortedListTest {

   @Test
   public void sortedListCreation() {
      for (List<Integer> any : someLists(integers())) {
         SortedList sortedList = new SortedList(any);
         
         List<Integer> expected = sort(any);
         assertEquals(expected, sortedList.toList());
      }
   }

   private List<Integer> sort(List<Integer> any) {
      ArrayList<Integer> sorted = new ArrayList<Integer>(any);
      Collections.sort(sorted);
      return sorted;
   }
}

Releases

Quickcheck 0.4 June 3, 2009

Quickcheck 0.3 March 23, 2008

Quickcheck 0.2 July 23, 2007

Quickcheck 0.1 July 7, 2007

Snapshots

QuickCheck 0.4. snapshot November 23, 2008

QuickCheck 0.3. snapshot March 21, 2008

QuickCheck 0.2. snapshot July 15, 2007

Blog

Blog

Help

• General development help.
• Sample code.
  • The test code in Robert C. Martin's Clean Code offers plenty of opportunities to show Quickheck's strengths.
  • One possibility is to port the code from Alberto Savoia's Beautiful code chapter Beautiful Tests and make it beautiful with quickcheck.
• Documentation.

If you are interested please send an email to blob79 at dev.java.net.