davidedelvento / qc Goto Github PK

Provide the option to not stop in case of failures, and instead logging and continue -- or better handled by the underlying test framework?? See a discussion at http://stackoverflow.com/questions/4732827/continuing-in-pythons-unittest-when-an-assertion-fails

At the moment, the printed report assumes that each test method is run once, whereas it is run tries time. Is it possible to have an (optional) summary, e.g. including the total number of tries per method and other infos. It needs to work in at least nose and py.test, better if with unittest too

$ nosetests examples/ex4_bookcase.py

..

Ran 2 tests in 0.039s

OK

The bisection-based test case reduction is great for finding a small failing case automatically. However, delta-debugging is better. See the following links for details:

        http://www.st.cs.uni-saarland.de/dd/
        http://delta.tigris.org/
        http://classes.soe.ucsc.edu/cmps290g/Winter04/lectures/flanagan-290g-8.pdf

Check if qc is compatible with py.test
Make it so, if it isn't
Make examples for py.test

The Python 3 support seems to be broken.

Just pulling this here from the README.

Review the default range (inherited from dbravender) some of which seems inappropriate, e.g. integer currently are by the default [0, 100] so no negatives are tested. Investigate if
[-sys.maxint-1, sys.maxint] would be a sensible default (may need to be
different for long integers).