awesome! Just in time too, since we need complete tests for the py3k work happening now.
Committed revision 2009. One thing I noticed is that the collidedict moved from checking the key of the rect, to checking the value of the dict. The docs didn't specify if the key or the value was used. This makes sense since, rects can not be keys. However keys can be rect-like objects - tuples etc. I found only some code which used this(using google codesearch) - so we should still keep it using the keys for backwards compatibility. I added an argument to collidedict(adict, values=0) . So you can use the values, or the keys of a dict. It defaults to keys for backwards compatibility. What I didn't do yet, was change the src/rect.doc to add new argument to collidedict(adict, values=0) /collidedictall arguments. Will also need to clear up what they do with the keys/values - since currently the docs don't address that at all. Once the docs are changed you run the makeref.py script to update the docs. cheers, On Mon, Apr 13, 2009 at 6:19 AM, Sean Berry <[email protected]> wrote: > Hey, > > I finished up the 'todo' test cases for Rect. > > When doing them, I found a bug for rect.collidedict and > rect.collidedictall, it was checking the keys instead of the value, so > I did that quick fix. > > I did a svn diff > rect.patch (attached) > > Let me know if this is how one contributes, this is my first time and > I feel kind of awkward. > > -Sean >
