minfeaturesize support at layer or class level

[woodbri@…]

I would like to be able to set a MINFEATURESIZE at the LAYER or CLASS level and
have it act to filter out small features. It would be acceptable to implement
the test on the features extents for performance and simplity of implemetation.

This is needed for example with a water polygon layer so as you zoom out and the
polygons reduce to 1 or 2 or N pixels, you can filter them from drawing.
Otherwise you see a cloud of pixels the make it look like there is a lot of
water when there is not. This will allow us to make more professional looking maps.

[sdlime]

Good idea. Makes most sense to me at the class level, but I can see where you 
might want in both places...

Steve

[sdlime]

Sorry, forgot about the feature freeze for 4.6. Marking as FUTURE. I would try 
then to add immediately to the 4.7 development version in a month or so and 
then backport to 4.6 once tested, and assuming the implementation is trivial 
(which it should be).

Steve

[dmorissette]

Let's try to get this in 6.0

[aboudreault]

I have added the support of minfeaturesize in the LAYER and CLASS scope in r10982. I think there is only the following case that might not work: when the layer has a different unit than the map. I will produce a test case tomorrow to test it.

[aboudreault]

I improved the last commit to reflect the proper behavior in r10998. Stephen, it would be nice if you could take a few minutes to test it.