[ImageJ-devel] Code duplication...

[Stephan Preibisch]

Hi,

Tobias and me have been thinking we could establish a new package

net.imglib2.function

in which (image) data classes are located that are not Img. This includes RealRandomAccessibles, RandomAccessibles, Iterables, IterableIntervals or InterableRealIntervals. 

The first example would be indeed the Constant ... anything including its Cursors/RandomAcesses.

What do you think?

Bye bye,
Steffi


On Mar 13, 2012, at 12:32 , Stephan Preibisch wrote:

> Hi Lee,
> 
> this was exactly the course of our discussion. It is not an Img, so we need to find a place where this stuff will end up. Actually it is a specialization of any "Function" that can assign a value to any position in space. It should therefore also be RealRandomAccessible ... 
> 
> I would think about it a bit more before we start ...
> 
> Nice greetings,
> Steffi
> 
> On Mar 13, 2012, at 11:45 , Lee Kamentsky wrote:
> 
>> Hmmm... mine is certainly a candidate for removal. It looks like it is used in a couple of places and one is pretty pedantic. It's nice to have the ConstantRandomAccessible for places that take that instead of a ConstantRandomAccess.
>> 
>> I have a ConstantRealRandomAccess too, in the net.imglib2.sampler.special package. That doesn't fit into net.imglib2.img.constant. What happens to it?
>> 
>> If someone else wants to make the modifications, it's fine with me. If I were to do it, I'd add ConstantRandomAccessible to net.imglib2.img.constant, remove ConstantRandomAccessible and fix the two places where it's used. Any volunteers? If not, I'll mod.
>> 
>> --Lee
>> 
>> On 3/13/2012 11:04 AM, Tobias Pietzsch wrote:
>>> Hi,
>>> 
>>> I just noticed that currently there are 3 (three!) ConstantRandomAccess implementations in ImgLib2
>>> 
>>> One by Albert in
>>> net.imglib2.script.img.fn
>>> 
>>> One by Steffi in
>>> net.imglib2.img.constant
>>> 
>>> And one by Lee (as an inner class) in
>>> net.imglib2.sampler.special.ConstantRandomAccess
>>> 
>>> I didn't look in detail, but essentially all of those are doing the same thing.  Could we maybe get rid of two of those?
>>> 
>>> best regards,
>>> Tobias
>> 
> 
> 
> _______________________________________________
> ImageJ-devel mailing list
> ImageJ-devel at imagej.net
> http://imagej.net/mailman/listinfo/imagej-devel