[Gdal-dev] Compositing warped images

[Robert Osfield]

Hi Frank,

On Thursday 29 January 2004 15:00, Frank Warmerdam wrote:
> No, and it is unlikely that RasterIO() (already a pretty complicated
> monster) will have that added.  The warper does have a concept of "nodata"
> values but I am loath to add that complication to RasterIO().
>...

I havn't dabbled with the code but I can certainly understand that its a 
complex task changing something that targets to many different source and 
destination forms.

> Actually, at some point I suppose I might be willing to add a
> RasterIOWithNoData() that just calls down to the RasterIO() method.  It
> wouldn't be any more efficient than what you are doing, but we could ensure
> all cases are taken care of properly and it would save implementors some
> work.

Wouldn't RasterIOWithNoData() be a no op? :-)

I'll need to code something similar to RasterIOWithNoData() at my end, do you 
think it woul make sense to convert desitination line by line or to just 
convert the whole region at once.   I'm inclined to do a single RasterIO call 
and then copy and filter in one go across to the final destination.  Whats 
you thoughts on the tradeoffs on performance/memory usage.

When I'm finished if I have something reusable I'll post it in.

Regards,
Robert.