Subject: [ggl] spacial index construction interface
From: Bruno Lalande (bruno.lalande)
Date: 2011-03-08 04:54:35
> At the beginning it was jus a functor but later I realized that there
> should be equals() function. For other spatial indexes there will probably
> be other methods doing different things. Default 'translator' will have some
> number of methods used by different spatial indexes. So maby implementing it
> as a functor isn't the best idea. This is something like run-time traits.
> The name should probably refer to the 'index' or 'indexing'.
There are 2 possible approaches I think: either the translator defines the
whole set of specific functions which are necessary for the indexing
structure(s) to work, or those operations are defines at Indexable level and
all the translator has to do is to convert an index to its indexable. And
then the indexing structure calls whatever operation it needs on the
indexable itself. My approach was rather the latter. What would prevent us
from going that way?
-------------- next part --------------
An HTML attachment was scrubbed...
Geometry list run by mateusz at loskot.net