Components
The core part of gdalcubes is implemented as a C++ library, linking to GDAL, NetCDF, SQLite, and libcurl.
Among others, the library provides data structures for image collections, raster data cubes, and implements the creation of data cubes and some functions on data cubes.
The R package targets at data scientists and provides an easy to use interface to the gdalcubes library. It is, however, mostly a wrapper, i.e. does not add much functionality. The package includes the sources of the core library and does not link to the separately built library.
Starting with gdalcubes version 0.6.4, the C++ library has been completely integrated into the R package source tree (under src/gdalcubes). There is no separate repository anymore and the URL https://github.com/appelmar/gdalcubes now points to the R package repository (including the C++ library).
An executable command line interface links to the gdalcubes library and provides functionality to create and modify image collections, and to execute chains of data cube operations. It is mainly used for debugging and profiling.
gdalcubes includes an experimental server application, exposing parts of the library functions as a REST-like web service. The web service is mainly used for distributed processing where server instances participate as worker nodes.
Figure 1 summarizes components of gdalcubes, how they are related, and their external dependencies.
Some external libraries are included in the sources, i.e., they will be compiled together with gdalcubes, and do not need to be built separately.
