Use the scripts contained here to create a MinGW docker image from which you can
cross compile Libgda. The source files are "mounted" into the container in /srv/libgda and the compiled files
are installed (when running "make install") in the path provided as the 1st argument to the start.sh script.

* ./build.sh builds the docker image (named "libgda-mingw32"), and first checks and if necessary download the
  missing binary files (Win32/). Run without any argument.
* ./start.sh starts a docker container from the built image, the 1st argument being the installation directory (where
  the files are copied when running "make install"). It opens a shell where you
  can run (as root inside the container):
  # ./configure
  # make
  # make install
  # ./do_packages

WARNING: if you get a "permission denied" while running the configure script, or while runnin "make install",
	 then you probably need to change the SELinux context of Libgda's source directory or of the installation
	 directory (from the host and not the docker container) using: chcon -Rt svirt_sandbox_file_t <dirname>
