Building notes.

Steps for making the shared library (now integrated into setup.py)

The below steps are now integrated into the setup.py file included and can be ignored, but for those who want to try for themselves I left the following sequences for you.

First the user must generate their own shared library of CSPICE. In the Lib subdirectory in CSPICE run the following commands:

ar -x cspice.a
ar -x csupport.a

This will generate a large collection of *.o files. Next compile the shared library, last I checked this was correct on my system.

gcc -shared -fPIC -lm *.o -o spice.so

A Note About Windows Support

Windows support is currently highly experimental and difficult for the author to test locally. If attempting to install on windows platforms, please ensure you have a recent version of Visual Studio is installed and ensure cl.exe and link.exe is available on the path. Given the variability of systems, I will not be able to diagnose most issues encountered with running SpiceyPy on Windows. Below is a semi-complete list of instructions for getting SpiceyPy built and installed. For a more complete but less readable guide follow the appveyor.yml file included in this distribution.

1. Ensure Visual Studio is properly installed and that cl.exe and link.exe are available on the path.

• If you run cl or link you should see some indication that you have done this correctly.

1. Ensure you have pip, numpy, pytest, and six installed.
2. Call vcvarsall.bat from your visual studio with the option “amd64” for 64 bit builds (I have not tested 32bit yet)
3. Run python setup.py install to install SpiceyPy (this will take a few minutes.)
4. You are done!

• You can run tests by running py.test test, ensure the root directory of SpiceyPy does not have a lengthy path as the spice function furnsh fails with long absolute paths.