Add LCIO support

[kilian]

After discussion with Frank Gaede, this should be similar to HepMC support.

There is the opportunity to extend or adapt LCIO if it is not capable to hold the complete event-record information from the generator.

[Juergen Reuter]

I was pushed now several times (even by the speaker in the Simulation/Software? Summary talk) to do that. So I think we got the message.

[Juergen Reuter]

Always producing bytecode garbage. Do not know how to wrap the code. Handing over the baton.

[kilian]

Let's see ...

[Juergen Reuter]

The problem with the bit garbage came from the fact that I had compiled LCIO with standard cmake settings (using the system compiler of MAC OS X). This is clang 6 for MAC OS X Yosemite. For the C++ Wrapper inside WHIZARD for LCIO I used g++ (4.9.2). This apparently gets the OSX hiccuped over the bind C string conversion from Fortran -> C -> C++ (LCIO-Wrapper) > clang (LCIO library) > C. Compiling all C++ and C code with the GNU compilers seems to work now. (I never tried to use clang for the C and C++ code inside WHIZARD). I'll take over again.

[Juergen Reuter]

As of r6514 LCIO is mainly implemented, the configure detects the LCIO and is correctly linked. Most classes are wrapped inside WHIZARD and - as it appears - valid LCIO files are written out. Missing details are the treatment of polarization, and the read in of LCIO files, as well as a possible ASCII shower for LCIO events from WHIZARD. Most of these points seem doable until. 2015-02-06.

[Juergen Reuter]

In r6535, LCIO is implemented in its basic concepts. Polarization has been included (up to now) in a similar way than for HepMC. Scale and alpha_s can be added to the events as specific parameters. LCIO can be written out and read in. There will most likely be more issues coming up when LCIO is tested by the experimental community. But closing this general ticket.