[Rivet] problem with yaml

[Ben Waugh]

Ah yes, there were a few more setup steps now that I consult my notes. 
The following worked for me for SLC5 for Rivet 1.2.0, where I had 
previously installed newer versions of Python and Swig in the given 
locations. Not saying this is best practice, but it is the point at 
which I stopped hacking:

--------------------------------------------------------------------------------
umask 022
cd /tmp/waugh
source /afs/cern.ch/sw/lcg/contrib/gcc/4.3/x86_64-slc5-gcc43-opt/setup.sh
export PYTHONHOME=/usr/local/python-2.5.5
export PATH=$PYTHONHOME/bin:$PATH
export LD_LIBRARY_PATH=$PYTHONHOME/lib:$LD_LIBRARY_PATH
export PYTHONPATH=$PYTHONHOME/lib/python2.5:$PYTHONPATH
export PATH=$HOME/local/swig-1.3.40/bin:$PATH
svn co http://svn.hepforge.org/rivet/tags/rivet-1.2.0
cd rivet-1.2.0/
autoreconf -iv
./configure --prefix=/unix/cedar/software/slc5/rivet-1.2.0 
--with-boost=/afs/cern.ch/sw/lcg/external/Boost/1.42.0_python2.5/x86_64-slc5-gcc43-opt 
 
--with-boost-incpath=/afs/cern.ch/sw/lcg/external/Boost/1.42.0_python2.5/x86_64-slc5-gcc43-opt/include/boost-1_42 
 
--with-hepmc=/afs/cern.ch/sw/lcg/external/HepMC/2.03.11/x86_64-slc5-gcc43-opt 
 
--with-fastjet=/afs/cern.ch/sw/lcg/external/fastjet/2.4.2/x86_64-slc5-gcc43-opt
make
make install
--------------------------------------------------------------------------------

Cheers,
Ben

On 01/02/11 14:00, Andy Buckley wrote:
> You have a mix of SLC4 and SLC5 paths in your setup. That's the sort of 
> thing that the LCG GCC setup script that Ben pointed at is meant to sort 
> out, but I think you will also need to source a similar setup script for 
> the SLC5 Python build, as the "slc4_" is appearing in Python paths in 
> the error message below.
> 
> Andy
> 
> 
> On 01/02/11 13:34, Robert King wrote:
>> Hi again,
>>
>> I have tried your suggestion and the compiler error is as follows:
>>
>> c4_ia32_gcc34/include/python2.5
>> In file included from
>> /afs/cern.ch/sw/lcg/external/Python/2.5.4p2/slc4_ia32_gcc34/include/python2.5/Python.h:57 
>>
>> <http://cern.ch/sw/lcg/external/Python/2.5.4p2/slc4_ia32_gcc34/include/python2.5/Python.h:57>, 
>>
>>                   from conftest_wrap.cxx:149:
>> /afs/cern.ch/sw/lcg/external/Python/2.5.4p2/slc4_ia32_gcc34/include/python2.5/pyport.h:761:2 
>>
>> <http://cern.ch/sw/lcg/external/Python/2.5.4p2/slc4_ia32_gcc34/include/python2.5/pyport.h:761:2>: 
>>
>> error: #error "LONG_BIT definition appears wrong for platform (bad
>> gcc/glibc config?)."
>>
>> If you have more suggestions I'd be happy to try them out, but I think I
>> have a working solution (use lxplus!) so this isn't high priority.
>>
>> Cheers
>>
>> Robert
>>
>> On 31 January 2011 18:30, Andy Buckley <andy.buckley at ed.ac.uk
>> <mailto:andy.buckley at ed.ac.uk>> wrote:
>>
>>     On 31/01/11 17:10, Robert King wrote:
>>      > Hi Andy, (& ccd to Rivet list)
>>      >
>>      > On 31 January 2011 17:30, Andy Buckley <andy.buckley at ed.ac.uk
>>     <mailto:andy.buckley at ed.ac.uk>
>>      > <mailto:andy.buckley at ed.ac.uk <mailto:andy.buckley at ed.ac.uk>>> 
>> wrote:
>>      >
>>      >     On 31/01/11 16:11, Robert King wrote:
>>      >
>>      >     Hi again Robert,
>>      >
>>      >     The lines above are the relevant ones from your config.log
>>     file -- it
>>      >     looks like SWIG is working but is not compatible with your
>>     C++ compiler.
>>      >     As I don't have that compiler I can't check, but if you want
>>     to chase it
>>      >     up, try doing a manual test of SWIG with a mini interface
>>     file. The test
>>      >     being run by the configure script in your case is roughly to
>>     make a file
>>      >     called conftest.i, containing the following:
>>      >
>>      >     %module conftest
>>      >     %{
>>      >      void foo();
>>      >     %}
>>      >     void foo();
>>      >
>>      >
>>      >     Then run
>>      >
>>      >
>>     
>> /afs/cern.ch/sw/lcg/external/swig/1.3.40/x86_64-slc5-gcc43-opt/bin/swig
>>     
>> <http://cern.ch/sw/lcg/external/swig/1.3.40/x86_64-slc5-gcc43-opt/bin/swig> 
>>
>>      >
>>     
>> <http://cern.ch/sw/lcg/external/swig/1.3.40/x86_64-slc5-gcc43-opt/bin/swig> 
>>
>>      >     -c++ -python conftest.i
>>      >
>>      >     and then
>>      >
>>      >
>>     /afs/cern.ch/sw/lcg/contrib/gcc/4.3.2/x86_64-slc5-gcc34-opt/bin/g++
>>     
>> <http://cern.ch/sw/lcg/contrib/gcc/4.3.2/x86_64-slc5-gcc34-opt/bin/g++>
>>      >
>>     
>> <http://cern.ch/sw/lcg/contrib/gcc/4.3.2/x86_64-slc5-gcc34-opt/bin/g++>
>>      >     -c
>>      >     conftest.i -Wno-format
>>      >
>>      >
>>      > I tried this, but it doesn't like line 1 of conftest.i
>>      > conftest.i:1: error: expected unqualified-id before ‘%’ token
>>      > Do I need to head this up as a proper C++ code? Sorry if this is
>>      > obvious, I'm not sure what we're doing here.
>>
>>     Oh, sorry: my cut & paste mistake. Substitute "conftest_wrap.cxx" for
>>     "conftest.i" in the C++ compiler command! That file is generated from
>>     conftest.i by SWIG: the output should be valid C++, but the input is
>>     not ;)
>>
>>     You'll find the prototype for this suggestion in the Rivet
>>     configure.ac <http://configure.ac>
>>     file, which defines the tests that are failing for you.
>>
>>     Good luck,
>>     Andy
>>
>>     --
>>     Dr Andy Buckley
>>     SUPA Advanced Research Fellow
>>     Particle Physics Experiment Group, University of Edinburgh
>>
>>     The University of Edinburgh is a charitable body, registered in
>>     Scotland, with registration number SC005336.
>>
>>
> 
> 

-- 
Dr Ben Waugh                                   Tel. +44 (0)20 7679 7223
Dept of Physics and Astronomy                  Internal: 37223
University College London
London WC1E 6BT