[Rivet] rivet and GENSER

[Andy Buckley]

On 19/02/10 07:53, Hendrik Hoeth wrote:
> Thus spake Anton Karneyeu (Anton.Karneyeu at cern.ch):
> 
>> I take a look on http://svn.hepforge.org/rivet/trunk/bin/ .
>> As I see you use xml.etree.cElementTree or xml.etree.ElementTree modules.
> 
> Or plain cElementTree.
> 
>> Unfortunately these are available only from Python 2.5 (see 
>> http://docs.python.org/library/xml.etree.elementtree.html ).
>>
>> SLC4 has Python 2.3.4, SLC5 has 2.4.3:
> 
> Oh joy. The two SL5 boxes which I use on a regular basis both have
> Python 2.4.3 but cElementTree is available. Why it's not on lxplus is
> beyond my understanding.
> 
>> So, the only way to use new Rivet 1.2.0 "from the box" on SLC, which I
>> see, is to change xml parser to xml.dom or xml.sax. Do you think this
>> is possible?
> 
> Do you think you could send us a patch that fixes this? You seem to know
> better what's needed than I do -- XML isn't really my field of
> experience.

I would rather not migrate to these (certainly not the SAX parser) since
they are *much* more cumbersome to use than ET, and we have to maintain
the code.

I see there is an old LCG AA area on AFS for Python packages, including
elementtree:
/afs/cern.ch/sw/lcg/external/python-packages/lcgaa3_python24/slc3_amd64_gcc344/site-packages/

Is there an updated version of this for > slc3? Alternatively, I note
that my ATLAS software setup gives me a copy of Python 2.5.4 -- as long
as there is an LCG version of Python which can be set up to run the
Rivet scripts, I'm not too bothered about the system copy not being
sufficiently up to date. After all, even the system *compilers* on SLC5
are too out of date for the experiments!

Andy

-- 
Dr Andy Buckley
SUPA Advanced Research Fellow
Particle Physics Experiment Group, University of Edinburgh