[Rivet] Metadata: getting ahead of the game

[Andy Buckley]

Andy Buckley wrote:

> When you consider an analysis done, please strike-through its entry on
> the wiki page so we have some idea of how much progress has been made.
> Hopefully I'll be committing the Analysis.hh change in the next day or
> two, at which point no excuses will be accepted for tardiness ;)

Hendrik and Lars have admirably taken up the challenge of getting ahead
of the metadata deadline and we now have about 1/3 of the analyses done.
Since this is a great way to avoid actually breaking the build, please
get on with adding your metadata *now*, and I will only commit the
uncommented Analysis.hh when all analyses are done or someone is
obviously stalling ;)

I've made an executive decision to use restructuredText markup for the
strings returned by description() and runInfo(). I hope this isn't
overengineering --- reST is the standard doc format being used by the
Python project and others, and has flexible enough output that it should
make generating HTML, LaTeX and plain text easy, while still being very
readable as source. Check out http://sphinx.pocoo.org/rest.html and
related links for info: it's very simple. It's likely to only affect you
when formatting the bullet lists to be returned from runInfo().

Andy

PS. If you want to check the compatibility of your changes with the
interface-to-be, you'll have to uncomment the pure virtual methods in
Analysis.hh, and probably disable building unmigrated analyses in
src/Analyses/Makefile.am. Linking libRivetAnalysisStd will probably
still break, so just run 'make' from within the src/Analyses directory.
Make sure you don't accidentally commit these locally hacked files!

-- 
Dr Andy Buckley
Institute for Particle Physics Phenomenology
Durham University
0191 3343798 | 0191 3732613 | www.insectnation.org