patchchecker: trying to improve patches management


Warning: this code relies on the use of
  - a mail archive for the project list where patches are carried out
    and using MHonarc for the indexing (can probably be extended)
  - use of GIT as the SCM with the availabilitity of a local and
    kept up to date checkout of the project

How this works:
===============

  The indexer fetches the mail archives from the web and try to keep
  up to date, best it to run it in a crontab like:
  1 * * * * (cd $HOME/patchchecker ; ./indexer.py ) > $HOME/tmp/pck.log 2>&1
  this will progressively grow up a local XML database of all the
  mail informations, referers, and which ones have ACKs and reviews

  The patchchecker.py command itself is expected to be run from the command
  line manually will update informations from the git checkout, load
  the messages data from the indexer and will cross the informations
  from both and extract the useful data like patches waiting reviews

  The config modules tries to isolate the dependancies, it will check
  that a configuration file is present either as patchchecker.conf
  in the local dir or a .patchchecker one in the home directory

To get started
==============

  Copy the patchchecker.conf.exmpl to patchchecker.conf, edit
  [git]tree (and possibly [mail]archives if not for libvirt)
  launch the ./indexer.py a few times to verify it actually can
  load and parse messages.
  Then run patchchecker.py, it should fetch data from the git
  checkout, do some computations and give back the informations.

  For the curious have a look at the content of the XML database
  files. They are mostly used for caching, and should be easilly
  extendable to monitor more informations.

To submit reports/patches:
==========================

  Best is to use the libvirt list, see:
    https://www.redhat.com/mailman/listinfo/libvir-list

Thu Jun  9 16:07:24 CST 2011
Daniel Veillard <veillard@redhat.com>