|
Boost Testing : |
From: Rene Rivera (grafikrobot_at_[hidden])
Date: 2007-09-18 15:33:16
Sebastian Redl wrote:
> That's a pretty trivial change, but I'd need some data to actually test
> it. Where do I get regression XML from, and how do I run the
> transformations with all paths correct?
If you are Unix you can use the recent shell script I'm using to build
the full results
<http://svn.boost.org/trac/boost/browser/trunk/tools/regression/xsl_reports/build_results.sh>.
You could probably just use it to see how to run the python script that
builds the results. Also, there's a way to run the generation on local
XML results. Looking at the boost_wide_report.py script it looks like
you pass the --dont-collect-logs option. So if you want to run things
manually:
* Create a work dir, "boost-results", and cd to it.
* Get the trunk code (a link to your usual Boost checkout might work as
well), svn co https://svn.boost.org/svn/boost/trunk boost
* Create the ZIP results subtree, mkdir -p "incoming/trunk"
* Download some ZIP results from the ftp location
<ftp://fx.meta-comm.com/boost-regression/trunk> into the
"incoming/trunk" dir.
* Run the generation script:
python boost/tools/regression/xsl_reports/boost_wide_report.py \
--locate-root=. \
--tag=trunk \
--expected-results="boost/status/expected_results.xml" \
--failures-markup="boost/status/explicit-failures-markup.xml" \
--comment="" \
--user="" \
--dont-collect-logs
Note, the result files will not have the 'revision' property Beman
mentioned so you might want to manually edit the result file(s) to put
in a fake one for testing. But we also would need the XSL to work even
if the property isn't there since they will be a period before all tests
results have the property.
HTH.
-- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo