Merge lp:~deeptik/lava-test/openposix into lp:lava-test/0.0
Proposed by
Deepti B. Kalakeri
Status: | Merged |
---|---|
Merged at revision: | 33 |
Proposed branch: | lp:~deeptik/lava-test/openposix |
Merge into: | lp:lava-test/0.0 |
Diff against target: |
71 lines (+67/-0) 1 file modified
abrek/test_definitions/posixtestsuite.py (+67/-0) |
To merge this branch: | bzr merge lp:~deeptik/lava-test/openposix |
Related bugs: |
Reviewer | Review Type | Date Requested | Status |
---|---|---|---|
Paul Larson (community) | Approve | ||
Deepti B. Kalakeri (community) | Needs Resubmitting | ||
Review via email: mp+36014@code.launchpad.net |
Description of the change
This patch automates the installation, execution, and parsing of posix test suite. This patch has been tested on Manderick for installation/
To post a comment you must log in.
Please add the boilerplate gpl3 + linaro copyright at the top of the file. You can just copy/paste out of another file.
18 +URL="http:// sourceforge. net/projects/ posixtest/ files/posixtest /" \ -%s/posixtestsu ite-%s. tar.gz" % (VERSION, VERSION)
19 + "posixtestsuite
You may actually want to look at grabbing the LTP source rather than openposix. One thing I noticed is that openposix hasn't had a release since 2006, but several fixes have gone into the LTP version since then.
23 +PATTERN = "((?P<test_ case_id> \A(\w+[ /]+)+(\ d+[-]\d+ ):) (?P<message>\D+:)" \
24 + " (?P<result>\D+))"
this results in build, link, execution, etc being the "message". Is that what you intend? You don't have to have a message, and it might make more sense for this to be combined with the test_case_id. Not sure. I'll leave it to your judgement here as to what you think makes sense, based on looking at posixtest results.
46 + results[key] = results[ key].rstrip( "\n") key].rstrip( ":") key].rstrip( )
47 + results[key] = results[
48 + results[key] = results[
All this should not be necessary, adjust your regexp so that extraneous characters are outside of the ()'s
51 + self.results[ 'test_results' ].append( results)
This is causing you to get double entries for each test result
50 + results[ 'test_case_ id'] = test_case_ id.replace( "/", ".")
For bonus points, this is something we're going to have a problem with in other tests as well. I was just looking at this last week in relation to another test. I think we may be better off here making some changes to the way results are gathered to check for things like this, and handle bad characters automatically. That way, fewer tests would have to extend the AbrekResultsParser class. Feel free to take a look at this if you like, otherwise we can come back and fix this test once it's done.