public inbox for openembedded-core@lists.openembedded.org
 help / color / mirror / Atom feed
From: Joshua Lock <joshua.g.lock@linux.intel.com>
To: jwang <jing.j.wang@intel.com>, openembedded-core@lists.openembedded.org
Subject: Re: [PATCH 2/4] meta: implement key baserunner features
Date: Thu, 15 Sep 2016 14:15:41 +0100	[thread overview]
Message-ID: <1473945341.3558.33.camel@linux.intel.com> (raw)
In-Reply-To: <1473729455-32649-2-git-send-email-jing.j.wang@intel.com>

On Tue, 2016-09-13 at 09:17 +0800, jwang wrote:
> From: zjh <junhuix.zhang@intel.com>
> 
> Baserunner contains three features:
> 1. load cases from a manifest file
> 2. load cases from a package such as "oeqa.runtime"
> 3. create runner engine based on pyunit textrunner


I think this and 1/4 should probably be squashed together?

> 
> Signed-off-by: zjh <junhuix.zhang@intel.com>
> ---
>  meta/lib/base/baserunner.py | 44
> ++++++++++++++++++++++++++++++++++++++------
>  1 file changed, 38 insertions(+), 6 deletions(-)
> 
> diff --git a/meta/lib/base/baserunner.py
> b/meta/lib/base/baserunner.py
> index 56b838e..d59872f 100755
> --- a/meta/lib/base/baserunner.py
> +++ b/meta/lib/base/baserunner.py
> @@ -31,30 +31,62 @@ class FakeOptions(object):
>  class TestRunnerBase(object):
>      '''test runner base '''
>      def __init__(self, context=None):
> -        self.tclist = []
> +        self.testslist = []
>          self.runner = None
>          self.context = context if context else TestContext()
> +        self.test_options = None
>          self.test_result = None
>          self.run_time = None
>  
> +    def __del__(self):
> +        """
> +        Because unittest.TestCase is a class object, it will exist
> as long as the python virtual machine process.
> +        So tc can't be released if we don't release them explicitly.
> +        """
> +        if hasattr(unittest.TestCase, "tc"):
> +            delattr(unittest.TestCase, "tc")
> +
> +    @staticmethod
> +    def get_tc_from_manifest(fname):
> +        '''get tc list from manifest format '''
> +        with open(fname, "r") as f:
> +            tclist = [n.strip() for n in f.readlines() \
> +                                if n.strip() and not
> n.strip().startswith('#')]
> +        return tclist

It might be nice to handle open() failing here? If open() fails we're
trying to return an undefined instance.

>  
>      def configure(self, options=FakeOptions()):
>          '''configure before testing'''
> -        pass
> +        self.test_options = options
> +        self.runner = unittest.TextTestRunner(stream=sys.stderr, \

There's no need for a backslash here, we can rely on Python's implied
continuation.

> +                                                  verbosity=2)
>  
>      def result(self):
>          '''output test result '''
> -        pass
> +        return self.test_result
>  
>      def loadtest(self, names=None):
>          '''load test suite'''
> -        pass
> +        if names is None:

It's much more idiomatic to write these like:

	if not names:

> +            names = self.testslist
> +        testloader = unittest.TestLoader()
> +        tclist = []
> +        for name in names:
> +            tset = testloader.loadTestsFromName(name)
> +            if tset.countTestCases() > 0:
> +                tclist.append(tset)
> +            elif tset._tests == []:

variable names prefixed with an underscore are, by convention,
internal/private to the object.

Is there a case where countTestCases() might not be > 0 and _tests[] !=
[] ? i.e. can we just use an else here?

> +                tclist.append(testloader.discover(name, "[!_]*.py",
> os.path.curdir))
> +        return testloader.suiteClass(tclist)
>  
>      def runtest(self, testsuite):
>          '''run test suite'''
> -        pass
> +        starttime = time.time()
> +        self.test_result = self.runner.run(testsuite)
> +        self.run_time = time.time() - starttime
>  
>      def start(self, testsuite):
>          '''start testing'''
> -        pass
> +        setattr(unittest.TestCase, "tc", self.context)
> +        self.runtest(testsuite)
> +        self.result()
>  
> -- 
> 2.1.4
> 


  reply	other threads:[~2016-09-15 13:15 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-13  1:17 [PATCH 1/4] meta: introduce a small baserunner framework jwang
2016-09-13  1:17 ` [PATCH 2/4] meta: implement key baserunner features jwang
2016-09-15 13:15   ` Joshua Lock [this message]
2016-09-13  1:17 ` [PATCH 3/4] meta: use baserunner in oetest jwang
2016-09-15 13:15   ` Joshua Lock
2016-09-13  1:17 ` [PATCH 4/4] meta: modify runexported script to inherit the features from baserunner jwang
2016-09-15 13:15 ` [PATCH 1/4] meta: introduce a small baserunner framework Joshua Lock

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1473945341.3558.33.camel@linux.intel.com \
    --to=joshua.g.lock@linux.intel.com \
    --cc=jing.j.wang@intel.com \
    --cc=openembedded-core@lists.openembedded.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox