[Webkit-unassigned] [Bug 147800] Use python 'with' statement to make cleanup more robust and minor code cleaning

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Sun Aug 9 17:54:38 PDT 2015


https://bugs.webkit.org/show_bug.cgi?id=147800

--- Comment #6 from dewei_zhu at apple.com ---
Comment on attachment 258543
  --> https://bugs.webkit.org/attachment.cgi?id=258543
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=258543&action=review

>>>> Tools/Scripts/webkitpy/benchmark_runner/benchmark_runner.py:108
>>>>                      result = self._http_server_driver.fetch_result()
>>> 
>>> It looks like _browser_driver is not accessed anywhere else so I don't think it needs to be be on _runner at all.
>>> Ideally, I'd like see the code like this here:
>>> 
>>> with TestEnvironment(benchmark) as env, timeout(self._plan['timeout'])
>>>     result = env.fetch_results()
>> 
>> Browser driver is specified by platform and browser which are provided by the parameter of BenchmarkRunner.__init__. The solution I can think of currently is storing those values in a structure, and supplying this structure as  parameter of TextEnvironment.__init__ . Does that sounds good?
> 
> In that case, I don't think we want test_environment object all.
> It's better for both browser_driver and http_server_driver to independently implement __enter__ and __exit__ (without a wrapper class)
> so that we can directly instantiate those classes in this function.
> Otherwise, the extra data structures and abstractions will only increase the code complexity.

Yes, we should avoid introducing unnecessary abstractions. The problem with both browser_driver and http_server is that they are instantiated by the factory function, are we expecting something like "with BrowserDriverFactory.create(....), HttpServerDriverFactory.create(....): "?

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150810/83686261/attachment.html>


More information about the webkit-unassigned mailing list