Bug 65877 – SVNTest.test_svn_apply depends on the timezone of the computer executing the test

• Component: Tools / Tests
• Status: RESOLVED FIXED

The test webkitpy.common.checkout.scm.scm_unittest.SVNTest.test_svn_apply depends on the timezone to execute correctly. It compares the modifications to ChangeLogs by svn-apply (which sets the date in PST8PDT http://trac.webkit.org/browser/trunk/Tools/Scripts/VCSUtils.pm#L1364) to the current local date. If the computer is not on the same day as the current date in PST8PDT, the test fails

Created attachment 103766 [details]
Patch

Comment on attachment 103766 [details]
Patch

just so my nice patch isn't lost :)

[Eric Seidel (no email)]

Comment on attachment 103766 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=103766&action=review

> Tools/Scripts/webkitpy/common/checkout/scm/scm_unittest.py:560
> +        if hasattr(time, 'tzset'):
> +            has_tz = 'TZ' in os.environ
> +            old_tz = os.environ.get('TZ', '')
> +            os.environ['TZ'] = 'PST8PDT'
> +            time.tzset()

I mean, we could make this less ugly by using with and an object.  We'd have to be careful to do it so it works on Python 2.5.  But something like the following:

with TimezoneOverride('PST8PDT'):
    changelog_entry = changelog_entry.replace(....)

class TimezoneOverride(object):
   def __init__(self, timezone_string):
        self.timezone_string = timezone_string

    def __enter__(self)
       // All the gunkto set up the timezone.

    def __exit__(self,type, value, traceback):
      // teardown.

It doesn't get rid of the ugly code, but at least it puts it in a place we can clean it up later if wefigure out how to.

Created attachment 104049 [details]
Patch

(In reply to comment #3)
> (From update of attachment 103766 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=103766&action=review
> 
> > Tools/Scripts/webkitpy/common/checkout/scm/scm_unittest.py:560
> > +        if hasattr(time, 'tzset'):
> > +            has_tz = 'TZ' in os.environ
> > +            old_tz = os.environ.get('TZ', '')
> > +            os.environ['TZ'] = 'PST8PDT'
> > +            time.tzset()
> 
> I mean, we could make this less ugly by using with and an object.  We'd have to be careful to do it so it works on Python 2.5.  But something like the following:
> 
> with TimezoneOverride('PST8PDT'):
>     changelog_entry = changelog_entry.replace(....)
> 
> class TimezoneOverride(object):
>    def __init__(self, timezone_string):
>         self.timezone_string = timezone_string
> 
>     def __enter__(self)
>        // All the gunkto set up the timezone.
> 
>     def __exit__(self,type, value, traceback):
>       // teardown.
> 
> It doesn't get rid of the ugly code, but at least it puts it in a place we can clean it up later if wefigure out how to.

Done

the necessary magic to get the with statement for python 2.5 was already part of scm_unittests.py

[Eric Seidel (no email)]

Comment on attachment 104049 [details]
Patch

Thank you!  Making the hack self-contained like this ia huge benefit for the long-term health of this code. :)

[WebKit Review Bot]

Comment on attachment 104049 [details]
Patch

Rejecting attachment 104049 [details] from commit-queue.

Failed to run "['./Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '--bot-id=ec2-cq-02', '--port..." exit_code: 2

Last 500 characters of output:
svg-fonts-word-spacing.html = IMAGE+TEXT

Regressions: Unexpected image mismatch : (5)
  fast/text/atsui-multiple-renderers.html = IMAGE
  fast/text/international/danda-space.html = IMAGE
  fast/text/international/thai-baht-space.html = IMAGE
  fast/text/international/thai-line-breaks.html = IMAGE
  platform/chromium-linux/fast/text/international/complex-joining-using-gpos.html = IMAGE

Regressions: Unexpected text diff mismatch : (1)
  editing/selection/collapse-selection-in-bidi.html = TEXT



Full output: http://queues.webkit.org/results/9420075

[Adam Barth]

Comment on attachment 104049 [details]
Patch

Lies!

[WebKit Review Bot]

Comment on attachment 104049 [details]
Patch

Rejecting attachment 104049 [details] from commit-queue.

Failed to run "['./Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '--bot-id=ec2-cq-02', '--port..." exit_code: 2

Last 500 characters of output:
nt): Merge conflict in Tools/ChangeLog
Failed to merge in the changes.
Patch failed at 0001 <rdar://problem/9971293> Mid-word breaks can occur in the middle of a surrogate pair

When you have resolved this problem run "git rebase --continue".
If you would prefer to skip this patch, instead run "git rebase --skip".
To restore the original branch and stop rebasing run "git rebase --abort".

rebase refs/remotes/origin/master: command returned error: 1

Died at Tools/Scripts/update-webkit line 146.

Full output: http://queues.webkit.org/results/9422140

Comment on attachment 104049 [details]
Patch

cq, why do you hate me?

[WebKit Review Bot]

Comment on attachment 104049 [details]
Patch

Clearing flags on attachment: 104049

Committed r93257: <http://trac.webkit.org/changeset/93257>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Tim 'mithro' Ansell]

*** Bug 31391 has been marked as a duplicate of this bug. ***