Misc. cleanup:

Misc. cleanup:

* accessibility_win_browsertest.cc: Remove "using", simplify, fix style, eliminate auto_ptr, L"" -> std::wstring()
* browser_accessibility_win_unittest.cc: Add OVERRIDE, fix style, use scoped_ptr, declare variables just above first use, use (expected, actual) ordering

BUG=none
TEST=none

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=194257

[Peter Kasting, 2013-04-15 20:00:56]

There are a lot of diffs, but the change should have no functional effect.

[dmazzoni, 2013-04-15 21:51:43]

lgtm

Thanks!

https://codereview.chromium.org/14264015/diff/1/content/browser/accessibility...
File content/browser/accessibility/browser_accessibility_win_unittest.cc
(right):

https://codereview.chromium.org/14264015/diff/1/content/browser/accessibility...
content/browser/accessibility/browser_accessibility_win_unittest.cc:142:
ASSERT_EQ(initial_instances + 3,
This change looks like it will make it harder to understand test failures,
because there's no way to know what initial_instances was.

How about: ASSERT_EQ(3, CountedBrowserAccessibility::num_instances() -
initial_instances)

[Peter Kasting, 2013-04-15 23:33:40]

Landing.

https://codereview.chromium.org/14264015/diff/1/content/browser/accessibility...
File content/browser/accessibility/browser_accessibility_win_unittest.cc
(right):

https://codereview.chromium.org/14264015/diff/1/content/browser/accessibility...
content/browser/accessibility/browser_accessibility_win_unittest.cc:142:
ASSERT_EQ(initial_instances + 3,
On 2013/04/15 21:51:44, Dominic Mazzoni wrote:
> This change looks like it will make it harder to understand test failures,
> because there's no way to know what initial_instances was.
> 
> How about: ASSERT_EQ(3, CountedBrowserAccessibility::num_instances() -
> initial_instances)

I liked that idea and changed to that and then decided "heck, at this point the
original model makes sense" so I added a reset() function on
CountedBrowserAccessibility which sets the instance count to zero, thus leaving
the code looking similar to the original version.

[Peter Kasting, 2013-04-15 23:55:04]

Committed patchset #2 manually as r194257 (presubmit successful).