Add a test for checking that refreshing a page does load the same WebUI page.

Add a test for checking that refreshing a page does load the same WebUI page.
BUG=57485
TEST=this one.
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=82950

[David Tseng, 2011-04-06 04:42:35]

Sending this out to see if it applies to the referenced bug.
•  I know that the Navigate calls within each test seems redundant, but when I
place the call in the SetUpTestFixture method, I'm getting a crash that I've not
yet investigated.
•  The RunJavascriptFunction() method now mirrors RunJavascriptTest because we
need for the calls to the js side be sync.  I'm open to adding async version of
RunJavascriptFunction(Async).

[Evan Stade, 2011-04-07 19:29:34]

http://codereview.chromium.org/6805003/diff/13/chrome/browser/ui/webui/settin...
File chrome/browser/ui/webui/settings_browsertest.cc (right):

http://codereview.chromium.org/6805003/diff/13/chrome/browser/ui/webui/settin...
chrome/browser/ui/webui/settings_browsertest.cc:69: virtual void
SetUpInProcessBrowserTestFixture() {
it seems this is called before BrowserMain and there is no good place to do
common test setup that depends on the browser already being initialized. Maybe
you can ask phajdan why this is.

http://codereview.chromium.org/6805003/diff/13/chrome/browser/ui/webui/settin...
chrome/browser/ui/webui/settings_browsertest.cc:101:
ASSERT_TRUE(RunJavascriptFunction("openUnderTheHood"));
these should all be one javascript test

[David Tseng, 2011-04-07 21:46:07]

http://codereview.chromium.org/6805003/diff/13/chrome/browser/ui/webui/settin...
> chrome/browser/ui/webui/settings_browsertest.cc:101:
> ASSERT_TRUE(RunJavascriptFunction("openUnderTheHood"));
> these should all be one javascript test
>
> http://codereview.chromium.org/6805003/
>
I'm concerned about the history.go(0) call; does that not refresh the
page (thereby wiping out the injected javascript tests)?

[Evan Stade, 2011-04-08 17:17:45]

http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
File chrome/test/data/webui/settings.js (right):

http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
chrome/test/data/webui/settings.js:17: }
can you also add a test to make sure the proper page is showing 

assertEquals(AdvancedPage.getInstance(), OptionsPage.getTopMostVisiblePage())

and might as well make sure document.location is correct

[David Tseng, 2011-04-19 17:16:55]

PTAL; wasn't sure how to get the claimed "document.location" from
class OptionsPage.

Otherwise, changed it so
-  we don't call console.error in test_api.js since that will
immediately failus on the C++ end.
-  pass all non-errors to regular processing.

On 4/8/11, estade@chromium.org <estade@chromium.org> wrote:
>
>
http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
> File chrome/test/data/webui/settings.js (right):
>
>
http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
> chrome/test/data/webui/settings.js:17: }
> can you also add a test to make sure the proper page is showing
>
> assertEquals(AdvancedPage.getInstance(),
> OptionsPage.getTopMostVisiblePage())
>
> and might as well make sure document.location is correct
>
> http://codereview.chromium.org/6805003/
>

[arv (Not doing code reviews), 2011-04-19 18:08:22]

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
File chrome/test/data/webui/settings_util.js (right):

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
chrome/test/data/webui/settings_util.js:6: function refreshPage() {
Can this be put in the same file as the test?

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
chrome/test/data/webui/settings_util.js:7: history.go(0);
This is strange. Please do not submit

[Evan Stade, 2011-04-19 18:08:33]

On 2011/04/19 17:16:55, David Tseng wrote:
> PTAL; wasn't sure how to get the claimed "document.location" from
> class OptionsPage.

just document.location.href

> 
> Otherwise, changed it so
> -  we don't call console.error in test_api.js since that will
> immediately failus on the C++ end.
> -  pass all non-errors to regular processing.
> 
> On 4/8/11, mailto:estade@chromium.org <mailto:estade@chromium.org> wrote:
> >
> >
>
http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
> > File chrome/test/data/webui/settings.js (right):
> >
> >
>
http://codereview.chromium.org/6805003/diff/13/chrome/test/data/webui/setting...
> > chrome/test/data/webui/settings.js:17: }
> > can you also add a test to make sure the proper page is showing
> >
> > assertEquals(AdvancedPage.getInstance(),
> > OptionsPage.getTopMostVisiblePage())
> >
> > and might as well make sure document.location is correct
> >
> > http://codereview.chromium.org/6805003/
> >

[David Tseng, 2011-04-19 19:05:36]

Also, refined the assertEquals() check; it now checks equality first with
non-strict (!=) and then checks strict (!==) type.  This gives better debug
output.

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
File chrome/test/data/webui/settings_util.js (right):

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
chrome/test/data/webui/settings_util.js:6: function refreshPage() {
On 2011/04/19 18:08:22, arv wrote:
> Can this be put in the same file as the test?

Done.

http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
chrome/test/data/webui/settings_util.js:7: history.go(0);
On 2011/04/19 18:08:22, arv wrote:
> This is strange. Please do not submit

I need a way to refresh a page within js; what's the proper way to do this?

[arv (Not doing code reviews), 2011-04-19 19:09:01]

On Tue, Apr 19, 2011 at 12:05,  <dtseng@chromium.org> wrote:
> Also, refined the assertEquals() check; it now checks equality first with
> non-strict (!=) and then checks strict (!==) type.  This gives better debug
> output.
>
>
>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
> File chrome/test/data/webui/settings_util.js (right):
>
>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
> chrome/test/data/webui/settings_util.js:6: function refreshPage() {
> On 2011/04/19 18:08:22, arv wrote:
>>
>> Can this be put in the same file as the test?
>
> Done.
>
>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
> chrome/test/data/webui/settings_util.js:7: history.go(0);
> On 2011/04/19 18:08:22, arv wrote:
>>
>> This is strange. Please do not submit
>
> I need a way to refresh a page within js; what's the proper way to do
> this?

window.location.reload()

is self documenting but the question still stands why this is needed?

>
> http://codereview.chromium.org/6805003/
>

[David Tseng, 2011-04-19 19:16:03]

It's necessary as this is a regression test:

BUG=57485
wherein we want to ensure that after refreshing a WebUI page, we are
directed properly to the *same* WebUI page/subpage
(chrome://settings/advanced) and not to chrome://settings.

On 4/19/11, Erik Arvidsson <arv@chromium.org> wrote:
> On Tue, Apr 19, 2011 at 12:05,  <dtseng@chromium.org> wrote:
>> Also, refined the assertEquals() check; it now checks equality first with
>> non-strict (!=) and then checks strict (!==) type.  This gives better
>> debug
>> output.
>>
>>
>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>> File chrome/test/data/webui/settings_util.js (right):
>>
>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>> chrome/test/data/webui/settings_util.js:6: function refreshPage() {
>> On 2011/04/19 18:08:22, arv wrote:
>>>
>>> Can this be put in the same file as the test?
>>
>> Done.
>>
>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>> chrome/test/data/webui/settings_util.js:7: history.go(0);
>> On 2011/04/19 18:08:22, arv wrote:
>>>
>>> This is strange. Please do not submit
>>
>> I need a way to refresh a page within js; what's the proper way to do
>> this?
>
> window.location.reload()
>
> is self documenting but the question still stands why this is needed?
>
>>
>> http://codereview.chromium.org/6805003/
>>
>

[David Tseng, 2011-04-19 19:20:57]

The scope might be larger than appropriate for the framework, but I'm
trying to simulate a user pressing f5/reload (through js).

On 4/19/11, David Tseng <dtseng@chromium.org> wrote:
> It's necessary as this is a regression test:
>
> BUG=57485
> wherein we want to ensure that after refreshing a WebUI page, we are
> directed properly to the *same* WebUI page/subpage
> (chrome://settings/advanced) and not to chrome://settings.
>
> On 4/19/11, Erik Arvidsson <arv@chromium.org> wrote:
>> On Tue, Apr 19, 2011 at 12:05,  <dtseng@chromium.org> wrote:
>>> Also, refined the assertEquals() check; it now checks equality first
>>> with
>>> non-strict (!=) and then checks strict (!==) type.  This gives better
>>> debug
>>> output.
>>>
>>>
>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>> File chrome/test/data/webui/settings_util.js (right):
>>>
>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>> chrome/test/data/webui/settings_util.js:6: function refreshPage() {
>>> On 2011/04/19 18:08:22, arv wrote:
>>>>
>>>> Can this be put in the same file as the test?
>>>
>>> Done.
>>>
>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>> chrome/test/data/webui/settings_util.js:7: history.go(0);
>>> On 2011/04/19 18:08:22, arv wrote:
>>>>
>>>> This is strange. Please do not submit
>>>
>>> I need a way to refresh a page within js; what's the proper way to do
>>> this?
>>
>> window.location.reload()
>>
>> is self documenting but the question still stands why this is needed?
>>
>>>
>>> http://codereview.chromium.org/6805003/
>>>
>>
>

[David Tseng, 2011-04-19 23:51:48]

ping.  I'd like to at least get the changes to the framework in (if
not hte test itself).

On 4/19/11, David Tseng <dtseng@chromium.org> wrote:
> The scope might be larger than appropriate for the framework, but I'm
> trying to simulate a user pressing f5/reload (through js).
>
> On 4/19/11, David Tseng <dtseng@chromium.org> wrote:
>> It's necessary as this is a regression test:
>>
>> BUG=57485
>> wherein we want to ensure that after refreshing a WebUI page, we are
>> directed properly to the *same* WebUI page/subpage
>> (chrome://settings/advanced) and not to chrome://settings.
>>
>> On 4/19/11, Erik Arvidsson <arv@chromium.org> wrote:
>>> On Tue, Apr 19, 2011 at 12:05,  <dtseng@chromium.org> wrote:
>>>> Also, refined the assertEquals() check; it now checks equality first
>>>> with
>>>> non-strict (!=) and then checks strict (!==) type.  This gives better
>>>> debug
>>>> output.
>>>>
>>>>
>>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>>> File chrome/test/data/webui/settings_util.js (right):
>>>>
>>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>>> chrome/test/data/webui/settings_util.js:6: function refreshPage() {
>>>> On 2011/04/19 18:08:22, arv wrote:
>>>>>
>>>>> Can this be put in the same file as the test?
>>>>
>>>> Done.
>>>>
>>>>
http://codereview.chromium.org/6805003/diff/13001/chrome/test/data/webui/sett...
>>>> chrome/test/data/webui/settings_util.js:7: history.go(0);
>>>> On 2011/04/19 18:08:22, arv wrote:
>>>>>
>>>>> This is strange. Please do not submit
>>>>
>>>> I need a way to refresh a page within js; what's the proper way to do
>>>> this?
>>>
>>> window.location.reload()
>>>
>>> is self documenting but the question still stands why this is needed?
>>>
>>>>
>>>> http://codereview.chromium.org/6805003/
>>>>
>>>
>>
>

[arv (Not doing code reviews), 2011-04-25 18:39:24]

LGTM